Volume 22, Issue 3 e8498
SCIENTIFIC OPINION
Open Access

Pest risk assessment of Leucinodes orbonalis for the European Union

EFSA Panel on Plant Health (PLH)

Corresponding Author

EFSA Panel on Plant Health (PLH)

Correspondence:[email protected]Search for more papers by this author
Claude BragardPaula BaptistaElisavet Chatzivassiliou

Elisavet Chatzivassiliou

Search for more papers by this author
Francesco Di SerioPaolo GonthierJosep Anton Jaques Miret

Josep Anton Jaques Miret

Search for more papers by this author
Annemarie Fejer Justesen

Annemarie Fejer Justesen

Search for more papers by this author
Alan MacLeodChrister Sven MagnussonPanagiotis MilonasJuan A. Navas-CortesStephen ParnellRoel PottingPhilippe Lucien Reignault

Philippe Lucien Reignault

Search for more papers by this author
Emilio StefaniHans-Hermann ThulkeAntonio Vicent CiveraJonathan YuenLucia ZappalàRichard MallyEwelina CzwienczekAlex GobbiJúlia López MercadalAndrea MaioranoOlaf Mosbach-SchulzMarco PautassoEugenio RossiGiuseppe StancanelliSara TramontiniWopke Van der Werf
First published: 12 March 2024
Adopted: 30 November 2023

Abstract

Following a request from the European Commission, the EFSA Panel on Plant Health performed a quantitative risk assessment of Leucinodes orbonalis (Lepidoptera: Crambidae), the eggplant fruit and shoot borer, for the EU. The assessment focused on potential pathways for entry, climatic conditions favouring establishment, spread and impact. Options for risk reduction are discussed but effectiveness was not quantified. L. orbonalis is a key pest of eggplant (aubergine/brinjal) in the Indian subcontinent and occurs throughout most of southern Asia with records mostly from India and Bangladesh. The main pathway of entry is fruit of solanaceous plants, primarily exotic varieties of eggplant, Solanum melongena and turkey berry, S. torvum. The trade in both commodities from Asia is small but nevertheless dwarfs the trade in other Solanum fruits from Asia (S. aethiopicum, S. anguivi, S. virginianum, S. aculeatissimum, S. undatum). Other Solanum fruits were therefore not further assessed as potential pathways. The trade in eggplant from Asia consists of special fruit types and caters mostly to niche markets in the EU, while most eggplant consumed in Europe is produced in southern European and northern African countries, where L. orbonalis does not occur. Using expert knowledge elicitation (EKE) and pathway modelling, the Panel estimated that approximately 3–670 infested fruit (90% certainty range, CR) of S. melongena or fruit bunches of S. torvum enter into regions of the EU that are suitable for L. orbonalis establishment each year. Based on CLIMEX modelling, and using two possible thresholds of ecoclimatic index (EI) to indicate uncertainty in establishment potential, climates favouring establishment occur mostly in southern Europe, where, based on human population, approximately 14% of the imported produce is distributed across NUTS2 regions where EI ≥ 30; or 23% of the produce is distributed where EI ≥ 15. Escape of adult moths occurs mostly from consumer waste. By analysing results of different scenarios for the proportion of S. melongena and S. torvum in the trade, and considering uncertainties in the climatic suitability of southern Europe, adult moth emergence in areas suitable for establishment is expected to vary between 84 individuals per year and one individual per 40 years (based on 90% CR in different scenarios). In the baseline scenario, 25% of the solanaceous fruit from Asia is S. torvum, 75% is S. melongena and EI ≥ 30 is required for establishment. After accounting for the chances of mating, host finding and establishment, the probability of a mated female establishing a founder population in the EU is less than 1 in 100,000 to about 1 event per 622 years (90% CR in baseline scenario). The waiting time until the first establishment is then 622 to more than 100,000 years (CR). If such a founder population were established, the moth is estimated to spread at a rate of 0.65–7.0 km per year after a lag phase of 5–92 years. The impact of the insect on the production of eggplant is estimated to be 0.67%–13% (CR) if growers take no specific action against the insect and 0.13%–1.9% if they do take targeted actions. Tomato (S. lycopersicum) and potato (S. tuberosum) are hosts of L. orbonalis, but the insect does not develop to maturity in tomato fruit, and it does not feed on potato tubers under field conditions; hence, damage to potato can only occur due to feeding on shoots. Tomato and potato are not preferred hosts; nevertheless, impact can occur if populations of L. orbonalis are high and preferred hosts are not available. The Panel did not assess this damage due to insufficient information.

SUMMARY

Following a request from the European Commission, the EFSA Panel on Plant Health performed a quantitative risk assessment of Leucinodes orbonalis (Lepidoptera: Crambidae), the eggplant fruit and shoot borer, for the EU. The assessment focused on potential pathways for entry, climatic conditions favouring establishment, spread and impact. Options for risk reduction are discussed but effectiveness was not quantified.

Leucinodes orbonalis is a key pest of eggplant (brinjal) in the Indian subcontinent and occurs throughout most of southern Asia with records mostly from India and Bangladesh. The main pathway of entry is fruit of Solanaceous plants, primarily exotic varieties of eggplant, Solanum melongena, turkey berry, S. torvum, but also including other Solanum species (S. aethiopicum, S. anguivi, S. virginianum, S. aculeatissimum, S. undatum). Interceptions have been reported on all these species, but S. melongena and S. torvum are the species with the greatest number of interceptions.

The Panel assessed information in the literature on hosts of L. orbonalis concluding that only species in the genus Solanum provide a pathway as the fruits of these species can be imported under current regulation, while viable larvae and pupae may be transported with the fruit of most species, though not with tomato (S. lycopersicum) as the fruit is too wet for the larvae to develop in. Also, the damage is restricted to species in the genus Solanum.

The trade in eggplant from Asia consists of special fruit types and caters mostly to niche markets in the EU. S. torvum is imported as bunches of small fruit that are used as a spice in exotic dishes. The trade in both commodities from Asia is small but nevertheless dwarfs the trade-in other Solanum fruit from Asia. These other Solanum fruits were therefore not further assessed as a pathway. The pathway most likely to provide a route for entry of L. orbonalis into the EU was judged to be fresh eggplant and fresh turkey berry from Asia.

Using expert knowledge elicitation (EKE) and pathway modelling, the Panel estimated that in the order of hundreds of thousands of fruit enter the EU each year. In the model, these fruits are distributed across the EU according to population, as the niche markets receiving these products are assumed to be homogeneously distributed across populations in the EU. NUTS regions where climatic conditions are conducive for establishment of L. orbonalis (median estimate with EI ≥ 15) receive approximately 427,000 transfer units; 90% CR approximately 237,000–715,000. With an EI threshold of 30, the number of fruits entering NUTS2 regions where parts are suitable for establishment drops to approximately 260,000 (90% CR approximately 144,000–436,000).

Infested fruit represent a small proportion of the total number of fruits entering the EU (in the order of one in a 2000 fruit are infested). The number of transfer units infested with live L. orbonalis entering NUTS2 areas with EI ≥ 15 is estimated to be approximately 175 per year (90% CR approximately 6–1100); using an EI threshold of 30, the median number of infested transfer units drops to approximately 105 per year (90% CR approximately 3–670). In the scenario where only turkey berry is imported, and using EI ≥ 15, the 95 percentile estimates 84 adults emerge in areas suitable for establishment. In contrast, where only eggplant is imported and using EI ≥ 30, the 5 percentile estimates one adult emerging in 40 years.

Climatic conditions are most suitable for establishment in parts of the southern EU, especially in Spain (Andalucía, Comunidad Valenciana and Extremadura), Portugal (Alentejo), Italy (Sicilia, Calabria, Puglia), Greece (Kriti) and Cyprus. When imports are allocated in proportion to the human population, between 14% and 23% of transfer units enter regions of the EU suitable for establishment (lower estimates based on EI ≥ 30, higher estimate based on EI ≥ 15). Of the infested units entering NUTS regions where EI ≥ 15 approximately 12% are discarded before reaching the final consumer and approximately 50% of infested units are discarded by the consumer (Appendix D). Further, 1.0% (median; 90% CR, 0.2%–1.9%) of larvae survive to adulthood and escape from commercial waste while a median of 5.2% (90% CR 0.98%–12.2%) escape from consumer household waste. When the resulting numbers of adults emerge across NUTS2 regions, the likelihood that a female will find a mate depends on the window of encounter in space and time. In combination with the likelihood that the subsequent progeny survives to initiate a founder population, the number of established founder populations was estimated to be 0.00014 per year (90% CR 0.00000–0.00264). Thus, the Panel would not expect new founder populations within the foreseeable future or the time horizon of 5 years of this assessment. The predicted waiting time between new founder populations is in the Panel's estimation at least approximately 380 years. Given such low estimates, the Panel did not proceed to quantitatively assess the effectiveness of risk reduction options targeting L. orbonalis. However, such options exist, for example, the production of eggplant and turkey berry fruit in pest-free places of production and designation of L. orbonalis as a quarantine species. The Panel found four notifications in Europhyt (in 2011 and 2012) mentioning that product infested with L. orbonalis was granted entry to the EU. The temporary emergency measures for L. orbonalis that were instated in October 2022 (Commission Implementing Regulation (EU) 2022/1941 of 13 October 2022) will have stopped this practice.

If L. orbonalis would be introduced into the EU, the Panel estimates that it would take between 5 and 92 years (90% CR; median 34.5 years) for populations to grow sufficiently before a steady rate of spread of approximately 2.3 km/year (90% CR 0.65–7.02 km/year) was reached.

In a scenario where L. orbonalis enters, establishes and spreads within the EU and the population reaches an approximate equilibrium such that EU farmers consider the organism a member of the general pest fauna, median eggplant yield losses are estimated to be 4.5% (90% CR 0.67%–13.0%) when no specific control measures are in place, and 0.54% (90% CR 0.13%–1.94%) when growers apply targeted pest control against L. orbonalis.

The Panel did not assess the potential of damage to potato and tomato, major hosts of L. orbonalis that are widely grown in the potential area of establishment. There is sparse information in the literature on damage to these two crops, even though they are widely grown in countries where L. orbonalis is a serious pest, particularly India and Bangladesh. This suggests that the damage is unimportant, though there are few papers that state the contrary. Potato and tomato are known to be incidental hosts of L. orbonalis, accepted in case insects cannot find their favoured hosts, in particular S. melongena. However, with proper control of L. orbonalis in its main host, important spillovers to potato and tomato are not expected. Based on the scant information available, the Panel judges there to be insufficient evidence to regard L. orbonalis as a threat to the production of potato and tomato in the EU.

This PRA on L. orbonalis has several uncertainties as the Panel was unable to find information on (i) specific trade data on the commodities that are a pathway for L. orbonalis, (ii) information on consignment sizes and inspection practices for all the EU countries importing S. melongena and S. torvum, (iii) production practices in the countries of origin for eggplant or turkey berry destined for the European market and (iv) specific data demonstrating the potential for damage to potato and tomato.

In conclusion, L. orbonalis arrives with current measures in the EU with produce from Asian countries exporting eggplant and turkey berry to the EU. The numbers of insects entering are so low that establishment is a very rare event and unlikely in the foreseeable future. Were the insect to establish, it would spread and after it would reach an equilibrium population in the potential area of establishment, which includes a major part of the production area of eggplant in the EU, it would cause damage and add to the pest complex in this crop. Measures are available to reduce the likelihood of entry and consequently establishment, spread and impact.

1 INTRODUCTION

1.1 Background and terms of reference as provided by the requestor

1.1.1 Background

The new Plant Health Regulation (EU) 2016/2031, on the protective measures against pests of plants, is applying from 14 December 2019. Conditions are laid down in this legislation in order for pests to qualify for listing as Union quarantine pests, protected zone quarantine pests or Union regulated non-quarantine pests. The lists of the EU regulated pests together with the associated import or internal movement requirements of commodities are included in Commission Implementing Regulation (EU) 2019/2072. Additionally, as stipulated in the Commission Implementing Regulation 2018/2019, certain commodities are provisionally prohibited to enter in the EU (high-risk plants, HRP). EFSA is performing the risk assessment of the dossiers submitted by exporting to the EU countries of the HRP commodities, as stipulated in Commission Implementing Regulation 2018/2018. Furthermore, EFSA has evaluated a number of requests from exporting to the EU countries for derogations from specific EU import requirements.

In line with the principles of the new plant health law, the European Commission with the Member States are discussing monthly the reports of the interceptions and the outbreaks of pests notified by the Member States. Notifications of an imminent danger from pests that may fulfil the conditions for inclusion in the list of the Union quarantine pest are included. Furthermore, EFSA has been performing horizon scanning of media and literature.

As a follow-up of the above-mentioned activities (reporting of interceptions and outbreaks, HRP, derogation requests and horizon scanning), a number of pests of concern have been identified. EFSA is requested to provide scientific opinions for these pests, in view of their potential inclusion in the lists of Commission Implementing Regulation (EU) 2019/2072 and the inclusion of specific import requirements for relevant host commodities, when deemed necessary.

1.1.2 Terms of reference (ToR)

EFSA is requested, pursuant to Article 29(1) of Regulation (EC) No 178/2002, to provide scientific opinions in the field of plant health.

EFSA is requested to deliver 50 pest categorisations for the pests listed in Annex 1A, 1B and 1D. Additionally, EFSA is requested to perform pest categorisations for the pests so far not regulated in the EU, identified as pests potentially associated with a commodity in the commodity risk assessments of the HRP dossiers (Annex 1C). Such pest categorisations are needed in the case where there are not available risk assessments for the EU.

When the pests of Annex 1A are qualifying as potential Union quarantine pests, EFSA should proceed to phase 2 risk assessment. The opinions should address entry pathways, spread, establishment, impact and include a risk reduction options analysis.

Additionally, EFSA is requested to develop further the quantitative methodology currently followed for risk assessment, in order to have the possibility to deliver an express risk assessment methodology. Such methodological development should take into account the EFSA Plant Health Panel Guidance on quantitative pest risk assessment and the experience obtained during its implementation for the Union candidate priority pests and for the likelihood of pest freedom at entry for the commodity risk assessment of high-risk plants.

Annex 1. List of pests.

A)

1. Amyelois transitella

2. Citripestis sagittiferella

3. Colletotrichum fructicola

4. Elasmopalpus lignosellus

5. Phlyctinus callosus

6. Resseliella citrifrugis

7. Retithrips syriacus

8. Xylella taiwanensis

E)

List of pests identified to develop further the quantitative risk assessment (phase 1 and phase 2) methodology followed for plant pests, to include in the assessments the effect of climate change for plant pests (for more details, see Annex 3).

1. Leucinodes orbonalis

2. Leucinodes pseudorbonalis

3. Xanthomonas citri pv. viticola

1.2 Interpretation of the terms of reference

Leucinodes orbonalis is one of the three plant pest species listed in Annex 1E of the terms of reference. The pest categorisation of L. orbonalis concluded that the species satisfies the EU criteria that are within the remit of EFSA to assess, for it to be regarded as a potential Union quarantine pest (EFSA PLH Panel, 2021). Hence, EFSA is to proceed to conduct phase two of the risk assessment.

Adults of L. orbonalis are relatively small moths with predominantly white wings, featuring a characteristic triangular brown patch at half the forewings' length and a dark grey patch at the centre of the outer forewing margin (Figure 1); their wingspan is approximately 25 mm. The species was scientifically described based on specimens from Bangladesh and Java (Indonesia) and occurs widely in southern Asia (Figure 2), confirmed by DNA data (Chang et al., 2014; Sagarbarria et al., 2018; Shashank et al., 2015). These DNA data, however, indicate the existence of three still undescribed species belonging to the genus Leucinodes in the Austral-Asian region: one in the northern part of Vietnam and two in the northern part of Australia (see Appendix A). Nonetheless, the Panel is almost certain that, in the available literature from Asia, L. orbonalis is correctly identified as this species.

Details are in the caption following the image
Leucinodes orbonalis adult female. © Donald Hobern – Flickr: Leucinodes orbonalis, CC BY 2.0, https://commons.wikimedia.org/w/index.php?curid=18001695
Details are in the caption following the image
Genetically and morphologically confirmed country-level occurrences of Leucinodes orbonalis (blue), and of the four African Leucinodes species intercepted in Europe, i.e. L. africensis (red), L. rimavallis (yellow), L. pseudorbonalis (green) and L. laisalis (grey). A single specimen of L. orbonalis has been found in Côte d'Ivoire (blue arrow and dot) outside its native Asian range. Map © Richard Mally.

Asia currently comprises seven species of moths placed in the genus Leucinodes: Leucinodes diaphana, L. labefactalis, L. melanopalis, L. orbonalis, L. perlucidalis, L. sigulalis and L. unilinealis (Nuss et al., 2003–2023). However, all species except L. orbonalis are misplaced in this genus and need to be transferred to different genera in Crambidae (R. Mally, pers. obs.). Adults of the misplaced species can be distinguished externally from L. orbonalis (R. Mally, pers. obs.), so that misidentification is unlikely. Host plants of the misplaced Asian Leucinodes species are unknown except for L. melanopalis, whose larvae are reported to feed on Ficus religiosa (Moraceae) and Anacardium occidentale (Anacardiaceae) (Robinson et al., 2023).

Historically, all African specimens morphologically appearing like L. orbonalis have been attributed to that species. However, Mally et al. (2015) did not find specimens of L. orbonalis in museum material originating from Sub-Saharan Africa. It is therefore extremely unlikely that L. orbonalis is present in Africa or only locally present through potential unintentional introduction from Asia (Figure 2). Mally et al. (2015) further discovered that the African specimens externally looking like L. orbonalis are in fact a complex of previously undescribed species. Several of these newly discovered species cannot be distinguished from the Asian L. orbonalis based on external morphology of the larvae or adults, and dissection of the male genitalia or analysis of the ‘DNA Barcode’ sequence is necessary for species identification. None of the literature reporting ‘L. orbonalis’ from Africa mentions identification efforts of the investigated African specimens by means of genitalia dissection and/or DNA sequences, and their correct identification is therefore almost impossible; the Panel is currently almost certain that all African specimens identified as L. orbonalis in African literature are misidentifications of the species first described by Mally et al. (2015) or of still undiscovered species. The literature on African Leucinodes published since 2015 appears to be largely unaware of the African species complex described by Mally et al. (2015).

Four of the African Leucinodes species have been intercepted in Europe and identified to species based on male genitalia morphology and/or DNA ‘Barcode’ sequences: L. africensis, L. pseudorbonalis and L. rimavallis, which are externally indistinguishable from each other and from the Asian L. orbonalis, and the greyish-brown L. laisalis, which was reported as Sceliodes laisalis or Daraba laisalis in earlier literature. The latter species has established in the south of Spain and Portugal, with observations from 1958 to 2023 (see Appendix E, Spread). The Asian L. orbonalis has mainly been intercepted from eggplant/brinjal (Solanum melongena), whereas the African Leucinodes species have mainly been intercepted from bitter tomato (S. aethiopicum).

The Leucinodes (Lepidoptera: Crambidae) taxonomic specialist (Dr Richard Mally) is almost certain that Leucinodes orbonalis from Asia is correctly identified as such, and that the Asian literature on L. orbonalis can reliably be attributed to this species. It is extremely likely that African Leucinodes species looking like (and identified as) L. orbonalis are misidentifications of L. pseudorbonalis, L. africensis, L. rimavallis and L. kenyensis (all described as new in Mally et al., 2015, EFSA PLH Panel, 2021) or of still undescribed species (see ‘Leucinodes spp.’ in Mally et al., 2015); in the larval stage, L. orbonalis can furthermore be confused with the African L. laisalis. It is impossible to determine in retrospect which of the four known African species resembling L. orbonalis were studied in African literature; consequently, host plants and information on the biology reported in African literature cannot be reliably attributed to any of these four African species (unless voucher specimens were kept from the studies, which, however, none of the African studies indicated). Therefore, host plant records for L. orbonalis from countries outside of Asia might be incorrect, and in the case of African records may refer to any of the four known African look-alike species.

Given this situation, the request to conduct a quantitative pest risk assessment on L. orbonalis has been interpreted as a request to assess the risk from L. orbonalis using the very comprehensive literature from Asian populations. Literature from Africa purporting to concern L. orbonalis will not be considered but will be assessed in a separate opinion. Entry pathways in the current opinion will exclude Africa as a possible source of L. orbonalis as there is a lack of reliable evidence that the species occurs in Africa.

Taxonomic experts know of a single specimen of L. orbonalis, identified based on the male genitalia, that was intercepted on Solanum sp. imported in 2011 from Cote d'Ivoire to France (J.-M. Ramel, personal communication). The Panel considers the possibility that L. orbonalis may have been unintentionally introduced to Cote d'Ivoire through fruit imports from Asia, and that it may have established a founder population there. At the moment, however, this is speculation and requires further investigation.

Entry, establishment, spread and impact are to be quantitatively evaluated. An analysis of risk reduction options is also required. The Panel will therefore undertake a quantitative pest risk assessment according to the principles laid down in its guidance on quantitative pest risk assessment (EFSA PLH Panel, 2018) while recognising the need of the Commission for an express (i.e. as fast as possible) risk assessment.

In addition, as agreed with the Commission, the effect of climate change will not be examined for L. orbonalis as it has instead been assessed previously for Elasmopalpus lignosellus Zeller (Lepidoptera: Pyralidae) (EFSA PLH Panel, 2023). A brief discussion on consequences of climate change is given in Section 53.

2 DATA AND METHODOLOGIES

To obtain a deeper understanding of the organism and to inform the necessary steps in the risk assessment, a literature review was conducted using the Web of Science databases. The review built on the information collected for the pest categorisation (EFSA PLH Panel, 2021). The scientific and common names of the pest were used as search terms, no filters (limits) for either time of publication nor language were implemented and all Web of Science databases were selected. The following search string was used to retrieve results: Leucinodes orbonalis OR Leucinodes pseudorbonalis.

The Web of Science search resulted in 1293 hits after removal of duplicates. An additional search was conducted via the Google Scholar search engine to specifically find literature published in French, with the following French names inserted individually (with number of results in parentheses): foreuse des solanacées (8), perceuse de l'aubergine (2). The Web of Science search was conducted on 23 March 2021, and the Google Scholar search in August 2023. Of the altogether 2164 references found to mention Leucinodes, full texts of 583 references could not be retrieved.

Additional searches to retrieve additional specific documents cited in other literature were run when developing the opinion. The available scientific information, including the previous EFSA pest categorisation (EFSA PLH Panel, 2021) and the relevant literature and legislation, e.g. Regulation (EU) 2016/2031, Commission Implementing Regulation (EU) 2019/2072 and Commission Implementing Regulation (EU) 2022/1941 were taken into account.

In performing the risk assessment, the following assessment steps were distinguished after identifying appropriate pathways:
  • Estimating the number of infested host fruit that enter the EU,
  • Identifying the areas where L. orbonalis can establish in the EU,
  • Quantifying the number of host fruit entering NUTS2 areas of the EU where climatic conditions are suitable for establishment and where the pest could reproduce resulting in transfer to a host in those areas, leading to the initiation of a founder population,
  • Estimating the duration of the lag period before a founder population begins to spread as well as the steady rate of spread,
  • Estimating the potential loss in yield of solanaceous host crops in situations with and without specific pest management of L. orbonalis being used by farmers.

Judgements made in each assessment step were based on a combination of literature review, meta-analysis, information collected during interviews with hearing experts and expert knowledge elicitation (EKE) involving Panel members and EFSA staff to assess quantities that could not be well identified from the literature or databases alone (EFSA, 2014). To link commodity entry volumes into the EU with the assessment of establishment, imported commodities were distributed by apportioning relevant imported plant products to NUTS2 regions on the basis of the human population in each NUTS2 region, on the assumption that consumer demand is proportional to population size. Human population data were sourced from Eurostat (EFSA PLH Panel, 2018).

According to ISPM 5 (FAO, 2018), entry is ‘movement of a pest into an area where it is not yet present, or present but not widely distributed and being officially controlled’ while establishment is ‘perpetuation, for the foreseeable future, of a pest within an area after entry’. Introduction, according to the same ISPM 5, is ‘the entry of a pest resulting in its establishment’. In the assessment of entry, the Panel first identified pathways for entry of L. orbonalis into Europe, finding the main pathways to be Solanum produce, specifically exotic/special cultivars of eggplant, S. melongena and turkey berry, S. torvum. The volume of imports into the EU was estimated based on past imports, as well as the proportion of host fruit infested (Section 7). A pathway model was developed. Attention then shifted from pathway modelling of entry to identifying and mapping areas of the EU where establishment is possible following entry. Methods are described in detail in Rossi et al. (2023) – available on the Zenodo platform – and summarised in Sections 10, 14. After identification of the areas at risk using CLIMEX, SDM and Köppen–Geiger climate mapping, the pathway modelling was continued in Section 15. In this section, the entry flow is partitioned to parts of Europe suitable for establishment and not suitable for establishment. Transfer is modelled using a stochastic pathway model only for the areas where establishment is likely, assuming that no populations of L. orbonalis will be founded in areas that are not suitable for its establishment. Section 7 presents the overall pathway model for introduction, encompassing both entry and establishment.

2.1 Entry

2.1.1 Identifying pathways

Leucinodes orbonalis is an oligophagous pest that feeds on different plant species in the nightshade family (Solanaceae), with eggplant (Solanum melongena) being by far the most important and impacted plant species (EFSA PLH Panel, 2021). The larvae bore into the stems and fruits, weakening the host plant and rendering the fruits unfit for sale (Appendices B and F). The Panel compiled a list of host plants that are imported into the EU that could plausibly act as vehicles for entry (e.g. Table C.1 in Appendix C). Entry would require the importation of fruits or stems with eggs or pupae attached to the outside or with larvae feeding in or on the fruit. Stems are not a pathway because they are not traded as a commodity. Furthermore, it is not allowed to introduce plants for planting of Solanum or tubers of Solanum spp. into the EU from countries where L. orbonalis is present (EU regulation 2019/2072 Annex VI parts 15–18). EU regulation 2019/2072 Annex VI part 18 bans imports of plants for planting of Solanaceae. Imports of tubers of potato, S. tuberosum and other tuber forming species of Solanum are regulated by EU regulation 2019/2072 Annex IV, parts 15, 16 and 17. No such imports are allowed from countries in which L. orbonalis is present. Hence, efforts to identify plausible pathways focussed on (i) commodities on which interceptions had been found and (ii) hosts that are imported into the EU as fruits from countries (Bangladesh, Brunei Darussalam, China, Indonesia, India, Japan, Cambodia, Lao People's Democratic Republic, Sri Lanka, Myanmar-Burma, Malaysia, Nepal, Philippines, Pakistan, Singapore, Thailand, Taiwan, Vietnam) where L. orbonalis is known to occur.

Saudi Arabia, the United Arab Emirates and Australia were excluded from the list of countries where L. orbonalis is considered present. In Australia, L. orbonalis appears to be misidentified and likely represents one or two still undescribed species of Leucinodes (Appendix A). Leucinodes in the two Arabian countries are potentially one or several of the African species described in Mally et al. (2015).

