4 CONCLUSIONS
A number of general themes emerge from these data. First, they provide clear support for the hypothesis that foraging environment influences the genotype-phenotype map for craniofacial skeletal traits (Parsons et al., 2016; Navon et al., 2021; Zogbaum et al., 2021). More specifically, our data suggest that pelagic foraging “drives” species- and environment-specific DE. This may seem counterintuitive as diets that involve large/hard prey items are generally considered to be the more mechanically demanding compared to small/soft food (Muschick et al., 2011; Gunter et al., 2013; Hulsey et al., 2020). However, Navon et al (2020) showed that in MZ bone matrix was deposited at a fast rate under pelagic foraging conditions, and speculated that suction feeding imposes mechanical load on the feeding apparatus as animals repeatedly open and protrude their jaws. Our data support this assertion, and thus we consider the foraging treatments utilized here to challenge the feeding apparatus in two distinct ways (compared to a “standard” flaked food diet); our benthic treatment was designed to impose high amplitude but low frequency loading onto the feeding apparatus as animals scrapped food from rocks, whereas our pelagic treatment translated to higher frequency but lower amplitude loading as animals repeatedly protruded their jaws to gather small food items.
Our data also detected evidence for genetic assimilation. In particular, when considering loci that were DE between species + environments, patterns in MZ benthic fish resembled those across TRC. Tropheopsspecies, including sp. “red cheek”, are generally found in a benthic environment (Ribbink et al., 1983), and may have lost a degree of plasticity as they evolved to specialize on benthic food items. MZ on the other hand are true generalists in the sense that they routinely foraging from both the benthic and pelagic zones (Ribbink et al., 1983). While plasticity has been noted in TRC (Parsons et al., 2014; Navon et al., 2020), our data suggest that MZ may be more plastic than TRC in that they mount a more pronounced transcriptional response, at least at the time point analyzed in this study.
Cell cycle regulation consistently appeared in GO analyses, describing species differences, as well as plasticity within MZ. This implicates cell proliferation as an important biological mechanism of species- and environment-specific bone growth in cichlids. This observation is notable as our previous work has implicated Hedgehog signaling in the evolution and plasticity of the cichlid jaw, including the IOP-RA complex (Hu & Albertson, 2014; Parsons et al., 2016; Navon et al., 2020). While canonical members of the Hedgehog signaling pathway were not significantly DE or DA in this dataset (although KIAA0586regulates the signal, Schock et al., 2016), cell proliferation is well-known to be regulated by this pathway (St. Jacques et al., 1999; Tiet et al., 2006; Sun & Deng, 2007; Zaman et al., 2019), providing a potential cellular mechanism through which variation in Hedgehog signaling leads to differences in bone shape among and within cichlid species.
Finally, with these large overlapping genome-wide datasets, we were able to narrow down thousands of DEGs to roughly two dozen that were both DE and DA. Given that each experiment was conducted at a different time point, this reduced dataset points to loci whose expression is important for species divergence over extended periods of time. Among these were genes that were both sensitive and robust to the environment. Notably, nearly all of these genes are new to the field of bone biology, and while some encode known effectors of well studied signaling pathways (e.g., interleukin/Wnt, Talpid/Hh) and cell behaviors (e.g., Casp6/apoptosis, Impdh1b/cell-cycle), others implicate largely novel mechanisms (e.g., Gnmt/methionine cycle). Thus, this work establishes a robust foundation for future studies into how genotype and the environment combine to influence bone formation, remodeling, and evolution.