Fitness and mating compatibility of Chrysodeixis includens (W.) (Lepidoptera: Noctuidae) populations collected in different provinces and crops in Argentina

Chrysodeixis includens, is an important Plusiinae pest that feeds on soybean, beans, and cotton. This species has a genetic structure determined by molecular markers in populations collected in soybean crops in Brazil and has developed resistance to insecticides. This had led to an increased interest in the genetic diversity of its populations in Argentina. This study aimed to characterize the fi tness parameters and mating compatibility of geographically distant C. includens populations and on different host plants in Argentina. Intraand inter-population crosses revealed that C. includens populations from Tucumán and Santiago del Estero provinces did not present any evidence of prezygotic and postzygotic incompatibilities, suggesting that both populations collected from soybean plants belonged to a single wideranging species. Prezygotic and postzygotic incompatibilities were detected between populations collected from different crops (soybean and bean). These incompatibilities may be caused by the different nutrients that each plant species renders, or the specifi c pest control strategies which may affect C. includens behavior. Studies of the genetic structure of C. includens populations based on mating isolation on other crops, as well as of the impact of management strategies on the specie, will increase understanding of the reproductive compatibility revealed by the present study.


INTRODUCTION
The soybean looper Chrysodeixis includens Walker 1857 (Lepidoptera: Noctuidae), is a Plusiine pest that feeds of numerous crops plants of great economic value (Specht et al. 2015). Its origin is unclear, but it is recognized as a migratory species restricted to the Western hemisphere, from The United States of America (USA) in the north to southern South America (Navarro et al. 2009, Barrionuevo et al. 2012). Chrysodeixis includens is a polyphagous species that attacks beans (Phaseolus vulgaris L.), cotton (Gossypium hirsutum L.), tobacco (Nicotiana tabacum L.), fl ax (Linum usitatissimum L.), sunfl ower (Helianthus annuus L.), alfalfa (Medicago sativa L.), and horticultural crops, among others (Navarro et al. 2009, Barrionuevo et al. 2012, Specht et al. 2015, but it shows preference for legumes, such as soybean (Glycine max L.). The period of greatest leaf consumption of C. includens in soybean, takes place after growth stage R3, which is a critical phase that directly impact on yield (Carter-Wientjes et al. 2004, Casmuz et al. 2015. In The USA, C. includens is the most destructive defoliation pest that attacks the soybean crop (Moonga & Davis 2016). In Brazil, it has also become an important phytosanitary problem in soybean and cotton production since early 2000's (Moscardi et al. 2012, De Lima Viana et al. 2014, Yano et al. 2015, and their populations have been considered to be of growing importance for bean , due to the damage caused in this crop (Baldin et al. 2014).
Chrysodeixis includens is an important pest in Argentina from 33° south latitude northwards being present in the northwest and the northeast regions, Córdoba and Buenos Aires provinces (Navarro et al. 2009, Barrionuevo & San Blas 2015). In the 2013/2014 growing season, in soybean field of San Agustín (Tucumán province, Argentina) this species caused a 20.2% defoliation and a 19.7% yield reduction . In the last season 2017/2018 farmers of Rosario de la Frontera county (Salta province, Argentina), recorded outbreaks of this pest in bean crops above the level of the economic damage and have had to resort to three agrochemical sprays to their control. Currently, insecticides and genetically modified soybean and cotton that express the insecticidal protein Cry1Ac of Bacillus thuringiensis are used in Argentina to control C. includens. In The USA, resistance to insecticides (organophosphates, carbamates and pyrethroids) and control failures with IGR (insect growth regulator) insecticides (methoxyfenozide) (Owen et al. 2013) have been reported for C. includens (Felland et al. 1990, Higley & Boethel 1994, Mascarenhas & Boethel 2000.  also indicated that the mixture of pyrethroid insecticides with organophosphates had no control over C. includens in soybean in Tucumán (Argentina). Moreover, Mascarenhas et al. (1998) found that C. includens field-collected larvae from U.S. had a lower susceptibility to the Bacillus thuringiensis variety kurstaki insecticide as compared to a reference laboratory population.
Additionally, some insect populations identified as the same species may yet display different biological and genetic traits and show reproductive isolation (cryptic species), preventing complete gene flow among populations as consequence of the presence of prezygotic and postzygotic barriers (Drès & Mallet 2002, Bickford et al. 2007. Therefore, identifying cryptic species is important not only to explain the evolution of phytophagous insects, but also to understand the effect of susceptibility of pest populations to different control tactics (Pérez Contreras 1999, Rull et al. 2012, Joyce et al. 2014. It has led to an increased interest in fully understanding C. includens genetic diversity and gene flow between geographically distant populations of C. includens, and between populations collecting in different crops. Palma et al. (2015) carried out the first molecular characterization of C. includens populations in soybean crops from Brazil with molecular markers of the Inter Simple Sequence Repeat (ISSR) type, determining the existence of genetic structuring in populations which was not related to geographical distances. Also, structured genetic populations have been reported for other Plusiinae species, the cabbage looper Trichoplusia ni Hübner 1800-1803 (Lepidoptera: Noctuidae) (Zhu et al. 1997, Evenden et al. 2002, Franklin et al. 2010. These studies revealed variation in their pheromone composition and the presence of haplotypes.
The objective of this study was to characterize biological, reproductive parameters, and mating compatibility of C. includens populations collected in different host plant and geographical distance in Argentina, to determine whether this pest comprises a complex of cryptic species, or simply consist of a single wide-ranging species.
Thus, C. includens specimens were collected from soybean crops in Tucumán and Santiago del Estero provinces, to establish the influence of geographic distance and from bean crop in Salta province, to establish the influence of host plant association on reproductive isolation. This research provides information about reproductive compatibility of soybean loopers, which has remained undocumented until now. This information will be useful for understanding reproductive compatibility among C. includens populations, which is necessary for designing sustainable strategies to control this species in Argentina.
At each sampling site, a minimum of 300 larvae (instars 3-6) were collected using a vertical cloth as described by Drees & Rice (1985). Each larva was placed in a plastic tube (12 cm tall and with a 1.5 cm diameter) containing one soybean or bean leaf. The larvae were taken to the laboratory and kept in breeding chambers under controlled conditions (at 27 ± 2 ºC, 70-75% relative humidity and 14:10 h light:dark photoperiod). All larvae were examined following procedures described by Barrionuevo & San Blas (2015) to confirm the species. After the emergence of adults a voucher specimens from each population were deposited in the insect collection of Sección Zoología Agrícola, at Estación Experimental Agroindustrial Obispo Colombres (EEAOC) (Tucumán, Argentina).