Except for an isolated population in northern Vietnam that is genetically distinct from L. orbonalis but has to this date not been described as a separate species (Srinivasan Ramasamy and R. Mally, personal communication), all Leucinodes from Asian countries feeding on solanaceous species are considered to be L. orbonalis. The Asian species of Leucinodes that do not feed on solanaceous species are distinct from Leucinodes are incorrectly placed in this genus (R. Mally, pers. obs.); however, that does not affect this opinion as these species are not reported to feed on Solanum spp.

Interceptions: EU data of interceptions are shown in the pest categorisation (EFSA PLH Panel, 2021). However, some EU member states only make Europhyt notifications for interceptions of quarantine pests and L. orbonalis was listed as a quarantine pest in October 2022. Future trade flow of goods on which interceptions were found in the EU was estimated from Eurostat data. Eurostat aggregates some types of vegetable commodity imports and the accessible eight-digit codes do not specifically identify S. torvum. Nevertheless, the WG was able to identify and focus on the pathways most likely to lead to pest entry after excluding hosts whose import practice was judged unlikely to provide a pathway.

2.1.2 Scenario definitions for entry

An evidence dossier to support judgements of entry was developed based on literature review. The collected evidence is summarised in Appendix C: Entry was reviewed during the EKE to develop a pathway model for entry.

Estimates of the probability of units of the imported commodity being infested with L. orbonalis were made and uncertainties identified using expert judgement following EFSA guidance (Annex B.8 of EFSA Scientific Committee, 2018).

Scenario description: considering existing practices and phytosanitary measures

To estimate the number of host commodity units entering the EU infested with the pest, the Panel developed a general scenario with the following description:
  • The vegetable fruit of eggplant (S. melongena) and turkey berry (S. torvum) are considered the only significant possible pathway for introduction of L. orbonalis.
  • Most of the eggplant (S. melongena) consumed in the EU is produced and traded within the EU, especially in Spain and Italy.
  • Special and exotic varieties of eggplant (S. melongena) and turkey berry (S. torvum) are imported from Asian countries over the next few years in similar volumes and frequency to imports during the period 2010–2019. Data from 2020 to 2022 were not considered due to potential trade disruption during the Covid-19 pandemic. Data were sourced from Eurostat.
  • The proportion of infested fruit is based on information on production practices in countries of origin, literature on impact in countries of origin and the frequency of interceptions in the past.
  • In tropical and subtropical regions of Asia, L. orbonalis is reproducing year-round, with the potential of 10 and more overlapping generations per year (Appendix B: Biology).
  • Production and pest management: Eggplant is grown primarily in the open field, very rarely in protected conditions (greenhouses); in India, there is heavy use of pesticides (personal communication dr S. Ramasamy, A. Jovanovic); population control furthermore uses pheromone mass trapping, but light traps are avoided as these also attract other pests (personal communication Dr S. Ramasamy).
  • Post-harvest management: For local markets, post-harvest treatment is minimal, and fruits are sold within a few hours to days. Fruits determined for export: grading of fruits according to colour and size, storage for 1–4 days in shade at ambient temperature, or 7–10 days at 7–10°C and 85%–95% relative humidity (National Horticulture Board, 2023); sorting is done meticulously by farmers to avoid rejection at market (personal communication Dr S. Ramasamy).
  • Transport to EU: In containers via airplane, mostly in small quantities (too small to further split up before distribution in the EU) and in mixed consignments; purchasers are mostly restaurants and ethnic food shops (personal communication A. Jovanovic).
An estimate of imports of eggplant and turkey berry was determined from previous trade data. Unfortunately, Eurostat HS codes (8-digit resolution) do not discriminate between eggplants and turkey berry and they are combined within code HS 0709 3000. As such the PLH Panel cannot estimate the amount of each commodity imported individually. Instead, three simple scenarios were imagined:
  • Scenario 1: 75% of imports are eggplants, 25% turkey berry. This percentage split is based on number of interceptions on eggplants (222) and number of interceptions on turkey berry (66).
  • Scenario 2: 100% of imports are turkey berry, 0% are eggplants. While this is an unrealistic scenario it gives the highest likelihood of entry, representing a worst-case scenario.
  • Scenario 3: 0% of imports are turkey berry, 100% are eggplants. Again, this is an unrealistic scenario but allows the lower limit of entry to be determined.

In reality, the true proportions of imports are somewhere between Scenario 2 and 3. Scenario 1 was felt reasonable.

EKE was applied to answer a sequence of questions according to the pathway model of Figure 3:
  • What is the mean weight of each imported commodity (single S. melongena fruit or bunch (cluster) of S. torvum) (kg)? Estimates are then used to provide an estimate of the range of transfer units imported.
  • What is the proportion of infested transfer units entering the EU? The risk assessment used individual host fruit of S. melongena or bunches (clusters) of S. torvum as the most suitable unit for transfer unit because data are available on sampling procedures used for inspection at entry in the EU. The sampling protocols use eggplant (S. melongena) fruit as a unit of sampling.
Details are in the caption following the image
Conceptual diagram of pathway model to quantitatively estimate the likelihood of introduction of L. orbonalis into the EU. Blue is a parameter and grey is a variable (Numbers in brackets correspond to numbering of parameters in spreadsheet model (see Supporting materials – Annex A)). A mathematical description of the pathway model is given in Appendix C section Analysis and Appendix I. The Excel implementation of the pathway model, with a user-friendly presentation of the parameters and intermediate results of the calculation, is available in the supplementary materials to this opinion.

The uncertainties associated with the EKE were taken into account and quantified in the probability distribution applying the semi-formal method described in Section 3.5.2 of the EFSA-PLH Guidance on quantitative pest risk assessment (EFSA PLH Panel, 2018).

When results from the modelling of introduction were obtained which showed the very small likelihood of introduction in the foreseeable future for any of the three scenarios (import is composed of 100% S. melongena, 100% S. torvum or 75% S. melongena and 25% S. torvum), no specific scenarios for risk-reduction options were quantitatively evaluated. Available options are briefly mentioned (Sections 42 and 52).

2.2 Establishment

A detailed description of the methods used to assess the area in the EU where climatic conditions could support establishment of L. orbonalis is provided in Rossi et al. (2023). The assessment of establishment considered outdoor conditions only. Four methods were used to inform assessment of the area of potential establishment. (1) Mapping of Köppen–Geiger zones in Europe and in the area where the organism is distributed to evaluate whether the insect is present in climate types that are also present in the EU, (2) mapping of accumulated degree days and derived number of generations for Europe and Asia to evaluate whether the insect is present in Asia in areas with cumulated degree days that occur in Europe, (3) the CLIMEX modelling system (Kriticos et al., 2015), to analyse the potential for growth and persistence (i.e. establishment) of the insect and compare model output with known occurrence in Asia and then use the same parameters to show where growth and establishment could occur in the Europe-Mediterranean region and (4) species distribution modelling (SDM) which uses correlative methods to assess correspondence between predictor variables (e.g. climate variables) and the distribution of the insect in Asia, allowing to derive a climate suitability index for Europe. Developing SDM for the case of L. orbonalis provides a potential valuable comparison of SDM to CLIMEX which will inform its usefulness for exploring areas at risk for African Leucinodes species.

2.2.1 Literature search on the distribution and ecophysiology of Leucinodes orbonalis

An extensive literature search for pest distribution was conducted in Web of Science (all databases, excluding Data Citation Index and Zoological Record) and Scopus on 14 September 2022 (Rossi et al., 2023). The search string was based only on the scientific and common names of the pest. Other keywords such as ‘biology’, ‘physiology’ and ‘temperature’ were not used, so as not to limit the retrieval of distribution data, often reported as secondary information. The review followed a two-step approach for selecting relevant papers, the first step was based on screening the title and abstract of the paper, while the second step was based on the full-text analysis. A full description of the literature search methodology is available in Rossi et al. (2023).

2.2.2 Köppen–Geiger climate classification analysis

The SCAN-Clim tool (EFSA and Maiorano, 2022) was used to produce climate suitability maps based on the Köppen–Geiger climate classification approach. The re-analysis of Rubel et al. (2017) of the Köppen–Geiger climate classification from Kottek et al. (2006), considering the period 1986–2010 (available at http://koeppen-geiger.vu-wien.ac.at/present.htm), was used. The climate types present in the observed locations of L. orbonalis were identified and mapped. Because the PRA area is the EU, the output maps considered only climate types that are also present in the EU (Rossi et al., 2023).

2.2.3 Degree days and number of generations maps

Degree days and number of generations maps were developed for South-East Asia and Iran (area of distribution of L. orbonalis), and Europe and the Mediterranean Basin. Degree days were calculated by accumulating the positive daily differences between the daily mean temperature and a base temperature (BaseT) of 15°C. If daily mean temperature was below the base temperature, that day did not contribute to temperature accumulation. A linear regression model combining data from life-table studies (Dhaliwal & Aggarwal, 2021; Islam et al., 2020) was used to estimate the number of degree days to complete one generation. Degree days were calculated using the Copernicus ERA5-Land data (Muñoz Sabater et al., 2021) for the 30-year period from 1993 to 2022.

2.2.4 Species distribution modelling

The suitability of the EU territory to the establishment of L. orbonalis was analysed using a species distribution model ensemble developed in rStudio (rStudio Team) using the R sdm package (Naimi and Araújo, 2016). The bioclimatic variables from WorldClim (Table 1), for the period 1970–2000, at the resolution of 10 arcmin (~ 18 km ix 18 km), were used as predictor variables (Fick & Hijmans, 2017) (https://www.worldclim.org/data/worldclim21.html).

TABLE 1. Bioclimatic variables from WorldClim used in SDM (Fick & Hijmans, 2017).
Name Description
BIO2 Mean Diurnal Range (Mean of monthly (max temp–min temp))
BIO3 Isothermality (BIO2/BIO7) (×100)
BIO8 Mean Temperature of Wettest Quarter
BIO9 Mean Temperature of Driest Quarter
BIO13 Precipitation of Wettest Month
BIO15 Precipitation Seasonality (Coefficient of Variation)
BIO17 Precipitation of Driest Quarter
BIO18 Precipitation of Warmest Quarter
BIO19 Precipitation of Coldest Quarter

Pest distribution data at the point level were used in the analysis as presence-only data. These were thinned to include only one point, selected randomly in each grid cell with the same resolution of the predictor variables (10 arcmin or one sixth degree). The area for the training of the SDM model was between the geographical bounding-box 40° E − 165° E and 10° S–45° N.

Pseudo-absence data were generated inside the study area with two approaches. In the first approach, a modified convex hull polygon was created including an area that was assumed to be suitable for the organism based on known distribution. In the second approach, pseudo-absences were created in the entire training area, with the only limitation of a buffer of 10 km around each distribution point.

For the convex-hull approach, three series of simulations were run based on different number of pseudoabsence points: 4800 pseudo-absence points (10x number of observations), 2400 (5x) and 480 (1x) random points. To avoid collinearity among predictors, the variance inflation factor (VIF) method was used to exclude all the Bioclimatic variables with collinearity. Then, the species distribution modelling methods were fitted to the predictor variables.

Ten models were used to fit the data: bioclim, brt, cart, domain.dismo, gam, mars, maxent, rf, rpart and svm. Data splitting was achieved through a fivefold cross-validation process repeated for five times. Therefore, 25 simulations per model were created, yielding a total of 250 model runs. Ensemble modelling of the 250 simulations, based on the weighted average of the True Skill Statistics (TSS), was used to produce the final output.

2.2.5 CLIMEX

CLIMEX model (version 4.1.0.0, Kriticos et al., 2015) was used to investigate the climate suitability of the EU to the establishment of L. orbonalis. CLIMEX is based on the organism distribution and on its ecophysiological requirements to survive and complete the life cycle across a geographic region, given historic climate data. CLIMEX can assess the influence of weather-related stress factors (cold, heat, drought, humidity), and their interactions, on survival and growth through the calculation of growth-related indices and stress-related indices. The two groups of indices are combined into an Ecoclimatic Index (EI), which quantifies suitability for establishment of the pest. Simulations were run using the climate data set CM30 1995H V2 WO (Kriticos et al., 2012), package v4.1 (available at: https://www.climond.org/). This data set is based on the 0.5° world grid of historical meteorological data (30 years centred on 1995) originating from the Climate Research Unit (Norwich, UK), and transformed using the methods of Kriticos et al. (2012). Rossi et al. (2023) give more detail on how the parameters used in CLIMEX were determined.

The ecoclimatic index EI spans the integers from 0 to 100, where 0 means that a place is unsuitable for the organism, whereas 100 means a place is highly suitable. It is expected that with increasing EI, the density and impact of an organism will increase. According to Kriticos et al. (2015), a value of EI greater than 30 demarcates areas where climate is (very) favourable for the species whereas areas where EI < 30 are less favourable. They state, ‘An EI of more than 30 represents a very favourable climate for a species, as it means that during the (say) six months suitable for growth with a maximum Growth Index (GI) of 50, the species has achieved 60% of the potential population growth’. However, a precise threshold value for establishment and impact cannot be given and any cut-off value of EI may be species-specific and should be operationally defined on the basis of additional evidence. The Panel used two EI thresholds (≥ 15 and ≥ 30) to identify areas where climate suitability favoured establishment. See also assessment Section 48 on Impact.

2.2.6 Transfer and initiation of a founder population

While most fresh eggplant and turkey berry fruit imported will be sold, cooked and consumed, a proportion is discarded at various steps along the supply chain by importers, wholesalers, retailers and the final consumers, e.g. due to damage during handling and transport, physical quality problems, market conditions and pest finds (Gould & Maldonado, 2006). There is a possibility that live larvae in discarded host fruits will develop to adulthood, escape from the discarded material and find a mate resulting in fertilised eggs being laid on a host plant in the neighbourhood of the discarded material, a process referred to as transfer. Should the subsequent progeny develop and reproduce, a potential founder population would have been initiated. The process of transfer and initiation of a founder population was broken down into four steps:
  • Estimating the proportion of imported host-plant material discarded by commercial stakeholders in the supply chain due to e.g. infestation, physical damage, substandard quality or oversupply;
  • Estimating the proportion of infested material discarded by consumers;
  • The proportion of larvae that develop to adulthood and escape from discarded material;
  • The proportion of females that find a mating partner and find a suitable host plant in the surrounding environment and lay fertilised eggs;
  • The likelihood that adults develop from the eggs to reproduce and initiate a founder population.

Information to support judgements relating to these steps, necessary for establishment, was sought within the literature review. The collected evidence was reviewed during EKE and is summarised in Appendix D: Establishment.

2.2.7 Scenarios for establishment

A parameter in the model for pest introduction is the area of the EU where climatic conditions are suitable for L. orbonalis development. The area considered is within NUTS2 regions. Such information was determined using CLIMEX modelling (2.2.5). To capture uncertainty about the threshold for establishment, two thresholds were considered a lower Eco-climatic Index (EI) threshold of 15, and a higher EI threshold of 30.

2.2.8 Overall model for introduction (entry and establishment)

The pathway model for introduction is a product of the following components:
  • Mean annual EU import quantity of potential transfer units (eggplant and turkey berry) from countries where L. orbonalis occurs;
  • Inverse weight of a single transfer unit (to calculate the number of imported fruits as the volume of trade (kg) divided by the weight of a single transfer unit);
  • Proportion of infested units entering the EU;
  • Proportion of infested units imported to suitable NUTS2 regions;
  • Proportion of infested units disposed of as waste;
  • Probability of larva in discarded unit surviving to become an adult;
  • Probability of a female mating;
  • Probability of a mated female initiating a founder population that persists.

Figure 3 illustrates the model for introduction.

With three scenarios for entry, based on the proportion of turkey berry and eggplant imported, and two scenarios for area of suitable establishment, based on different EI thresholds, six scenarios were considered for introduction, Table 2.

TABLE 2. Key to introduction scenarios.
% of turkey berry and eggplant imported Threshold for Ecoclimatic Index
EI 15 EI 30
25% turkey berry; 75% eggplant Scenario 1 Scenario 4
100% turkey berry; 0% eggplant Scenario 2 Scenario 5
0% turkey berry; 100% eggplant Scenario 3 Scenario 6

2.2.9 Distribution of imported infested eggplant and S. torvum in the EU

The Panel did not find information on the final destination of eggplant and S. torvum imported from Asian countries of origin. The Panel therefore developed the pathway model on the assumption that consumers of eggplant and S. torvum from Asia are equally represented across Europe. Thus, it was assumed that the imported product is apportioned to NUTS regions according to the population in each NUTS region.

2.2.10 Identifying NUTS2 regions with suitable climate for establishment

The fractions of CLIMEX grid cells in each NUTS2 region with EI ≥ 15, or EI ≥ 30, were determined. Each full grid cell approximates to 2500 km2. All grids and grid fractions with EI ≥ 15 or ≥ 30 were summed for each NUTS2 area and multiplied by 2500 to give an approximate area where climate may be suitable for establishment. NUTS2 areas were then ranked by suitable area.

2.3 Spread

The area of the colonised territory occupied during spread is expected to follow a sigmoid curve (Figure 4). After an initial lag phase of slow spread during which the founder population builds up, spread accelerates and reaches a constant rate for some time before declining again as the suitable area gets fully colonised (saturation phase). Rather than estimate the parameters for logistic spread (i.e. Figure 4), this assessment followed the method of EFSA PLH Panel (2018) to estimate the duration of the lag phase and the linear rate of range expansion when spread is at its fastest. In this way, the spread assessment is simplified.

Details are in the caption following the image
Stages of conceptual logistic spread: Following the lag phase (lag period) spread accelerates, becomes almost linear then slows.

Spread is not just expansion of a contiguous area, but it may also include the generation of distant satellite populations (Herms & McCullough, 2014; Muirhead et al., 2006; Robinet et al., 2009). The Panel assessed the rate of natural spread as both processes together (spilling over at the edge and generation of satellite populations that later merge) determine the spatial expansion of a population.

An evidence dossier on spread was assembled from the literature review. Because no data was available to estimate the natural spread capacities of L. orbonalis, particular attention was given to available data on the spread of L. laisalis, an African species closely related to L. orbonalis that established a population in Southern Spain in the 1950s and has spread further on the southern Iberian Peninsula in the past decades. These spread data were used to estimate the expected spread rate of L. orbonalis. The assessment of spread of L. orbonalis considered both natural dispersal and farm-scale human-assisted spread with agricultural equipment (Section 44 Spread). Assessors took part in the semi-formal EKE using behavioural aggregation (EFSA, 2014). The collected evidence was reviewed during the EKE and is summarised in Appendix E: Spread.

2.3.1 Scenario definition for spread

Scenario for spread: considering existing practices and phytosanitary measures

To estimate the lag period and rate of linear range expansion, the Panel developed a general scenario with the following description:
  • The pest initiates a founder population at a single point somewhere within the area of potential establishment (where the CLIMEX EI is greater than the minimum threshold (EI 15 or 30) (see Sections 12 and 35 Establishment).
  • L. orbonalis is a specialised feeder of Solanum spp. For the EU, 48 species of Solanum are reported (Valdés, 2012), all of which can be considered putative host plants for L. orbonalis.
  • During the lag period, the population size increases until it reaches a local steady state in the centre of the population (determined by the habitat-carrying capacity).
  • By reaching the local habitat-carrying capacity (saturation), the population enters the spread phase, pushing the outer edge of the saturated population at a constant rate into suitable, unoccupied neighbouring habitats.
  • The spread assessment considers the outcome of the combined contributions of natural and local human-assisted spread. The human-assisted component only includes operations related to production and local movement (e.g. common agricultural practices) but no post-harvest movements, such as the trade in commodities (EFSA, 2019).
  • Spread occurs within regions where the CLIMEX EI Index is greater than the thresholds (EI 15 or 30) (see Establishment).
Uncertainties:
  • Allee effects (already considered during the assessment of establishment) might have an important impact on the survival/extinction of small founder populations.

2.4 Impact

The scientific literature on L. orbonalis was screened for information on impact of the pest on host plants. An evidence dossier on impact was assembled by EFSA staff and Working Group members. It was analysed to conceptualise the impact elements of risk and to inform the assessment of impact using EKE. Two scenarios were considered in the information retrieval from the literature: (1) yield lost under pesticide-free treatments, (2) yield lost despite the use of pesticides. Data were extracted and analysed separately in meta-analyses to determine the damage done by L. orbonalis with and without chemical control. The results of meta-analysis were used as input for the EKE on impact of L. orbonalis on host plants in the EU in NUTS2 regions where L. orbonalis is able to establish, with and without specific controls in place.

The collected evidence was reviewed during the EKE and is summarised in Appendix F: Impact.

2.4.1 Scenario definition for impact

Scenario Impact-1 (baseline): assuming no pest control is applied (i.e. artificial situation, akin to experimental no treatment ‘control’ plots in an experimental trial).

To estimate potential impact in terms of yield loss, a scenario with the following characteristics was defined:
  • The pest has spread to its maximum geographic extent in EU NUTS2 regions with climate suitable for establishment and equilibrium population levels.
  • Within the area of potential establishment, pest presence depends on the heterogeneity of the patches where the host occurs. It is therefore not necessarily the case that the pest is present in all suitable patches.
  • In each location where the pest occurs, its abundance is in equilibrium with the available resources (e.g. host plants) and environmental conditions (including climate, ecosystem resistance and resilience) (e.g. Grimm & Wissel, 1997).
  • No action is taken for pest control – yield loss data (% of fruit yield) in control plots of field trials were extracted from the literature and the subsequent meta-analysis used to inform losses when no control options are applied (representing worst-case conditions).
  • Current crop production practices (e.g. chemical insecticides targeted at L. orbonalis are not used).
  • The assessment of impact assumes a situation in which L. orbonalis has been established in a climatically suitable area (EI ≥ minimum threshold) for a long enough period of time to have reached carrying capacity and maximum impact.
  • Potential impact of transient populations was not considered i.e. in NUTS2 regions with low suitability for establishment (EI < minimum threshold) or production of eggplants in greenhouses.
  • Different susceptibilities of host plants (e.g. eggplants, potatoes and tomatoes), and the detailed biological characteristics of L. orbonalis (e.g. dispersal, feeding activity) were not considered in the assessment of impact.
  • The focus was on eggplant in the southern EU, largely Mediterranean coastal areas.

Scenario Impact-2 (with pest management in place): considering existing practices and any additional pest management by farmers to target the pest.

To estimate potential impact in terms of yield loss under scenario 2, the Panel envisaged scenario 1 with the following additional conditions:
  • Pest control practices would be applied by farmers.
  • Cropping practices and management options are those currently in place in the area of potential pest distribution, considering differences with those applied in countries where L. orbonalis is present (and evidence was collected).
  • The effect of currently applied control against other pests is taken into account (e.g. yield losses in EU crops given existing pest pressures were considered – how much more would Leucinodes add to the existing burden of pests in the EU?).
  • In a scenario where the pest is widely established and there would be no statutory action by NPPOs in the EU against Leucinodes.

2.5 Evaluation of risk reduction options/risk mitigation measures

As noted in Section 5, the EFSA PLH Panel planned to evaluate how additional risk mitigation measures (Appendix G) may reduce the likelihood of pest entry. However, results from the entry and establishment modelling reveal that the initiation of L. orbonalis founder populations in the EU is already unlikely given the relatively small quantities of product imported and taking existing practices into account (see Section 41). Consequently, options for further risk reduction are discussed, but their effectiveness was not quantified. Therefore, options for further risk reduction are discussed, but their effectiveness was not quantified. This opinion therefore presents an assessment of pest risk based on historic trade volumes, existing practices and generic phytosanitary measures.

2.6 Temporal and spatial scales

The pathway model calculates the trade flow of relevant commodities per year, on average, over the next 5 years (2024–2028).

The distribution of potentially infested plant material entering the EU was assessed using NUTS2 spatial resolution using EU census data from 2021 (Eurostat, accessed 31/12/2022). The CLIMEX model used 30 years of climate data, ranging from 1981 to 2010.

3 ASSESSMENT

A description of taxonomic issues relating to the genus Leucinodes is provided in Appendix A. As L. orbonalis occurs only in Asia, all pathways were considered to originate in Asia.

A synthesis of the biology of L. orbonalis based on the literature review is provided in Appendix B together with some exemplary pictures of the pest and the damage it causes. A list of cultivated and wild hosts is provided in the pest categorisation for L. orbonalis (EFSA PLH Panel, 2021). Female adults lay eggs singly or in small clusters of two to four on the lower leaf surfaces of the topmost and middle leaves of its preferred host plant, S. melongena (Ardez et al., 2008). The eggs are oval, about 0.5 mm in diameter, and turn from their initial creamy white colour to a deep orange towards larval hatching, when the black head capsule becomes visible through the eggshell (Ardez et al., 2008; Lall & Ahmad, 1965). The larvae are internal feeders, boring into the shoots and fruits, the latter being the preferred host tissue (Navasero & Calilung, 1990) and usually harbouring one or two larvae (Shukla, 1986). The entry hole into the plant tissue is closed by a plug of excreta. The concealed larval feeding makes infestation difficult to detect. However, in host plants with fruits too small to harbour the growing larva (e.g. Solanum nigrum), the larva exits the fruit and webs together three or four fruits to continue feeding from inside the web (Das & Patnaik, 1971). Such behaviour is seen with infested S. torvum.

3.1 Entry

3.1.1 Analysis of interceptions on produce

The Panel searched for interceptions of L. orbonalis in Europhyt (1995 until May 2020) and TRACES (June 2020 to ongoing database, last check 7 March 2023).

In the years 2004–2023, there were 350 notifications of interceptions of L. orbonalis in Europhyt and TRACES from consignments of plant products originating from Asia. The majority of plant species with interceptions are in the Solanaceae, but a number of species from other families are also reported (Appendix C, Table C.1). The Panel analysed the host status of plants on which interceptions have been reported to distinguish actual pathways of introduction (commodities in which the insect can develop to a viable pupa and adult) and incidental interceptions due to movement of larvae from an actual host to other plant material in the same shipment. The Panel judged all interceptions on plant material from non-solanaceous plant products as incidental. Details on the underlying evidence are provided below.

The primary potential pathways for introduction based on the numbers of interceptions are fruit of S. melongena (222 interceptions) and S. torvum (66 interceptions). Additional pathways are fruit of S. virginianum (8 interceptions), S. aculeatissimum, S. aethiopicum, S. anguivi and S. macrocarpon (2 interceptions in each of the four species), S. stramoniifolium and S. undatum (1 interception in both species) and undetermined Solanum spp. (26 interceptions). All these Solanum spp. are confirmed hosts of L. orbonalis based on literature reports (Appendix C, Table C.1).

Four interceptions were made on Momordica spp. A comment in the interception data indicates a misidentification, the actual species intercepted being Diaphania indica (Saunders) (Lepidoptera: Crambidae), which is, like L. orbonalis, a predominantly white moth with similar-looking larvae that primarily feed on Cucurbitaceae such as Momordica. Maureal et al. (1982) observed in a no-choice trial that third-instar L. orbonalis larvae did not feed on Momordica charantia fruit offered for 24 h. Momordica was therefore disregarded as a pathway.

