Identification and life history of aphids associated with chili pepper crops in southwestern Colombia

Viral diseases, transmitted by aphids, are the most limiting problems in chili pepper crops. Understanding the demographic features of these aphids, may thus assist the design of better disease control strategies in chili peppers. Aphid species found in chili pepper crops in southwestern Colombia were identified as Aphis gossypii Glover and Myzus persicae (Sulzer). An array of life-history parameters of both aphid species were investigated at 25 ◦C ± 0.5, 75 ± 1.75 % r.h., L12:D12, and LS 5-Light Storm in chili pepper crops. Both aphid populations consisted only of parthenogenetic females, showing a similar average development time—from the first nymphal instar to the post-reproductive adult—, female longevity, and daily average fertility values. The length of the reproductive period was higher for M. persicae. A. gossypii reached its adult state significantly faster than M. persicae. The intrinsic rate of population growth (rm) was lower for M. persicae (0.39) compared to A. gossypii (0.43). Results showed a potential for fast population growth in both species, which would enhance their role as virus vectors. The information acquired is essential to develop pest management initiatives for these two aphid species.


Introduction
Chili pepper viral diseases, caused by Potyvirus, Cucumovirus, and Poleovirus, are transmitted by several aphid species (Hemiptera: Aphididae) [1] and severely affect crop production; losses may reach 100 % depending on the level disease incidence. Worldwide, 16 aphid species have been specifically reported in chili pepper crops [2]. In Colombia, 3 941 ha are cultivated with of chili peppers, of which 167 are located in the department of Valle del Cauca (southwestern Colombia) [3]. A fraction of the product is destined to local consumption, and most of it is exported as chili paste. Local chili pepper producers spray insecticides to control both viral symptoms and aphids regardless of aphid species identity. The lack of species-related information on chili pepper aphids hinders the development of proper pest management plans [4].
A precise taxonomic identification of aphids, especially in the field, is challenging because of morphotypes with discreet morphometric differences within the same species [5] and chromatic variations unleashed by environmental stimuli [6] and intrinsic factors such as overpopulation or the type of host [7]. Molecular identification approaches, overcome the difficulties posed by morphological traits in telling aphid species apart. More precisely, the Cytochrome c Oxidase Subunit I (COI) gene region provides barcoding data complementing the morphologic identification of aphid species at any developmental stage [4,8], namely by discriminating cryptic species [9], allowing the assignment of morphotypes of different development stages within a given species [10], and clarifying the confounding effect of host-related variation among aphid morphotypes [11].
Aphid populations can grow profusely because they exhibit cyclic parthenogenesis which combines sexuality and parthenogenesis, and, depending on crowding conditions or plant quality, they are capable of producing either winged or wingless parthenogenetic females [12]. However, in hot tropical and sub-tropical climates, aphids usually reproduce by thelytokous parthenogenesis [2], resulting in fast population growth. Besides, aphid populations entail overlapping generations, parental groups, as well as their older descendants contribute to population increase [13]. One way to measure the potential population growth of an aphid species is via life tables. These constitute crucial tools to comprehend population dynamics, estimate the insects' potential and reproductive growth parameters [14], which, in turn, influence how aphid-borne viruses spread [15].
To broaden the knowledge about aphids in chili pepper crops in southwestern Colombia, we first identified the aphid species occurring in these crops, using morphological and molecular tools. Next, we established and compared some demographic and life history traits of the aphid species, in the hopes of understanding their impact on aphid-borne virus transmission in chili pepper crops. The results of this study can be applied to develop suitable aphid management plans in chili peppers. quantifications were done with a NanoDrop 2000 spectrophotometer. COI amplification was carried out with the primers described by Folmer et al. [17], LCO1490 (5' ggtcaacaaatcataaagatattgg-3') and HCO2198 (5'-taaacttcagggtgaccaaaaaatca-3'), under the amplification conditions used by Duque-Gamboa et al. [18]. Amplified products were displayed in agarose gels and sequenced at MACROGEN (Seoul, South Korea). The sequence files obtained were edited and aligned using MEGA6 [19].
Molecular identification of aphid species: Species recognition was carried out employing the neighbor-joining (N-J) grouping method, calculating genetic distances between haplotypes, and comparing the sequences obtained with those reported in public databases [20,21]. The N-J analysis under the Kimura's two-parameter model (K2P) and 1 000 bootstrap replicates were performed with the purpose of discriminating haplotype clusters and genetic distances among pairs of haplotypes within every cluster. Sequences showing genetic intra-group distances below 2 % were assigned to the same taxonomic unit. Interspecific genetic distances were estimated among clusters. All DNA sequence analyses were done with the software MEGA6 [19]. Finally, the sequences of different taxonomic units were compared with the ones available in the BOLD public databases (Barcode of Life Data Systems) and NCBI (National Center for Biotechnology Information), which allowed for species identification [22,20]. Each reared aphid stage and the morphotypes found due to polychromy, during their immature development, were identified via COI gene sequencing. Three specimens were used per morphological form and color, adding to 33 COI gene sequences for A. gossypii. Individuals of M. persicae showed a stable coloration pattern during their development period and six COI gene sequences were obtained.