Insect rearing
Approximately 180 adults (90 females and 90 males) were randomly selected among the larvae collected, and reared to establish an experimental colony for each population. Approximately 30 female and 30 males were placed in cylindrical oviposition cages (40 cm tall and with a 20 cm diameter), lined with polyethylene bags. For aeration, both ends of the cage were covered with a nylon cloth. The cages were placed on plastic trays lined with moistened polyurethane foam. Each population was maintained in the same chamber under identically controlled conditions (at 27 ± 2 ºC, 70 -75% relative humidity and a 14:10 h light:dark photoperiod). Food was provided via a cotton plug saturated with a mixture of honey and water (1:1), which was replaced every day. Cages were checked daily for oviposition and adult mortality. Eggs were collected daily from the cages and placed in 1000 ml plastic containers lined with moistened filter paper. Once they had emerged, neonate larvae were placed individually in glass tubes and fed on an artificial larval diet that included bean flour (Grandiet®, Buenos Aires, Argentina), wheat germ (Grandiet®, Buenos Aires, Argentina), brewer's yeast (Calsa®, Tucumán, Argentina), vitamin C (Anedra®, Buenos Aires, Argentina), sorbic acid (Anedra®, Buenos Aires, Argentina), sodium benzoate (Todo Droga ® , Córdoba, Argentina), vitamin supplement amino acids (Ruminal®, Buenos Aires, Argentina), and methylparaben (Todo Droga®, Córdoba, Argentina) (Murúa et al. 2003). Artificial diet was replaced every 2-3 days. As larvae pupated, pupae were sexed and placed in cup containers with filter paper, until adults emerged. Adults were used to initiate a new generation. After establishing a colony for each population, F1 generation of field-collected strains was used for the fitness study, and F2 generation was used to begin the parental crosses in the mating compatibility study. The following generations for the mating compatibility studies (F3-F5) were originated from the parental crosses.