Another case considered as potential pathway among non-Solanaceae interceptions was Ipomoea aquatica (Convolvulaceae). L. orbonalis has been reported to feed on tubers of the related sweet potato, I. batatas, but lab-rearing experiments showed that L. orbonalis cannot successfully complete its life cycle from egg to adult on this plant (Ardez et al., 2008). I. batatas was therefore judged not to offer a pathway.

One interception was reported on Capsicum annuum. Maureal et al. (1982) showed in a 24-h no-choice test that the larvae of L. orbonalis do not feed on C. annuum fruit; hence, C. annuum was not considered a pathway.

An overview of the interceptions on Solanum spp. is provided in Figures 5, 67 below.

Details are in the caption following the image
Interceptions of Leucinodes orbonalis on Solanum spp. imported from Asia according to records in Europhyt and TRACES, 2004–2023. The total number of interceptions is 324. While some records in Europhyt and TRACES do not identify the species, the product originates from countries where only L. orbonalis is present (except perhaps Vietnam, which has a genetically distinct Leucinodes population which has to date not been described as a species). This indicates that for all countries (including Vietnam until the species is described), the species is L. orbonalis.
Details are in the caption following the image
Interceptions of Leucinodes orbonalis and Leucinodes spp. from Asia on different Solanum spp. according to records of Europhyt and TRACES, 2004–2023. The figure highlights that the countries of origin of interceptions vary substantially by date. This variation reflects both variation in trade volume from those countries and variation in the frequency of interceptions per unit product imported (Appendix C). While some records in Europhyt and TRACES do not identify the species, the product originates from countries where only L. orbonalis is present (except perhaps Vietnam, which has a genetically distinct Leucinodes population which has to date not been described as a species). This indicates that for all countries (including Vietnam until the species is described), the species is L. orbonalis.
Details are in the caption following the image
Interceptions of Leucinodes orbonalis and Leucinodes sp. from different countries of origin in Asia on different Solanum spp. according to records of Europhyt and TRACES, 2004–2023. While some records in Europhyt and TRACES do not identify the species, the product originates from countries where only L. orbonalis is present (except perhaps Vietnam, which has a genetically distinct Leucinodes population which has to date not been described as a species). This indicates that for all countries (including Vietnam until the species is described), the species is L. orbonalis.

3.1.2 Identifying pathways (plants for planting)

Plants for planting of Solanaceae, other than seeds, are largely prohibited from entering the EU except from a few European and Mediterranean countries and parts of European Russia (Commission Implementing Regulation (EU) 2019/2072, Annex VI, 18). Annex VI prohibitions also concern potato (S. tuberosum), which has more detailed prohibitions. Thus, the Panel concludes that, under the current regulation, plants for planting are not a pathway for entry of L. orbonalis into the EU.

3.1.3 Identifying pathways: Plant hosts of L. orbonalis

In Asia, the single most important larval host plant of L. orbonalis is Solanum melongena, commonly known as eggplant, brinjal or aubergine. The larvae feed in the stems and fruits of the plant. When given free choice for ovipositing among eggplant, tomato (S. lycopersicum), potato (S. tuberosum), black nightshade (S. nigrum) and three non-Solanaceae crops, females chose exclusively eggplant (Ardez et al., 2008).

The next most-preferred host plant in India is S. aethiopicum (= S. gilo, S. integrifolium) (Dr S. Ramasamy, personal communication), commonly known as gilo, or as garden egg in Africa (dr L. C. Nwosu, personal communication). Tejavathu et al. (1991) report an infestation rate of 3.5% of shoots and 4.4% of fruits in India. The fruits have a diameter of 10–50 mm but are usually 15–25 mm in size (Vorontsova, 2023).

Apart from its primary host (eggplant), L. orbonalis is able to successfully complete its life cycle on several other Solanum species such as S. torvum (turkey berry), S. nigrum (black nightshade), S. tuberosum (potato), S. lycopersicum (tomato), S. anomalum, S. macrocarpon, S. myriacanthum (Himalayan nightshade), S. viarum (tropical soda apple) and S. virginianum (yellow-fruit nightshade); see Appendix C for more details about the biology of L. orbonalis on these plants. In S. lycopersicum, the larvae can only develop to pupation in the stems, but there is uncertainty whether they can develop into viable adults in the fruit due to the high-water content; hence, tomato fruit are not considered a pathway (personal communication Dr S. Ramasamy; Ardez et al., 2008; Boopal et al., 2013; Das & Patnaik, 1971).

The EU does import a small amount of ware potatoes from Asia (e.g. Israel) and although literature reports L. orbonalis infesting stems of S. tuberosum in India and the larvae being able to feed and complete their development on potato tubers in the laboratory, the larvae do not infest underground tubers under field conditions. Given lack of imports and that larvae do not infest tubers under practical growing conditions, potato tubers were not considered a realistic pathway (Appendix C, Table C.1 provides further details).

Physalis and Capsicum have been cited as host plants in the literature, but the Panel could not find any reference confirming that L. orbonalis larvae are actually feeding on these plant genera; Maureal et al. (1982) demonstrated that the larvae do not feed on Capsicum. Other Solanaceae plants of economic importance, such as tobacco (Nicotiana) and thornapples (Datura), have neither been reported as hosts of L. orbonalis, nor has the species been intercepted in the EU on plants of these genera.

Records of L. orbonalis on non-Solanaceae host plants appear to be incidental, and the species is unlikely to complete its life cycle on these plants. Ardez et al. (2008) conducted no-choice experiments including some non-Solanaceae crops and found that either the larvae could not survive feeding on sweet potato (Ipomoea batatas, Convolvulaceae) and okra (Abelmoschus esculentus, Malvaceae), or that feeding on cowpea (Vigna unguiculata = syn. Vigna sinensis, Fabaceae) was successful, but no adults emerged from the formed pupae.

The only report of L. orbonalis feeding on mango (Mangifera indica) is from an abstract of Hutson (1931), stating: ‘the shoot-borer, Leucinodes orbonalis, Gn., and the pentatomid, Coptosoma siamica, Wlk. (Hemiptera: Pentatomidae), on mango’. This abstract refers to a 17-page typescript article that was apparently never published. Mango has not been confirmed as larval host of L. orbonalis by any other resource, and the Panel therefore considers it an erroneous host plant.

Momordica (bitter melon, Goya) was repeatedly mentioned as host plant in the literature, however, always without a reference. In the Europhyt database, an interception L. orbonalis on Momordica was probably a misidentification of Diaphania indica, another species of moth in the Crambidae family. The larvae of Diaphania species commonly feed on Cucurbitaceae such as Momordica (Solis, 2006). The Panel therefore considers Momordica an erroneously reported host for L. orbonalis.

In conclusion, the Panel identified Solanum species in the Solanaceae plant family as potential produce pathway for L. orbonalis to enter the EU. The Panel focused on S. melongena and S. torvum (Table 3) as main entry pathways, but considered that fruit of any species of Solanum might act as a pathway. However, due to less trade and fewer interceptions compared to S. melongena and S. torvum, these pathways were not considered during the quantitative assessment.

TABLE 3. Host plants considered potential entry pathways for Leucinodes orbonalis.
Binomial name Common English names Interceptions in the EU
Solanum melongena Eggplant, aubergine, brinjal Yes
Solanum torvum Turkey berry, pea eggplant Yes

3.1.4 Pathway evaluation (EKE results)

Key results from the entry pathway model are shown in Table 4 below. Results represent model outputs for scenario 1 where 25% of pathway imports are turkey berry and 75% are eggplants; imports are distributed for consumption across the EU according to human population; NUTS2 regions where EI > 15 are considered NUTS regions where establishment is possible. Details of the source of the data used for the estimation and the calculations can be found in Appendix C.

TABLE 4. Model output results illustrating the range in estimates of mean imports and subsequent range in number of infested host transfer units entering the EU each year into areas where climate may be suitable in scenario 1 (EI > 15).
Percentile (%) 1 5 25 50 75 95 99
Mean import of fresh turkey berry from Asia into the EU (t/year) 18.6 25.1 34.5 41.0 47.5 56.7 63.1
Import of fresh eggplant fruit from Asia into the EU (t) 55.8 75.4 103.2 122.6 142.0 170.0 189.8
Turkey berry allocated to NUTS2 (t) 4.3 5.8 8.0 9.5 11.0 13.1 14.6
Eggplant allocated to NUTS2, (t) 12.9 17.4 2.9 28.4 32.9 39.3 43.8
Range of weight in turkey berry clusters (g) 18.0 20.3 27.3 34.5 42.4 51.4 55.0
Range in weight of eggplant fruit (g) 100.1 112.6 151.7 196.6 253.3 342.8 400.6
Number of turkey berry transfer units entering NUTS2 107,313 146,230 211,587 271,691 354,590 511,462 626,254
Number of eggplant transfer units entering NUTS2 50,103 69,456 105,016 141,054 189,933 275,507 339,899
Number of total transfer units entering NUTS2 174,555 237,386 340,186 426,956 533,138 715,591 863,327
Number of infested transfer units per 10,000 units imported 0.10 0.13 0.98 4.20 11.13 23.73 30.07
Number of infested transfer units entering NUTS2 3.3 5.5 40.4 172.6 468.8 1100.9 1627.0
  • Notes: Scenario 1 considers an establishment index, EI > 15 and turkey berry and eggplant representing 25% and 75%, respectively, of total imports of the commodity code HS 0709 3000. Imports are from the following countries: Bangladesh, Brunei, China, Indonesia, India, Japan, Cambodia, Laos, Sri Lanka, Myanmar, Malaysia, Nepal, Philippines, Pakistan, Singapore, Thailand, Taiwan and Vietnam.

Results from Scenario 4 are shown in Table 5.

TABLE 5. Model output results illustrating the range in estimates of mean imports and subsequent range in number of infested host transfer units entering the EU each year into areas where climate may be suitable (EI > 30).
Percentile (%) 1 5 25 50 75 95 99
Import of fresh turkey berry from Asia into the EU (t) 18.6 25.1 34.5 41.0 47.5 56.7 63.1
Import of fresh eggplant fruit from Asia into the EU (t) 55.8 75.4 103.2 122.6 142.0 170.0 189.8
Turkey berry allocated to NUTS2 (t) 2.6 3.5 4.9 5.8 6.7 8.0 8.9
Eggplant allocated to NUTS2 (t) 7.9 10.6 14.6 17.3 20.0 240. 26.7
Range of weight in turkey berry clusters (g) 18.0 20.3 27.3 34.5 42.4 51.4 55.0
Range in weight of eggplant fruit (g) 100.1 112.6 151.7 196.6 253.3 342.8 400.6
Number of turkey berry transfer units entering NUTS2 65,420 89,145 128,987 165,627 216,164 311,796 381,775
Number of eggplant transfer units entering NUTS2 30,543 42,342 64,019 85,989 115,786 167,954 207,208
Number of total transfer units entering NUTS2 106,412 144,714 207,383 260,279 325,010 436,236 526,299
Number of infested transfer units per 10,000 units imported 0.10 0.13 0.98 4.20 11.13 23.73 30.07
Number of infested transfer units entering NUTS2 2.0 3.4 24.6 105.2 285.8 671.2 991.8
  • Note: Scenario 4 considers an establishment index, EI ≥ 30, and turkey berry and eggplant representing 25% and 75%, respectively, of total imports of the commodity code HS 0709 3000.

3.1.5 Unquantified uncertainties affecting the assessment of entry

  • The proportion of imports which are turkey berry (S. torvum) and which are eggplant (S. melongena) is unknown.
  • Volumes of eggplant imports from Asia may change in future; some Asian type varieties are being grown in the EU already, hence imports may fall.
  • Growing conditions for material intended for export to the EU; amount of infestation at origin and effectiveness of cleaning/sorting.
  • On arrival in the EU, there is uncertainty on the distribution of consignment sizes, the percentage of consignments inspected in each country, the sample size at inspection and the chance of detection of infestation if an inspector examines an eggplant or turkey berries.
  • Estimation of interceptions based on incomplete information (not all EU member states notify interceptions of non-quarantine pests).

3.1.6 Conclusion on the assessment of entry

The pathway most likely to provide a route for entry of L. orbonalis into the EU was judged to be fresh eggplant and fresh turkey berry from Asia. The number of fruits expected to enter the EU each year and be distributed across NUTS regions where climatic conditions are conducive for establishment of L. orbonalis is expected to be in the order of hundreds of thousands (median estimate with EI≥15 approximately 430,000 transfer units; 90% CR approximately 175,000–865,000). With an EI threshold of 30, the number of fruits entering NUTS2 regions where parts are suitable for establishment drops to approximately 260,000 (90% CR approximately 106,000–526,000).

Infested fruits represent a small proportion of the total number of fruits entering the EU. The number of transfer units infested with live L. orbonalis entering NUTS2 areas with EI ≥ 15 is estimated to be approximately 175 per year (90% CR approximately 6–1100); using an EI threshold of 30, the median number of infested transfer units drops to approximately 105 per year (90% CR approximately 3–670). In the scenario where only turkey berry is imported, and using EI≥ 15, the 95 percentile estimates 84 adults emerge in areas suitable for establishment. In contrast, where only eggplant is imported and using EI ≥ 30, the 5 percentile estimates one adult emerging in 40 years.

3.2 Establishment

Climatic mapping is a common approach to identify new areas that might provide suitable conditions for the establishment of alien organisms (Baker, 2002; Venette, 2017). Climatic mapping is based on combining information on climate in the known distribution of a poikilothermic organism, the organisms' physiological responses to environmental conditions and the climate in the risk area. The current distribution of L. orbonalis is presented in Section 35. The results of climatic mapping are presented in Sections 36-3839. SDM results were not included in the current opinion because the results were found to be not mature for interpretation. The details about L. orbonalis climate suitability modelling are presented in Rossi et al. (2023) and available online on the ZENODO platform.

3.2.1 Global distribution of Leucinodes orbonalis

L. orbonalis is a tropical and sub-tropical species native to Asia with India and Bangladesh thought to be its centre of origin (Karthika et al., 2019). Figure 8 shows the distribution of L. orbonalis. Until Hayden et al. (2013) and Gilligan and Passoa (2014) reported that L. orbonalis was restricted to Asia, it was thought that L. orbonalis also occurred in sub-Saharan Africa. Literature previously reporting L. orbonalis from Africa should be regarded as referring to members of a complex of other species of Leucinodes native to Africa and not as referring to L. orbonalis (see also Section 5).

Details are in the caption following the image
Detailed distribution of Leucinodes orbonalis in Asia based on a systematic literature search showing (1) points and (2) administrative regions where the organism occurs.

3.2.2 Degree days model

Data from Dhaliwal and Aggarwal (2021) were used to determine the base temperature of 15°C and a thermal constant of 438.6 degree days above the threshold to complete a generation. Based on this, between one and four generations could theoretically be possible in Europe each year. This compares to the 8–10 or more generations per year across large parts of India and southeast Asia (Figures 9 and 10) where L. orbonalis is a recognised pest.

Details are in the caption following the image
Estimated mean number of generations for Leucinodes orbonalis in Europe and the Mediterranean Basin (1) and southeast Asia and India (2). Number of generations were calculated considering a minimum number of accumulated degree days to complete one generation of 438.6 Degree days. Number of generations were calculated for each year in the period 1993–2022 and then averaged.
Details are in the caption following the image
Estimated mean number of Leucinodes orbonalis generations in India and neighbouring regions.
Mountainous regions in northern India are much cooler than central and southern India. The accumulated heat (measured as DD) is therefore much less in the mountains leading to fewer generations. Note that L. orbonalis presence has been reported in areas where degree day mapping suggests a temperature sum supporting less than one is possible. Possible explanations for this include:
  • Findings could be from a transient population.
  • Mapping data are based on average temperatures over several years and reported findings could be from unusually warm periods.
  • L. orbonalis surviving in a ‘hot spot’ above the average of the grid cell of mapping resolution.
  • The reports are from trap catches of individuals moving from suitable neighbouring areas.

3.2.3 CLIMEX

A more sophisticated modelling approach to heat accumulation is the use of CLIMEX which takes organism ecophysiological requirements into account across a geographic region. Results from the CLIMEX model outputs for EI across India and southeast Asia are shown in Figure 11A,B. Green dots in Asia indicate point locations where L. orbonalis has been reported.

Details are in the caption following the image
(A, B) CLIMEX Ecoclimatic Index for Leucinodes orbonalis (i) India and southeast Asia, and (ii) the Euro-Mediterranean Basin.

CLIMEX modelling indicates EI values ≥ 30 across most of India and southeast Asia. In northern cooler areas, and at higher altitudes, EI values drop (Figure 11A). EI values that can support establishment are found in southern Europe, especially around the Mediterranean coast (Figure 11B).

The details about L. orbonalis climate suitability modelling are presented in Rossi et al. (2023) and available online on the ZENODO platform.

3.2.4 Establishment

Pest interceptions indicate that L. orbonalis can enter the EU (Appendix C). Modelling estimates that, depending on the EI threshold used for establishment, the number of transfer units infested with live L. orbonalis entering NUTS2 areas where establishment may be possible varies from a median of approximately 175 per year (EI ≥ 15) to approximately 105 per year (EI ≥ 30). To initiate a founder population, larvae entering the EU with infested S. melongena or S. torvum must complete development, locate a partner, mate and the subsequent progeny must survive, complete development and mate, thus initiating a founder population (Appendix D).

Table 6 provides key results from the EKE modelling. It shows the likelihood that a potential founder population will be initiated in the EU each year. Using an EI threshold of 15, the median number of founder populations establishing in the EU annually is 0.00014 (90% CR 0.00000–0.00264). This equates to a median estimate of one founder population approximately every 7000 years (90% CR approximately one every 380–355,600 years).

TABLE 6. Model output results illustrating the range in estimates for key model steps from entry to initiation of founder population.
Percentile (%) 1 5 25 50 75 95 99
Estimated mean annual number of total transfer units entering NUTS2 174,555 237,386 340,186 426,956 533,138 715,591 863,327
Number of infested transfer units per 10,000 units imported 0.10 0.13 0.98 4.20 11.13 23.73 30.07
Estimated mean number of infested transfer units entering NUTS2 3.3 5.5 40.4 172.6 468.8 1100.9 1627.0
Number of infested transfer units discarded in NUTS 2 by trade (pest may survive) 0.3 0.6 4.6 19.9 55.5 145.0 237.3
Number of infested transfer units discarded by consumer in NUTS2 (pest may survive) 1.1 2.2 15.8 68.3 185.5 488.4 784.4
Number of emerged adults surviving in NUTS2 areas 0.03 0.09 0.72 3.16 9.87 32.29 61.25
Number of founder populations in NUTS2 areas per year 0.00000 0.00000 0.00003 0.00014 0.00056 0.00264 0.00626
Expected number of years between founder populations 160 379 1782 7006 35,590 355,600 1,760,100
  • Note: Scenario 1 considers an establishment index (EI) ≥ 15 and turkey berry and eggplant representing 25% and 75%, respectively, of total imports of the commodity code HS 0709 3000.(Scenario 1 in Table 2).

Using EI ≥ 30 as a threshold for establishment, Table 7 key results from the pathway modelling. It shows that the median likelihood that a founder population will be initiated in the EU is 0.00009 per year (90% CR 0.00000–0.00161). This equates to a median estimate of one founder population approximately every 11,500 years (90% CR approximately one every 620–583,300 years).

TABLE 7. Model output results illustrating the range in estimates for each model step from entry to initiation of founder population.
Percentile (%) 1 5 25 50 75 95 99
Number of total transfer units entering NUTS2 106,412 144,714 207,383 260,279 325,010 436,236 526,299
Number of infested transfer units per 10,000 units imported 0.10 0.13 0.98 4.20 11.13 23.73 30.07
Number of infested transfer units entering NUTS2 2.0 3.4 24.6 105.2 285.8 671.2 991.8
Number of infested transfer units discarded in NUTS 2 by trade (pest may survive) 0.2 0.4 2.8 12.1 33.8 88.4 144.7
Number of infested transfer units discarded by consumer in NUTS2 (pest may survive) 0.7 1.3 9.7 41.6 113.1 297.8 478.2
Number of emerged adults surviving in NUTS2 areas 0.02 0.06 0.44 1.93 6.02 19.69 37.34
Number of founder populations in NUTS2 areas per year 0.00000 0.00000 0.00002 0.00009 0.00034 0.00161 0.00381
Expected number of years between founder populations 262 622 2923 11,493 58,380 583,316 2,887,223
  • Note: Scenario 4 considers an establishment index (EI) ≥ 30, and turkey berry and eggplant representing 25% and 75%, respectively, of total imports of the commodity code HS 0709 3000 (scenario 4 in Table 2).

3.2.5 Conclusions on identifying NUTS2 regions suitable for establishment

The fractions of CLIMEX grid cells in each NUTS2 region with EI ≥ 15 or ≥ 30 (Figure 12) were determined and the area where EI ≥ 15 or 30 calculated. Table 8 lists the NUTS2 regions in descending order of area.

TABLE 8. Area of NUTS2 regions where establishment may be possible, based on EI thresholds of 15 or 30.
NUTS2 code Country (NUTS name) % of EU population Area EI ≥ 15 (thousands km2) Area EI ≥ 30 (thousands km2)
ES61 ES (Andalucía) 1.93 67.50 47.50
PT18 PT (Alentejo) 0.16 35.00 17.50
ITG1 IT (Sicilia) 1.09 25.00 17.50
ES52 ES (Comunidad Valenciana) 1.15 15.00 12.50
EL43 EL (Kriti) 0.14 12.50 12.50
CY00 CY (Kýpros) 0.20 10.00 10.00
ES43 ES (Extremadura) 0.24 37.50 7.50
ITF6 IT (Calabria) 0.42 15.00 7.50
ITF4 IT (Puglia) 0.89 17.50 6.40
EL42 EL (Notio Aigaio) 0.07 5.17 5.16
ITG2 IT (Sardegna) 0.36 22.50 5.00
EL41 EL (Voreio Aigaio) 0.04 5.00 5.00
EL65 EL (Peloponnisos) 0.12 7.03 4.58
ES53 ES (Illes Balears) 0.28 5.18 3.69
ES62 ES (Región de Murcia) 0.34 10.00 3.31
EL30 EL (Attiki) 0.86 3.88 3.02
EL64 EL (Sterea Elláda) 0.11 5.00 1.96
EL63 EL (Dytiki Elláda) 0.15 7.50 1.88
ITI4 IT (Lazio) 1.29 12.50 1.43
PT17 PT (Área Metropolitana de Lisboa) 0.65 5.00 1.36
ES51 ES (Cataluña) 1.74 17.50 1.25
ITF3 IT (Campania) 1.27 3.90 1.20
ITF5 IT (Basilicata) 0.12 2.36 1.07
EL62 EL (Ionia Nisia) 0.05 5.00 0.70
PT20 PT (Região Autónoma dos Açores) 0.05 5.00 0.44
MT00 MT (Malta) 0.12 0.31 0.31
PT15 PT (Algarve) 0.11 7.50 0.25
FRM0 FR (Corse) 0.08 5.00 0.22
PT30 PT (Região Autónoma da Madeira) 0.06 0.76 0.04
PT16 PT (Centro (PT)) 0.52 20.00
FRI1 FR (Aquitaine) 0.81 12.50
ES24 ES (Aragón) 0.30 10.00
FRL0 FR (Provence-Alpes-Côte d'Azur) 1.18 7.50
FRJ1 FR (Languedoc-Roussillon) 0.67 5.00
ITI1 IT (Toscana) 0.84 5.00
PT11 PT (Norte) 0.82 3.73
ES42 ES (Castilla-La Mancha) 0.47 3.06
EL54 EL (Ipeiros) 0.07 2.67
HR03 HR (Jadranska Hrvatska) 0.29 2.25
ES11 ES (Galicia) 0.63 2.17
EL52 EL (Kentriki Makedonia) 0.41 1.45
ITC3 IT (Liguria) 0.34 0.98
FRJ2 FR (Midi-Pyrénées) 0.73 0.87
ITF1 IT (Abruzzo) 0.29 0.66
ES12 ES (Principado de Asturias) 0.23 0.52
ITF2 IT (Molise) 0.07 0.45
EL61 EL (Thessalia) 0.17 0.21
ITI2 IT (Umbria) 0.19 0.09
Sum % EU population in regions containing EI ≥30 14.09
Sum % EU population in regions containing EI ≥15 23.12
Sum area (km 2 ) 517.69 248.26
Details are in the caption following the image
EU NUTS 2 regions that contain any areas where EI ≥ 15 (pale yellow) or EI ≥ 30 (orange). These areas account for 23% (EI ≥ 15) or 14% (EI ≥ 30) of EU human population and where we assume 23% and 14% of imported commodities are distributed.

3.2.6 Introduction of L. orbonalis into the EU

For successful introduction, a pest that enters the EU must initiate a founder population. For a species that reproduces sexually and enters the EU in a virgin state, establishing a founder population requires mate finding and finding a host on which progeny can develop. There should be some spatial and temporal synchrony in emergence of mating partners and with hosts for success. The Panel used Monte Carlo simulations with a probabilistic pathway model to assess the number of eggplant fruit or bunches of turkey berry entering each year into those parts of the EU that are suitable for establishment. The model then quantifies the subsequent steps of waste production, escape of adult insects from waste, mating and initiation of a founder population by an egg-laying female. Within the EU imported S. melongena and S. torvum are distributed in proportion to human population in NUTS2 areas.

3.2.7 Unquantified uncertainties affecting the assessment of introduction

What happens to organisms invading a new area is a field of invasion biology that is little known or understood (Puth and Post, 2005); hence, there is uncertainty regarding the effect of stochasticity of behavioural events (development, emergence, mating, host finding) leading to successful pest transfer to suitable hosts and initiation of a founder population that is sustainable. This is because such steps are largely unobserved and there is little empirical evidence around the processes involved although successful invasion is often attributed to propagule pressure (Leung et al., 2004; Simberloff, 2009). The following uncertainties were identified but are not reflected in the uncertainty of the introduction model:
  • Although the assessment focused on eggplants and turkey berry from Asia, other potential pathways exist, thus adding to the likelihood of entry, however, they were considered less relevant after detailed consideration of the small volumes entering; any additional entry would be marginal.
  • No literature could be found detailing the growing practices used in Asia to produce eggplant or turkey berry for export to the EU.
  • The annual quantity of imports is relatively small; some EU growers are already producing more exotic varieties of eggplants of the type currently sourced from Asia; imports from some sources have declined and imports might drop further in future or even cease.
  • Changes in pest management regime (e.g. ban of insecticides, ineffectiveness of pesticides due to development of resistance, increased use of biological control) can affect probability that export consignment is infested and future likelihood of establishment in opposing directions.
  • What happens to organisms invading a new area is a field of invasion biology that is little known or understood (Puth and Post, 2005); hence, there is uncertainty regarding the effect of stochasticity of behavioural events (development, emergence, mating, host finding) leading to successful pest transfer to suitable hosts and initiation of a founder population that is sustainable. This is because such steps are largely unobserved and there is little empirical evidence around the processes involved although successful invasion is often attributed to propagule pressure (Leung et al., 2004; Simberloff, 2009).

