Synopsis of fruit-piercing moths of the genus Eudocima (Lepidoptera, Erebidae) from Colombia

Abstract In order to provide information about the diversity and distribution of Eudocima species in Colombia, 261 specimens deposited in entomological collections were examined and identified. We found seven of the eight species of Eudocima recorded in the Neotropics: E. anguina, E. colubra, E. collusoria, E. memorans and E. serpentifera, all being recorded for the first time from the country. We provide a list of the species, comments on the biology and distribution data, illustrations of the adults, and keys for species identification.


Introduction
The fruit-piercing moth genus Eudocima Billberg, 1820 (Erebidae, Calpinae) encompasses approximately 50 species distributed throughout tropical and subtropical re-gions of the world Branham 2008, Zilli et al. 2017), with eight species occurring in the Neotropics (Zilli and Hogenes 2002). They are generally large-sized and with variably colored patterns, mainly cryptic on the forewings and with bright yellow-orange hindwings, and at least in the Neotropics the species always bear dark spots or bands on the hindwings. Historically, neotropical species of Eudocima have been placed in several genera according to differences in their habitus, e.g., Elygea Billberg, 1820, Othreis Hübner, [1823, Trissophaes Hübner, [1823] and Ophideres Boisduval, 1832, all now subsumed under Eudocima. Like other genera of the subfamily Calpinae, they possess sclerotized and apically sharpened proboscis with tearing hooks, with which they pierce fruits to feed on their juices. Accordingly, unlike other groups of agriculturally important Lepidoptera, it is the adults that damage crops, which in this case takes place due to rotting agents such as fungi and bacteria that penetrate the holes that they leave onto the fruit skin.
In Asian countries and islands of the Pacific, fruit-piercing Eudocima are frequently reported as damaging crops, while in the Americas they are only sporadically mentioned as pests , and information about this group is generally scarce.
In Colombia, E. apta (Walker, [1858]) and E. procus (Cramer, 1777), two species widely distributed in the Neotropics, were recently reported as occasional citrus pests ), but the diversity and distribution of this genus in the country are unknown. Vouchers in biological collections can provide important information about the spatial and temporal distribution of species. In Colombia, many universities and research centers are maintaining biological collections where specimens from monitoring programs and ecological sampling are regularly being deposited. It was therefore expected that these colorful large moths would be well represented in such collections.
The aim of this work is to report information on Eudocima from specimen data preserved in collections and produce a checklist and an identification key to species occurring in Colombia. Additionally, we provide information about the environmental variables determining species distribution. This information will facilitate a baseline for planning ecological studies and taking phytosanitary actions in case of the detection of pest species in fruit orchards. Furthermore, the checklist could assist with the resolution of environmental factors determining presence of these moths in cultivations and enable the development of models to forecast their occurrence in agroecosystems.

Material and methods
The checklist presented here collates literature records for Colombia based on Walker ([1858]) and Montes et al. (2018) with specimen data drawn from the following entomological collections:   Distribution. Widespread in the New World, from southern United States and the Caribbean to Brazil, the South Atlantic Islands and north of Chile (Angulo and Jana-Sáenz 1983;Brou 1994, Zilli and Hogenes 2002, Brou and Núñez 2013. Powell and Brown (1990) recorded E. apta up to an elevation of 3900 m. In Colombia, it has been recorded in several localities within the eastern cordillera and eastern slope of the central cordillera in a wide elevational range.

CEUA
Remarks. Traditionally it has been incorrectly identified as Eudocima materna (Linnaeus, 1767) (e.g., Costa Lima 1950: fig. 158). However, E. materna is distributed in the Old World. Zilli and Hogenes (2002) provided a rationale for considering E. apta as a valid species and not a synonym of E. materna. Comments. Marked sexual dimorphism in the coloration of forewings. The species has a sinuous band on the posterior wings in the form of an "m", similar to those of E. memorans and E. collusoria. Janzen and Hallwachs (2009) record larvae of E. colubra as feeding on Disciphania calocarpa Standl. (Menispermaceae) in Costa Rica.

