Fishes affect aquatic nutrient cycling both by storing important nutrients for long periods and by releasing bioavailable nutrients in their wastes. Evolution of nutrient-rich traits impacts these processes both by increasing nutrient storage in the body and by altering dietary nutrient acquisition and nutrient release in wastes. In this study, we evaluate how evolution of bone, a phosphorus (P) rich trait, influences P uptake, storage, and release by threespine stickleback (a well-known model species for bone evolution).

We analyzed how intraspecific bone variation along an environmental gradient affected the P-specific elemental phenotype of stickleback (i.e., nutrient acquisition from the diet, nutrient absorption in the gut, nutrient assimilation into tissues, nutrient storage in tissues, and nutrient excretion and egestion). We measured these traits in 5 stickleback populations within a single drainage. We hypothesized that those with more bone would store more P in their bodies and so would need to compensate for increased P demands by increasing dietary P intake or intestinal P absorption. We therefore predicted that increasing bone content would increase dietary P intake, whole-body %P, and P absorption but would decrease P egestion and P excretion.

Contrary to our predictions, we found that although increasing bone content increased whole-body %P, bone content did not clearly impact diet, absorption, or egestion. Furthermore, P excretion actually increased with whole-body %P, contradicting the fundamental predictions of ecological stoichiometry. We also found that although P egestion was unrelated to whole-body %P, all stickleback showed higher fecal %P than expected, suggesting their demand is relatively low. Our results instead suggested that stickleback absorb relatively more C and N in their guts. These effects varied depending on site conditions and sex.

Our results suggest that bone content increases whole-body %P without increasing overall P demand, perhaps decoupling whole-body %P and P release in wastes. This potentially makes bonier vertebrates an important source of recycled P. Our overall findings therefore support the vertebrate bone hypothesis but counters the traditional predictions of ecological stoichiometry.  

We collected stickleback from five populations in the Cowichan drainage and measured their morphological characteristics, some bone characteristics, and the P content of bodies, bones, diet, and wastes. We also measured excretion in a subset of individuals.

1. Morphology: We measured stickleback morphology using the methods used by Durston and El-Sabaawi (2018). See: https://doi.org/10.1111/1365-2435.12919 
2. We measured %P using the acid-molybdate method. 
3. We measured excretion using incubation methods (2h incubation time). See: https://aslopubs.onlinelibrary.wiley.com/doi/abs/10.4319/lom.2009.7.1
4. We measured foregut %P and hindgut %P using foregut-hindgut analyses. See: https://aslopubs.onlinelibrary.wiley.com/doi/10.1002/lom3.10516#
5. We estimated bone mineral content with the proxy "bone %P."