Trade-offs among life history traits are of key interest in evolutionary biology due to their potential impacts on maintaining phenotypic variation or constraining adaptation under changing environments (Reznick and Bryga 1987; Futuyma and Moreno 1988). On one hand, trade-offs may act as evolutionary constraints for adaptation among individuals; on the other hand, long-term selection may overcome such constraints and reshape trait combinations (Agrawal 2020). Therefore, investigating whether trade-offs scale across different levels of biological organization, from individuals to populations and species, will shed light on the evolutionary processes underlying life history traits (Bolnick et al. 2003; Agrawal 2020). For example, if trait divergence between species originates from trade-offs at the intraspecific level, we would expect to find trade-offs among traits both within and between species. Alternatively, if recently diverged species evolve independently in response to divergent environmental conditions, trade-offs might only be detected at the interspecific level due to environmental constraints while not being evident at the intraspecific level. Our study addresses this potential connection between inter- and intraspecific variation by testing whether there is a trade-off between dispersal capacity and reproductive output in two sister species of gall wasps belonging to the genus Belonocnema, which exhibit different dispersal capabilities. We also examined whether such trade-offs exist within each monomorphic species.
We discovered that trade-offs between flight capability and reproductive output are evident as a consistent difference between the two wasp species, but no trade-offs were detected at the intraspecific level. Specifically, B. fossoria exhibits reduced flight capability and smaller wing size, while demonstrating significantly higher reproductive effort compared to B. treatae (Fig. 4). Our finding of an interspecific trade-off, but the absence of intraspecific trade-offs, aligns with previous studies exploring trade-offs between life history traits such as flower size and flower number (Worley et al. 2000; Sargent et al. 2007; Hahn and Maron 2016). One possible explanation for the lack of intraspecific trade-offs in these studies is that individual variation in resource acquisition masks the correlation among traits. Individuals with greater resource availability may develop both larger and more numerous flowers, whereas individuals with limited resources but larger flowers would have fewer flowers (van Noordwijk et al. 1986; King et al. 2011; Agrawal 2020). However, in the case of our wasp species, individual resource acquisition variation is unlikely to explain the absence of trade-offs between flight-related traits and reproductive output. This is because our analysis already partially controls for the effect of body size, which indirectly accounts for individual resource variation (Fig. 7). An alternative explanation is that these two wasp species have recently diverged and specialized within a narrow niche (only one or two host plant species), resulting in long-term and strong selection that has shaped divergence in resource allocation between the species. Within each species, considering the narrow niche, further specialization along these life history dimensions may not be expected, thereby leading to the absence of trade-offs among life history traits (Futuyma and Moreno 1988). We posit that the discrepancy in trade-off patterns between intra- and interspecific levels is more likely to occur among closely related species inhabiting different environments compared to species inhabiting similar environments. This is based on the common assumption that phylogenetically closely related species tend to share similar life history traits and evolutionary constraints, resulting in a consistent trade-off pattern within and among sister species (Webb et al. 2002). However, if a particular life history trait plays a crucial role in adaptation to divergent environments, long-term and intense selection pressure could disrupt the genetic constraints on the evolution of life history traits (Agrawal 2020). Further studies employing a comparative framework similar to this study are needed to systematically test this hypothesis.
For the two wasp species studied here, B. fossoria and B. treatae, the divergence in life history traits is not solely due to changes in body size but also involves alterations in the relationship between body size and multiple traits, including wing size, abdomen size, and total reproductive effort (Fig. 5). This means that the adjustments in flight capability and reproductive effort, as depicted by Scenario 4 in Fig. 1, are independent of body size. Specifically, the independent adjustment of flight capability between the sister species is reflected in variation in wing size, while changes in reproductive effort are achieved through modifications in abdomen size. The independent adjustment of these life history traits is further supported by a common garden rearing experiment: the divergence in the linear relationship between body size and wing size, as well as between body size and reproductive effort, is maintained even when the wasps are reared on a non-native host (Fig. 6). Therefore, the divergence in life history traits is likely attributable, at least in part, to genetic differences between the two wasp species rather than being solely a plastic response to the rearing host plants (i.e., environmental effects and phenotypic plasticity).
Interestingly, the linear relationship between body size and wing size differs in intercepts, while the linear relationship between body size and other reproductive-related traits, including abdomen volume, potential fecundity, total reproductive effort, differs in both intercepts and slopes (Table 1, Fig. 5). According to quantitative genetic theory, linear relationships between traits are more prone to change in intercepts under selection, while stronger and longer-term selection is required for changes in slopes (Roff et al. 2002). Thus, it is possible that reproductive traits experience stronger selection pressure than flight-related traits. Notably, similar changes in both intercepts and slopes between potential fecundity and body size have been observed in the alternating asexual and sexual generations of another species in the genus, B. kinsey (Hood and Ott 2017), where each generation faces unique environmental challenges. Hence, both Hood and Ott (2017) and the present study provide evidence that reproductive traits are highly adaptable phenotype under selection in this group of Belonocnema gall wasps. Since dispersal and reproduction often exhibit trade-offs either within species or among species across many study systems (Carroll et al. 2003; Guerra 2011; King et al. 2011; Nasu and Tokuda 2021), more studies should link changes in linear relationship among these life history traits to the variation of selection pressures each of these traits experience (e.g., populations at the range edge).
We argue that B. fossoria individuals with higher reproductive output but lower flight capability may gain a fitness advantage in the Q. geminata environment, leading to selection favoring the divergent phenotypes between B. fossoria and B. treatae. This is because B. fossoria's host plant, Q. geminata, is typically found as short shrubs no taller than 10 m in small clusters in sandy soil, whereas B. treatae's host plant, Q. virginiana, is a much taller tree species (with a mean height of ~ 21m) and has a patchier distribution (Cavender-Bares and Pahlich 2009). Dense, short trees may require less flight capability for the asexual generation of B. fossoria to locate suitable sites for gall induction on the roots, which host the next generation (Zera and Denno 1997). The reduced flight capability observed in B. fossoria, is likely a product of relaxed selection on flight associated with more persistent, shorter, and smaller host trees, as suggested by other studies highlighting the role of habitat structure in the evolution of flight capability (Roff 1990; Roff 1994; Denno et al. 1996; Zera and Denno 1997; Dell'Aglio et al. 2022). Additionally, a third allopatric species in this genus, B. kinseyi, which shares host plant Q. virginina with B. treatae, exhibits life history traits similar in size to the B. treatae populations analyzed here. These include similar flight capability (Zhang et al. 2021c) and correlations between body size and wing length (wasp species: t = -0.03, p = 0.981; body size x wasp species: t = -1.04, p = 0.303) and between body size and abdomen length (wasp species: t = -2.88, p = 0.103; body size x wasp species: t = 0.56, p = 0.581; see Supplement). Therefore, the divergent phenotypes between B. treatae and B. fossoria are likely the result of adaptation to divergent host plant-related environments. Future studies should apply a similar comparative work to other gall wasp species that also utilized the same set of host plant species Q. virginiana and Q. geminata. If similar divergent phenotypes are observed repeatedly between host-associated populations in other gall wasp species, this would strongly suggest the similar host plant-related environments play a critical role in shaping the divergence among these traits (Egan et al. 2013; Zhang et al. 2019, 2022).