Eudocima collusoria (Cramer, 1777) (Phalaena (Noctua))
Distribution. Costa Rica, Peru (Schaus 1911) and Colombia. Remarks. Three specimens were examined from the departments of Antioquia and Chocó. It is recorded from Colombia for the first time.  Fig. 3D Comments. The hindwings show an "m"-shaped band with a distinctly sinuous inner margin. Unlike E. serpentifera, this band reaches the wing margin. Forewing with oblique pale bands. Hostplant and life cycle unknown.
Remarks. This species is recorded for the first time from Colombia.  Hogenes 2002, Zaspel andBranham 2008). In the present work, specimens from several localities, mainly of the eastern and central cordilleras, were found. Widespread in Colombia. Gallego (1946) reported this species (as Othereis procus, genus misspelled) to be frequently found in buildings of Medellín during the first half of the twentieth century. This species seems to be adapted to urban ecosystems as it is frequently attracted to city lights or even to boats near the Brazilian coast (Alves et al. 2019). Fig. 4E chicayá;3.6186, -76.9133;28 Nov. 1975-20 Nov. 1976MUSENUV 14846-14848. Comments. This species has a sinuous "m"-shaped band on the hindwings. Unlike other species with similar pattern on the hindwings such as E. memorans, E. collusoria and E. colubra, in E. serpentifera the "m"-shaped band does not reach the wing margin. Janzen and Hallwachs (2009) reported D. calocarpa (Menispermaceae) as its hostplant. Adults have been found feeding on C. papaya (Caricaceae) and Citrus in Mexico Hernández-Ruiz et al. 2017). In Mexico, adults are active from April to November and are commonly collected with light traps (Chamé-Vásquez and Jiménez 2009).

Eudocima serpentifera (Walker, [1858]) (Ophideres)
Distribution. Widely distributed in Tropical America. Walker ([1858]) describes this species from the Dominican Republic and Brazil. Additionally, there are occasional records of this species from the southern United States (Brou 2006). In Mexico, it occurs in an elevational range between 150 and 3000 m (Chamé-Vásquez and Jiménez 2009).
Remarks. This species is recorded for the first time from Colombia.

Spatial and temporal distribution
The collections examined essentially consist of holdings from the Andean region. It is no surprise then that 94% of records are from the Andes, mainly the eastern mountain chain, with 65%, and 35% solely from Bogotá city. The Caribbean and Pacific regions have only three records each, and most of the Amazon region and Orinoquia are not represented in the sample; Eudocima moths are known only from three locations in the Amazonian foothills in the departments of Putumayo and Meta.
Based on collection data from the city of Bogotá, the most common location represented in our sample, it is evident that seasonality of these moths is mainly determined by precipitation. Captures appear to be low in December and January, which are the months of lowest rainfall, and sharply increase during March, when the rainy season begins. Both the annual distribution of precipitation and that of moths show a bimodal pattern (Fig. 5). The relationship between moths and precipitation has frequently been reported (e.g., Bhumannavar and Viraktamath 2012), since with the onset of rainfall the sprouting of host plants increases, and oviposition of hundreds of eggs per female is triggered (Cochereau 1977, Magar et al. 2015.
The known geographical distribution of species of Eudocima is considerably expanded with our data. For instance, E. anguina was only known from Costa Rica, making the present record the first of this species in South America; E. collusoria was only known from Suriname and French Guiana; the record of E. colubra was predictable as this was known previously from Costa Rica and Peru; and E. memorans, described from the western coast of the Americas, was also found in the eastern mountain chain of Colombia and the Caribbean coast.
The wide distribution of Eudocima species is related to both their strong flight capacity (Bhumannavar and Viraktamath 2012) and close relationship with plants of the family Menispermaceae (Fay 1996). The larvae of Eudocima apta feed on Cissampelos pareira and those of E. procus have been recorded from Odontocarya tamoides. Both plants are widely distributed latitudinally and at elevations from 0 to 2800 m in the Andean, Pacific and Caribbean regions (Parr et al. 2014), overlapping with the distribution of these moths. On the other hand, the known host plant of E. serpentifera and E. colubra, namely Disciphania calocarpa, is mainly found in Central America and has only been recorded from Colombia in the humid montane forest of Dagua, Valle del Cauca, on the Pacific coast (Parr et al. 2014).

