From basic to applied knowledge: larval development and sex morphology to control Leucoptera coffeella (Lepidoptera: Lyonetiidae)

The coffee leaf miner Leucoptera coffeella (Guérin-Mèneville & Perrottet), is a key pest in coffee producing countries. During their development, the larvae feed on the palisade parenchyma of the leaves forming mines. As a result, the photosynthetic area of the plant decreases, affecting coffee production. Despite the severity of the damage caused by coffee leaf miner (CLM), morphological aspects of the larval development and the adult genitalia remain unknown. It is important to identify more susceptible targets to an efficient and narrow control by natural or synthetic approaches, relies on determining the larval instars. Equally important is the sexing of the adult, in experiments aiming efficient ways to control CLM, as the study of pheromone-based control methods. This work presents the first morphological description of the four larval instars and the adult genitalia of L. coffeella . In each larval instar, we measured the Mean ± SD (mm) of the cephalic capsules (1 st 0,14±0,03; 2 nd 0,25±0,04; 3 rd 0,32±0,03; 4 th 0,42±0,03) and observed the following morphological details: primary setae, prolegs, crochets and ecdysial line of the cephalic capsule. In the adults, we observed the sexual structures present in both genitalia: male - bulbus ejaculatorius, valva, anellus, gnathos and aedeagus and female - ovipositor, sclerite and corpus bursae. The dissection of the adult specimens confirmed that the external morphology corresponds to the correct sex attribution in CLM adults. These results may support innovative and improved control strategies for CLM Integrated Pest Management (IPM). for the identification of the four larval instars and description of male and female genitalia of L. coffeella


INTRODUCTION
Annually, more than 160 million bags of coffee are consumed on all continents. Coffee is produced in the countries of the neotropical region in America, where the coffee leaf miner Leucoptera coffeella (Guérin-Méneville and Perrottet 1842) occurs (Mey 1994, Pereira et al. 2007 . Even though Brazil is the world's largest coffee producer, the coffee leaf miner (CLM) is a key pest causing damages every year (Medina Filho et al. 1977, Vieira et al. 2006, Consórcio de Pesquisa do Café 2020. The CLM infestation impacts severely the coffee yield and bean quality (Gallo et al. 1988) in all Brazilian producing regions (Parra et al. 1981, Pantoja-Gomez et al. 2019 (Fig. 1). The CLM is a monophagous pest (Reis and Souza 1986) which feeds exclusively on coffee leaves, seriously compromising the health of the plant. L. coffeella has been reported as responsible for losses that can reach up to 87% drop in productivity. Depending on the season, the defoliation of the coffee plant can reach up to 75% Souza 1996, Neves 2006).
The life cycle of L. coffeella is holometabolic, including the stages of egg, larva, pupa, and adult in its winged version (Box 1923) (Fig. 2). After being described by Guérin-Mèneville & Perrottet in 1842, the immature forms (larva and pupa) were only described in 1923 by Harold Box. The larval phase penetrates the palisade parenchyma of the leaves and feeds there forming mines, which gives rise to the common name of this pest. The lesions caused by the mines cause necrosis, thus reducing the photosynthetic area, which can cause severe defoliation in more sensitive coffee genotypes, such as Coffea arabica (Medina Filho et al. 1977, Ramiro et al. 2004 . The symptoms of this plague attack consist of mines that progress from light green to brown, as the larva develops and moves to new feeding sites (Fig. 3).  The knowledge of life cycle phases is critical to allow the development of pest control methodologies, e.g., oviposition assays, infestation challenge to test pesticide products or resistant cultivars. Studies aimed at chemical control, biological control, transcriptome, and plant resistance to Lepidoptera species pests are reported the use of specific larval instars. Wheatley and Crowe (1964) conducted a control study of mining larvae, Leucoptera meyricki, in the field with the application of insecticides sprayed on mined leaves in order to kill larvae of the 2 nd , 3 rd and 4 th instar, decreasing the development of the infestation. Fragoso et al. (2002) conducted a study, in the laboratory, with the use of organophosphate insecticides in larvae of the third instar of L. coffeella. Biological control agents of mining larvae have been studied by Draganova and Tomov (1998), in which the fungus species Beauveria bassiana, is a pathogen of last instar larvae of the species Leucoptera malifoliella. Camargo et al. (2015) conducted a transcriptome study with the larval stages of the tomato leaf miner Tuta absoluta. Perthuis et al. (2005) tested the resistance of transformed C. canephora 126 using fourth instar larvae of CLM and Meriño-Cabrera et al. (2017) conducted a study with plant defense as an alternative to control CLM, using larvae of the fourth instar.
Adult sexing is used in experiments aimed at the mortality rate of females and males, laboratory breeding, sexual behavior, and pheromone employment. Informations about the male and female mortality rate of L. coffeella was observed by Notley (1948). Katiyar and Ferrer (1968) conducted a study with CLM adult breeding techniques, presenting a cage model and suggesting feeding to keep specimens in laboratory conditions. The sexual behavior of CLM adults has been described in detail by Michereff et al. (2007). Malo et al. (2009) conducted a study on the chemical composition of pheromones released by females of L. coffeella. Although the sexual dimorphism in adults of L. coffeella is mentioned in the literature (Parra 1985), there is no information available about the male and female genitalia of this species or the external aspects to differentiate them.
Despite the economic importance and geographical scope of the CLM, further studies on the larval morphological characteristics and sexual dimorphism of L. coffeella are lacking. In this work, we present a detailed study on the morphology of larvae for the identification of the four larval instars and description of male and female genitalia of L. coffeella.