Duration of the immature and adult stages
To establish the duration of aphid immature and adult stages, newborn nymphs were observed daily until they completed their development. The presence of the exuvia marked the duration of the nymphal instar. Aphid development was divided into four stages: i) presence of immature stages, ii) presence of pre-reproductive adults, iii) presence of reproductive adults, and iv) presence of post-reproductive adults. The evaluated time lapse for all stages in every identified aphid species was recorded in days.

Development time, survival rate, and proportion of females
Adult females of each aphid species (n = 148 for A. gossypii and n = 150 for M. persicae) were placed individually in Petri dishes with agar on chili pepper leaflets. Nymphs were fed with new chili pepper leaves every three days and females were withdrawn as soon as they delivered their first nymph. Development time (days) was established from the first nymphal stage until the emerged female delivered the first descendant nymph (pre-reproductive period). Survival rate of immature individuals and proportion of females of each aphid species were recorded.

Aphid longevity and reproduction
Newly emerged adult females were moved to Petri dishes with agar on chili pepper leaflets. To determine longevity and fertility values, the number of nymphs produced by each female (n = 45 for A. gossypii and n = 49 for M. persicae) was registered every 24 hours until the female died.

Aphid demographic traits
Development time and survival of immature individuals were combined with experimental reproduction data to create life tables "l x − m x " and calculate aphid demographic traits. The following parameters were calculated for each species: net reproductive rate (R 0 , is the number of daughters replacing an average female in the course of a generation), and generational time (T , describes the time between each generation). To calculate the intrinsic rate of population growth (r m ) per species, Eq. 1 was used [23]: Where: x = age (days), l x = female age-specific survival, and m x = proportion of female offspring of a female at age x. Following Carey [23], the pivotal female age was used, i.e. x + 0.5 to calculate the r m value.

Statistical analysis
Average values were accompanied with standard errors (± SE). The normality of data was checked using the Shapiro-Wilk test. Differences among species

Taxonomic identification of chili pepper aphids
Two aphid species were collected in the chili pepper crops assessed in 29 commercial farms from 18 localities in ten municipalities of Colombia. The two species were Aphis gossypii Glover, 1877 and Myzus persicae (Sulzer, 1776). Out of the 29 locations sampled, aphids were found at 20 sites, with A. gossypii detected at 19 sites (65.5 %), M. persicae at 5 sites (17.2 %,) and both species at 4 sites (13.8 %). Aphis gossypii dwelled mostly in tabasco pepper (66.7 %), followed by cayenne pepper (14.3 %), jalapeño pepper (9.5 %), and habanero pepper (9.5 %). Likewise, the main host for M. persicae was tabasco pepper (80 %), followed by jalapeño pepper (20 %). The distribution of both aphid species in the geographic valley of the Cauca River (southwestern Colombia) in all chili pepper species is shown in Fig. 1 and expanded in Suppl. 1.

Identification of aphids with DNA barcoding
COI gene sequences from 40 aphids, collected in jalapeño, habanero, tabasco, and cayenne chili pepper crops from different localities, were obtained. The N-J analysis on the COI sequences produced two distinct groups (Fig. 2). The genetic distance within every cluster was 0 %, revealing the existence of a unique haplotype in each group. Haplotypes were deposited into the GenBank under the unique accession codes MH203408 for A. gossypii and MH203409 for M. persicae.
The distance between clusters was 10 %, which matches interspecific differentiation. When comparing these haplotypes with those available in public databases, only matches with 100 % similarity and coverage were considered to identify sequences. Thus, sequences in clade A agreed with the species A. gossypii, whereas sequences in clade B belonged to the species M. persicae (Fig. 2). Aphids reared under laboratory conditions matched the GenBank sequences for A. gossypii and M. persicae, and corresponded to the morphological identification of each development stage of both species (Fig. 3). Additionally, the Barcoding technique established that yellow, yellow-light green, yellow-dark green, and black/black-green aphids observed in the field and in the laboratory are phenotypical variations of the species A. gossypii.     A difference regarding the duration of each of the nymphal instars was found between A. gossypii and M. persicae (Table 1) Table 3).

