Body size is one of the most determining features in the fitness of insects due to its close relationship with various physiological and ecological characteristics (Kalinkat et al. 2015). For programs that use the sterile insect technique to control pest species of fruit flies (Diptera: Tephritidae), body size is one of the main indicators of the quality of mass-reared insects because a large male and large female are strongly correlated with greater mating success and fecundity in some species, respectively (Churchill-Stanland et al. 1986; Liedo et al. 1992; Fernandes-da-Silva & Zucoloto 1997; Tejeda et al. 2020). However, the processes of domestication and mass-rearing might lead to inadvertent behavioral, life history, and morphometric changes. These changes can result in positive attributes for mass-rearing, but in adverse aspects for field performance of the laboratory reared individuals (Cayol 2000; Weldon 2005; Sørensen et al. 2012; Hoffmann et al. 2005; Hoffmann & Ross 2018).

Because the effectiveness of sterile insect technique depends on the ability of sterile males to compete against their wild congeners for matings with wild females (Knipling 1955), it is crucial to understand how laboratory colonization and mass-rearing conditions can contribute to differences between laboratory and wild populations, especially in relevant traits such as body size or morphometric features related to male competitiveness (Clarke & McKenzie 1992; Gómez Cendra et al. 2014; Meza et al. 2014; Parker et al. 2021).

The assessment of mass-reared sterile insect quality allows the identification of possible causes of low field performance of the released flies (Chambers 1977). In tephritid fruit fly sterile insect technique programs, quality control protocols use pupal weight and diam, an indicator of pupal size, as the first indicators of sterile insect quality. Pupal size is an indicator of adult size, as well as an indirect measure of nutritional status, such as energy reserves, which are essential for the dispersal of insects and desiccation resistance that affects longevity (Weldon et al. 2013, 2021; FAO 2019). However, the relevance of producing large males is based on evidence that male size influences female choice (Aquino & Joachim-Bravo 2014; Benelli et al. 2015).

Because the adult size is determined by pupa size, and this in turn by the larva size, the larval diet (quality and amount) is considered one of the main factors of size variation in the mass-rearing of fruit flies. Nevertheless, this phenotype also reacts to different environmental conditions, such as adaptation to artificial rearing conditions or stressful environments (Clarke & Mckenzie 1992; Nijhout et al. 2014). For example, in Anastrepha fraterculus (Wiedemann) (Diptera: Tephritidae), it was found that laboratory flies were larger than wild flies, laboratory males had greater head width and eye length, although both laboratory males and females had narrower wings than their wild congeners (Gómez Cendra et al. 2014). Recently, morphometric differences between laboratory and wild males from similar pupal size were determined in A. ludens and A. obliqua (Sánchez-Rosario et al. 2021). These differences could be due to environmental or genetic factors, or as a response to adaptation to artificial conditions.

Ecological and evolutionary processes act differently in each sex, thus the relevance of studying size variation in the context of sexual size dimorphism (Fairbairn 1997; Stillwell et al. 2010; Testa et al. 2013). In Anastrepha suspensa (Loew) and other species of tephritids of economic importance, such as: A. ludens (Loew), A. obliqua (Macquart), Bactrocera dorsalis (Hendel), Bactrocera cucurbitae (Coquillett), Bactrocera oleae (Rossi), Ceratitis capitata (Wiedemann), Rhagoletis pomonella (Walsh) (all Diptera: Tephritidae), females are larger than males (Sivinsky & Dodson 1992). Sex-specific allometry was reported in B. dorsalis, particularly in weight relative to tibial length. Females gained disproportionally more weight than males with the increase of hind-tibial length (Zhou et al. 2016). In Bactrocera tryoni (Froggatt) (Diptera: Tephritidae), sexual dimorphism in body size and wing shape also has been reported, but with similar symmetrical growth. Females had a larger body size but shorter and wider wings than males, which may be associated with their reproductive biology or locomotion (Sivinsky & Dodson 1992; Zhou et al. 2020).

Given the evidence of morphometric differences (size and shape) between wild and laboratory populations, improved understanding is needed on how mass-rearing conditions affect intraspecific size-frequency distributions in tephritid fruit flies. Therefore, the objective of our research was to compare the pupal size distributions and dimorphism of 2 populations, wild and laboratory, of 2 economically important species, A. ludens and A. obliqua. In this study we used a pupal sorter machine, and we used the same calibration for studying the size distribution of 2 populations. We measured the length, width, and weight of male and female pupae of wild and laboratory populations of A. ludens and A. obliqua, 2 species used in the sterile insect technique.