Fitness of soybean looper populations
From each experimental C. includens colony, 200 eggs were randomly selected to analyze the following parameters: adult sex ratio, duration of egg, instars larval and pupal stages, and pupal mass (obtained 24 h after pupation). Adult sex ratio was determined in pupal stage according to the Angulo & Weigert's methodology (1985).
Among the adults obtained, couples (females and males) were randomly selected to determine reproductive parameters. One virgin female and one virgin male (24 hours old) were selected from each population and placed in cylindrical oviposition cages, similar to those described above (40 cm tall and with a 20 cm diameter).
Dead females were dissected to establish the presence and number of spermatophores present in their reproductive tract, to determine whether mating had occurred. Duration of preoviposition, oviposition, and post-oviposition periods, fecundity (number of eggs deposited by a female during her entire life period), fertility (percentage of hatched eggs) and adult longevity were recorded.

Reproductive compatibility between soybean looper populations
Two crossing experiments were performed to determine reproductive compatibility between C. includens populations: 1) between populations distant geographically and collected from soybean crop (Tucumán and Santiago del Estero provinces); and 2) between populations collected from different host plants (soybean and bean crops from Tucumán and Salta provinces, respectively).
The crossings were performed following the methodologies described by Herrero et al. (2017).
To determine compatibility, one virgin female and one virgin male (both being 24 hours old) were used. Four different types of crosses were performed: i) parental crosses using parents from the same population (in this case, single pair mating from the fitness studies were used); ii) hybrid crosses using one parent from different populations; iii) backcrosses, some with the female progenitor as the recurrent parent, and others with the male progenitor as the recurrent parent; and iv) interhybrid mating crosses between F1 hybrids from different populations. A subset of larval progeny (F1) from each fertile cross was monitored for survival until pupation, and then reared to adulthood. The parameters measured to determine compatibility were as follows: presence and number of spermatophores per female, preoviposition, oviposition and postoviposition period duration, total fecundity, and total fertility.

Data analysis
Fitness data from different populations were compared by Kruskal-Wallis (1952) test (P > 0.05), because they did not show normal distribution. For the reproductive compatibility studies, due to the high number of combinations, the performance of all parental crosses was compared with that deriving from other types of crosses (hybrid crosses, backcrosses, and interhybrid mating), using Kruskal-Wallis (1952) test (P > 0.05). For all of these studies, the data of duration of preoviposition, oviposition and postoviposition periods were compared in the case of females that had laid eggs, whereas fecundity values were compared considering all females, including those that had not laid eggs. For fertility, females that had laid eggs but had no spermatophores were not included. For the number of spermatophores, females that had none in their tract were included as a value representing copula failure in all cross types. All data were analyzed using InfoStat 2016 software (Di Rienzo et al. 2008).

RESULTS
All adults from the three sampled populations were identified as C. includens based on the morphology of their larvae (Barrionuevo & San Blas 2015).

