Thermally induced chronic developmental stress in coho salmon: Integrating measures of mortality, early growth and fluctuating asymmetry

Developmental stability, or homeostasis, facilitates the production of consistent phenotypes by buffering against stress. Fluctuating asymmetry is produced by developmental instability and is manifested as small random departures from bilateral symmetry. Increased fluctuating asymmetry is thought to parallel compromised fitness, in part, because stress promotes energy dissipation. Compensatory energy expenditures within the organism are required to complete development, thus promoting instability through reductions in homeostasis. Increased heterozygosity may enhance developmental stability by reducing energy dissipation from stress through increased metabolic efficiency, possibly by providing greater flexibility in metabolic pathways. Traditionally, fluctuating asymmetry has been used as a bioindicator of chronic stress, provided that selective mortality of less fit individuals did not reduce stress-mediated increases in fluctuating asymmetry to background levels produced by natural developmental error, or create data inconsistencies such as higher asymmetry in groups exposed to lower stress. Unfortunately, absence of selective mortality and its effects, while often assumed, can be difficult to substantiate. We integrated measures of early growth, mortality, fluctuating asymmetry (mandibular pores, pectoral finrays, pelvic finrays, and gillrakers on the upper and lower arms of the first branchial arch) and directional asymmetry (branchiostegal rays) to assess chronic thermal stress (fluctuating temperatures as opposed to ambient temperatures) in developing eggs from two different coho salmon (Oncorhynchus kisutch) stocks and their reciprocal hybrids. Hybridization provided insight on the capacity of heterozygosity to reduce stress during development. Although egg losses were consistently higher in crosses exposed to fluctuating temperatures, egg mortality was predominantly a function of maternal stock of origin. Post-hatch losses were higher in crosses exposed to ambient temperatures than in crosses exposed to fluctuating temperatures during embryogenesis. Observed patterns of early growth revealed no heterosis, but instead reflected maternal effects, with some crosses slowing growth and yolk utilization when exposed to fluctuating temperatures. Analyses of fluctuating asymmetry also showed no effects from heterosis. While analyses of composite asymmetry scores and branchiostegal rays were inconclusive, analyses of individual characters showed significantly higher fluctuating asymmetry in pelvic finray counts and a marginal change in the numbers of fish asymmetric for this character in crosses exposed to chronic thermal stress. In contrast, the fluctuating asymmetry in lower gillraker counts was significantly higher in crosses exposed to ambient temperatures and there were significantly more fish asymmetric for this character. Data on mortalities and fluctuating asymmetry indicate pelvic finray development was thermally stressed, while the heightened fluctuating asymmetry in lower gillraker counts under ambient temperatures was due to a greater frequency of less fit fish that had not been culled by thermal stress. Changes in early growth patterns in response to developmental stress yielded no parallel responses in meristic characters. We conclude that chronic thermal stress produced both selectively lethal and sublethal effects that directly shaped fluctuating asymmetry and fitness profiles in these crosses. Implicit in this conclusion is that developmental instability analyses can detect more than just chronic sublethal stress, thus providing substantial credence for using instability studies as proactive bioassessment methodologies.