3.2.8 Conclusion on entry and establishment (Pest introduction)

Based on the evidence regarding host plants of L. orbonalis, and on interception data, the importation of eggplant (S. melongena) and turkey berry (S. torvum) from Asia were judged to be the most relevant pathways for entry of L. orbonalis into the EU. In a scenario where 25% of imports are turkey berry and 75% are eggplant, and with a threshold EI of 15, then using EKE and pathway modelling, the median number of infested transfer units (individual eggplant fruit or bunches of turkey berries) entering the EU annually, where climate is suitable for establishment, is estimated to be approximately 175 (90% CR approximately 6–1100). Each infested transfer unit is likely to contain only one larva (Appendix B). Assuming a minimum EI of 30 for successful establishment, the number of infested units entering into suitable regions would be 105 as a median estimate and 3–670 as a credible range.

Climatic conditions are most suitable for establishment in parts of the southern EU, especially in Spain (Andalucía, Comunidad Valenciana and Extremadura), Portugal (Alentejo), Italy (Sicilia, Calabria, Puglia), Greece (Kriti) and Malta and Cyprus. If imports are allocated in proportion to human population, between 14.09% and 23.12% of transfer units enter regions of the EU suitable for establishment (lower estimates based on EI ≥ 30, higher estimate based on EI ≥ 15). Of the infested units entering NUTS2 regions where EI ≥ 15 approximately 12% are discarded before reaching the final consumer and approximately 50% of infested units is discarded by the consumer (Appendix D). Furthermore, 1.0% (median; 90% CR, 0.2%–1.9%) of larvae are expected to survive to adulthood and escape from commercial waste while a median of 5.2% (90% CR 0.98%–12.2%) escape from consumer household waste. When the resulting numbers of adults emerge across NUTS2 regions the likelihood that a female will find a mate depends on the window of encounter in space and time. In combination with the likelihood that the subsequent progeny survives to initiate a founder population, the number of established founder populations was estimated to be 0.00014 per year (90% CR 0.00000–0.00264). Thus, the Panel would not expect new founder populations within the foreseeable future or the time horizon of 5 years of this assessment. The predicted waiting time between new founder populations is in the Panel's estimation at least approximately 380 years assuming an EI of 15 required for establishment (Table 6: 5 percentile) and 622 years (Table 7: 5 percentile) when an EI of 30 is assumed to be required for successful establishment. Table 9 gives quantiles of the average number of years between subsequent founder populations in different scenarios. Given the low frequency of new founder populations across all scenarios, the Panel did not proceed to quantitatively assess effectiveness of risk reduction options targeting L. orbonalis. However, such options exist, e.g. the production of eggplant and turkey berry fruit in pest-free places of production. The Panel found four notifications mentioning that product infested with L. orbonalis was granted entry to the EU. This should no longer be the case after the instatement of temporary emergency measures against L. orbonalis in October 2022.

TABLE 9. Range of estimates in expected number of years between establishment of founder populations according to scenario assumptions.
Scenario % of turkey berry: % eggplant Threshold EI Expected number of years between establishment of founder populations
Percentile (%)
5 25 50 75 95
1 25:75 15 379 1782 7006 35,590 355,600
2 100:0 15 144 686 2709 13,957 137,574
3 0:100 15 836 4007 15,845 81,222 831,525
4 25: 75 30 622 2923 11,493 58,380 583,316
5 100:0 30 236 1125 4444 22,895 225,673
6 0:100 30 1371 6573 25,991 133,235 1,364,012

Scenario 2 presents the greatest likelihood of L. orbonalis establishing in the EU although the median estimate is that a founder population would occur once in 2709 years (CR once in 144–137,574 years). However, this assumes only turkey berry is imported, which the Panel knows is not the case. The frequency of founder populations being introduced will be less than in scenario 2. The ratio turkey berry: eggplant of 25:75 (indicated with grey shading in the table and as scenario 4 in table 9) is considered the most plausible among the three tested compositions of the trade; the other two scenarios are extreme assumptions.

3.3 Spread

3.3.1 Assessment of spread

The spread of a species introduced into a new environment (provided that it finds an ecological niche to occupy) is characterised by a lag phase, a spread phase and a saturation phase (Figure 4). The lag period is the time from the first introduction and reproduction of the pest, i.e. the initiation of a founder population, to its establishment with a constant rate of range expansion into pest-free areas.

To make an informed estimate of the presumed spread rate of L. orbonalis if it were to establish in suitable regions of the EU, the Panel investigated occurrence records of the related African species L. laisalis. This species was first reported in Europe from southern Spain in the 1950s and has been spreading in the southern Iberian Peninsula since then (Appendix E, Spread). L. orbonalis and L. laisalis are closely related (Mally et al., 2015) and share many similar life traits: They are native in the tropics and subtropics, the larvae of both species feed on Solanum spp. and the adults are of similar size, indicating similar flight- and therefore spread capacities.

The duration of the lag period in the regions where L. orbonalis could potentially establish was estimated to be approximately 34.5 years (90% CR 4.8–92.2 years). After the lag period, L. orbonalis is estimated to spread at a rate of 2.3 km/year (90% CR 0.6–7.0 km/year). More details are available in Appendix E (Spread).

3.3.2 Uncertainties affecting the assessment of spread

  • The African L. laisalis has different larval host plant preferences, with Solanum sodomeum (a synonym of Solanum linnaeanum) being its preferred host (Huertas-Dionisio, 2000), although it has also been found to feed on S. melongena in Spain (Huertas-Dionisio, 2000). S. linnaeanum appears to be of no agricultural interest, and its spatial availability will impact the spread rate of L. laisalis. Leucinodes orbonalis with its preferred host plant S. melongena – a widely cultivated crop in southern Spain – might therefore spread more easily, as its host plant is presumably much more abundant.
  • No studies were found investigating the flight capability of Leucinodes moths; Chang et al. (2014) state that the adults ‘only fly for short distances’.
  • It is unclear whether the spread of L. laisalis in the southern Iberian Peninsula originates from a single founder population, or whether spread from multiple founder populations.
  • Based on the occurrence records of L. laisalis, a mean spread rate of 1.6 km/year was calculated; in contrast, the second occurrence record in 1975 is at a distance of 196 km from the first record in 1958, resulting in a much higher spread rate of 11.5 km/year.
  • The duration of the lag period is mainly driven by the effect of EU agricultural practices and by the presence of natural enemies and control measures targeted at other Lepidopteran species, e.g. the Tomato leaf miner Tuta absoluta (Meyrick) (Gelechiidae), the potato tuber moth Phthorimaea operculella (Zeller) (Gelechiidae), and the Beet armyworm Spodoptera exigua (Hübner) (Noctuidae). These factors were not evaluated but are sources of uncertainty.
  • The expansion rate is driven by the dispersal ability of the insect, which is not well known, and by the effect of the host species communities in terms of species composition, patchiness and distance among suitable patches and availability in the EU environments compared to the observations collected from the area of origin. Information is lacking to assess this in detail.

3.3.3 Conclusions on spread

Were L. orbonalis to be introduced into the EU, the Panel estimates that it would take between 4.9 and 92.2 years (90% CR; median 34.5 years) for populations to grow sufficiently before a steady rate of spread of approximately 2.28 km/year (90% CR 0.65–7.02 km/year) was reached.

3.4 Impact

Leucinodes orbonalis larvae are oligophagous and feed on different species of Solanum. Its preferred host plant is eggplant, S. melongena, but it can also feed on other Solanum plants of economic relevance, such as tomato (S. lycopersicum) and potato (S. tuberosum). In tomato and potato, the larvae would mostly be confined to stem-boring, as tomato fruit are reported as suboptimal plant tissue for feeding due to their high-water content, and potato tubers are not fed on under natural conditions. With tubers developing underground, the larvae do not access the tubers.

3.4.1 Assessment of impact

Literature on impact of L. orbonalis in Asia is heavily skewed towards eggplant, the pest's primary host plant, with 800+ studies focussing on L. orbonalis feeding on this crop; in contrast, only very few papers with actual observations of L. orbonalis feeding on potato and tomato exist, with six papers and a single paper, respectively (see Appendix 88).

In India, the cultivation area of eggplant in 2013 was 0.53 million hectares (Indian Council of Agricultural Research, 2016), whereas that of tomato amounts to 0.79 million hectares (Indian Council of Agricultural Research, undated). The area of potato cultivation was more than three times larger than that of eggplant, with 1.79 million hectares averaged over the years 2007–2009 (Scott & Suarez, 2011). These large cultivation areas of both tomato and potato in India vs. the low and infrequent infestations reported in the literature indicate that the expected impact on tomato and potato crops in the EU is similarly negligible. Nevertheless, the lack of data on yield losses in potato and tomato in the literature makes an assessment on the potential losses in the EU unfeasible at the moment.

The main impact of L. orbonalis on solanaceous crops in the EU is to be expected on eggplant. In India and Bangladesh, baseline infestation rates are 25%–30%, often due to poor agricultural sanitation where plant parts infested with L. orbonalis larvae are not immediately destroyed but instead collected at a corner of the field, enabling re-infestation of the crop from adults emerging from the discarded plant parts.

Literature reporting eggplant yield losses in Asia was generally from regions with a CLIMEX Environmental Index (EI) > 30. Impacts in the EU were therefore considered to be largely limited to regions where EI > 30. Such locations coincide with the area for establishment, hence transient populations outside regions where EI < 30 were judged not to be able to cause measurable impacts. In southern European countries with EI > 30, the Panel estimates that approximately 4.5% (90% CR 2%–13%) of eggplant fruit grown outdoors may be lost to L. orbonalis infestation when no specific pest control measures are applied (see Appendix F for details and reasoning). In a scenario with specific pest control measures, the yield loss of eggplant was estimated to range from 0.28% to 1.9% (90% CR; median 0.54%).

3.4.2 Uncertainties affecting the assessment of impact

  • The main uncertainties affecting the impact assessment are related to the transferability of reports from Asia on impacts caused by L. orbonalis to the situation in the EU. Reports from Asia are predominantly on eggplant, which is grown in substantial quantities in Spain and Italy, the main sources of eggplant consumed within the EU.
  • The future likely loss of registered chemical insecticides in the EU and the increasing resistance of L. orbonalis against pesticides could constrain the effectiveness of control measures.
  • The development of non-chemical alternatives (e.g. mating disruption, mass trapping and especially conservation biological control with generalist predators) could potentially improve control.
  • Climate change would be expected to influence the pest cycle, with higher temperatures increasing the population growth and the number of generations per year. How climate change will affect EU production of S. melongena is unknown.

3.4.3 Conclusions on impact

Were L. orbonalis to enter, establish and spread within the EU with the population reaching an approximate equilibrium such that EU farmers consider the organism a member of the general pest fauna, estimated median eggplant yield losses are estimated to be 4.47% (90% CR 0.67%–13.0%) when no specific control measures are in place, and 0.54% (90% CR 0.13%–1.94%) when growers apply targeted pest control against L. orbonalis.

3.5 Risk reduction options

Recognising the very low likelihood of establishment in the EU (0.00014 founder populations per year, 90% CR 0.00000–0.00264 per year), the Panel did not quantitatively evaluate the effectiveness of additional specific phytosanitary measures targeted at L. orbonalis. However, options for risk reduction exist, e.g. the production of eggplant and turkey berry fruit in pest-free places of production in the countries of origin, and designation of L. orbonalis as a quarantine species. The Panel found four notifications mentioning that product infested with L. orbonalis was granted entry to the EU. L. orbonalis has a temporary EU emergency quarantine status since October 2022 (Article 30(1) of Regulation (EU) 2016/2031), which is an effective measure to stop this practice.

3.6 Consequences of climate change

The Panel did not conduct quantitative climate change scenarios for L. orbonalis, but considers that in the baseline scenario (EI ≥ 30; 25% turkey berry and 75% eggplant), 23% of the imported infested eggplant and turkey berry would be moved to areas suitable for establishment. With climate warming, the area suitable for establishment would enlarge, and hence, a greater proportion of the imported infested fruit would end up in areas suitable for establishment. If the number of infested fruits imported to areas suitable for establishment would increase by a certain percentage, then the number of moths emerging in those areas, the number of moths finding a mate, and the number of population founding events per year would increase by the same percentage in the pathway model used in this opinion. The average waiting time until a population founding event would then be decreased by this percentage. Based on this reasoning, the Panel concludes that climate warming would not qualitatively change the conclusions of the analysis under the assumption of current climate. Quantitatively, the risk area would be larger, and the time until the next founding event would be shorter, but founding events would still remain rare. In a worst-case scenario where 100% of imports were distributed across areas with suitable climate, the frequency of founder populations would increase more, approximately four- to sevenfold (100/23 to 100/14). A sevenfold increase in likelihood of introduction for the baseline scenario (Table 7, Scenario 4) would give a median frequency of a founder population establishing in approximately 1600 years (90% CR 90–83,000 years).

3.7 Overall uncertainty

Entry pathways

This assessment focused on special and exotic varieties of eggplant (S. melongena) and on turkey berry (S. torvum) as pathways, imported from different tropical and subtropical Asian countries. Both commodities are part of a niche market and enter the EU in small, but largely uncertain quantities. The majority of eggplant consumed in the EU is produced within the EU, primarily in Spain and Italy. A wide range of varieties is grown in the EU, and niche products such as mini-eggplants and white eggplants are added to this spectrum. Fruit of other Solanum plants may become popular and imported in larger quantities. This may change the composition and relative importance of pathways over time.

Transfer & establishment of founder populations

The imported consignments of eggplant and turkey berry from the countries where the pest occurs are generally small and in the order of a few kg to tens of kg per consignment (Appendix C). These consignments are too small to be further divided and distributed in the EU but will instead go as a whole to the customer. Where in the EU these final customers are is unknown, and the Panel therefore made the assumption that imported eggplants and turkey berry are distributed according to population in the EU (EFSA PLH Panel, 2021 Elasmopalpus lignosellus Quantitative Pest Risk Assessment) – an assumption that may be incorrect. However, the EFSA PLH Panel has no other simple and efficient basis on which to distribute imported produce that will aid the identification of points for pest introduction.

There is no specific information on production of waste of aubergines and turkey berry pre-consumer (during distribution and retail) and by the consumer. The pathway model assumes that 10% of the product is turned into waste in the supply chain pre-consumer, in line with an earlier assessment for Elasmopalpus lignosellus in asparagus for which empirical estimates were available from an American study (EFSA PLH Panel, 2021 Elasmopalpus lignosellus Quantitative Pest Risk Assessment; EFSA PLH Panel, 2023b). Waste by the consumer was based on expert knowledge elicitation, with a wide distribution of uncertainty. Escape from waste was also assessed by EKE, taking into account the experience gained during the earlier opinion on Taumatotibia leucotreta in cut roses, for which the escape from waste was analysed in detail (EFSA PLH Panel, 2023a Taumatotibia leucotreta Quantitative Pest Risk Assessment). Assessments of the escape probability of adult moths from discarded waste were made without the support from pest- or commodity-specific empirical data, which contribute to wide ranges reflecting high uncertainty on the true values. There are no empirical data on the establishment of new persisting founder populations by single mated female moths, and this negatively affects the certainty of the estimates.

Climatic modelling of establishment

Establishment modelling results in maps of relative likelihood of establishment, and thresholds such as the threshold of EI ≥ 30 for establishment in the CLIMEX model should not be taken absolutely. In reality, there is an increase in suitability for establishment, number of generations and likely level of impact as EI increases. At which EI population densities are high enough for the insect to establish or to cause impact, is unknown. The threshold EI ≥ 30 is recommended in the CLIMEX documentation as a practically useful guide (Kritikos et al., 2015), but it cannot be used to distinguish areas where the insect can and cannot establish. Degrees of establishment potential exist. Maps can also not be translated 1:1 in expected population densities, but zones with higher EI are more likely to have higher population densities and higher impacts than zones with lower EI. Cross continents, zones with similar EI are expected to have similar pest pressure if the organism establishes outside the native range, but other factors than those accounted for in the model may affect pest pressure, e.g. natural enemies and cultivation practices. The Panel considered a threshold EI ≥ 15 to account for uncertainty about the precise value of the threshold.

Spread

Estimates of lag phase and constant rate of range expansion were made based on the spread of a related species, Leucinodes laisalis, in the southern Iberian Peninsula. It is not known whether the spread of L. laisalis was from one or multiple founder populations, which affects the certainty of the estimates.

Impact

Many studies have investigated the impact of L. orbonalis, especially in India. The majority of studies expressed impact as infestation rate (number of shoots and/or fruit affected), but yield loss is more related to severity than infestation rate. This requires estimation of yield losses and causes uncertainty.

Studies on impact are done under conditions that are conducive to impact to increase the power of comparing treatments in experiments, but this reduces the representativeness of the resulting data. Experiments that result in low impact may never be published because treatment comparisons would likely be inconclusive. Again, this lowers the representativeness of the studies.

Uncertainty decomposition

The decomposition of uncertainty with the pathway model (Table 10) indicates that the largest uncertainty is within the estimate of the amount of infested produce at the origin of the pathway (51% of model uncertainty). The level of pest infestation is often the largest uncertainty in quantitative pest risk assessments (e.g. EFSA PLH Panel, 2016a, 2016b, 2017a, 2017b2023). The next largest uncertainty in the model is the estimate of likelihood that larvae would escape from consumer waste and complete their development (15.0%), followed by the probability that eggs oviposited in the EU survive, hatch and the resulting larvae develop and complete development to reproduce and initiate a founder population (13%). Combining the factors involved in transfer, 45% of the model uncertainty is due to lack of information about transfer which is an area of invasion biology that typically lacks empirical evidence on the detailed steps involved (Leung et al., 2004; Simberloff, 2009).

TABLE 10. Decomposition of explained variance in the pathway model for introduction of Leucinodes orbonalis. R2 gives the partial R2 of each regressor in a linear regression meta-model of pathway model results in which the number of founder populations is the response variable and the parameter values in the model are regressors. The final column indicates the relative contribution of each parameter to explained variance. Here, variance represents the uncertainty in pathway model calculations, and the contribution of each parameter is the contribution to uncertainty.
Rank Model parameter R2 decomposition Relative sensitivity (%)
#1 Infestation rate on pathway on arrival into EU 0.24 51
#2 Escape from consumer waste 0.07 15
#3 Survival and development of eggs to initiate founder population 0.06 13
#4 Mating in EU 0.06 13
#5 % discarded by consumers 0.02 4
#6 Quantity of imports 0.01 2
#7 Weight of turkey berry cluster 0.01 1
#8 Weight of exotic eggplant varieties 0.00 1
#9 % discarded by trade 0.00 0
Sum 0.46 100

4 CONCLUSIONS

Following a request from the European Commission, the EFSA Panel on Plant Health performed a pest risk assessment of Leucinodes orbonalis for the EU. The quantitative assessment focused on pathways and likelihood of entry, climatic conditions allowing establishment, the distribution of imported material within the EU after entry, the likelihood of establishment, the rate of spread following a lag period and potential impacts to Solanaceous crops in the EU.

L. orbonalis is an oligophagous pest feeding mainly on Solanaceae family crops (such as eggplants, potatoes and tomatoes). The main pathway (S. melongena and S. torvum from Asian countries where L. orbonalis presence is recorded) was deduced from the potential combinations between crops and countries of the origin. L. orbonalis is not known to occur outside of Asia and Oceania and the majority of records in Africa is most probably invalid. This species has been intercepted in the EU over 300 times since 2004. Based on the size and frequency of imports, and with evidence of interceptions in the EU, the interceptions were mainly on S. torvum and S. melongena fruits. The CN code for eggplants HS 070930 unifies the two above mentioned species in the category called ‘eggplants’ or ‘aubergines’.

The import data for those goods were downloaded from the Eurostat database for the years 2010–2019 from the Asian countries with the records of L. orbonalis extracted from the scientific literature. Based on the size and frequency of imports, and with evidence of interceptions in Europe, the importation of eggplants from Asian countries was identified as the most likely pathways for entry. Tomato and potato are hosts of L. orbonalis, but were not considered as pathways as the larvae cannot develop to maturity in tomato fruit while import of potato from Asia is forbidden and the insects do not enter the tuber under field conditions.

Climate matching and CLIMEX modelling indicate that conditions are most suitable for establishment of L. orbonalis in parts of the southern EU, especially around the Mediterranean Sea. Two possible scenarios for establishment were considered based on two EI thresholds. Using EI ≥ 15 approximately 23% of imported S. melongena and S. torvum are distributed to NUTS2 regions in which climatic conditions are suitable for establishment. Using EI ≥ 30 approximately 14% of imports reach such areas.

Each infested eggplant entering the EU is likely to contain only one larva, as such an important limiting factor in establishing a founder population is the likelihood of a male and a female emerging in temporal and spatial proximity to locate each other and mate. With respect to the need of larval development to adulthood from discarded infested produce, then mating and the progeny surviving, the number of newly established founder populations developing was estimated to be 0.00014 per year (90% CR 0.00000–0.00264). Thus, the Panel would not expect new founder populations within the foreseeable future. Nevertheless, if a founder population were to establish, it would likely remain local for a number of years and the lag period before sustained spread was estimated to be 34.5 years (90% CR 5–92 years) following the establishment of a founder population. L. orbonalis is not considered to be a strong flyer. Were L. orbonalis to establish, the median rate of natural spread was estimated to be 2.3 km/year (90% CR 0.65–7.0 km per year).

Impact assessment focused on Solanaceae host plant species, mainly eggplants. In a scenario where L. orbonalis has spread and is managed by farmers as part of the general pest fauna, i.e. no specific official phytosanitary measures are in place against it, and growers apply targeted pest control against L. orbonalis, median yield losses in eggplant were estimated to be 0.54% (90% CR 0.13%–1.94%). The Panel found insufficient evidence to consider tomato and potato production to be at risk from infestation by L. orbonalis if the insect were to establish in the EU because tomato and potato are unpreferred hosts. In the presence of eggplant, L. orbonalis will feed on eggplant. The Panel did not consider the possibility of shifts to previously unreported host plants because it falls outside the scope of a PRA (EFSA PLH Panel, 2018).

Concluding overall, this opinion shows that the EU encompasses regions with climate suitable for the establishment of L. orbonalis, and that this species could cause damage if it established. However, it is unlikely to be introduced in the foreseeable future because of the relatively low volume of commodity providing a pathway, and the low likelihood that adults emerging in the EU will successfully mate and initiate a founder population.

ABBREVIATIONS

  • CN
  • Combined nomenclature (8-digit code building on HS codes to provide greater resolution)
  • CR
  • certainty range
  • DD
  • degree days
  • DNA
  • Deoxyribonucleic acid
  • EI
  • ecoclimatic index (an index of climatic suitability used by CLIMEX)
  • EKE
  • Expert Knowledge Elicitation
  • EPPO
  • European and Mediterranean Plant Protection Organization
  • HRP
  • High Risk Plants
  • HS
  • Harmonised System (6-digit World Customs Organization system to categorise goods)
  • IPM
  • Integrated Pest Management
  • ISPM
  • International Standard for Phytosanitary Measures
  • MS
  • Member state (of the EU)
  • NUTS
  • Nomenclature Units for Territorial Statistics
  • RRO
  • risk reduction option
  • SDM
  • Species Distribution Model
  • ToR
  • Terms of Reference
  • ACKNOWLEDGEMENTS

    The EFSA PLH Panel wishes to thank the following for the support provided to this scientific output: Irene Pilar Munoz Guajardo and Marika De Santis from the EFSA library for providing an easy access to the crucial references; ISA Expert: Ivana Majić (University of Osijek, Faculty of Agrobiotechnical Sciences) for the first draft of the Expert Knowledge Elicitation evidence dossier for the Spread and Impact session. The EFSA PLH Panel wishes to thank Hearing Experts: Cherubino Leonardi (University of Catania, Italy), Aleksandar Jovanovic (Export Marketing Consultant in ‘Autentika’), Srinivasan Ramasamy (the Flagship Program Leader for Safe and Sustainable Value Chains and Lead Entomologist of the World Vegetable Center (Taiwan), Luke Chinaru Nwsou (University of Port Harcourt, Nigeria).

      CONFLICT OF INTEREST

      If you wish to access the declaration of interests of any expert contributing to an EFSA scientific assessment, please contact [email protected].

      REQUESTOR

      European Commission

      QUESTION NUMBER

      EFSA-Q-2023-00069

      PANEL MEMBERS

      Claude Bragard, Paula Baptista, Elisavet Chatzivassiliou, Francesco Di Serio, Paolo Gonthier, Josep Anton Jaques Miret, Annemarie Fejer Justesen, Alan MacLeod, Christer Sven Magnusson, Panagiotis Milonas, Juan A. Navas-Cortes, Stephen Parnell, Roel Potting, Philippe L. Reignault, Emilio Stefani, Hans-Hermann Thulke, Wopke Van der Werf, Antonio Vicent, Jonathan Yuen, and Lucia Zappalà.

      COPYRIGHT FOR NON-EFSA CONTENT

      EFSA may include images or other content for which it does not hold copyright. In such cases, EFSA indicates the copyright holder and users should seek permission to reproduce the content from the original source.

      MAP DISCLAIMER

      The designations employed and the presentation of material on any maps included in this scientific output do not imply the expression of any opinion whatsoever on the part of the European Food Safety Authority concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.

      APPENDIX A: Identity of Asian Leucinodes

      There is the possibility of currently undiscovered Asian Leucinodes species similar to the situation uncovered by Mally et al. (2015) in Sub-Saharan Africa. The vast majority of reviewed Asian literature does not report any efforts or methods of identifying the investigated specimens to species. In order to investigate this possibility of undiscovered Leucinodes species, the available literature for published DNA sequence data were screened, that could help answering this question. A part of the mitochondrial cytochrome oxidase I (COI) gene was proposed to serve as ‘DNA Barcode’ to identify animal species based on their unique group of DNA Barcode sequences (Hebert et al., 2003). Therefore, ‘DNA Barcode’ sequences were compiled from several publications on the Austral-Asian L. orbonalis (Chang et al., 2014; Mally et al., 2015; Sagarbarria et al., 2018; Shashank et al., 2015; Murali et al., 2017; Palraju et al., 2018; Karthika et al., 2019; Natikar et al., 2022; Padwal et al., 2022), DNA Barcode sequences available in the Barcode of Life Database and sequences not published in the literature but submitted as ‘unpublished’ to GenBank. We assembled a total of 398 sequences into a DNA alignment (Data S1) and used the R packages APE (Paradis et al., 2004) and SPIDER (Brown et al., 2012) as well as the FigTree v1.4.4 software (Rambaut, 2006–2018) to analyse the data and illustrate a neighbour-joining tree (Figure A.1).

      Details are in the caption following the image
      Neighbour-joining tree of Austral-Asian Leucinodes partial cytochrome oxidase I (COI) sequences. The coloured groups represent putative undescribed species from Vietnam (in blue) and Australia (in red and orange); the large remainder of sequences (marked in black) refers to L. orbonalis and exclusively originates from Asia. The scale bar represents 0.02 nucleotide substitutions per site (Creator and Copyright©: R. Mally, R packages ape and spider, FigTree v1.4.4).