MATERIAL AND METHODS
Insect Collection and Breeding: L. coffeella collections were carried out at Embrapa Cerrados (Federal District,Brazil,15º36'15.1" S,47º42'44.4" W) and Embrapa Genetic Resources & Biotechnology (Federal District,Brazil,15º43'52.8 "S,47º54'10.2" W) between May and August 2020. Leaves with mines and pupae were collected directly from the coffee plant or leaves fallen on the ground around the plant. The leaves with larvae and pupae were taken to the laboratory where they were placed in breeding cages until adult emergence (Fig. 4). A sugary solution of water with industrialized sugar was offered ad libidum for adult feeding. Larvae: To examine the larvae and determine the number of instars, the infested leaf mines were opened using tweezers and scalpels and observed with a Leica EZ-4 © stereomicroscope. Cephalic capsule width and body length measurements were made using a Leica M205C © stereomicroscope and processed with the Leica LAS © V3.8 to morphometry and image acquisition software. The width of the cephalic capsule and the body length of 50 larvae were measured in each of the four larval instars. Larval morphological characteristics, associated with cephalic capsule measurements, follow Nielsen and Common (1991). To identify morphological characteristics to distinguish the larval stages, we used (1) presence of primary body setae in the first instar, (2) presence or absence of prolegs in the third instar (3) presence or absence of crochets in the fourth instar (4) presence or absence of ecdysial line in the frontal suture of the fourth instar. To observe these structures, the cephalic capsules were processed through diaphanization with 10% KOH heated for about 2 minutes, and larvae were processed through diaphanization with 10% KOH heated for about 6 minutes and then washed in water. The images were taken with a Leica DM-2000 © microscope coupled to a photographic camera and analyzed with Leica LAS © V3.8 software.
Genitalia: Male and female genitalia were processed through diaphanization of the abdomen with 10% KOH heated, for about five minutes. Subsequently, they were washed in water and dissected in 70% alcohol. The photographs were taken with light microscopy (LM) using a Leica DFC295 camera mounted on a Leica M205C © stereomicroscope and Leica DM2000 © microscope. The specimens were measured (mm) with the Leica LAS © V3.8 software. The terminology of Nielsen and Common (1991) was used for the external morphology of the immature; Box (1923) for external adult morphology; Bradley and Carter (1982), Mey (1994) and Solis and Metz (2016) for male genitalia; Schmitt et al. (1996) and Matthews and Miller (2010) for female genitalia.
Map: The incidence map of the coffee leaf mining in Brazil was built using the SimpleMappr © program.

RESULTS
Morphological characteristics and chaetotaxy of the head and thorax were used to identify the four larval instars of L. coffeella. Also, we present a detailed study of the genitalia of males and females to better characterize both sexes.
Head. White scales on the dorsal apex. Base of the antenna white, with brownish length. Thorax. Covered with white scales. Legs. Long white scales. Wings. Long and thin, with two rounded black dots surrounded by yellowish scales. Apex of the wings with a row of black scales, laterally widened and with a "V" shape facing the posterior region of the body (Fig. 7C). Abdomen. White colored tergites. In ventral view, the last bipartite segment (Fig. 7B). Genitalia. Tergite 8 with coremata, forming a pair of elongated limbs with setae at the apex (Fig. 8A). Bilobate sternite 8, with an oval bulbus ejaculatorius (Fig. 8A). Bipartite valva, setose, and with small teeth in the internal region of the apex (Fig. 8B). Anellus fused medially and split at the apex, covering the valva (Fig. 8B). Gnathos elongated, forming a thin pair of arms (Fig. 8C). Short, cylindrical, and slender aedeagus at the apex (Fig. 8D).  Female Imago (Fig. 9A): External morphology. Similar to the male except by: average body length of 2,11±0,10 mm. Apex of the wings narrower and with a "C" shape facing the anterior region of the body (Fig. 9C). In ventral view, the last segment with a tubular shape (Fig. 9B). Genitalia. Bilobed ovipositor with long apical setae (Fig. 10A). Scleritis of the ductus bursae apically tapered and enlarged in the anterior region; extremities of the anterior region are tapered and have a concave shape facing the apex (Fig. 10B). Oval membranous corpus bursae with small spines (Fig. 10C).