Materials and Methods

Results

Pupal size distribution data for A. ludens are based on the emergence record of 184 wild pupae (103 males and 81 females), and 316 laboratory pupae (123 males and 193 females). For A. obliqua, the pupal size distribution data is based on 224 wild pupae (123 males and 101 females) and 243 from laboratory pupae (133 males and 110 females).

Fig. 1.

Size distribution of laboratory and wild Anastrepha ludens and Anastrepha obliqua male and female pupae. The proportion of male and female pupae is shown in 10 pupal size classes (pupal diam mm) on the x-axis.

The highest proportion of males and females of both populations and species was distributed in categories 5 to 7, although this distribution tended to be smoother in the wild population. In A. ludens, the peak of distribution in wild females was skewed more towards the large size relative to the laboratory flies. Regarding males, both distributions were quite similar. In both the laboratory and the wild population, a higher proportion of males and females was observed in A. ludens compared to A. obliqua. In A. obliqua, male and female pupae from the laboratory population were distributed principally in the middle categories (5 to 7), whereas the wild population showed a wider distribution. In this species, unlike A. ludens, the distribution was skewed more to the right (Fig. 1).

According to Pearson's correlation, all variables (width, weight, and length) were positively and significantly correlated with each other in laboratory and wild pupae of both species (P < 0.0001). In A. ludens, the association between width and weight was stronger in the laboratory population than in the wild population (0.9651 and 0.8339, respectively), in such a way that the weight was more correlated with the width in laboratory pupae, whereas in the wild pupae the correlation between the parameters was more uniform. In A. obliqua, the correlation values were quite similar in both populations.

The canonical analysis performed for males and females of each species was significant (Figs. 2 & 3). In A. ludens, there were significant differences between the populations (Pillai test, F_3,494 = 54.73; P < 0.0001) and between the sexes (F_3,494 = 46.31; P < 0.0001), but there was no significant interaction between populations and sexes (F_3,494 = 2.44; P = 0.063). According to the canonical analysis, wild males and females were longer than laboratory pupae, whereas laboratory males and females were wider and heavier in comparison to wild congeners (Fig. 2). For A. obliqua, there also were significant differences between populations (Pillai test, F_3,461 = 33.52; P < 0.0001) and between sexes (F_3,461 = 26.18; P < 0.0001); the interaction between population and sex also was statistically significant (F_3,461 = 3.90; P = 0.008) (Fig. 3). Wild males and females were larger in pupal weight, width, and length than their laboratory congeners.

Fig. 2.

Canonical analysis for pupae size parameters of males and females of laboratory and wild populations in Anastrepha ludens. The canonical analysis is represented on the first canonical axis (can 1), where the boxplots indicate the populations (laboratory and wild) and sexes (males and females) (left). The variables of pupae size: length (mm), width (mm), and weight (mg) are indicated by vectors (right).

Discussion

In this study we compared the pupal size distributions of 2 populations, laboratory and wild, of 2 economically important species: A. ludens and A. obliqua. We evaluated 3 parameters associated with pupal size: length, width, and weight. Given the correlation with the adult size (male or female), we inherently considered the sexual size dimorphism in the analysis of pupal size distribution.

Our findings showed differences in size distribution between laboratory and wild populations within each species, as well as between males and females. We confirmed sexual size dimorphism in pupae of both species. Although sexual dimorphism has been demonstrated in some species of tephritids, this study constitutes the first reference regarding sexual size dimorphism in pupae (Sivinski & Dodson 1992; Tejeda et al. 2014; Zhou et al. 2016, 2020). In some experiments where body size was evaluated, size selection was performed through manipulation of larval densities in the diet. However, this leads to different nutritional status of the adult, and therefore an impact on size and fitness components (Aluja et al. 2001; Kaspi et al. 2002; Nash & Chapman 2014). Here, all laboratory-reared males were obtained at controlled larval densities (Orozco-Dávila et al. 2017; FAO 2019) and wild flies were obtained from infested fruits and collected randomly. This approach allowed us to reduce the effects of other causes of variation in male size.