      In the Barcode of Life Database (https://v4.boldsystems.org), L. orbonalis is represented by the Barcode BIN BOLD:AAE7334. However, based on examination of DNA sequences of the mitochondrial COI gene, there appear to be three potentially undescribed Leucinodes species: one from Hanoi (Vietnam) (Barcode BIN BOLD:ACH6349; Chang et al., 2014), and two from Queensland (Australia; Barcode BINs BOLD:AAO2620, BOLD:AAO9130), potentially occurring sympatrically with L. orbonalis.

      APPENDIX B: Biology of Leucinodes orbonalis

      Leucinodes orbonalis is a species of Lepidoptera. The life cycle comprises an egg stage, six larval stages, a pupal stage and an adult stage.

      Biology. The biology of L. orbonalis has been described in more detail from populations in India (Lall & Ahmad, 1965; Saxena, 1965), the Philippines (Ardez et al., 2008; Navasero & Calilung, 1990) and China (Li et al., 2014). The pre-imaginal development of L. orbonalis is temperature-dependent (Koundinya et al., 2019; Li et al., 2014; Mannan et al., 2015).

      Egg. Eggs are usually laid in four subsequent nights, with 48–53 eggs per night totalling to 192–212 eggs (Saxena, 1965); in the Philippines, Navasero and Calilung (1990) found the females to lay 8–295 eggs, with an average fecundity of 118 eggs per female. Eggs are laid during night, singly or in small clusters of two to four on the lower leaf surfaces of the topmost and middle leaves of S. melongena (Ardez et al., 2008); under lab conditions, eggs are laid in larger egg masses of 40–60 eggs (Mannan et al., 2015). The eggs are oval and creamy white, turning orange towards larval hatching, with the black head capsule becoming visible through the eggshell (Ardez et al., 2008; Lall & Ahmad, 1965). The diameter of a single egg is on average 0.48 mm (Lall & Ahmad, 1965). Egg development time is inversely correlated with temperature, with higher temperatures shortening the time for development (Li et al., 2014). Larval hatching tends to occur primarily in early mornings (Mannan et al., 2015) and late afternoons (Li et al., 2014).

      Larva. Freshly hatched larvae wander around on their host plant in search of a suitable place to penetrate the plant tissue; this searching lasts from a few minutes (Das & Patnaik, 1971) to about an hour (Saxena, 1965). The entry hole into the fruit/shoot is plugged with excreta, presumably to keep the plant part from drying out, and to keep pathogens, parasitoids and predators out. However, Navasero and Calilung (1990) report that the entry holes and excreta facilitate the development of secondary infections by microorganisms and insect scavengers (Staphylinidae, Phoridae). The entry holes may also partly heal while the fruit is further growing in size (Navasero & Calilung, 1990). Usually, a single fruit is inhabited by one or more rarely two larvae, but sometimes up to four larvae can be found in one fruit (Akulov et al., 2015; Shukla, 1986).

      While Saxena (1965) states that the larva does not leave the inside of the plant shoot or fruit until it starts building the cocoon for pupation, Navasero and Calilung (1990) assume that larvae from dried-up, fallen leaves, flowers and shoots climb back up the plant and infest other plant parts, which is evidenced by exit holes (which lacked a plug of excreta) and entry holes larger than those made by first- and second-instar larvae. According to Li et al. (2014), the larvae feed in as many shoots as necessary for development, but once they enter fruit, they remain inside until mature enough for pupation. In host plants such as Solanum nigrum with fruits too small to harbour the growing larva, it exits the fruit and webs together three to four fruits to feed on them externally from inside the web (Das & Patnaik, 1971).

      Information about larvae feeding on leaves are contradictory: Saxena (1965) expressed doubts about earlier reports of the caterpillar boring into the leaves' petioles and mid-ribs (Banerjee & Basu, 1955), ‘because the entry of the caterpillar inside the plant tissue brings about a withering effect on the tender parts of plants which dry up and fall down’. Li et al. (2014) also state that L. orbonalis does not damage the leaves of its host plant. However, Navasero and Calilung (1990) confirmed that larvae bore into the petioles and mid-ribs, causing the leaves to dry and eventually fall off. Especially the young instars feed on the petals of ovaries of flowers, causing the flower to fall off or to develop into an abnormally shaped fruit that may or may not reach maturity (Li et al., 2014; Navasero & Calilung, 1990).

      When feeding in shoots, the larva enters this part of the host plant through tender parts, e.g. where the leaf buds are located. Stem-boring leads to the drooping of the topmost leaves due to the disruption of the water-conducting tissues in the stem. Fruits are entered directly and not through the shoots (Saxena, 1965), and apparently usually from below the calyx where the neonate larvae hide (Das & Patnaik, 1971; Akulov et al., 2015). When fruits are present on the host plant, they are preferred by the larvae over the plant's shoots (Navasero & Calilung, 1990), potentially due to the shoots becoming tough and hard in older plants (Lall & Ahmad, 1965).

      From the fifth instar on, the caterpillar feeds less voraciously than in earlier instars (Saxena, 1965). If disturbed in the fourth instar, or if the fruit starts to rot, the caterpillar will construct a silken cocoon and for the remainder of its larval development stay in there, passing the remaining instars and eventually pupation in there (‘cocooning behaviour’; Saxena, 1965). This has important implications for the potential accidental introduction of L. orbonalis to other parts of the world: The larva only requires a time window of about 15–16 days of undisturbed feeding from hatching from the egg to the fourth larval instar in order to get to a stage of development where no more food is required and the life cycle can be completed. This cocooning behaviour, with the larva residing in its pupation cocoon from the fifth or sixth larval instar on, might explain why some authors (e.g. Atwal, 1976; Boopal et al., 2013; Mathur & Jain, 2006) report only four or five larval instars.

      Three days after the fifth moult, the larva leaves the fruit/shoot through a relatively large exit hole to find a suitable place on the lower branches of the plant among dry leaves for pupation (Saxena, 1965).

      In tropical climate, the development of second- to sixth-instar larvae is faster compared to that in subtropical climate (Figure B.1). Egg, pre-pupal and pupal stages, on the other hand, appear to hardly vary in their duration between different climates (Figure B.1).

      Details are in the caption following the image
      Relative (x-axis) and absolute (y-axis) duration of the phases of development of L. orbonalis from egg to pupa for Northern India (green area, from Saxena 1965) and the Philippines (blue area, from Navasero & Calilung, 1990); the shaded areas between the point lines indicate the range between minimum and maximum developmental time for each stage. Baseline for relative durations of developmental phases based on Saxena (1965) (Creator and Copyright©: R. Mally, data from Table B.1, created with R package collection ‘tidyverse’).
      TABLE B.1. Pre-imaginal stages, their durations and body measures; from (a): Saxena (1965; Patna, Bihar state, India), (b): Navasero and Calilung (1990; Los Baños, Laguna province, Philippines).
      Stage Duration of instar (days) Length Width
      Egg

      (a): 3 in summer, 3–4 days in winter

      (b): 4.12 ± 0.22 days

      (a): 0.8–0.85 mm

      (b): 0.82 mm

      (a): 0.5–0.6 mm

      (b): 0.59 mm

      1st larval

      (a): 2–3 days

      (b): 2.19 ± 0.25 days

      (a): 0.7–3.2 mm

      (b): 1.3 ± 0.094 mm

      (a): 0.2–0.5 mm

      (b): 0.26 ± 0.02 mm

      2nd larval

      (a): 3–4 days

      (b): 2.16 ± 0.35 days

      (a): 4.5–6.2 mm

      (b): 2.92 ± 0.15 mm

      (a): 0.8–1.3 mm

      (b): 0.41 ± 0.05 mm

      3rd larval

      (a): not stated

      (b): 1.77 ± 0.34 days

      (a): 6.5–8.7 mm

      (b): 5.32 ± 0.61 mm

      (a): 1.4–1.7 mm

      (b): 0.61 ± 0.05 mm

      4th larval

      (a): 4–5 days

      (b): 1.8 ± 0.33 days

      (a): 9.3 mm

      (b): 8.91–9.79 mm

      (a): 1.8 mm

      (b): 0.93–1.19 mm

      5th larval

      (a): 5–6 days

      (b): 1.61 ± 0.45 days

      (a): 16.8 mm

      (b): 13.41 ± 0.82 mm

      (a): 2.9 mm

      (b): 1.8 ± 0.2 mm

      6th larval

      (a): 3 days

      (b): 1.36 ± 0.54 days

      (a): 19.3 mm

      (b): 18.2 ± 1.51 mm

      (a): 3.6 mm

      (b): 3.62 ± 0.37 mm

      Prepupa

      (a): 46–48 h

      (b): 1.34 days

      (a): 11.3 mm

      (b): Not stated

      (a): Not stated

      (b): Not stated

      Pupa

      (a): 8–9 days, regardless of the season

      (b): 8.1 ± 0.2 days

      (a): 11.6 mm

      (b): 12.21 ± 0.63 mm

      (a): 3.8 mm

      (b): 4.2 ± 0.25 mm

      Adult

      (a): ~ 7 days (females), 4–5 days (males)

      (b): 4–8 days (mean 5.84 days)

      (a): Not stated

      (b): 13 mm

      (a): 22 mm wingspan

      (b): 23.5 ± 0.5 mm wingspan

      Larval host plants. The species is part of the tribe Lineodini, a lineage of Crambidae moths that has its highest diversity in the American tropics, and that has specialised on Solanaceae (nightshades) as larval food plants (Mally et al., 2019). The preferred host plant is eggplant/aubergine (Solanum melongena), where usually 1–3 (but up to 20) larvae feed inside a single fruit, which primarily depends on the number of seeds in the fruit (the fewer seeds, the more larvae), but also on the fruit's size (personal communication dr S. Ramasamy). Despite numerous citations of non-Solanaceae larval host plants in the literature, L. orbonalis appears to be unable to complete its life cycle on plants other than nightshades (Alam et al., 2003). Ardez et al. (2008) studied the host plant preference in no-choice trials in the laboratory and found that L. orbonalis larvae were unable to develop on fruits of okra (Abelmoschus esculentus, Malvaceae) or tubers of sweet potato (Ipomoea batatas, Convolvulaceae). Larvae were feeding on cowpea fruits (Vigna unguiculata, Fabaceae), but formed abnormal pupae, from which no adults emerged (Ardez et al., 2008).

      In free-choice trials, L. orbonalis only laid eggs on Solanum melongena, but on none of the other six available plants (three Solanaceae: S. tuberosum, S. nigrum, S. lycopersicum, and three non-Solanaceae: Vigna unguiculata, Abelmoschus esculentus, Ipomoea batatas), indicating that eggplant is the preferred larval host plant (Ardez et al., 2008). However, no-choice trials showed that the life cycle can also be completed on potato tubers (S. tuberosum) and fruits of black nightshade (S. nigrum), although the moths reared from potato tubers had a shorter adult lifespan compared to moths reared from fruits of eggplant and black nightshade (Ardez et al., 2008). Under natural conditions (i.e. not in laboratory rearings), L. orbonalis larvae appear to only attack the shoots of potato but not the tubers (Nair, 1967; Natikar & Balikai, 2021). The larvae can successfully complete their life cycle in the shoots of potato plants (Navasero & Calilung, 1990), and the species has become the most harmful organism for potato crops in the Indian state of Karnataka (Natikar & Balikai, 2018a). In plants with fruits being too small for the growing caterpillar to effectively feed inside (as is the case e.g. in S. nigrum), the larva in later instars exits the fruit and spins several fruits together with a silken web, lives in the fruit cluster and feeds on the fruits externally (Das & Patnaik, 1971; Navasero & Calilung, 1990). Outside the Indian growing season, when eggplant is not planted, L. orbonalis uses the black nightshade, S. nigrum, and the yellow-fruit nightshade, S. virginianum (= S. xanthocarpum) as larval host plants, feeding in shoots and fruits (Lall & Ahmad, 1965).

      Ardez et al. (2008) also report that larvae successfully fed on tomato fruits (Solanum lycopersicum, Solanaceae) and reached the pupal stage, but as in cowpea, no adults emerged from the pupae. Contrary to the observation of Ardez et al. (2008), Das and Patnaik (1971) found the species boring into tomato fruits and shoots in the wild, and, like Boopal et al. (2013), successfully reared larvae to adults on tomato fruits in the lab. Lab-reared L. orbonalis also successfully completed their life cycle in the study of Jethva and Vyas (2009), but it is not clear from the study on which part of the tomato plant (stem/fruits) the larvae were feeding. Due to the paucity of literature reporting L. orbonalis infestations on tomato in Asia as well as the absence of interceptions of L. orbonalis from tomato fruit imported from Asia, tomato is not considered as a pathway for entry (Table B.2).

      TABLE B.2. Leucinodes orbonalis larval host plants: summary of information from literature about biology.
      Hosts described in the literature Number of references (refIDs) Life stage of L. orbonalis Where on the plant detected Detailed description from the cited reference
      Solanum aethiopicum (Ethiopian eggplant), gilo cultivar group (Scarlet eggplant); synonym S. integrifolium 2 (Kariyanna et al., 2020a, Kariyanna et al., 2020b) Egg, larva, pupa, adult Shoots, fruits
      Solanum incanum × Solanum melongena F1 hybrid variety 1 (Baksh, 1979) Larva Fruits
      Solanum lycopersicum (tomato) 5 (Ardez et al., 2008; Boopal et al., 2013; Das & Patnaik, 1971; Jethva & Vyas, 2009; Maureal et al., 1982) Egg, larva, pupa Leaves, fruits

      ‘Bores into tender shoots’ (Shrestha & Lee, 2020)

      ‘unclear where the eggs where laid, and on which plant tissue the larvae fed’ (Boopal et al., 2013)

      Solanum macrocarpon (African eggplant) 2 (Dhaliwal & Aggarwal, 2021; Hanur et al., 2011) Larva, pupa (in the soil) Fruits, shoots

      ‘BSFB was able to infect both the shoot meristem tips as well as developing fruits, suggesting that S. macrocarpon is vulnerable to BSFB in all stages. The extent of susceptibility, however, was more in the latter case. Also, larvae of BSFB were found to search for the fruit rather than the tougher calyx region when challenged onto the fruits per se.’ (Hanur et al., 2011)

      ‘Larvae of L. orbonalis have an internal feeding habit, which results in withered shoots and destruction of fruit tissues.’ (Dhaliwal & Aggarwal, 2021)

      Solanum melongena (eggplant, aubergine, brinjal) 147 references Egg, larva, pupa, adult Flowers, shoots, stems, fruits, leaves

      ‘The pest overwinters in larval stage under a hardy, leathery and dirty coloured cocoon attached to the host plant usually at about 1–3 cm below the soil surface’ (Lal, 1975)

      ‘larvae bore into tender shoots during vegetative growth, subsequently attacking flowers and fruits’ (Islam et al., 2020)

      ‘SFB larvae enter the internal tissue of the plant and become inaccessible to the external environment by sealing their entry points with their silk and excreta’ (Talukder et al., 2021)

      ‘Larvae grow to 2 cm in length with segmented legs, mandibles and the body is divided into three parts – head, thorax and abdomen. It moults 4 times and passes through 5 instars (Singla et al., 2018).’ (Talukder et al., 2021)’

      ‘Female can lay up to 250 white-coloured with 0.5 mm diameter and flat eggs singly on lower surface of leaves, shoots, flower buds and sometimes on calyces of fruits. Incubation period is 3–6 days; larva becomes mature in 10–15 days. Larva is stout, pinkish in colour with brown head (Plate 7.2L). Pupation takes place in greyish boat-shaped tough cocoons (Plate 7.2 m) among fallen leaves or on plant itself. Pupal period lasts for 6–10 days, after which adult emerges. The whole life cycle from egg to adult emergence takes 21–45 days with up to five overlapping generations in a year’ (Kunjwal & Srivastava, 2018)

      ‘Being an internal borer the early instar larvae of this pest feed exclusively on the tender shoots, petiole and flower buds, while the later instars bore into the fruits’ (Nayak, 2014)

      ‘Third- and fourth-instar ESFB larvae are most responsible for the damage to eggplant, according to reports. The larvae of the eggplant shoot and fruit borer eat the vulnerable shoots of infected plants’ (Islam et al., 2022)

      ‘Because the larvae hide in fruits and shoots, insecticides used to get rid of them are typically unsuccessful’ (Islam et al., 2022)

      ‘lay its eggs on the lower surface of fresh leaves in clusters of 80–253 eggs on average’ (Ullah et al., 2022)

      ‘Larvae also feed on flowers, which results in flower drop or misshapen fruits’ (Shelton et al., 2019)

      ‘Numbers of larvae per leaf ranged among the five compared cultivars from 0.27 to 0.43 larvae per leaf’ (Kassi et al., 2019)

      ‘Once the larva bores into petiole and midrib of leaves and tender shoots, it causes dead hearts. In later stages, it also bores into the flower and fruits’ (Singh et al., 2014)

      Solanum nigrum (black nightshade) 2 (Ardez et al., 2008; Das & Patnaik, 1971) Egg, larva Fruits, leaves

      ‘The eggs of EFSB [...] reared on black nightshade which laid their eggs on the underside of the leaflets, hatched in 4 days. The EFSB eggs are usually laid singly or in a cluster of 2–4.’ (Ardez et al., 2008)

      ‘As the larvae grows, it cannot accommodate itself in the fruit of Solanum nigrum. Therefore, it comes out and webs 3–4 such fruits together and feeds on them by remaining inside the web.’ (Das & Patnaik, 1971)

      Solanum sp. 2 (Mannan et al., 2015; Vinod et al., 2001) Larva Fruits
      Solanum tuberosum (potato) 11 (Lall & Ahmad, 1965; Nair, 1967; Maureal et al., 1982; Hanapur & Nandihalli, 2004; Ardez et al., 2008; Tribikram & Patnaik, 2010; Boopal et al., 2013; Kariyanna et al., 2020a; Kariyanna et al., 2020b; Natikar & Balikai, 2018b, 2022) Egg, larva, pupa Shoots, leaves, roots

      ‘The larva of shoot borer, L. orbonalis, attacks the shoots of potato, causing withering and wilting of the stem ultimately resulting in retardation of the plant growth’ (Natikar & Balikai, 2022)

      ‘The newly hatched larvae entered the lamina of the leaf, fed on the parenchyma and then entered into the leaf stalk, as a result of which the leaf withered. It has been learnt (in litt.) from Dr. K. K. Nirula, Nematologist, Central Potato Research Institute, Shimla (Punjab) that L. orbonalis is found boring the shoots of potato in Bihar, but the incidence of the borer is very low.’ (Nair, 1967)

      ‘Up to 4 larvae were observed boring the shoots and leaf stalks on each plant of this variety [Alpha]’ (Nair, 1967)

      Solanum viarum (tropical soda apple) 1 (Udayagiri & Mohan, 1985) Larva Fruits ‘The caterpillars were noticed feeding on the contents of the developing berries. The entry holes were plugged with excreta.’ (Udayagiri & Mohan, 1985)

      Pupa. The sixth-instar larva (or fourth- to fifth-instar larvae in the case of ‘cocooning behaviour’, see under ‘Larva’) leaves the plant part it was feeding in and searches for a dry place for pupation, usually near the stem base (Lal, 1975), or on twigs or leaves (Mathur & Jain, 2006), where it constructs a boat-shaped pupation cocoon, leaving a small exit hole in the cocoon from which it will later emerge as an adult (Saxena, 1965). Construction of the cocoon takes about a day (Navasero & Calilung, 1990). The fresh cocoon is completely white and turns into a dark brown in the course of a few hours (Saxena, 1965), matching the surroundings and making it difficult to spot the cocoons (Alam et al., 2003). According to Ardez et al. (2008), the pupal cases' colour depends on the larval host plant, ranging from (light) brown (reared on eggplant) over light to yellowish brown (tomato) to pinkish brown (on black nightshade). The size of the cocoon is determined by the nutrition status of the larva (Patel & Basu, 1948; Saxena, 1965). The actual pupation is, as all other stages of the insect, temperature-dependent (Alam et al., 2003; Mannan et al., 2015) and takes 8–9 days on average (Ardez et al., 2008; Lall & Ahmad, 1965; Navasero & Calilung, 1990; Saxena, 1965). The maculation of the adult becomes slightly visible about 2 days before the moth emerges (Navasero & Calilung, 1990).

      Adult. After emergence form the pupal cocoon, the adult rests for about 15 min upon the cocoon in order to dry, and to pump haemolymph into the wings, thus completely unfolding them (Saxena, 1965). Adult emergence takes primarily place in the first part of the night (Alam et al., 2003; Mannan et al., 2015).

      Female moths can be distinguished from males by their somewhat larger appearance due to a thicker abdomen and a larger wingspan. Furthermore, the female bents its thick abdomen upwards above the body in resting position (Figure B.2), releasing sex pheromones from the abdomen tip to attract males for mating (‘calling behaviour’), whereas males keep their slender abdomen straight at rest. The proboscis is well-developed, but according to Saxena (1965) the adults do not appear to feed; nonetheless, in lab rearing studies, a sugary liquid is generally provided for the adults to feed on (Lall & Ahmad, 1965; Navasero & Calilung, 1990).

      Details are in the caption following the image
      Adult Leucinodes orbonalis female in its typical resting position with the abdomen bent up, releasing sex pheromones from the abdomen tip to attract males for mating. Photo: © Md Jusri, under CC-BY-NC 4.0 licence (https://creativecommons.org/licenses/by-nc/4.0/), iNaturalist observation https://www.inaturalist.org/observations/15009501.

      The adults are purely nocturnal, with emergence from the pupal cocoon, mating and egg laying all taking place at night, while remaining inactive during the day, resting on the underside of the host plants' leaves (Mannan et al., 2015). They are attracted to light and can be caught and monitored using light traps (Yousafi et al., 2016a2016b).

      The adult stage is short-lived: approximately 3 days after its emergence form the pupal cocoon, and after mating, the female lays eggs for four consecutive nights and then dies; the males live for 4–5 days and die shortly after copulation (Saxena, 1965). Other sources state that mating occurs already in the night of emergence from the cocoon or the following night (Kavitha et al., 2008; Mannan et al., 2015). The entire copula lasts approximately 30–45 min (Mannan et al., 2015). Females usually mate more than once (Mathur & Jain, 2006). An artificial sex pheromone mimicking that of the females has been developed to mass-trap males and disrupt mating (Cork et al., 2001; Gunawardena et al., 1989). Females produce their sex pheromone mainly during a time window from 10 p.m. to midnight, and male response is highest during the first night (Gunawardena et al., 1989), potentially due to the female still being virgin. The female sex pheromone has a maximum effective range of 10 m, and moths released at 15 or 20 m from a pheromone source did not reach the source in a mark-and-recapture study (Prasad et al., 2005). The female starts laying eggs in the night after the (first) copulation, although egg laying in the night of copulation has also been reported (Mannan et al., 2015 and references therein). The number of eggs laid decreases with each night, and the entire duration of oviposition is 2–4 days (Mannan et al., 2015 and references therein).

      Generations per year. In tropical climate (Laguna province, the Philippines) under lab rearing conditions, the life cycle of L. orbonalis takes about 26–36 days (Navasero & Calilung, 1990), so that 10–14 generations per year can be completed. The species also reproduces year-round under subtropical conditions in regions such as Bihar state (India), with the entire life cycle ranging from 39 days in January/February to 26 days in May, with an average of 30.61 days per generation (Lall & Ahmad, 1965). In such a climate, 11–12 generations per year can be completed. Jethva and Vyas (2009) extrapolated from lab studies of L. orbonalis reared on tomato that the population would multiply by a factor of 2.358 per week.

      Hibernation. In the Kullu Valley of Himachal Pradesh, India, the species occurs up to the middle part of the valley (1200–2000 m above sea level), where cold winter temperatures (down to −6.5°C) and snow cover (up to 200 cm) require L. orbonalis to hibernate (Lal, 1975). Here, the species overwinters as fully grown larva in a cocoon attached to the base of the main shoot, just above the roots between 3 cm beneath to 1.5 cm above the soil. There are usually 1–2, and up to 5 cocoons per plant (Lal, 1975). In lower altitude regions of India, where hostplants are available year-round, L. orbonalis supposedly does not hibernate (Lal, 1975).

      Natural enemies. Srinivasan (2008) lists the lacewing Chrysopa kulingensis (Neuroptera: Chrysopidae), an unidentified species of the true bug genus Campyloneura (Hemiptera: Miridae) and three species of Coccinellidae (Cheilomenes sexmaculata, Coccinella septempunctata, Brumoides suturalis) as predators of L. orbonalis, presumably especially of the eggs and pupae, and Navasero and Calilung (1990) report an unidentified earwig (Dermaptera) present in about 15% of damaged fruits that likely preys on L. orbonalis larvae, as these were either missing or dead in those fruits. Fathi (2022) found three predatory insect species on eggplant in Northern Iran: Chrysoperla carnea (Neuroptera: Chrysopidae), Orius niger (Heteroptera: Anthocoridae) and Coccinella septempunctata.

      Srinivasan (2008) lists six parasitic wasp species of Ichneumonidae, three of Braconidae, one Chalcididae and one Eulophidae species, as well as one species of Tachinidae (Diptera) as parasitoids of immature stages of L. orbonalis. In the Philippines, five species of parasitic wasps are reported by Navasero and Calilung (1990), with an unidentified Apanteles species (Braconidae) being the most abundant parasitoid. Parasitisation rates of L. orbonalis larvae by Apanteles sp. were found to be higher in shoots (11.11%) as compared to fruits (3.12%), presumably because host larvae can be detected and reached more easily by the parasitoid in shoots than in the more voluminous fruits (Navasero & Calilung, 1990). In Northern Iran, two main parasitoids are reported: Trichogramma brassicae (Hymenoptera: Trichogrammatidae) and Bracon hebetor (Hymenoptera: Braconidae) (Fathi, 2022).

      The fungus Curvularia spicifera (Ascomycota: Pleosporaceae) is a plant pathogen but is reported (as synonym Bipolaris tetramera) as an L. orbonalis entomopathogen by Srinivasan (2008), citing Tripathi and Singh (1991).

      APPENDIX C: Pest Introduction: Identification of main pathway

      To be considered a genuine host, a phytophagous insect must be able to complete its development and produce viable progeny following mating by feeding only on the plant regarded as a host. Such information is seldom found in literature, which often only summarises the information on pests and plants they are associated with. Reports of different life stages being found on the plant can often be a good indicator of a plant being a true host. Table C.1 lists the plant species on which L. orbonalis has been intercepted, as well as plants recorded in the literature and amount of information available on that (sometimes none, or scarce).