Discussion
Globally, sixteen aphid species have been reported in chili pepper (Capsicum spp.) crops [2]. Our research indicated that the aphid pest community associated with chili pepper crops in southwestern Colombia is much less diverse than expected [15]. The only aphid species found were M. persicae and A. gossypii, the latter being the one with the highest incidence. Both species have previously been reported in chili pepper crops [25,26]. Specifically, in the tropics, a low aphid species diversity is a common situation due to the short period in which aphids can survive without food. Three related factors contribute to this situation, (i) aphids must meet the high food demands of the embryos developing within them, (ii) they exhibit a high degree of host specificity, and (iii) they are not efficient in locating host plants [27]. Furthermore, our results show that the lack of genetic variation within A. gossypii and M. persicae could be a consequence of the observed situation; females were genetically identical since they reproduced exclusively by parthenogenesis.
The DNA barcoding technique confirmed and supported the morphological identification of A. gossypii and M. persicae [4,10]. Barcoding was useful to overcome species diagnostic difficulties in aphids. These difficulties are due to loss of useful taxonomical characters, phenotypic plasticity related to host plant interactions, stress, morphological particularities within each development stage, as well as polychromy [11]. In fact, most of these were observed specifically in A. gossypii.
Climate variables, especially temperature, influence the demographic traits of aphids [28]. The highest incidence of A. gossypii, and, concomitantly, the lowest incidence of M. persicae in the assessed chili pepper crops (18.6-30.1 • C range) can be partially explained by the adaptation capacity of each species to tropical temperature. M. persicae is greatly affected by high temperatures, with an optimal development between 15-20 • C [29], thus being mostly found in temperate regions [30]. On the contrary, A. gossypii, has its best performance between 26-30 • C [31] and prefers tropical zones [32]. Nevertheless, the average immature development time (from the first nymphal stage up to the post-reproductive adult) was 25.7 days for A. gossypi and 24.8 days for M. persicae. These values were similar in our study, since temperature [33] and host plant conditions [34] affect the length of the development of aphids. Both species showed four nymphal stages, as found by Dixon et al. [27], lasting between 1.1-1.5 days, on average. This range matches what has previously been reported in chili pepper crop aphids [35].
The development time from nymph 1 (Stage 1) to adult was short for both species (<6 days). This makes it possible to obtain females rapidly, which will start to reproduce the day after emerging. These features, as well as their reproduction via obliged thelytoky parthenogenesis, are common in tropical regions [12], and allow both species to develop larger populations in a short amount of time. Our results show that their descendants are clones, with adaptive characteristics for local environmental conditions [36]. Both species delivered around three or four nymphs per day although the length of the reproductive period was significantly higher for M. persicae, than for A. gossypii. This finding clashes with the idea of a less adapted M. persicae to hot climates; namely this species showing higher fertility and longevity at temperatures below 15 • C [37]. Nevertheless, the trait values we observed seem to reflect the adaptation of the aphid to chili pepper crops. For instance, a maximum overall average fertility of 62.7 individuals at 25 ± 1 • C [25] has been reported in bell pepper (C. annuum) aphids, which is a lower value than what we observed in our work.
The post-reproductive period was shorter for M. persicae (<2 days), indicating that females of this species die quickly after they reproduce [38], whereas in A. gossypii we observed a longer post-reproductive period. M. persicae compensates its shorter survival time with a broader reproduction period that could be interpreted as an adaptive advantage in this species. These biological differences can influence virus transmission dynamics in the assessed chili pepper crops [39], given that A. gossypii stays and feeds longer in the crop, thus having more chance to spreading the virus.
Population parameters needed to create life tables for aphids are highly influenced by the host plant [34,14] and by the environmental temperature [31]. In fact, the Capsicum r m values found in our study are in the range reported by other authors. In C. frutescens and C. annuum, the r m values of 0.27 and 0.48 for A. gossypii were reported by Singh and Singh [40] and Satar et al. [41], respectively. For M. persicae in C. annuum, r m values of 0.31 [29], 0.33 [25] and 0.41 [41] were reported.
According to our results, net reproductive rate (R o ) and generational time (T ) were superior in M. persicae (72.8 and 13.8, respectively) than in A. gossypii (65.5 and 11.7, respectively). The R o value found in A. gossypii is superior to the one reported in cotton crops, i.e. 21.7 and 24.8 at a temperature of 27.5 ± 1 • C [34] and 40 at a temperature of 23 • C [42]. This depicts, once again, the adaptative capacity of A. gossypii to chili pepper crops.
Biological and demographic features show the adaptation of both aphid species to Capsicum spp., although A. gossypii is more frequent and abundant. Therefore, future studies are needed to establish if this asymmetry is a product of direct competitive interactions in these crops. Precisely, Denno et al. [43] mention the following facts as competition triggers between the two species: both are sap feeders, live in aggregates, and inhabit managed environments (i.e. agroecosystems). On the other hand, top-down cascades represented by the diversity of the natural enemies of aphids (for instance, Coccinellid beetles)  [44] found in chili pepper crops and their companion plants (unpublished data) could be shaping the incidence of the aphid species [45]. In addition to these trophic cascades, the insecticides sprayed in the crops sampled could have affected the incidence of aphids, either decreasing their presence [46] or favoring it through the elimination of their natural enemies [47].
Although the presence of aphids causes demonstrable yield losses in several crops [48], for chili pepper crops, the greatest concern is the economic loss caused by the viruses transmitted by both aphid species. These viruses, transmitted through the stylet when feeding, appear in a non-persistent manner (unpublished data). The higher population growth of A. gossypii, a consequence of better adaptation to tropical climates [32] than M. persicae [37], could be reflected in a larger virus spread. However, our demographic results for M. persicae suggest that this species could be adapting to warm climates.

