• O‘Connor, Constance M.
  • Barthel, Brandon L.
  • Gilmour, Kathleen M.
  • Philipp, David P.
  • Van Der Kraak, Glen
  • Cooke, Steven J.


Using a long-term study population of wild smallmouth bass Micropterus dolomieu in a connected river-lake system, we investigated whether circulating glucocorticoid (cortisol) and androgen (testosterone) concentrations are influenced by reproductive investment and nesting environment in fish providing nest-guarding paternal care. For all individuals, we collected measures of reproductive history and the value of current parental care. We assessed nest environment and monitored individuals to quantify seasonal reproductive success. Finally, we measured circulating cortisol concentrations following a standardized stressor and circulating testosterone concentrations. Using general linear models, we found that poststress circulating cortisol concentrations were positively related to water temperature and were higher in fish nesting in the river than in the lake. Circulating testosterone concentrations were negatively related to water temperature and were higher in reproductively experienced fish. When considering the factors that influence reproductive success, we found that only parental size was positively related to current nest success. In summary, the results demonstrate that nesting environment is correlated with parental stress responses during parental care, while reproductive history and nesting environment are correlated with circulating androgen concentrations. Collectively, these results offer insight into the roles of both glucocorticoid and androgen steroid hormones during parental care in teleost fish.


Study Site and Animals

The study was conducted in a short reach of the Mississippi River of Ontario (Frontenac County, Ontario, Canada; 44°57′N, 76°43′W). For a full description of the study site, see Barthel et al. (2008). In brief, the study site is composed of distinct upstream riverine (Mississippi River) and downstream lacustrine (Miller’s Lake) habitats separated by a <1-m waterfall that is not a barrier to smallmouth bass movement. The study site is bordered upstream by a >2-m waterfall that is a barrier to smallmouth bass and downstream by a series of rapids and waterfalls. Radio telemetry studies have not detected fish movement out of the study site in either direction (Barthel et al. 2008). Although smallmouth bass show high nest-site fidelity to either the river or the lake during the reproductive period, most fish overwinter in the lake (Barthel et al. 2008). The study site is relatively isolated, and the system receives negligible fishing pressure.

Selection of Sampling Protocol

The fish used in this study were important as individuals within a long-term (>15 yr) study population on the reproductive ecology of smallmouth bass. To minimize the effect of our sampling protocol, we were limited to collecting only a single blood sample per fish. To understand the information provided by a single blood sample, we first obtained both baseline and poststress cortisol and testosterone values from a separate population of fish. These fish were originally captured as part of a separate study; for full details, please refer to O’Connor et al. (2011b). Briefly, fish were captured while guarding fresh eggs and immediately sampled for blood by caudal puncture. They then were subjected to a 3-min standardized capture-and-restraint protocol followed by a 25-min recovery before being sampled for blood again by caudal puncture. The 3-min standardized capture-and-restraint protocol, blood sampling protocols, and the handling and assaying of blood samples for cortisol concentration were all identical to the procedures used for this study and are fully described below. For the purposes of this study, the relationships between baseline and poststress cortisol and testosterone concentrations were explored using regression models to assess the relationship between baseline and poststress steroid concentration and the relationship between poststress steroid concentration and the steroid response to the standardized stressor (i.e., the difference between poststress and baseline concentrations). Only values obtained from unmanipulated control smallmouth bass providing care to eggs <24 h old were used ().

These data demonstrated that in smallmouth bass providing parental care to fresh eggs, baseline circulating cortisol concentrations were consistently low with little variation (fig. 1A). The very low variation in baseline circulating cortisol concentrations may be the result of a “floor” effect of the 3-ng mL−1 detection limit of the cortisol assay (O’Connor et al. 2011b). Poststress cortisol concentration was unrelated to baseline circulating cortisol concentration (fig. 1A) but was an accurate measure of the cortisol stress response (i.e., the change in cortisol concentration from baseline to poststress; fig. 1B). On the other hand, poststress testosterone concentrations more closely reflected baseline testosterone concentrations (fig. 1C) than the ability of parents to maintain testosterone concentrations following a stressor (fig. 1D). Thus, being limited to a single sample, we opted to collect only a poststress sample because it provided an accurate measure of the cortisol stress response (fig. 1B) and a reasonable measure of baseline circulating androgen concentration (fig. 1C).