Fitness of C. includens populations
In total, 135, 137 and 124 individuals from Tucumán, Salta and Santiago del Estero populations, respectively, were used in this study. The life stage and reproductive parameters are presented in Table I. Significant differences were detected in the duration of the L1, L2, L3, L5 and L6 instars larval, duration of overall larval stage and pupal stages. The duration of L1, L2 and L6 instars larval from Salta population collected in bean crop (4.04, 3.05 and 5.55 days, respectively) were significantly longer than those from Tucumán ( 38 days), was significantly longer than others population collected from soybean crop (14.7 days for both populations) (H= 208.22, P= <0.0001). Duration of the pupal stage from Santiago del Estero and Salta populations (8.6 days for both) was significantly longer than from Tucumán populations (7.78 days) (H= 59.90, P= <0.0001). The duration of the preoviposition period from Santiago del Estero population (5.13 days) was significantly longer than from Salta and Tucumán populations, in decreasing order (4.58 and 3.25 days, respectively) (H= 9.39, P= 0.0074). No significant differences were detected between the remaining fitness and reproductive parameters (Table I).
Reproductive compatibility between C. includens populations distant geographically and collected from the same crop In total, 43 parental crosses, 43 hybrid crosses, 153 backcrosses and 63 interhybrid matings were made between Tucumán and Santiago del Estero populations (Table II). Significant differences were found in the duration of preoviposition (H= 9.07, P= 0.0225), and postoviposition periods (H= 7.91, P= 0.0407), fecundity (H= 20.58, P= 0.0001) and fertility rates (H= 18.83, P= 0.0003) (Tables II  and IV). The results obtained in the four types of crosses, show a successful matings between Tucumán and Santiago del Estero populations.

Reproductive compatibility between C. includens populations collected from different crops
In total, 37 parental crosses and 60 hybrid crosses were made between Tucumán and Salta populations collected in soybean and bean crops, respectively (Table III). Significant differences were found in preoviposition period duration (H= 5.84, P= 0.0141), fecundity (H= 8.27, P= 0.0033) and fertility rates (H= 12.22, P= 0.0003),

