MEPS 526:101-112 (2015)  -  DOI: https://doi.org/10.3354/meps11204

ABSTRACT: Condition (the relationship between individual weight and length) has been researched in fisheries science for over 100 yr and is claimed to be an integrated measure of physiological status for fishes. Spatial or temporal variation in condition can contribute to otherwise unexplained variation in the relationship between spawning biomass and recruitment. Individual condition is also included in age-structured population models, which use weight at age to convert population estimates between numbers and biomass. However, no study has analyzed spatial and temporal variation in condition for multiple marine species. Here I apply recent improvements in spatial modeling to analyze coastwide variation in condition for 28 groundfishes in the California Current. I show that, on average, 22% of individual-level variation in condition can be explained via persistent (constant over time) and annually varying spatial differences in condition, and condition for many species varies 10 to 20% spatially and among years. While population density, bottom temperature, and calendar date are parsimonious descriptors of condition in several species, the sign of these coefficients varies, and their magnitude is small relative to the magnitude of residual spatial and temporal variation. Additionally, annually varying spatial differences have nearly twice the magnitude of persistent spatial differences in condition. I therefore conclude that dynamic habitat conditions contribute a substantial portion of variation in individual condition for these groundfishes. Spatial and temporal variation in condition will be important for population models that convert between numbers, fishery catch, and population biomass, and may also clarify unexplained variability in productivity for marine fishes.

KEY WORDS: Relative condition factor · Spatial analysis · Weight at length · Northeast Pacific groundfish · State-dependent life history theory · Fish growth · Density dependence · Gaussian random fields