      TABLE C.1. Compilation of plant commodities with L. orbonalis interceptions (2nd column) and host plants reported in the literature (3rd column), including confirmed feeding (4th column) and confirmed completion of the life cycle (5th column). Potential pathways (plant species) of introduction requiring action are marked in bold, potential pathways without action required are marked with an asterisk.*
      Plant species Number of interceptions on that commodity Plant part imported – Any life stage associated? Number of literature references reporting as host Is the larva feeding on this host? (only confirmed cases) Does it complete the life cycle? (only confirmed cases)
      Solanaceae
      Capsicum annuum (bell and chilli peppers) 1 Fruit – no 3 No No
      Physalis angulata (= P. minima; angular winter cherry) 0 Fruit – no 2 No information No information
      Physalis peruviana (cape gooseberry) 0 Fruit – no 3 No information No information
      Solanum aculeatissimum (love-apple) 2 Fruit – yes 2 No information No information
      S. aethiopicum (= S. integrifolium; gilo/garden egg(plant)) 2 Fruit – yes 9 Yes No information
      S. aethiopicum × S. melongena 0 Fruit – yes 3 No information No information
      S. anguivi 2 Fruit – yes No information No information
      S. anomalum 0 Fruit – yes 2 Yes No information
      S. donianum (mullein nightshade) 0 Fruit – yes 1 No information No information
      S. incanum 0 Fruit – yes 3 Yes No information
      S. incanum × S. melongena 0 Fruit – yes 2 Yes No information
      S. linnaeanum (= S. undatum; devil's apple) 1 Fruit – yes No information No information
      S. lycopersicum (tomato) 0 Fruit – yes 16 Yes Yes
      S. macrocarpon (African eggplant) 2 Fruit – yes 5 Yes Yes
      S. macrocarpon × S. melongena 0 Fruit – yes 1 No information No information
      S. melongena (eggplant/brinjal/aubergine) 222 Fruit – yes 467 Yes Yes
      S. melongena × S. viarum 0 Fruit – yes 1 No information No information
      S. melongena × S. violaceum (‘indicum’) 0 Fruit – yes 1 No information No information
      S. myriacanthum (Himalayan nightshade) 0 Fruit – yes 2 Yes No information
      S. nigrum (black nightshade) 0 Fruit – yes 14 Yes Yes
      S. stramoniifolium 1 Fruit – yes No information No information
      S. torvum (turkey berry) 66 Fruit – yes 8 Yes No information
      S. tuberosum (potato) 0 Tubers – no 28 Yes Yes
      S. viarum (tropical soda apple) 0 Fruit – yes 1 Yes No information
      S. violaceum 0 Fruit – yes 7 No information No information
      S. virginianum (= S. xanthocarpum schrad.; yellow-fruit nightshade) 8 Fruit – yes 5 Yes Yes
      Solanum sp. 26 Fruit – yes 2 Yes No information
      Anacardiaceae
      Mangifera indica/sp. (mango) 3 Fruit – no 3 No information No information
      Spondias dulcis (golden apple) 1 Fruit – no 0 No information No information
      Asteraceae
      Helianthus sp. (sunflower) 0 No information 1 No information No information
      Convolvulaceae
      Ipomoea batatas (sweet potato)* 1 (ipomoea aquatica) Tubers – no 1 Yes No
      Cucurbitaceae
      Momordica sp. 4 Fruit – no 0 No (m. Charantia) No
      Fabaceae
      Cajanus cajan (pigeon pea) 0 Pods – no information 1 No information No information
      Cicer arietinum (chickpea) 0 Pods – no information 1 No information No information
      Pisum sativum (pea) 0 Pods – no information 2 No information No information
      Vigna unguiculata (= V. sinensis; cowpea) 1 (Vigna sp.) Pods – yes 1 Yes No
      Lamiaceae
      Ocimum sp. (basil) 2 No information 0 No information No information
      Malvaceae
      Abelmoschus esculentus (okra) 0 Pods – yes 1

      Yes (Ardez et al., 2008);

      No (Maureal et al., 1982)

      No
      Gossypium sp. (cotton) 0 No information 1 No information No information
      Myrtaceae
      Psidium guajava (guava) 2 Fruit – no 0 No information No information
      Poaceae
      Zea mays (corn) 0 No information 1 No information No information
      Sorghum sp. (broomcorn) 0 No information 1 No information No information
      Rosaceae
      Rosa sp. (rose) 0 No information 1 No information No information
      Rutaceae
      Citrus hystrix (kafir lime) 1 Fruit – no 0 No information No information
      Murraya paniculata (orange jasmine) 1 Fruit – no 0 No information No information
      Total 350 604

      As Table C.1 shows, the majority of non-Solanaceae plant species listed as host are based either on a few interceptions, many of which were in mixed consignments with eggplant, or on a few reports in the literature that usually refer to older publications or don't even provide a reference at all. These plant associations were deemed incidental, and thus dismissed as principal pathways for introduction. However, two cases stand out in Table C.1: Momordica sp. with four interceptions, and mango (Mangifera indica) with three interceptions and three reports in the literature.

      The claim of mango as host plant can be traced back to Hutson (1931), which is solely based on an abstract of a 17-page typescript that apparently never has been published, and that states ‘the shoot-borer, Leucinodes orbonalis, Gn., and the Pentatomid, Coptosoma siamica, Wlk., on mango’. The association with mango therefore remains dubious, especially since the pentatomid true bug Coptosoma siamica is known to only feed on Desmodium novaehollandiae latifolium (Fabaceae) (Rider, 2015), and not on mango as stated in the Hutson (1931) abstract. In the case of Momordica sp., there is no information in the literature reporting bitter melon as a host. In 24-h no-choice tests, Maureal et al. (1982) found that third-instar larvae did not feed on Momordica charantia (bitter melon) fruit. Both mango and bitter melon are therefore also excluded as principal pathways for introduction.

      Among Solanaceae, Capsicum was demonstrated to not be eaten by the larvae (Maureal et al., 1982); Physalis was only cited in the literature, but never intercepted in association with L. orbonalis, nor shown to be fed on by the larvae.

      Isahaque and Chadhuri (1983) report S. torvum (turkey berry) to be an alternative host. Turkey berry is the second-most common interception commodity of L. orbonalis in the EU, after eggplant. In Solanum species with small fruits, like S. nigrum and potentially also S. torvum, the individual fruits do not provide enough space for the developing larva to feed inside; instead, when the fruit becomes too small to sustain the caterpillar, it exits the fruit, spins together several fruits with a web and feeds from within this web externally on the fruits (Das & Patnaik, 1971). Leucinodes orbonalis infestations on imported fruits of such Solanum species should therefore be easily recognised, at least when infested with older larvae too large to feed inside the fruits. Turkey berry fruits have a diameter of 10–13 mm (Vorontsova & Knapp, 2023a).

      Ardez et al. (2008) observed in their lab experiments that moths reared from S. nigrum (black nightshade) fruits did not differ in their longevity from those reared from eggplant, and they conclude that S. nigrum is the most suitable alternative host plant for L. orbonalis. At the same time, Das & Patnaik (1971) report that adults reared from S. nigrum were smaller in size, and females laid far fewer eggs than compared to eggplant- or tomato-reared specimens; nonetheless, all eggs were viable, and the insect can complete its life cycle on this plant. Das & Patnaik (1971) report that the fruits are too small for the developing larvae, and they therefore exit the fruit they were feeding in as small larva, and web three to four fruits together, feeding on them from inside the web. Navasero and Calilung (1990) confirmed this larval behaviour of feeding on spun-together fruit clusters. The diameter of black nightshade fruits is 5–13 mm (Coleman et al., 2020). Although no study investigated whether L. orbonalis larvae can also feed on other parts of the plant, Panel assumes that they also bore into the shoots of black nightshade.

      In India, L. orbonalis larvae were reported to occasionally bore into the shoots and leaf stems of potato (S. tuberosum) plants, and infestation can reach up to 40% (Nair, 1967). In the lab, L. orbonalis larvae can be reared with potato tubers (Ardez et al., 2008; Boopal et al., 2013; Kariyanna et al., 2020a2020b; Maureal et al., 1982). The longevity of L. orbonalis adults reared on potato tubers was significantly shorter than of those reared on fruits of S. melongena and S. nigrum (Ardez et al., 2008). Maureal et al. (1982) observed a longer developmental period, higher mortalities especially in later larval instars, and a shorter adult lifespan in specimens reared on potato tubers as compared to specimens reared on eggplant. There are no published reports of L. orbonalis larvae feeding on tubers in the wild, and the species appears to feed solely on plant parts above ground.

      Das & Patnaik (1971) successfully reared L. orbonalis on the fruits of S. lycopersicum (tomato), with a somewhat lower fecundity of the resulting females as compared to S. melongena; they furthermore state that in Bhubaneswar (India), the larvae were observed to also bore into the shoots of tomato plants. Jethva and Vyas (2009) conducted laboratory rearings on tomato, but they did not specify the plant tissue they provided the larvae for feeding. Maureal et al. (1982) observed a longer developmental period, higher mortalities especially in later larval instars, and a shorter adult lifespan in specimens reared on tomato fruits as compared to specimens reared on eggplant.

      Tejavathu et al. (1991) find S. anomalum to be ‘resistant’ to L. orbonalis, despite 5.9% shoot- and 11.3% fruit infestation. No information is available on whether the species can complete its life cycle on this plant. Solanum anomalum has comparably small fruits, with diameters of 5.5–9 mm (Vorontsova & Knapp, 2023b). Solanum anomalum is endemic to Africa, and the species reported by Tejavathu et al. (1991) might be a misidentification of S. aethiopicum or S. anguivi, the wild progenitor of S. aethiopicum (Vorontsova & Knapp, 2023b). Hanur et al. (2011) report that L. orbonalis can successfully develop all larval stages on S. macrocarpon before pupating in the soil, and that ‘Patterns and extent of infestation and damage on the affected parts were comparable in both S. melongena and S. macrocarpon […], indicating that S. macrocarpon is equally susceptible to infestation by BSFB’. Srinivasappa et al. (1998) report an infestation rate of 10.16% of fruits, whereas according to Kumar and Sadashiva (1991), fruit damage by L. orbonalis larvae is less than 1%. Isahaque and Chadhuri (1983) report feeding on S. myriacanthum (Himalayan nightshade), who observed the larvae boring into the shoots. It is unknown whether the small fruits (diameter 20–30 mm; Nee, 2023) are also consumed. Udayagiri and Mohan (1985) report L. orbonalis from S. viarum (tropical soda apple), stating ‘The caterpillars were noticed feeding on the contents of the developing berries. The entry holes were plugged with excreta. L. orbonalis infestation was observed to be low’. Since the fruits are fairly small (22–25 mm in diameter according to Nee et al., 2023), the later-stage larvae may also feed from the outside, as reported for S. nigrum by Das & Patnaik (1971). Similarly small fruits (15–25 mm; Knapp, 2023) are present in S. virginianum (= S. xanthocarpum Schrad.; yellow-fruit nightshade), which Lall and Ahmad (1965) report (as S. xanthocarpum) to be a wild host of L. orbonalis ‘during off season (April to August)’, i.e. when S. melongena is not grown.

      Baksh (1979), Behera and Gyanendra (2002), Shinde (2007) and Pugalendhi et al. (2010) report shoot- and/or fruit infestation of hybrids of eggplant (S. melongena) with S. aethiopicum, S. incanum, S. viarum and S. violaceum (as syn. S. indicum), but it remains unclear whether the insect's life cycle can be successfully completed on these hybrids.

      In conclusion, the principal pathways for the potential entry of L. orbonalis into the EU are judged to be with fruit of Solanum melogena (eggplant) and of Solanum torvum (turkey berry).

      Introduction model: Quantity of imports

      As noted above, many potential pathways for entry were considered although the commodities judged most feasible were the import of exotic and specialist varieties of eggplant fruit (Solanum melongena) and turkey berry fruit (S. torvum).

      Solanum melongena is a widely consumed solanaceous crop with many common names including aubergine, brinjal, eggplant, garden egg, melanzana and patlican. The related Solanum aethiopicum has a similar common name of Ethiopian eggplant, while Solanum torvum is known as turkey berry or pea eggplant. There is a wide range of cultivars of S. melongena with differences in growth habits, fruit size, shape and colour. Depending on cultivar, the length of the fruits varies from approximately 5 cm to more than 30 cm. Fruit may be elongated, cylindrical, ovoid, slender or spherical. Fruit can be purple, green, white, yellow and variegated. The most familiar varieties grown and consumed in Europe are deep purple with a very high gloss.

      The market for eggplant can be divided into three sectors: common, special and exotic varieties (Table C.2). A relatively small amount of eggplants are imported into the EU from countries where L. orbonalis occurs. Asian eggplants are niche varieties imported by specialist buyers (CBI, 2020).

      TABLE C.2. Market segmentation of fresh eggplant in the EU.
      Eggplant category Target outlets Examples
      Common varieties Supermarkets, General greengrocers Typical and common EU grown eggplant varieties
      Special varieties Supermarkets, Specialist shops, Street markets Graffiti eggplant, Japanese eggplant
      Exotic varieties Ethnic shops, Restaurants White eggplant, Thai eggplant, Chinese eggplant

      Exotic and special varieties of eggplants are typically imported in standard 5 kg boxes (Re:fresh Directory, 2018).

      Overview of model input estimates: Import quantity

      Based on EU imports of eggplants over the period of 2010–2019 from countries where L. orbonalis occurs (Table C.3), the mean annual import was calculated. These countries are Austria (AT), Belgium (BE), Bulgaria (BG), Cyprus (CY), Czech Republic (CZ), Germany (DE), Denmark (DK), Estonia (EE), Spain (ES), Finland (FI), France (FR), Greece (GR), Croatia (HR), Hungary (HU), Ireland (IE), Italy (IT), Latvia (LT), Luxembourg (LU), Latvia (LV), Malta (MT), the Netherlands (NL), Poland (PL), Portugal (PT), Romania (RO), Slovenia (SI), Slovakia (SK) and Sweden (SE).

      From 2010 to 2014 imports declined then recovered (Figure C.1); between 2016 and 2019 they were relatively stable, ranging from a minimum of 157 tonnes in 2018 to a maximum of 182 tonnes in 2016. An import ban was established on Indian eggplant from 2014 to 2016 but was not extended due to improved phytosanitary measurements (https://www.freshfruitportal.com/news/2017/01/04/eu-lifts-restrictions-on-indian-eggplant-and-other-vegetables/). Trade during 2020 and shortly afterwards was not considered as the Covid-19 pandemic had disrupted trade flows (Table C.3).

      TABLE C.3. Direct annual import of aubergines (CN 070930) from countries with reported presence of Leucinodes orbonalis (Bangladesh, Brunei Darussalam, China, Indonesia, India, Japan, Cambodia, Lao People's Democratic Republic, Sri Lanka, Myanmar [Burma], Malaysia, Nepal, Philippines, Pakistan, Singapore, Thailand, Taiwan, Viet Nam) to EU member states from 2010 to 2022 (in t, calculated from EUROSTAT table DS-045409, downloaded on 13th June 2023, status 16th May 2023). The years 2020–2022 (grey shading) were excluded from the calculation of the mean annual import volume, as these were affected by the Covid-19 pandemic.
      MS 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Mean CV (%) Prop. (%)
      Tot.(EU) 243.426 179.251 175.625 140.893 98.81 126.386 178.601 163.744 155.773 173.81 123.685 130.912 121.698 163.6 23% 100%
      AT 1.506 3.032 1.943 1.644 3.272 4.267 2.605 2.405 1.212 1.340 0.757 0.851 0.557 2.3 42% 1.4%
      BE 28.300 30.900 44.600 35.500 6.789 0.018 0.927 1.383 0.049 0.447 0.040 0.374 0.000 14.9 119% 9.1%
      BG 0.0 0.0%
      CY 0.0 0.0%
      CZ 0.050 0.092 0.049 1.006 0.141 3.030 48.631 41.791 42.358 49.365 29.880 47.146 46.610 18.7 125% 11.4%
      DE 129.686 73.579 76.056 48.083 30.439 24.318 27.137 28.422 27.050 27.101 24.987 13.251 6.573 49.2 70% 30.1%
      DK 29.016 11.843 5.317 9.204 12.218 14.226 12.979 11.451 10.799 11.351 11.243 10.751 6.036 12.8 48% 7.8%
      EE 0.0 0.0%
      ES 0.000 0.000 0.040 0.000 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 300% 0.0%
      FI 1.180 1.044 0.000 0.339 0.315 0.000 0.319 1.187 0.351 0.000 0.000 0.000 0.000 0.5 102% 0.3%
      FR 13.166 17.178 10.070 12.065 4.703 37.032 51.835 47.357 44.785 57.912 43.585 43.298 37.161 29.6 68% 18.1%
      GR 0.0 0.0%
      HR 0.0 0.0%
      HU 0.0 0.0%
      IE 0.216 0.000 0.435 1.049 0.000 0.000 0.000 0.000 0.000 0.304 0.045 1.146 1.266 0.2 169% 0.1%
      IT 0.0 0.0%
      LT 0.0 0.0%
      LU 0.0 0.0%
      LV 0.0 0.0%
      MT 0.0 0.0%
      NL 7.687 40.470 35.920 30.713 39.832 37.296 29.973 26.787 22.309 15.722 8.090 9.314 19.611 28.7 38% 17.5%
      PL 0.150 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 282% 0.0%
      PT 0.015 0.000 0.000 0.155 0.000 0.000 0.015 0.000 0.000 0.000 0.000 0.000 0.000 0.0 261% 0.0%
      RO 0.0 0.0%
      SE 32.454 1.113 1.195 1.135 1.084 6.197 4.180 2.961 6.860 10.268 5.058 4.781 3.884 6.7 142% 4.1%
      SI 0.0 0.0%
      SK 0.0 0.0%
      Details are in the caption following the image
      Tonnes of eggplants imported into EU 27 in 2010–2022 from countries where Leucinodes orbonalis is known to occur. (EUROSTAT, EU trade (ds-059322) by CN, status 15th November 2023).

      Based on a normal distribution and the annual imports from 2010–2019, a mean import volume of 163.6 tonnes per year was calculated, with a standard derivation of 38.37 t/year and a coefficient of variation of 23%.

      Uncertainties

      The Panel assumed that trade over the next 5 years would be similar to trade in the recent past (2010–2019) (Table C.3 and Figure C.2). In reality, trade may increase or decrease quickly in response to markets and consumer preferences. This uncertainty was not accounted for but was considered small compared to uncertainty in other parameters of the pathway model.

      Details are in the caption following the image
      Distribution of weight of exotic/special eggplant varieties fitted to EKE estimates. (Left hand chart shows probability density function density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      Introduction model: Weight of an eggplant fruit

      There is variation in the size and weight of exotic and special eggplant cultivars imported into the EU. Examples are provided in Table C.4. Eurostat data do not distinguish between eggplant cultivars and therefore an EKE was performed to estimate the mean weight of an exotic eggplant fruit. Eggplants from Bangladesh, India, and Thailand are typically exported in boxes of 5 kg; exports from Pakistan are more variable and do not have a standardised box weight (FPJ, 2011).

      TABLE C.4. Example dimensions and weights of exotic eggplant varieties (Source: https://plantura.garden/uk/vegetables/aubergine/aubergine-varieties).
      Type Shape Colour Radius (cm) Length (cm) Approximate volume (cm3) Weight (g) Fruits per 5 kg box
      Thai eggplant Spherical Green variegated 5.7 249.5 78 65
      Japanese eggplant Elongated Very dark purple 1.9 20.3 231.7 142 35
      White eggplant Elongated White 2.4 18.4 327.8 163 30
      Chinese eggplant Elongated Light purple 2.5 39.4 797.9 312 16
      Philippine eggplant Elongated Purple and green 3.7 24.1 1009.9 397 12

      TABLE C.5. Estimated mean weight of a fresh exotic/special eggplant fruit (grams).

      Question What is the mean weight of a fresh exotic/special eggplant variety (in grams) imported into the EU from countries where Leucinodes orbonalis occurs?
      Results Estimated mean weight of exotic/special fresh eggplant fruit imported into EU (kg)
      Percentiles % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 100 150 200 250 400
      Fitted results (g) 100 113 152 197 253 343 401
      Fitted distribution BetaGeneral (1.6549, 4.6322, 93,530)

      TABLE C.6. Estimated mean weight of a fresh turkey berry fruit cluster (grams).

      Question What is the mean weight of a fresh turkey berry fruit cluster (in grams) imported into the EU from countries where Leucinodes orbonalis occurs?
      Results Estimated mean weight of fresh turkey berry cluster imported into EU (kg)
      Percentiles % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 18 27 35 42 55
      Fitted results (g) 18.0 20.3 27.3 34.5 42.4 51.4 55.0
      Fitted distribution BetaGeneral (1.5622, 1.9533, 16.8, 57.8)

      TABLE C.7. Uncertainties associated with low and high fruit weight scenarios.

      Uncertainty Low weight scenario High weight scenario
      Fruit size Mainly small eggplant and turkey berries Mainly large eggplant fruit and turkey berries
      Cluster size Small number of turkey berries per cluster (infructescence) Large number of turkey berries per cluster (infructescence)

      TABLE C.8. Imports of eggplant from non-EU countries to the Netherlands in 2022, by country (kg).
      Country Minimum P25 P50 Mean P75 Maximum Number Sum
      Egypt 140 150 160 200.0 230 300 3 600
      Ghana 20 32.5 50 127.7 177.5 404 7 894
      Jordan 104 557 803 1081.0 1549 2587 38 41,079
      Kenia 12 14 16 16.0 18 20 2 32
      Laos 1 19.75 28.75 32.1 32.1 33.9 4 92
      Malaysia 30 33.45 36.9 36.9 40.35 43.8 2 74
      Morocco 20 1006 2156 2626.0 3520 14,520 58 152,304
      Mexico 15 390 629.5 725.0 910 3510 84 60,900
      Uganda 7 54 699 682.4 975 1890 8 5459
      Spain 126 136.5 154 207.0 294 308 6 1449
      Surinam 10 90 150 190.2 252.2 1708 420 79,873
      Thailand 0.291 21.2 24.7 23.4 27.9 32.55 31 724
      Turkey 60 450 900 1364.0 1350 11,170 42 57,297
      South Africa 1.2 19.2 42 54.2 72 1250 499 27,045
      Total 0.291 37.2 96 355.0 275 14,520 1205 427,822

      TABLE C.9. Estimated mean number of eggplant fruits infested with L. orbonalis when entering the EU (per 10,000 fruit).

      Question: How many out of 10,000 transfer units (fresh eggplant fruit) will be on average infested with L.orbonalis when entering the EU from countries where the pest occurs?
      Results Infestation rate of eggplant fruit when entering the EU (per 10,000 fruit)
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates: 0.1 1.5 3 12 30
      Fitted values: (fruit infested per 10,000) 0.100 0.127 0.983 4.20 11.1 23.7 30.1
      Fitted distribution BetaGeneral (0.46872, 1.8525, 0.099, 35)

      TABLE C.10. Scenario calculations on the likely level of infestation with L. orbonalis in eggplant fruit.

      Step Detail Data (source) Scenario
      1 2 3 4 5 6 7 8 9
      x1 Import into EU of aubergines (CN code 070930) from countries with L. orbonalis 2010–2022 (kg) From Eurostat 160,000 160,000 160,000 160,000 160,000 160,000 160,000 160,000 160,000
      x2 Weight of a single consignment (kg) Three scenarios: 12, 24, 48 12 24 48 12 24 48 12 24 48
      x3 Number of consignments Calculated: x1/x2 13,333 6667 3333 13,333 6667 3333 13,333 6667 3333
      x4 Percentage of inspected consignments Constant: 100% 100 100 100 100 100 100 100 100 100
      x5 Number of consignments inspected Calculated: x3 * x4/100 13,333 6667 3333 13,333 6667 3333 13,333 6667 3333
      x6 Number of consignments found contaminated with L. orbonalis per year 2010–2022 Reported to Europhyt 11.5 11.5 11.5 11.5 11.5 11.5 11.5 11.5 11.5
      x7 Number of fruit/bunches inspected per consignment Dutch NVWA info and scenarios 60 60 60 30 30 30 10 10 10
      x8 Estimated proportion of infested fruit/bunches Calculated: -ln(1-x6/x5)/x7 1.44E-05 2.88E-05 5.76E-05 2.88E-05 5.75E-05 0.000115 8.63E-05 0.000173 0.000346
      x9 Infested fruit/bunches per 10,000 Calculated: x8 * 10,000 0.144 0.288 0.576 0.288 0.575 1.152 0.863 1.726 3.456

      TABLE C.11. Uncertainties associated with low and high infestation rate scenarios.

      Uncertainty Low infestation rate scenario High infestation rate scenario
      Infestation in field Low pest pressure High pest pressure
      Sorting High quality for export Inadequate sorting
      Control at border Good border control: large sample size, visual control Poor border control: small sample size, only identity check
      Efficiency of visual control Clearly visible entry holes Difficult detection of eggs, old entry holes

      The median weight of an exotic/special eggplant variety is 197 g; (90% CR is from 113 to 343 g).

      Uncertainties

      There are more varieties than those given as examples in Table C.4.

      There is more variation in the dimensions and weights of exotic varieties than shown in Table C.4.

      Eggplants from Bangladesh, India and Thailand are typically exported in boxes of 5 kg; exports from Pakistan are more variable and do not have a standardised box weight (re:Fresh Directory, 2018).

      Reasoning
      • Weights of exotic eggplant varieties range from 78 to 397 g (Table C.4).
      • Turkey berries have a diameter of 10–13 mm (Vorontsova & Knapp, 2023a).
      • 5–40+ turkey berries per infructescence (Vorontsova & Knapp, 2023a).
      Lower limit:
      • Predominantly small-fruited eggplants with light weight imported;
      • Predominantly small clusters of turkey berry with small berries imported.
      Upper limit:
      • Predominantly large varieties of eggplant imported;
      • Predominantly large clusters of turkey berry with large berries imported.
      Median:
      • Imported eggplant fruit represents a mix of large and small varieties.
      • Imported turkey berries represent a mix of small and large berry clusters.
      Inter-quartile range:
      • Relatively high uncertainties below and above the median for both eggplant and turkey berry.

      Introduction model: Infestation rate of eggplant fruit when entering the EU

      According to hearing expert Dr Ramasamy, in subtropical and tropical Asia where eggplant is grown, fruit infestation with L. orbonalis usually ranges from 25% to 30% at the lower end and can reach up to 100%; extremely well-managed fields can have as little as 10% infestation. The current practice in many Asian eggplant-producing countries is to apply large amounts of (cheap) pesticides during the entire growth season to keep L. orbonalis infestations rates as low as possible, resulting in increasing pesticide resistance (e.g. Shirale et al., 2017). Pheromone mass trapping of males is very common and is part of a (still poorly adapted) integrated pest management (IPM) package that further includes healthy seedling production, the removal and prompt destruction of infested shoots and fruits, and the withholding of chemical pesticides to promote natural enemies of L. orbonalis. These IPM measures can reduce the infestation to 20%–30% and lower the intensity of pesticide application from 90 sprays to 15–20 sprays per growth cycle (Dr S. Ramasamy, personal communication).

      No methods are currently known to physically, chemically or spectrometrically detect L. orbonalis infestations of eggplant fruit in a non-invasive way.