DISCUSSION
This study compared biological, reproductive parameters and mating compatibility between C. includens populations from different geographically distance and from different hosts in Argentina. This study is the first report of mating compatibility among C. includens populations in Argentina.
The results of this study showed that under laboratory conditions, soybean looper populations from Salta collected in bean crop complete a single generation (from egg to adult) in approximately 35 days, longer than Santiago del Estero and Tucumán (31 and 32 days, respectively), collected in soybean crop. Similar results were reported by Combe & Pérez (1978). A life cycle duration of 43 days was recorded by Barrionuevo et al. (2012) and 38-42 days by Moonga & Davis (2016).
Egg stage duration was similar to those reported in previous studies (Canerday & Arant 1967, Combe & Peréz 1978, Navarro et al. 2009, Barrionuevo et al. 2012. Chrysodeixis includens generally completes larval stage in six instars (Moscardi et al. 2012). However, 56%, 76% and 88% of the insects from Tucumán, Santiago del Estero and Salta, respectively, presented only six instars. The population from Salta had the highest percentage of individuals that molted for six instars. Shour & Sparks (1981) observed a range from five to seven instars larval, with a predominance of six instars when the larvae were fed on an artificial diet. The differences recorded in larval stage duration and number of instars may be due to the components of the artificial diet, and the rearing conditions (Barrionuevo et al. 2012). The duration of the overall larval stage from Tucumán and Santiago del Estero populations was in agreement with Moonga & Davis (2016) and Canerday & Arant (1967) who used soybean leaves and artificial diet to breed the larvae, respectively. The duration of the overall larval stage from Salta population was similar to that reported by Combe & Pérez (1978) and Boldt et al. (1975). Besides, all the populations had shorter larval stage duration than reported by Barrionuevo et al. (2012) (2012) and Canerday & Arant (1967). Pupal stage from Salta and Santiago del Estero populations was similar to that reported by Barrionuevo et al. (2012). The pupal mass was similar to the one reported in other studies (Canerday & Arant 1967, Jensen et al. 1974, Boldt et al. 1975. By contrast, Kidd & Orr (2001) and Moonga & Davis (2016) reported a lower pupal mass using soybean, cotton, cowpea (Vigna unguiculata L.) and sweet potato (Ipomoea batatas L.) leaves as larval diet.
Adult longevity for three populations was lower than reported by Combe & Peréz (1978) and Canerday & Arant (1967), but similar to those recorded by other authors (Jensen et al. 1974, Barrionuevo et al. 2012. In this study, it was found that females emerged before males, suggesting that C. includens is a protogyny species. No previous studies have reported this female behavior. The sex ratio found was of 0.8/0.9 F -1 M. Barrionuevo et al. (2012) and Combe & Pérez (1978) reported a sex ratio with more females than males.
Number of spermatophores per female was generally 1 for the three populations, with a range from 1 to 5. This observation was similar to reported by Jensen et al. (1974). The duration of preoviposition period from Tucumán and Salta populations was similar to those recorded by Canerday & Arant (1967) and Combe & Pérez (1978). The duration of this period for Santiago del Estero population was higher than those reported (Canerday & Arant 1967, Combe & Pérez 1978. The oviposition period was similar to that reported by Barrionuevo et al. (2012). No previous studies have examined C. includens postoviposition period.
Fecundity was similar to those reported by Barrionuevo et al. (2012) and Jensen et al. (1974); but it was lower than those reported by other authors (Combe & Pérez 1978, Kidd & Orr 2001. Total fertility for all populations, was higher than reported by Moonga & Davis (2016) and lower than reported by Barrionuevo et al. (2012).
The results about mating compatibility between C. includens populations from Tucumán and Santiago del Estero) collected in soybean crop, did not show signs of geographical isolation. The presence of spermatophores in bursa copulatrix of females, and fecundity and fertility values, recorded in the different types of crosses, demonstrated successfully mating in both directions. This may account for the long-range migratory capacity of this species, whose presence has been widely documented in the American continent (Pogue 2005, Navarro et al. 2009). Palma et al. (2015 carried out the first molecular characterization of C. includens populations in soybean crops in Brazil and determined the existence a low but significant genetic differentiation among populations not related to geographical distance. The successful mating between C. includens populations from soybean crops presented in this study support the results reported by Palma et al (2015). The present study can contribute to the understanding of gene flow among C. includens individuals from different provinces, a knowledge that is necessary to plan management programs aimed at offsetting resistance to Bt crops and insecticides (Grimi et al. 2018). Transgenic cotton and soybean crops that express Cry1Ac insecticidal proteins have been commercialized in Argentina since 1998 and2012, respectively, and Chrysodeixis includens is one of the target pests of these technologies (Massoni et al. 2015). In 2017, Bt soybean was produced on 3.08 million hectares and Bt cotton on 250.000 hectares in Argentina. Currently the area cultivated with Bt crops is increasing rapidly and Argentina ranks third in planted biotech crops, with a total of 23.9 million hectares (ISAAA 2018). Furthermore, the soybeans that express a single insecticidal protein Cry1Ac, did not fully meet the high-dose concept for C. includens, which is one of the most important parameters in IRM programs (Bernardi et al. 2012).
Mating compatibility between Tucumán and Salta populations collected in soybean and bean crops, respectively, showed lower values of fecundity (103.4 eggs per female) and fertility (6.4%) parameters, spermatophores per female (0.5) and percentage mated female (41.9 and 48.3 %) in hybrid crosses than in parental crosses (Table IV). Also, these low values put off the possibility of making subsequent backcrosses and interhybrid mating of this study (Table  V). These results suggested the presence of a barrier of gene flow between C. includens populations from soybean and bean which may be based on the different nutrients each plant species (Pashley 1988, Murúa et al. 2008 or pest control strategies used in the fields that would affect the behavior of this species (Palma et al. 2015). Besides, in Argentina, bean production is located in the northwest region, with the Salta province covering 44% of total national bean production, which makes it the largest producer in the country (Ministerio de Agroindustria de Argentina 2016). This concentration of bean plantations may be promoting physiological adaptations of this pest to the crop. However, additional research based on reproductive compatibility studies of C. includens populations from other crops, and the effect of management strategies will further deepen our understanding of the mating isolation found in the present study, and clarify the possible existence of host races.