The size distribution showed some differences between the 2 populations. In general, in laboratory A. obliqua, a lower proportion of males and females was recorded in the first categories; this was due to low emergence in the first and second category. According to Tejeda et al. (2014), resistance to desiccation increases with size, so it is likely that small pupae lose more water, which in turn is due to a lower amount of lipids. This could explain the lower emergence observed in small size categories. Regarding the differences in the size distribution between populations, wild males and females showed a smoother normal distribution than laboratory congeners. Wild females of both species showed a higher frequency in the large size categories. The standardized environmental variables used in mass-rearing to increase production and quality of insects, such as quantity and quality of the diet, probably explains a higher frequency in the size distribution of the middle categories (5–7) (Orozco-Dávila et al. 2017). In C. capitata, analysis of body mass frequency distributions determined that this distribution tends to be normal to slightly skewed to the right (Gouws et al. 2011). Although we separated the pupae based on their diam, in both species we observed that the highest frequency of individuals was distributed in size categories 5 to 7, so they show a similar trend.

Sexual dimorphism in pupal size was evident in both species and populations. Females were bigger than males in all pupal size parameters, although these differences were more evident in A. ludens than in A. obliqua. These results on pupal size dimorphism were as expected due to body size dimorphism in adults of several tephritid species (Sivinski & Dodson 1992). The explanation for this marked dimorphism is closely related to the reproductive roles. In B. tryoni, the larger body size of the female is related to the increase in the number of ovarioles, whereas in A. ludens, Anastrepha serpentina (Wiedemann) (Diptera: Tephritidae), and A. obliqua, large flies presented greater fecundity (Fitt 1990; Liedo et al. 1992).

Pupal size parameters differed markedly according to sex between populations. However, in A. ludens we observed differences in the pupae shape. Wild pupae were longer, whereas laboratory pupae were wider and heavier. These differences in the pupae shape also explain that classification criteria discriminated more strongly in the laboratory than in the wild population. In A. obliqua, wild pupae were bigger than laboratory pupae in all size parameters.

Fig. 3.

Canonical analysis for pupae size parameters of males and females of laboratory and wild populations in Anastrepha obliqua. The canonical analysis is represented on the first canonical axis (can 1), where the boxplots indicate the populations (laboratory and wild) and sexes (males and females) (left). The variables of pupae size: length (mm), width (mm), and weight (mg) are indicated by vectors (right).

In wild populations, body size differences may be influenced by the host fruit and the season of the yr, whereas in laboratory populations, size differences probably were due to larval density in artificial diets, diet formula, texture, and rearing temperatures (Navarro-Campos et al. 2011; Shelly 2018). Although the factors that affect body size were not studied here, it is important to consider that differences among individuals contribute to disparities in body size, whereas differences in populations may be related to genetic or environmental factors, and their interactions (Chown & Gaston 2010; Nijhout et al. 2014). Environmental conditions can strongly affect the degree but not the direction of sexual size dimorphism within species. Female size appears to be more sensitive to environmental conditions than male size (Teder & Tammaru 2005). Although much of the body size variations are due to body size plasticity, it is recognized that many of the variations are adaptive, and therefore inherently linked to biological functions and behavior (Stillwell et al. 2010).

All measured variables were correlated highly with the pupae classification criteria (diam), thus verifying the feasibility of using the pupae sorting machine for size characterization, quality control, and selection of males as a colony management strategy to improve the performance of sterile males (McInnis 1987). Furthermore, we found evidence of the increase in size in mass-reared populations and reduction in the variation of pupal size.

Morphological differences between laboratory and wild populations have been documented in adult fruit flies (Gómez Cendra et al. 2014; Sánchez-Rosario et al. 2021). Although body size is a trait with genetic plasticity, it is also known that mass-rearing conditions represent selection agents that can lead to evolutionary changes in traits related to size.

This research constitutes a reference in the attempts to explore and characterize the potential differences between laboratory and wild populations in an immature stage, and that could be related to the differences described in adults. Therefore, our results can be used as a baseline in the quality control protocols in mass-rearing programs for these species of fruit flies. More research is needed to analyze the effects of mass-rearing at the morphological level, as well as its potential effects on the performance of mass-reared sterile insects.