      Dossier findings are summarised below.
      • Eggplant fruit can be harvested 3–4 months after propagation by seed.
      • Under favourable conditions, flowering and fruit production is continuous, consequently eggplants are repeatedly harvested by hand.
      • Fruits grow rapidly under optimum conditions and plants are harvested frequently.
      • Cotton gloves are often worn during harvest to protect fruit against physical injury.
      • Fruits are handled carefully to avoid bruising and compression injury.
      • Fruits are graded then individually wrapped with paper or plastic film.
      • Careful handling of each individual fruit provides opportunity to detect infested fruit and to reject these from export.
      • Fruits are packed in bags, crates or baskets.
      • Fruits undergo rapid (pre-)cooling to 10°C immediately after harvest to inhibit discolouration, weight loss, drying of calyx, and decay.
      • Eggplant fruit are sensitive to chilling (temperatures below 10°C). Damage such as surface pitting, surface bronzing, and seed and tissue browning will result if stored for a few days below 10°C.
      • Storage: Fruit are stored at 10–12°C with 90–95% RH. (WG interpretation: insufficient to cause larval mortality).
      • Eggplant can be stored for 2–3 weeks but are generally held for less than 14 days as visual and sensor qualities deteriorate rapidly. Decay is likely to increase after storage > 2 weeks, especially after removal to typical retail conditions.
      • Eggplants must have a phytosanitary certificate before being brought into the EU. Eggplants intended for the EU therefore have to be inspected and found free from quarantine pests, specifically Spodoptera frugiperda and Thrips palmi. Eggplant fruit should also be practically free from other pests (EU 2019/2072, Annex VII).

      However, there have been interceptions of L.orbonalis from a number of Asian countries.

      Pest issues

      The Panel analysed Dutch NVWA data for 2022 on the size of incoming shipments of Solanum fruit with CN code 070930, comprising a.o. S. melongena and S. torvum, here referred to as eggplant sensu lato (s.l.). Fourteen extra-EU countries imported eggplant s.l. to EU. Shipment sizes differed greatly between countries of origin, indicating that some countries send product in bulk, whereas other send small shipment of, probably, a specialty product, e.g. Solanum torvum berries. Shipments from Asian countries were on average small compared to shipments from most other countries of origin (Table C.8, Figure C.3).

      Details are in the caption following the image
      Average estimated shipment size of eggplants shipments of countries (based on imports by NL in 2022).

      The figure below (based on the same data) ranks the countries by the size of average shipments, showing that shipment sizes of aubergine s.l. from Asia are among the smallest of countries of origin of eggplant.

      Usually, the number of interceptions per 100 kg of product is quite low. The Panel estimated a level of infestation in product coming from Asia from 0.13 to 24 per 10,000 fruits (approximately 2000 kg for eggplant or 450 kg for turkey berry). However, occasionally, from a country of origin, this number can be higher, as is the case for Sri Lanka and Pakistan in Figure C.4. In such cases, the number of infestations will diminish over time as stricter phytosanitary measurements are implemented (dashed lines in Figure C.4), to reach the required level of pest freedom, i.e. the absence of the pest from the exported commodity.

      Details are in the caption following the image
      Number of Leucinodes orbonalis interceptions per 100 kg of eggplant (Solanum melongena) imported from Asian countries. Dashed lines indicate changes of interception proportions over time in imports from Pakistan (green) and Sri Lanka (dark blue).

      There are no data on the number of inspections conducted to produce the number of interceptions reported in Europhyt and TRACES. This severely weakens how Figures C.1 and C.4 (Interceptions plotted against time) can be interpreted.

      Post-harvest issues
      • After harvest, eggplant fruits are handled very carefully to avoid physical damage.
      • Eggplants are carefully cleaned in the field.
      • Fruit are gently rubbed to remove debris and soil particles.
      • Harvested fruit should be protected from environmental conditions and kept in shaded areas.
      • Plastic crates are recommended when transporting the eggplants from the field to minimise damage.
      • In cleaning, sorting, grading pack houses fruit can be washed and dried.
      • Eggplants are stored at 10°C or more (chill sensitive when stored below 10°C).
      • Eggplant is stored separately from tomatoes, bananas and melons – their high ethylene content causes eggplants to spoil quickly.

      Sources: Cargo Handbook.com; Majlis Ilmu, 2018.

      Overview of model input estimates: infestation rate on arrival into EU

      Elicited mean annual rates of infestation per 10,000 eggplant fruits are given in the table below with the probability distribution under the table. EKE estimates are values proposed by the expert working group as consensus estimates. Model inputs are derived from the distribution fitted to the EKE estimates.

      Figure C.5

      Details are in the caption following the image
      Distribution of infestation rate of eggplant fruit fitted to EKE estimates (left hand chart shows probability density function density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median rate of eggplant fruit infestation is 4.20 per 10,000 (= 42 per 100,000); (90% CR is from 12.7 per million to 2370 per million).

      Reasoning
      • The Panel used Europhyt and TRACES interception data to infer a likely level of infestation of the imported eggplant fruit.
      • The weights of consignments inspected are recorded as net weight (Europhyt) or commodity weight (TRACES) (kg).
      • Leucinodes orbonalis was most frequently intercepted on consignments of between 20.1 and 30 kg (Figure C.6).
      • Assuming that the consignment sizes on which interceptions occur are representative of all consignments, the number of consignments each year can be estimated.
      Details are in the caption following the image
      Consignment sizes vary from less than 10 kg to over 130 kg. (NL data).

      Analysis

      The Dutch NVWA import inspection data for 2022 include 37 records of imports of eggplant s.l. from Asian countries with presence of L. orbonalis, 31 shipments from Thailand, 2 from Malaysia and 4 from Laos.

      The mean weight of these 37 shipments was 24 kg, and the mean number of boxes (‘colli’) per shipment was also 24. Thus, the average contents of a box would be approximately 1 kg of fruit. This suggests that these must be small fruit.

      According to the inspection guidelines, inspectors should inspect 60 fruit per shipment as a minimum. With a shipment of 24 kg, if one fruit is 200 g, there would be 120 fruits in a shipment. If the commodity is S. torvum, assuming 35 g per bunch, there would be 24/0.035 = 685 bunches in a shipment.

      The average import of eggplant (CN code 709300) to EU is 160 t/year. With shipment sizes in the order of 24 kg, in means that about 7000 consignments with aubergine are imported each year to EU from countries with L. orbonalis. Each consignment should be inspected, and the prescribed sample size for inspection is 60 fruits, but Panel might take into account that in the actual practice, this number is not reached. On average, out of these approximately 7000 consignments per year, about 11.5 per year are found infested. The corresponding proportion of infested fruit can be worked out from the formula
      exp pN = 1 P ,

      where P is the proportion of consignments found infested (median estimate 11.5/6667 = 0.17%. It could be a larger or smaller proportion depending upon whether consignments tend to be smaller or larger than the mean estimate of 24 kg. N is the sample size per consignment which is prescribed to be 60 fruit or bunches but might be smaller in practice (not larger).

      The solution is:
      p = ln 1 P / N .

      Different scenarios are worked out in the table below. The conclusion is reached that the number of infested fruit/fruit clusters per 10,000 is in the order of 1 (range: 0.14–3.5). That is a similar order of magnitude (perhaps a bit less) compared to what was found for Elasmopalpus lignosellus, in asparagus from Peru (EFSA PLH Panel, 2023) but given the much smaller trade of eggplant compared to asparagus, the number of infested fruit/units imported to the EU will be much smaller for eggplant than for asparagus.

      Lower limit:
      • Estimation from the interception data with optimistic assumptions.
      Upper limit:
      • Estimation from interception data with low sample size;
      • High infestation level at country of origin, with no indications of proper measures.
      Median:
      • General low infestation level, as shown in the interception data;
      • Limitation of the detection at border control.
      Inter-quartile range:
      • High uncertainties below and above the median.

      APPENDIX D: Establishment

      This appendix summarises background information that was used to estimate those parameters relevant to assess the likelihood of pest establishment following entry. The establishment component of the overall model for introduction aimed to calculate the number of adult L. orbonalis emerging from imported infested fruit that go on to mate and transfer to hosts and start a founder population.

      A detailed description of how environmental parameters were used to inform the assessment of establishment is provided by Rossi et al. (2023). Following a thorough literature review, 573 references were used to provide data to assess establishment. Data from 334 papers were extracted regarding pest distribution for mapping. The assessment focussed on mapping the distribution of L. orbonalis in south Asia and projecting potential distribution in Europe and the around the Mediterranean. Output maps for the four methods used to inform assessment of the area of potential establishment are available on Zenodo, i.e. (i) Köppen–Geiger zones in Europe and in the area where the pest occurs, (ii) mapping of accumulated degree days and number of generations for Europe and Asia, (iii) maps produced using CLIMEX and (iv) the Species Distribution Modelling results.

      Degree days and number of generations

      Degree days were calculated by accumulating daily mean temperature above a base temperature of 15°C (Base T). A linear regression model using data from Dhaliwal and Aggarwal (2021) was used to estimate that 438.6 accumulated degree days above BaseT were necessary for one generation of L. orbonalis to be completed.

      Across large parts of Europe, the maximum degree days possible each year is 1500. Degree days in excess of 4000 per year occur widely across southern Asia (Figure D.1). See also Rossi et al. (2023).

      Details are in the caption following the image
      Mean degree days accumulation for Leucinodes orbonalis in Europe and the Mediterranean Basin (1) and South-East Asia and India (2) based on a base temperature of 15°C. Degree days were calculated for each year in the period 1993–2022 and then averaged.

      Figure D.2 indicates large parts of India accumulate sufficient degree days such that there can be 10 or more generations per year. In the EU degree days accumulations would theoretically sustain three to four life cycles of L. orbonalis per year.

      Details are in the caption following the image
      Estimated mean number of generations for Leucinodes orbonalis in Europe and the Mediterranean Basin (1) and South-East Asia and India (2). Number of generations were calculated considering a minimum number of accumulated degree days to complete one generation of 438.6 degree days. Number of generations were calculated for each year in the period 1993–2022 and then averaged.

      CLIMEX projection

      Results from the literature review (2.2.1) provided 64 papers that were used to inform the selection of CLIMEX parameters (Rossi et al., 2023). Parameters used in CLIMEX are shown in Table D.1.

      TABLE D.1. CLIMEX parameters for Leucinodes orbonalis. Parameters highlighted in grey were not adjusted and values already included in the CLIMEX ‘semi-arid’ template were used.
      Class Parameter Description No moisture index Moisture index Source
      Moisture index SM0 Lower soil moisture threshold 0.01 Iteration
      SM1 Lower optimal soil moisture 0.1 Iteration
      SM2 Upper optimal soil moisture 3 Iteration
      SM3 Upper soil moisture threshold 4 Iteration
      Temperature index DV0 Lower temperature threshold 15°C 15°C 315
      DV1 Lower optimal temperature 26°C 26°C 24, 315, 1861
      DV2 Upper optimal temperature 38°C 38°C 24, 315, 1833
      DV3 Upper temperature threshold 39°C 39°C 1683
      Cold stress TTCSa Cold stress temperature threshold 5.13°C 5.13°C a
      THCSa Cold stress accumulation rate −0.00237 week−1 −0.00237 week−1 a
      DTCS Cold stress will begin to accumulate when this threshold number of degree days above DVCS is not achieved 2°C 2°C Iteration
      DHCS Rate at which Cold Stress accumulates once the threshold number of degree days above DVCS (DTCS) is not achieved −0.001 week−1 −0.001 week−1
      TTCSA Average weekly temperature below which Cold Stress accumulates 0°C 0°C
      THCSA Rate at which Cold Stress accumulates once temperatures drop below the threshold value of TTCS1 0°C 0°C
      Heat stress TTHS Heat stress temperature threshold 39°C 39°C 1683
      THHS Heat stress accumulation rate 0.001 week−1 0.001 week−1
      DTHS Heat stress will begin to accumulate when this threshold number of degree days above DV3 is exceeded 0°C 0°C
      DHHS This is the rate at which Heat Stress accumulates once the threshold number of degree days above DV3 (DTHS) is exceeded 0°C 0°C
      Dry stress SMDSa Dry stress soil moisture value 0.09797 0.09797 a
      HDSa Dry stress soil moisture rate −0.00782 week−1 −0.00782 week−1 a
      Wet stress SMWS Soil moisture wet stress threshold 4 4 Iteration
      HDS Wet stress accumulation rate 0.001 week−1 0.001 week−1
      DD above DV0 DV0 Lower temperature threshold 15°C 15°C 315
      DV3 Upper temperature threshold 39°C 39°C 1683
      DD above DVCS DVCSa Degree days based cold stress 5.13°C 5.13°C a
      DD above DVHS DVHS Degree days based heat stress 39°C 39°C 1683
      Annual Heat sum PDDb Minimum degree day sum needed to complete a generation 438.6 438.6 b
      • a Values retrieved using the built-in CLIMEX genetic algorithm (version 4.1.0.0, Kriticos et al., 2015).
      • b PDD was calculated with a linear regression model using life table studies data (see section: Degree days and number of generations maps).

      TABLE D.4. Estimated proportion of eggplant fruit infested with Leucinodes orbonalis discarded as household waste, i.e. by the end consumer.

      Question: A transfer unit (aubergine fruits/grapes) is infested with L. orbonalis. Some transfer units will be discarded at the consumer level (private waste). What is the proportion of transfer units (aubergine fruits/grapes) that will be disposed at the consumer level?
      Results Proportion of infested transfer units discarded as household (private) waste
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates: 20% 35% 50% 60% 75%
      Fitted values: 20.0% 23.5% 35.8% 48.6% 61.0% 72.1% 74.9%
      Fitted distribution BetaGeneral (1.3262, 1.2334, 0.185, 0.76)

      Information from CLIMEX informed estimates of the lag period (Appendix E) and impacts (Appendix F).

      The majority of locations where L. orbonalis has been recorded from south Asia are in regions where EI exceeds 30 and a substantial number of records are from regions where EI is above 75 (Figure D.3.2). EI does not exceed 75 anywhere in the Euro-Mediterranean region (Figure D.3.1) Light orange regions in Figure D.3.1. indicate potential areas for establishment EI ≥ 15 but < 30). However, such regions would be sub-optimal for L. orbonalis. Darker orange regions (EI ≥ 30, < 75) indicate more favourable regions for establishment.

      Details are in the caption following the image
      CLIMEX Ecoclimatic Index for Leucinodes orbonalis. 1. Euro-Mediterranean region: 2 South Asia. EI range shown in five colour categories.

      Area of establishment for Leucinodes orbonalis

      Using the CLIMEX modelling system, Rossi et al. (2023) show that the majority of locations where L. orbonalis has been recorded from south Asia are in regions where EI exceeds 30 and a substantial number of records are from regions where EI is above 75. EI does not exceed 75 anywhere in the Euro-Mediterranean region. There is uncertainty about the minimum threshold for establishment so two thresholds were used; EI 30 and EI 15. The table below indicate the area where EI ≥ 30 and EI ≥ 15 in NUTS 2 regions of the EU.

      A total area of approximately 180,770 km2 of the EU has an EI ≥ 30 (Table D.2). Of this area, over 50% occurs in four NUTS 2 areas Andalucia (ES), Alentejo (PT) Sicily (IT) and Valencia (ES). Recognising the uncertainty of using an EI of 30 as a threshold, Table D.2 also shows the area of the EU where abiotic environmental factors considered within CLIMEX indicate where establishment may be possible if an EI threshold of 15 is used. In this case the total area suitable for establishment increases by a factor of approximately 2.5 to approximately 450,200 km2.

      TABLE D.2. Approximate area (thousands of km2) of NUTS2 regions of the EU where EI > 30 (where impacts thought most likely to occur) and where EI > 15 where establishment may be possible.
      A B C D E F G H
      NUTS2 code NUTS 2 name Area

      % of sum

      EI ≥ 30

      Cumulative % Area

      % of sum

      EI ≥ 15

      Cumulative %
      MS (name) EI ≥30 EI ≥ 15
      ES61 ES (Andalucía) 47.5 26.28 26.28 67.5 14.99 14.99
      PT18 PT (Alentejo) 17.5 9.68 35.96 35 7.77 22.77
      ITG1 IT (Sicilia) 17.5 9.68 45.64 25 5.55 28.32
      ES52 ES (Comunitat Valenciana) 12.5 6.91 52.55 15 3.33 31.65
      EL43 EL (Kriti) 12.5 6.91 59.47 12.5 2.78 34.43
      CY00 CY (Kýpros) 10 5.53 65.00 10 2.22 36.65
      ES43 ES (Extremadura) 7.5 4.15 69.15 37.5 8.33 44.98
      ITF6 IT (Calabria) 7.5 4.15 73.30 15 3.33 48.31
      ITF4 IT (Puglia) 6.4 3.54 76.84 17.5 3.89 52.20
      EL42 EL (Notio Aigaio) 5.16 2.85 79.69 5.17 1.15 53.35
      ITG2 IT (Sardegna) 5 2.77 82.46 22.5 5.00 58.35
      EL41 EL (Voreio Aigaio) 5 2.77 85.22 5 1.11 59.46
      EL65 EL (Peloponnisos) 4.58 2.53 87.76 7.03 1.56 61.02
      ES53 ES (Illes Balears) 3.69 2.04 89.80 5.18 1.15 62.17
      ES62 ES (Región de Murcia) 3.31 1.83 91.63 10 2.22 64.39
      EL30 EL (Attiki) 3.02 1.67 93.30 3.88 0.86 65.25
      EL64 EL (Sterea Elláda) 1.96 1.08 94.39 5 1.11 66.36
      EL63 EL (Dytiki Elláda) 1.88 1.04 95.43 7.5 1.67 68.03
      ITI4 IT (Lazio) 1.43 0.79 96.22 12.5 2.78 70.80
      PT17 PT (Área Metropolitana de Lisboa) 1.36 0.75 96.97 5 1.11 71.91
      ES51 ES (Cataluña) 1.25 0.69 97.66 17.5 3.89 75.80
      ITF3 IT (Campania) 1.2 0.66 98.32 3.9 0.87 76.67
      ITF5 IT (Basilicata) 1.07 0.59 98.92 2.36 0.52 77.19
      EL62 EL (Ionia Nisia) 0.7 0.39 99.30 5 1.11 78.30
      PT20 PT (Região Autónoma dos Açores) 0.44 0.24 99.55 5 1.11 79.41
      MT00 MT (Malta) 0.31 0.17 99.72 0.31 0.07 79.48
      PT15 PT (Algarve) 0.25 0.14 99.86 7.5 1.67 81.15
      FRM0 FR (Corse) 0.22 0.12 99.98 5 1.11 82.26
      PT30 PT (Região Autónoma da Madeira) 0.04 0.02 100.00 0.76 0.17 82.43
      PT16 PT (Centro (PT)) 20 4.44 86.87
      FRI1 FR (Aquitaine) 12.5 2.78 89.65
      ES24 ES (Aragón) 10 2.22 91.87
      FRL0 FR (Provence-Alpes-Côte d'Azur) 7.5 1.67 93.53
      FRJ1 FR (Languedoc-Roussillon) 5 1.11 94.64
      ITI1 IT (Toscana) 5 1.11 95.76
      PT11 PT (Norte) 3.73 0.83 96.58
      ES42 ES (Castilla-La Mancha) 3.06 0.68 97.26
      EL54 EL (Ipeiros) 2.67 0.59 97.86
      HR03 HR (Jadranska Hrvatska) 2.25 0.50 98.36
      ES11 ES (Galicia) 2.17 0.48 98.84
      EL52 EL (Kentriki Makedonia) 1.45 0.32 99.16
      ITC3 IT (Liguria) 0.98 0.22 99.38
      FRJ2 FR (Midi-Pyrénées) 0.87 0.19 99.57
      ITF1 IT (Abruzzo) 0.66 0.15 99.72
      ES12 ES (Principado de Asturias) 0.52 0.12 99.83
      ITF2 IT (Molise) 0.45 0.10 99.93
      EL61 EL (Thessalia) 0.21 0.05 99.98
      ITI2 IT (Umbria) 0.09 0.02 100.00
      Sum 180.77 100 450.2 100.00

      Introduction model: Allocation to NUTS regions

      Based on the results of CLIMEX shown in Rossi et al. (2023), the pathway model for entry detailed in Appendix C was followed up with a step for the within-EU distribution of infested host fruit, the allocation to areas with Eco-climatic Index (EI) ≥ 30, and ≥ 15 to reflect the uncertainty regarding the EI threshold for establishment. Following arrival in the EU, it is assumed that the produce is distributed across EU NUTS regions in proportion to the human population. Based on Eurostat reports of population in each NUTS2 area, 23.12% of the EU population are in NUTS2 regions in which EI ≥ 15 and 14.09% are in regions where EI ≥ 30.

      Introduction model: Proportion of fruit discarded

      An important factor to consider is that for larvae to complete their development, the contaminated eggplant fruit they arrive in should be discarded before it is processed, e.g. cooked and consumed.

      At this point in the pathway for introduction, the density of infested eggplant fruit is thought to be very low and only eggplants discarded together by importers, wholesalers or retailers will be of sufficient quantity for there to be a reasonable chance that a male and female may emerge in sufficiently close proximity for them to find each other and mate to potentially initiated a founder population. Eggplant is a sensitive fruit, and discards could be due to several reasons, such as damage during handling and transport, physical quality problems, market conditions and pest finds.

      Commercial waste of infested eggplant

      TABLE D.3. Estimated proportion of eggplant fruit infested with Leucinodes orbonalis discarded as commercial waste, i.e. between import and consumer.

      Question: A transfer unit (aubergine fruits/grapes) is infested with L. orbonalis. Some transfer units will be discarded between import and consumer (commercial waste). What is the proportion of transfer units (aubergine fruits/grapes) that will be disposed between import and consumer?
      Results Proportion of infested transfer units discarded as commercial waste
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates: 5% 9% 12% 15% 30%
      Fitted values: 4.99% 6.28% 9.10% 11.8% 15.2% 21.1% 26.0%
      Fitted distribution BetaGeneral (3.5888, 33.705, 0.032, 1)

      image

      FIGURE D.4. Distribution of proportion of infested eggplants disposed of as waste between import and consumer fitted to EKE estimates (left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median proportion of transfer units (eggplant fruits/grapes) infested with L. orbonalis and discarded after import but before reaching the consumer was estimated to be 11.8% (90% CR from 6.28% to 21.1%).

      Household waste of infested eggplant

      image

      FIGURE D.5. Distribution of proportion of infested eggplants disposed of at the consumer level as household waste fitted to EKE estimates (left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median proportion of transfer units (eggplant fruits/grapes) infested with L. orbonalis and discarded after import but before reaching the consumer was estimated to be 48.6% (90% CR from 23.5% to 72.1%).

      Reasoning
      • Infested eggplants are very likely detected by the end consumer.

      TABLE D.5. Uncertainties associated with low and high discard proportion scenarios.

      Uncertainty Low discard proportion scenario High discard proportion scenario
      Detection of infestation Importers, wholesalers and retailers are less likely to detect infestations from inspecting the fruit externally End consumers process the fruit by cutting them open, resulting in a high detection and disposal rate
      Quality standards Low standards, businesses want to sell as many fruit as possible High standards, businesses want to sell as high a quality product as possible, resulting in the sorting and disposal of unsatisfactory fruit
      Processing of fruit Small infested fruit may be cooked uncut Large fruit are cut open, and infestation will be detected

      Lower limit:
      • Eggplant is an expensive fruit, and importers, wholesalers and retailers want to sell as many fruit as possible.
      • Households will want to use as much of the expensive fruits as possible.
      • Infestation of turkey berries may be overlooked when cooking the berries uncut.
      Upper limit:
      • High proportion of disposal by importers, wholesalers and retailers to remove unsatisfactory fruit due to transport and handling damage or problems with physical quality;
      • The majority of infested fruit are detected by the end consumer and discarded.
      Median:
      • Infestation of fruit is difficult to detect between import and end consumer, i.e. before the fruit are processed for cooking.
      • While cutting the fruit for cooking, end consumers will detect infestation.
      • Cutting the fruit for cooking may kill larvae feeding inside.
      Inter-quartile range:
      • High uncertainties below and above the median

      Introduction model: Probability of larvae surviving to develop to adulthood then escape from discarded waste

      • L. orbonalis larvae exit the fruit prior to pupation (Navasero & Calilung, 1990; Saxena, 1965).
      • Eggs are laid on leaves and fruit, but larvae hatch within 3–4 days after egg deposition (Navasero & Calilung, 1990; Saxena, 1965), which is assumed to take place before a consignment of eggplant arrives in the EU; by the time of arrival, larvae hatching from eggs laid on fruit will have hatched and bored into the fruit.
      • Larval development (incl. egg phase) takes about 11–29 days, pupal development (incl. prepupal phase) 10–11 days at 21 C (Table B.1, Figure B.1; Navasero & Calilung, 1990; Saxena, 1965).
      • From the fourth larval instar on, when disturbed or the fruit is rotting, the larva can prematurely enter the prepupal phase by constructing the pupation cocoon, in which it passes the remaining larval instars without feeding until it pupates (‘cocooning behaviour’; Saxena, 1965).

      TABLE D.6. Estimated probability of an adult Leucinodes orbonalis emerging and escaping from eggplant fruit discarded with commercial waste (excluding household waste).

      Question: What is the probability that a larva in the discard waste will result in an adult emerging from the commercial waste?
      Results Likelihood that an adult will emerge from discarded eggplant commercial waste
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates: 0.1% 0.6% 1% 1.5% 2%
      Fitted values: 0.100% 0.195% 0.582% 1.03% 1.48% 1.90% 2.01%
      Fitted distribution BetaGeneral (1.1703, 1.2288, 0.00068, 0.0205)

      image

      FIGURE D.6. Distribution of probability of an adult emerging from infested eggplant fruit discarded with commercial waste fitted to EKE estimates (left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median likelihood that a larva discarded in commercial waste (i.e. prior to reaching the end consumer) will continue its development to emerge as an adult that escapes from waste was estimated to be 1.03% (90% CR from 0.195% to 1.90%).

      The median likelihood that a larva discarded in household waste (i.e. after the supply chain when the fruit have reached the end consumer) will continue its development to emerge as an adult that escapes from waste was estimated to be 5.13% (90% CR from 0.98% to 12.22%).

      TABLE D.7. Estimated probability of an adult Leucinodes orbonalis emerging and escaping from eggplant fruit discarded with household waste (excluding commercial waste).

      Question: What is probability that a larva in the discard waste will result in an adult emerging from the household waste?
      Results Likelihood that an adult will emerge from discarded eggplant household waste
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates: 0% 3% 5% 8% 15%
      Fitted values: 0.361% 0.98% 2.92% 5.13% 7.89% 12.2% 15.05%
      Fitted distribution BetaGeneral (1.6704, 4.5179, 0, 0.21)

      image

      FIGURE D.7. Distribution of probability of an adult emerging from infested eggplant fruit discarded with household waste fitted to EKE estimates (left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      Reasoning
      • Eggplant fruit discarded between import and consumer (commercial waste) likely end up in landfills or organic waste processing plants, where survival of the larva is unlikely.
      • A considerable amount of eggplant fruit discarded after arriving at the end consumer (household waste) is discarded on composts, where survival of the larva to adulthood is favourable.
      • From the fourth instar on, larvae can complete their life cycle in the prematurely constructed pupation cocoon without feeding (Saxena, 1965).