Conclusions
A. gossypii and M. persicae are currently the only aphid species present in Capsicum spp. crops in southwestern Colombia. A. gossypii has a higher incidence and shows sharp polychromy during its development.
Both aphid species reproduce via thelytoky parthenogenesis combined with a short immature development phase (less than 6 days), which allows them to produce large populations in a short period.

William Talaga Taquinas
Is an agricultural engineer from the Universidad Nacional de Colombia in Palmira who follows his MSc studies on Agricultural Sciences at the same institution. His research focuses on taxonomy and ecology of aphids and their role as virus vectors. He also does research about pest monitoring, pest insect spatial modeling and IPM on orchards and vegetable crops.

Clara Inés Melo Cerón
Is an agroforestry engineer from the Universidad de Nariño, Colombia. She finished her MSc studies on Agricultural Sciences at the Universidad Nacional de Colombia in Palmira and she follows her PhD studies on Agroecology at the same institution. Her research focuses on ecology, biological control and community structure of aphids and their natural enemies on chili pepper crops. Also she has been involved in research about biofertilization and agroforestry.

Yorley Lagos Alvarez
Is an agricultural engineer from the Universidad Nacional de Colombia in Palmira who follows her MSc studies on Agricultural Sciences at the same institution. Currently is a Visiting Researcher at the International Center of Tropical Agriculture (CIAT) in Cali, Colombia. Her research focuses on the evaluation of pesticide residues on Rubus glaucus fruits in Valle del Cauca, Colombia.

Diana Nataly Duque Gamboa
Is a biologist from the Universidad del Valle. She finished her MSc studies at the Universidad del Valle and currently follows her PhD studies on Biology (Sciences) at the same institution. Her research focuses on the use of molecular tools to identify pest insects, to determine virus presence on both plants and insects and also to characterize trophic interactions in biological control of pest insects.

Nelson Toro-Perea
Is a molecular biologist. He is currently an associate professor at the Universidad del Valle in Cali, Colombia. His research interest is focused on the use of molecular tools to know and characterize the evolutionary patterns of species in the neotropics, as well as the interactions between plants and their associated organisms, such as symbionts and pathogens.

Maria R. Manzano
Is a biologist and entomologist with graduated studies on crop protection and biological control of pest insects. Much of her work is collaborative, her research program is broad, combining biology, taxonomy and ecology of insects to develop basic knowledge that support both, the development of biological control and the integrated pest management (IPM) programs.