Perspectives
Research on Eudocima moths is intrinsically twofold and may develop along both conservationist and agricultural lines. The larvae of these moths in the Neotropics feed exclusively on wild lianas of the family Menispermaceae (Janzen and Hallwachs 2009). It is expected therefore that breeding populations of these moths are restricted to natural or semi-natural areas with sufficient extent of forest patches, which exposes them to high vulnerability due to the ongoing deforestation. Some species, such as E. anguina, E. collusoria, E. colubra and E. memorans, may even be facing a higher risk due to their trophic relationships with just one of few host plants. In fact, these species are already rare in collections.
On the other hand, Eudocima moths were recently recorded for the first time as occasional fruit pests in Latin America; E. apta and E. serpentifera on papaya in Mexico ) and E. apta and E. procus on citrus in Colombia . At least in Colombia, damage by these moths was previously unknown by farmers, so many questions now arise about their origin and frequency.
Records of occasional outbreaks of fruit-piercing moths affecting orchards such as those in Colombia and Mexico had already been reported by Cochereau (1977) for New Caledonia in Oceania. Cochereau (1977) observed and monitored changes in Eudocima phalonia (Linnaeus, 1763) populations for three years since 1968-1970 and recorded in 1969 that its population increased rapidly with the onset of rains, after a period of drought of several months, and caused damage of more than 90% in citrus production, while the normal rate was around 4% (Cochereau 1973(Cochereau , 1977. The drought event was prompted by the El Niño-Southern Oscillation (ENSO) episode of 1968-1969, which reduced rainfall on the island of New Caledonia (Benoit and Delcroix 2000).
Unusually dry periods such as those recorded by Cochereau (1977) are also known to occur during ENSO events in the Andean region, especially in the eastern mountain range and the Caribbean region of Colombia (Montealegre 2007), where they have the potential to boost populations of fruit-piercing moths. The outbreaks of these moths on citrus orchards recorded from several municipalities in Colombia  were most likely triggered by the 2014-2016 ENSO event. Van Bael et al. (2004) also reported an outbreak of E. apta in Panama on June 1998, which was apparently influenced by the ENSO episode of 1997-1998. In addition, outbreaks of E. serpentifera have been reported in Honduras during the rainy seasons in 2012 to 2014 and 2016 (Van Dort 2019). Outbreaks of other Lepidoptera taxa in the rainy season following ENSO events have also been recorded in Panama (Van Bael et al. 2004, Srygley et al. 2010. During such unusually dry periods several factors may act together and affect the natural control of moths, increasing their populations. In fact, the emergence of parasitoids is known to decrease with increasing temperature and drought (Romo and Tylianakis 2013), and also the rate of parasitoidism was shown to decrease with greater variability in rainfall between years (Stireman et al. 2005). This is likely an outcome of the uncoupling between cycles of hosts and parasitoids, which favors moth outbreaks. Another factor is the unusual sprouting of some plant species after an ENSO event.
With the onset of rainfall, the young plant tissue is also of better quality for herbivores, containing a greater amount of leaf nitrogen and lower concentration of secondary defensive compounds such as tannins and phenols (Shure et al. 1998). This allows the development of a greater number of larvae; therefore, the longer the dry season the more luxuriant the vegetation will be, to the advantage of moth populations (Srygley et al. 2010(Srygley et al. , 2014. Preliminary evidence therefore suggests a relationship between rainfall following ENSO-related drought and demographic increase of moth populations. That being the case, outbreaks of moth pests such as fruit-piercing Eudocima in orchards are expected to become commoner in future reflecting the increased frequency of ENSO events associated with climate change (Timmermann et al. 1999).
Although moth collections do not necessarily match exactly the distribution and abundance of species in the field, in the absence of strongly biasing factors (e.g., a 'maniac' collector selectively searching for particular species with exaggerate sampling effort) there is nonetheless an association between the commonness of a species in the field and the number of relevant vouchers deposited in collections. Accordingly, when several specimens of a species with the same locality and date are found, we expect such collection record to somewhat mark a natural population increase. Notably, when our records, which mostly originate from ecological sampling programs, are plotted along a timeline the increase of records matches the end of an ENSO event in most cases (Fig. 6). Remarkably, the high intensity ENSO event between 2014-2016 seems to have markedly increased the number of individuals in collections and outbreaks such as those reported in citrus ). Collections data provide invaluable information but there are some issues that cannot exclusively be addressed with these. Standardized long-term monitoring and sampling at night at selected sites with light traps, will be necessary to assess population dynamics over more ENSO cycles and to test the association between moth demography and climatic oscillations in the Neotropics. Surveys of host plants of fruit-piercing moths in natural areas will also shed light on several aspects of their biology, such as their life cycle and natural enemies. The importance of this information to preserve Eudocima diversity, especially regarding species exclusive of natural habitats, and reducing the damage caused to fruit orchards by pest species is evident.