      TABLE D.8. Uncertainties associated with low and high emergence probability scenarios.

      Uncertainty Low emergence probability scenario High emergence probability scenario
      Waste handling Low proportion of composting, primarily waste processing and/or landfills High proportion of composting
      Developmental stage of larva After disturbance, early instar larvae (1st to 3rd instar) may leave the fruit and starve in lack of alternative fruit to feed on Late instar larvae (from 4th instar on) can pass the remaining larval stages in the pupal cocoon without feeding

      Lower limit:
      • Low proportion of compost disposal as household waste in urban areas, where composting options are limited;
      • Majority of discarded fruit go to landfills, where they are quickly covered by additional waste, trapping or killing the developing larva.
      Upper limit:
      • High probability of disposal by end consumer on compost;
      • Household waste primarily containing late instar larvae that can survive to adulthood without food.
      Median:
      • Mix of waste disposal (landfills, waste processing, composting).
      Inter-quartile range:
      • High uncertainties below and above the median

      Introduction model: Proportion of adults mating

      An important limiting factor in establishing a founder population is the likelihood of a male and a female emerging in temporal and spatial proximity to locate each other and mate. Recall that border interceptions of insects are a poor predictor of successful establishment (Kenis et al., 2007; Caley et al., 2015).

      Panel adopted the probabilities of adult mating that were elicited for Elasmopalpus lignosellus (EFSA PLH Panel, 2023), but the actual probabilities are presumably even lower. One reason for this is the shorter adult lifespan of L. orbonalis, with males living 4–5 days and quickly dying after copulation, whereas females live 7 to maximally 8 days (Navasero & Calilung, 1990; Saxena, 1965). The time window for females and males to meet and mate is therefore narrower compared to E. lignosellus.

      TABLE D.9. Estimated probability that an adult female of Leucinodes orbonalis will mate, as elicited for Elasmopalpus lignosellus (EFSA PLH Panel, 2023).

      Question: What is the probability that an escaped adult female will find a mate either before flying off to find a host or at a host?
      Results Proportion of females successfully mating
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 0.00% 0.05% 0.08% 0.12% 0.20%
      Fitted values 0.0071% 0.018% 0.049% 0.081% 0.12% 0.17% 0.20%
      Fitted distribution BetaGeneral (1.8097, 3.2291, 0, 0.0024)

      image

      FIGURE D.8. Distribution of probability of a female successfully mating fitted to EKE estimates (left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median likelihood that an escaped adult female will successfully mate was estimated to be 0.081% (i.e. 8 in 10,000) (90% CR from 0.018% to 0.17%, approximately 2 in 10,000 to 17 in 10,000).

      Reasoning
      • Short range of attraction of the female sex pheromone, with a maximum range of 10 m (Prasad et al., 2005);
      • Shorter adult life span of L. orbonalis (males 4–5 days, females 7–8 days; Saxena, 1965; Navasero & Calilung, 1990) compared to E. lignosellus (9–13 days, up to 20 days under lab conditions; EFSA PLH Panel, 2023), creating a shorter time window for mate finding and mating.

      TABLE D.10. Uncertainties associated with low and high mating success scenarios.

      Uncertainty Low mating success scenario High mating success scenario
      Adult life span L. orbonalis adults more short-lived than those of E. lignosellus
      Mate attraction Short range of sex pheromone attraction Female sex pheromone is strongly attracting males
      Synchronisation of adult emergence Few or no adults of opposite sex emerge in temporal and spatial proximity Adult emergence from pupae is well synchronised

      Lower limit:
      • Only 3–4 days after adult emergence available for mate finding and mating;
      • Female sex pheromone is strongest in young female moths, loses attractiveness/potency with time.
      Upper limit:
      • Female sex pheromone is very attractive to males (an artificial version is used for mass trapping of L. orbonalis males in pest management).
      Median:
      • Effective range and potency of the sex pheromones of L. orbonalis and E. lignosellus are comparable.
      Inter-quartile range:
      • High uncertainties below and above the median.

      Introduction model: Likelihood of founder population initiation

      Leucinodes orbonalis is an oligophagous pest and feeds in the shoots and fruits of different species of Solanum in the plant family of nightshades (Solanaceae). The Mediterranean region is home to 48 species of Solanum (incl. lycopersicum; Valdés, 2012) (see table below with the presence recorded at the MS level from Valdes 2012 database online accessed on 28/11/2023), all serving as potential host plants for L. orbonalis larvae. Several European Solanum species are widely distributed weeds, such as S. nigrum, ‘a common weed of cultivated land, open spaces in forests and roadsides [that is] found in disturbed areas between 0 and 2200 (3500) m elevation;’ (Särkinen et al., 2023). As such, L. orbonalis has a good chance of finding a suitable host within areas suitable for establishment.

      AT BE and LU BG HR CY CZ DK EE FI FR DE GR NL ES HU IE IT LV LT PL PT RO MT SK SI SE
      S. abutiloides x
      S. alatum x x x x x x x x x x x x x x x x x x x
      S. americanum x x x x
      S. aviculare x
      S. bonariense x x x
      S. capsicoides x
      S. carolinense x x x
      S. chenopodioides x x x x x
      S. chrysotrichum x
      S. citrullifolium x x x
      S. dulcamara x x x x x x x x x x x x x x x x x x x x x x x x
      S. eleagnifolium x x x x x x x
      S. fastigiatum x
      S. giganteum x
      S. gracile x
      S. heterodoxum x x x
      S. hispidum x
      S. laciniatum x x
      S. laxum x x x
      S. lidii x
      S. linnaeanum x x x x x x x x x
      S. lycopersicum x x x x x x x x x x x x x x x x x x x x x x
      S. marginatum x x
      S. mauritanium x x
      S. melanocerasum x x x
      S. melongena x x x x x x x x x
      S. microcarpum x
      S. nava x
      S. nigrum x x x x x x x x x x x x x x x x x x x x x x x x x x
      S. patens x
      S. physalifolium x x x x x x
      S. pseudocapsicum x x x x x x
      S. pygmaeum x
      S. retroflexum x
      S. robustum x
      S. rostratum x x x x x x x x x x x x x
      S. sarachoides x x x x x x x x
      S. scabrum x x
      S. seaforthianum x
      S. sisymbrifolium x x x x x x x x
      S. torvum x
      S. triflorum x x x x x x x
      S. trisectum x
      S. tuberosum x x x x x x x x x x x x x x x x x x x x
      S. vespertilio x
      S. viarum x
      S. villosum x x x x x x x x x x x x x x x x x x x x x x x x
      S. virgatum x

      The female is carrying up to 300 eggs that are laid singly on the leaves and fruit of host plants (Navasero & Calilung, 1990). Specialised predators and parasitoids are expected to be absent in a potential founder population. Due to the internal feeding of the larvae and the closing of the entry hole into the plant with a plug of excreta, L. orbonalis is not very susceptible to generalist predators.

      Once the next generation emerges, males and females are likely to be present at the same location at a similar time, although the adult stage is short-lived, with 4–5 days for males and about 7 days for females (Navasero & Calilung, 1990; Saxena, 1965). A potent (though apparently short-ranged) female sex pheromone increases the likelihood of males and females meeting to mate (Prasad et al., 2005). Host plants in suitable stages are also likely to be found. In unfavourable conditions during winter, L. orbonalis is able to hibernate in the pupal stage (Lal, 1975). These factors promote the probability of a mated female founding a population that would persist. This probability was therefore estimated to not be small, with an elicited lower bound of 2.41% of mated females founding a persistent population and the upper bound 31.6% (Table D.11).

      TABLE D.11. Estimated probability that a founder population of Leucinodes orbonalis will be initiated following successful mating.
      Question: What is the proportion of mated females of Leucinodes orbonalis that will successfully establish a founder population?
      Results Likelihood that a founder population will be initiated following successful mating
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 1% 9% 17% 25% 33%
      Fitted values 1.00% 2.41% 9.00% 17.0% 25.0% 31.6% 33.0%
      Fitted distribution BetaGeneral (1.0456, 1.0456, 0.0062, 0.3338)
      image

      FIGURE D.9. Distribution of probability of a founder population being initiated following successful mating fitted to EKE estimates (left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median likelihood that a founder population will be initiated following successful mating was estimated to be 17.0% (90% CR from 2.41% to 31.6%).

      Reasoning
      • Gravid females of L. orbonalis and E. lignosellus lay similar amounts of eggs singly or in small clusters on or near the host plant.
      • Larvae of both species are internal borers and are thus fairly well protected from parasitoids and predators.

      TABLE D.12. Uncertainties associated with low and high founder population establishment scenarios.

      Uncertainty Low establishment scenario High establishment scenario
      Host plant availability Lower host plant availability compared to E. lignosellus Several Solanum weed species present in regions suitable for L. orbonalis establishment
      Natural enemies Higher biological control through natural enemies compared to E. lignosellus; eggs and especially young larvae more exposed to parasitation and predation as compared to E. lignosellus, whose larvae build a shelter in the upper soil level Similar natural enemy pressure as E. lignosellus; within a few minutes after hatching, young larvae bore into the host plant and plug the entry hole with excreta

      Lower limit:
      • L. orbonalis only has Solanum spp. available as potential host plants, compared to the polyphagous E. lignosellus that feeds on plants of several plant families.
      Upper limit:
      • L. orbonalis is oligophagous on several Solanum species, several of which are widely distributed weed species on cultivated and disturbed land.
      Median:
      • As a newly established species, biological control of L. orbonalis through natural enemies (parasitoids, predators) is likely comparable to that of E. lignosellus in the same situation.
      • The females of both L. orbonalis and E. lignosellus produce similar amounts of eggs, and both species lay their eggs singly or in small clusters, thus spreading the risk of parasitisation and mortality.
      Inter-quartile range:
      • High uncertainties below and above the median

      APPENDIX E: Spread

      To inform the assessment of spread, the Panel first estimated the duration of the lag phase before estimating the linear rate of range expansion during the phase in which spread is at its fastest.

      No information is available on the natural spread capacities or flight capabilities of L. orbonalis. Therefore, to aid our estimation of spread of L. orbonalis after its establishment in the EU, we analysed spread data of a related African species, L. laisalis (Guenée, 1854), which was first reported from the southern Iberian Peninsula in 1958 and has since then been spreading in this region (Figure E.1). Leucinodes laisalis adults (Figure E.1) are similar in size to L. orbonalis, with a forewing length of 7.0–11.5 mm and the females being somewhat larger than the males (Mally et al., 2015). From this, we assume that both Leucinodes species have similar spread rates based on their presumably comparable flight capability.

      Details are in the caption following the image
      Map of reported occurrences of Leucinodes laisalis (Guenée, 1854) on the southern Iberian Peninsula between 1958 (first record) and 2023; data compiled from GBIF, Lepiforum, iNaturalist and Mally et al. (2015). Photo: Female adult L. laisalis, Spain, Andalusia, 5 km N of Terifa, 2014-09-22, det. & phot. © Stephen McAvoy, map © Richard Mally.

      Occurrence data of L. laisalis on the Iberian Peninsula were compiled from different sources (GBIF.org, 2023; Huertas-Dionisio, 2000; iNaturalist, 2023; Lepiforum, 2023; Mally et al., 2015) and mapped in R. The data show that L. laisalis has been repeatedly reported in southern Spain and Portugal since 1958 (Figure E.1). Leucinodes laisalis appears to be established in the southernmost Iberian Peninsula and is slowly spreading at an average rate of approximately 1.6 km/year based on a linear regression of the available data (Figure E.2). Taking only the records into account that are consecutively furthest away from the first record, and assuming linear spread, an annual spread rate of 5.51–11.51 km/year is assessed for L. laisalis (Table E.1).

      Details are in the caption following the image
      Scatterplot of the spread of Leucinodes laisalis in Europe between 1958 and 2023.
      TABLE E.1. Years of record of Leucinodes laisalis in Spain and Portugal, geographic distance from the first record point in 1958 and annual spread rate calculated for the three furthest consecutive distances.
      Year of record Distance from first record (km) Annual spread based on furthest consecutive record (km/a)
      1958 0
      1975 195.63 195.63/(1975–1958) = 11.51
      1982 169.18
      1987 172.00
      2005 151.42
      2014 172.79
      2017 176.97
      2017 344.82 344.82/(2017–1958) = 5.84
      2019 178.02
      2019 169.15
      2020 175.61
      2020 173.17
      2021 145.74
      2021 347.06 347.06/(2021–1958) = 5.51
      2021 347.06
      2021 177.83
      2022 165.79
      2022 165.79
      2022 147.15
      2022 147.15
      2023 195.63
      2023 169.18

      E.1. Estimated duration of the lag phase

      Table E.2 shows EKE estimates for the duration of the lag phase based on the evidence summarised as bullet points below.

      TABLE E.2. Estimated duration of the lag phase of Leucinodes orbonalis (in years).
      Question What is the duration of the lag phase, this means the time between first establishment and the continuous expansion of the infested area (linear phase)?
      Results How long is the average duration of the lag phase? (years)
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 4 17 30 65 100
      Fitted values (years) 4.00 4.85 14.49 34.5 62.3 92.2 100.00
      Fitted distribution BetaGeneral (0.666, 1.1588, 3.92, 103)
      image

      FIGURE E.3. Distribution of the estimated mean duration of the lag phase of L. orbonalis (years) fitted to EKE estimates (left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      image

      FIGURE E.4. Distribution of the estimated mean spread rate of L. orbonalis (km/year) fitted to EKE estimates (Left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median estimate for the duration of the lag phase was 4.5 years (90% CR from 4.85 to 92.2 years).

      Reasoning
      • The occurrence and spread of the related species L. laisalis in southern Spain and Portugal were reviewed regarding the specific development at the beginning of the spread;
      • The spread behaviour of L. orbonalis is frequently described as stationary;
      • The lag phase is largely driven by the rate of population growth;
      • Fecundity: about 300 eggs per mated female;
      • Population development will depend on the availability and condition (e.g. state of nutrition) of host plants;
      • Population development will be faster on Solanum crops;
      • Host plant range includes wild plants such as Solanum torvum, S. nigrum, S. aethiopicum;
      • Mate attraction based on female sex pheromones with a short maximum range (10 m).

      TABLE E.3. Uncertainties associated with short and long lag phase scenarios.

      Uncertainty Short lag phase scenario Long lag phase scenario
      Life cycles 4–6 generations per year 1 generation per year
      Host Continuous host availability; Solanum crops Host plants die during winter; Solanum weeds
      Population dynamics Good conditions for development, incl. Crop monoculture, bad sanitary habits, staggered planting Poor conditions of development, esp. less preferred hosts available, wild environment, fewer fruit/biomass due to fewer nutrients

      Lower limit:
      • Favourable climatic conditions with up to six generations per year and continuous host plant availability;
      • Outbreaks in crop monocultures with poor phytosanitary measures, hygienic conditions and staggered planting.
      Upper limit:
      • Unfavourable climatic conditions with only one generation per year and host plants dying off during winter;
      • Outbreaks in natural environment, e.g. on Solanum weeds (Solanum nigrum) with poor nutritional conditions;
      • Pest is described as stationary.
      Median:
      • Good developmental conditions (agricultural areas with suitable crops) are more likely in the climatic suitable areas of the EU.
      • Over the years, Leucinodes orbonalis will adapt to poor conditions, which allows for better development.
      Inter-quartile range:
      • The lack of specific evidence indicates maximum uncertainties below/above the median.

      E.2. Estimated linear rate of range expansion (spread)

      Table E.4 shows the EKE estimates for the linear rate of spread based on the evidence summarised as bullet points below.

      TABLE E.4. Estimated rate of linear spread of Leucinodes orbonalis (km/year).
      Question What is the annual median rate of spread when the established population is spreading at a constant rate?
      Results
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 0.5 1.5 2.0 4.0 10.0
      Fitted values (km/year) 0.50 0.652 1.31 2.28 3.80 7.02 10.0
      Fitted distribution BetaGeneral (1.2452, 34.666, 0.445, 70)

      The median annual rate of spread for Leucinodes orbonalis was estimated to be 2.28 km per year (90% CR from 0.652 to 7.02 km per year) during the period of constant rate of spread.

      Reasoning
      • The spread rate is the outcome of the contribution of natural dispersal together with local human-assisted spread;
      • Spread due to post-harvest movement, such as the trade in commodities, was not included in the estimation;
      • There is no quantitative data on the flight capability of L. orbonalis;
      • Adults are weak fliers (Chang et al., 2014; Dr Ramasamy, personal communication);
      • The adults can move from field to field and hence can spread over an area over the course of several generations; maximum estimated field to field spread is 10 km in a single year (comprising several generations; see next bullet point) (Dr Ramasamy, personal communication);
      • To complete its life cycle (i.e. one generation), L. orbonalis requires 26–36 days in the tropics (Navasero & Calilung, 1990) and 26–39 days in the subtropics (Lall & Ahmad, 1965), which allows for 10–14 generations per year; under Mediterranean European conditions, the number of expected generations per year is significantly lower;
      • Jethva and Vyas (2009) extrapolated from lab studies of L. orbonalis reared on tomato that the population would multiply by a factor of 2.358 per week.

      TABLE E.5. Uncertainties associated with slow and faster spread rate scenarios.

      Uncertainty Slower spread rate scenario Faster spread rate scenario
      Flight capability extrapolated from the related L. laisalis Similar flight capacity of the two Leucinodes spp. assumed
      Range expansion extrapolation for L. laisalis Average spread rate (linear model): 1.6 km/year Maximum spread rate: 5.5–11.5 km/year; calculated average rate might include the lag phase, with the actual average rate being higher
      Larval host plants L. orbonalis spreads as fast as L. laisalis due to similar hosts L. orbonalis spreads faster than L. laisalis due to broader host range and higher host plant density, with S. melongena as primary host
      Plant production cycle Eggplants not grown year-round, therefore lower number of L. orbonalis generations per year in EU as compared to India (where a spread of 10 km/year is estimated) Host plants grown/available throughout the year
      Plant production conditions Low availability of host plant (eggplant) due to managed cultivation, especially in protected conditions Unmanaged, unprotected wild hosts available
      Climatic suitability The primarily tropical L. orbonalis is less well adapted to Mediterranean climate than L. laisalis, thus slower life cycle, reproduction and consequently spread L. orbonalis also occurs in regions with occasional winter frost (Lal, 1975), indicating a relatively broad climatic tolerance

      Lower limit:
      • Spread rate for L. laisalis is average not lower limit.
      • L. laisalis is more adapted to the Mediterranean climate, whereas L. orbonalis is more adapted to the tropics and subtropics.
      • Main crop production is on small, scattered plots.
      Upper limit:
      • L. orbonalis has larger host range with higher plant density in EU.
      • Lower host plant availability due to managed cultivation reduces the spread, esp. in protected conditions, but hosts in natural environment exist.
      • Spread is expected to be reduced compared to Asia.
      Median:
      • Lower suitability for EU climate than L. laisalis.
      Inter-quartile range:
      • High uncertainty below the median;
      • Medium uncertainty above the median.

      Singh et al. (2007) made catches of L. orbonalis in a landscape with only a single plot of brinjal (Solanum melongena) at different distances from the brinjal plot. Traps located at 50 and 100 m from brinjal field attracted fewer male L. orbonalis moths than those at 0, 150 and 350 m, indicating the feasibility of trapping male Leucinodes orbonalis moths even in a non-brinjal area. Trap direction did not significantly influence trap catch. Nearly 60% of male L. orbonalis moths were observed in traps placed against direction of the wind.

      APPENDIX F: Impact

      The Panel focused on the impact on eggplant, as it is evident from the large body of literature (467 studies; Table C.1) that this crop is heavily impacted by L. orbonalis in Asia, and vice versa that L. orbonalis prefers eggplant as host above all other solanaceous crops (Ardez et al., 2008). The Panel based its assessment on several Asian studies investigating impact of L. orbonalis on eggplant, which were compiled in meta-analyses (see F.3–F.5 below).

      F.1. Estimated impact with no pest management practice in place

      This is an artificial scenario for a commercial farmer but can be seen in experimental trials where pest management practices are excluded.

      Table F.1 shows EKE estimates for yield loss without specific control measurements against L. orbonalis, informed by a meta-analysis of losses (below) reported from control plots in pesticide trials and on the evidence summarised as bullet points below.

      TABLE F.1. Estimated mean reduction in yield (yield loss) caused by L. orbonalis in the absence of pest control.
      Question What is the likely mean reduction of annual yield of eggplant, grown outdoors in a southern European country, with EI higher than 30 without specific pest control measures?
      Results Annual eggplant yield loss without control measures
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 0.5% 2.0% 4.5% 8.0% 15.0%
      Fitted values (% losses) 0.504% 0.670% 2.01% 4.47% 8.03% 13.0% 15.1%
      Fitted distribution BetaGeneral (0.78643, 1.8133, 0.0048, 0.166)
      image

      FIGURE F.1. Distribution of mean reduction of annual eggplant yield without control measures fitted to EKE estimates. (Left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median annual yield loss in eggplant crops was estimated to be 4.47% (90% CR from 0.67% to 13.0%) in a scenario where no pest management is in place.

      Reasoning
      • No specific measures taken for L. orbonalis but insecticides might be (but not necessarily) used against other Lepidoptera.

      TABLE F.2. Uncertainties associated with low and high yield loss scenarios.

      Uncertainty Low yield loss scenario High yield loss scenario
      Agricultural practices Good agricultural practices, e.g. crop rotation, sanitation, clean planting material, regular inspections by grower Amateur gardens. Small scale. No pesticides. Staggered planting, some greenhouses, but not sanitated, so near-Indian opportunities for field-to-field movement of the insect and perpetuation. Lack of expertise and attention to pests. Improper waste disposal
      Climatic conditions More moderate summer temperatures (Mediterranean) High summer temperatures (above Mediterranean)
      Cultivars Varieties less susceptible; integrated nutrient management Susceptible varieties, high fertiliser input

      Lower limit:
      • Low-level persistent population that cannot be entirely eradicated.
      Upper limit:
      • Poor agricultural practices that allow for re-infestation from discarded infested material
      Median:
      • Overall, well under control in an ideal scenario.
      Inter-quartile range:
      • High uncertainties below and above the median.

      F.2. Impact with pest management practices in place

      Table F.3 shows EKE estimates for yield loss with specific pest control measurements against L. orbonalis, informed by a meta-analysis of losses (Appendix F, below) reported from control plots in pesticide trials and on the evidence summarised as bullet points below.

      TABLE F.3. Estimated mean reduction in yield (yield loss) caused by L. orbonalis with pest control in place.
      Question What is the likely mean reduction of annual yield of eggplant, grown outdoors in a southern European country, with EI higher than 30 with specific pest control measures?
      Results Annual eggplant yield loss with specific control measures
      Percentiles: % 1% 5% 25% 50% 75% 95% 99%
      EKE estimates 0.1% 0.3% 0.5% 1% 3%
      Fitted values (% yield loss) 0.102% 0.130% 0.281% 0.536% 0.964% 1.94% 2.91%
      Fitted distribution BetaGeneral (1.0297, 163.33, 0,00095, 1)
      image

      FIGURE F.2. Distribution of mean reduction of annual eggplant yield with specific control measures fitted to EKE estimates. (Left hand chart shows probability density function to describe the remaining uncertainties of the parameter; right hand chart shows cumulative distribution function (CDF) of the likelihood of the parameter).

      The median annual yield loss in eggplant crops was estimated to be 0.536% (90% CR from 0.13% to 2.91%) in a scenario with specific pest management measurements is in place.

      Reasoning
      • Dense pheromone trapping;
      • Public awareness campaigns.

      TABLE F.4. Uncertainties associated with low and high yield loss scenarios.

      Uncertainty Low yield loss scenario High yield loss scenario
      Agricultural practices Good agricultural practices, e.g. crop rotation, sanitation, clean planting material, regular inspections by grower, pheromone trapping and mating disruption, spraying when insects are caught in traps Amateur gardens; small scale; no pesticides. Staggered planting, some greenhouses, but not sanitated, so near-Indian opportunities for field-to-field movement of the insect and perpetuation; lack of expertise and attention to pests; improper waste disposal; resistance to pesticides lowers efficacy of control
      Climatic conditions More moderate summer temperatures (Mediterranean) High summer temperatures (above Mediterranean)
      Cultivars Resistant cultivars; integrated nutrient management Susceptible varieties, high fertiliser input

      Lower limit:
      • Low-level persistent population that cannot be entirely eradicated;
      • Effective mass trapping of males with dense network of pheromone traps;
      • Wide public awareness of effective integrated pest management.
      Upper limit:
      • Poor agricultural practices that allow for re-infestation from discarded infested material.
      • Poor public awareness of effective control measurements.
      Median:
      • Overall, well under control in an ideal scenario.
      Inter-quartile range:
      • High uncertainties below and above the median.

      F.3. Meta-analysis of Leucinodes orbonalis infestation reported in Solanum melongena in the literature (with control)

      The full texts of the publications on Leucinodes orbonalis, identified during the systematic search, were screened for information on L. orbonalis infestation. This meta-analysis is based on the results of 17 independent experiments. Publications reported L. orbonalis shoot and fruit infestation in Solanum melongena. Data on infestation were extracted and analysed. The datafile consisted of 453 records (observations) (Table F.5 and Figures F.3 and F.4).

      TABLE F.5. Number of records used to analyse the impact of Leucinodes orbonalis infestation on eggplant (Solanum melongena).
      Number of records (N = 453)
      Fruit 248
      Shoot 178
      Not specified 27
      Number of records (N = 453)
      India 429
      Pakistan 24
      Details are in the caption following the image
      Percentage of Leucinodes orbonalis infestation on Solanum melongena when pesticides applied.
      Details are in the caption following the image
      Leucinodes orbonalis infestation in India and Pakistan.

      In the data set, we have data from India and Pakistan, so we are going to plot the data separately as well. Pakistan is a not suitable area for L. orbonalis and that could be a reason that would explain the low density in this country (Figure F.4).

      The review on chemical pest (L. orbonalis) control in Asia on eggplants was completed and visualised in Table F.6.

      TABLE F.6. A–Z of insecticides (active compounds) tested in studies on the control of Leucinodes orbonalis on eggplant (Solanum melongena) in the places of origin.
      Active ingredient
      Acephate Carbosulfan Endosulfan Malathion Spinetoram
      Acetamiprid Cartap Endrin Monocrotophos Spinosad
      Avermectin Chlorantraniliprole Etofenprox Neem Sulphur
      Azadirachtin Chlorpyrifos Fenpropathrin Novaluron Thiacloprid
      Bacillus thuringiensis Cypermethrin Fenvalerate Parathion Thiamethoxam
      Beauveria bassiana Deltamethrin Fipronil Parathion-methyl Thiometon
      Bifenthrin Demeton-S-Methyl Flubendiamide Permethrin Triazophos
      Bt galleriae Diazinon Heptachlor Phosalone Trichlorfon
      Bt kustaki Dichlorvos Imidacloprid Phosphamidon Urea
      Carbaryl Dimethoate Indoxacarb Profenofos Zeta-cypermethrin