DISCUSSION
Leucoptera Hübner brings together leaf-mining moths that cause economic damage worldwide (Seven 2006, Magalhães et al. 2010. Despite the serious damage caused by L. coffeella to the coffee crop, knowledge gaps about the biology of the CLM remain to be fulfilled and made publicly available. The CLM damage causing agents are all the larval stages before pupation (Notley 1948). Although Notley (1948) declared that L. coffeella has four larval instars, to our knowledge, there is no description of the morphological characteristics of each instar. The determination of Lepidoptera larval instars can be done by measuring the cephalic capsule (Dyar 1890, Ecole et al. 1999. After measuring the 200 larvae head capsule, the mean values were divided into four groups, one referring to each instar (1 st instar 0,14±0,03 mm, 2 nd instar 0,25±0,04 mm, 3 rd instar 0,32±0,03 mm and 4 th instar 0,42±0,03 mm), corroborating with Notley (1948) regarding the number of larval development stages. However, we have identified other morphological traits which can be useful to develop morphological markers of larval instars.
Besides length and width, the larval instars may also have different morphological characteristics (Nielsen and Common 1991). We observed that the 1 st larval instar of L. coffeella has primary arrows and no prolegs; 2 nd instar has arrows longer than 1 st instar and no prolegs; 3 rd instar presents prolegs without defined crochets and 4 th instar presents prolegs with defined crochets and ecdysial line. The larval instars of L. coffeella can be determined by the union of the head capsule measurements and the external body morphological characteristics.
To our understanding, the instar determination by the larval age is less precise than the morphological method we described here. In Navarro-Gutiérrez and Gallardo-Covas (2009), a bioassay study that referred the larval instars according to their age, reported that, in laboratory at 27ºC, the 1 st instar correspond to 3 days, 2 nd to 7 days, 3 rd to 11 days and 4 th to 13 days. However, the CLM life cycle is shorter in dry climates combined with higher temperatures (Wolcott 1947, Dantas et al. 2020. Hence, it is highly probable that a natural variation is reflected in the life cycle timing due to the climatic conditions in the coffee crop (Costa et al. 2012).
In order to prevent economic losses the chemical control against CLM is widely used in areas of high incidence (Fragoso et al. 2001). One approach to CLM controlling is to refrain from oviposition, which often requires stronger chemicals to control adults. A huge drawback of using chemical pesticides is that this control can lead to pest resistance, as reported elsewhere (Leite et al. 2020). The selection and dispersion of insecticide resistance was correlated with a larger number of organophosphate applications and likely dispersion of resistant individuals on the CLM populations (Fragoso et al. 2003). Chemical control is also harmful to non-target organisms like wasps, long reported as natural CLM predators (Kay andCollins 1987, Pereira et al. 2007). Wasps are able to cut the coffee leaf epidermis prey on the CLM larvae (Fragoso et al. 2001, Tuelher et al. 2003. Sexing CLM adults is important to conduct studies aiming the control of L. coffeella specimens in plantations using pheromones. The identification of males and females is essential to understand sexual behavior, sex-related mortality rates, and pheromone production (Notley 1948, Katiyar, K. P. & Ferrer, F. 1968, Michereff et al. 2007, Malo et al. 2009). Semiochemicals can be used to alter the natural reproductive behavior of insects, decreasing pest population levels (Cardé andMinks 1995, Dantas et al. 2020). The CLM sexual pheromone was used to determine the pattern of pheromone production by females, as an indirect measure of the behavior calling for sexual intercourse (Lima et al. 2008). However, although several previous studies have used adult sexing, none has clearly explained how to differentiate males from females, so the methodology of sexing males and females has remained unclear.
The genitalia structures involved in the L. coffeella copulation are the male gnathos, for fixation with the female ovipositor, and the valva as both mechanical and sensory function. The aedeagus is inserted into the female to insemination, when the sperm, initially located in the bulbus ejaculatorius, is delivered in the female sclerite to reach the corpus bursae, causing the fertilization of the female (Powell andOpler 2009, Cordero andBaixeras 2015).
The elements in the male and female genitalia of L. coffeella corroborating the structures described for the other species of Lyonetiidae (Bradley andCarter 1982, Mey 1994) (Fig. 11). The L. coffeella male exhibits anellus, bulbus ejaculatorius, and aedeagus. These structures are common with the male genitalia of L. malifoliella. In L. coffeella, tergite 8 with setose coremata, bilobed sternite 8, bipartite and setose valve, and elongated gnathos were observed, forming two thin arms. In contrast, pleurallobus, vinculum, pedunculus, subanalplatte, tuba analis are observed in L. malifoliella and this species does not have a valva (Mey 1994). In the female genitalia of L. coffeella, the structures we observed were bilobed ovipositor, sclerite in a conical shape, and oval corpus bursae with small spines. The corpus bursae was identified by Mey (1994) in the general structure of the Leucoptera female genitalia. The sclerite structure was identified by Schmitt et al. (1996) in another genus belonging to the Lyonetiidae family, where the female genital structure was characterized by a cruciform "T" shape, while in L. coffeella the sclerite has a conical shape (Fig. 12).  In conclusion, there are pest controls designed to be ovicidal, larvicidal, repellent or attractive to adult traps (Dantas et al. 2020). Different types of products can act at different stages of pest development and can be targeted to a specific sex. Thus, the identification of larval instars and adult sexes of L. coffeella are valuable to designing control strategies, as the discovery of new assets to bioproducts, nanoformulations, RNAi, and other strategies to improve sustainable IPM methods to control the CLM.