Mysid population responses to resource limitation differ from those predicted by cohort studies

Abstract

Effects of anthropogenic stressors on animal populations are often evaluated by assembling vital rate responses from isolated cohort studies into a single demographic model. However, these models are difficult to translate into ecological predictions because stressor effects observed in isolated cohorts may differ from those occurring in populations with overlapping generations. This problem is evident in many areas of stressor-response research, including the burgeoning work on ocean acidification. To address this problem, we compared vital rates in experimental cohorts and populations of the mysid Americamysis bahia at 4 levels of resource limitation. This required development of a novel observational scheme that allows inverse estimation of stage-specific vital rates in mixed-age populations without the use of cohort isolation or marking. Using digitally imaged time series observations of stage abundances, the most strongly supported inverse models indicated opposing (i.e. compensatory) effects of resource limitation on adult survival and juvenile maturation. The model with adult survival response only (i.e. no compensation) was also strongly supported. This contrasts with cohort results, where feeding effects on fecundity were strongest. These results suggest that emphasis in stressor-response studies on early life stages and cohorts of uniformly aged individuals may miss important demographic responses and should be augmented by observations of intact populations, especially as methods such as ours become more available. KEY WORDS: Inverse demography · Zooplankton · Population model · Cohort · Survival · Americamysis bahia Resale or republication not permitted without written consent of the publisherMar Ecol Prog Ser 432: [115][116][117][118][119][120][121][122][123] 2011 now accumulated from such studies and has enabled meta-analytic reviews of general life history patterns and vital rate responses to large-scale stressors (e.g. Forbes et al. 2010, Hendriks et al. 2010. It is unclear whether the conclusions from these reviews are affected by reliance on cohort studies and matrix substitution analyses. If so, implications for prediction of extinction risk, loss of biodiversity and ecosystem change are significant. In our study of the mysid Americamysis bahia, we compared results from this substitution approach with analyses of observations on whole populations, where matrix models were parameterized using inverse estimation from stagestructured time series observations. Our comparison focused on demographic responses to resource limitation, which is a common complication in applied population ecology.Mysids are peracarid crustaceans that include >1000 species from freshwater and marine environments. Americamysis bahia (formerly Mysidopsis bahia) is commonly used as a test organism in ecotoxicology (reviewed in Verslycke et al. 2007), so substantial effort has been devoted to the standardization of laboratory culturing methods and production of uniformly conditioned individuals (Lussier et al. 1...

Translating crustacean biological responses from CO2 bubbling experiments into population‐level predictions

“…Food resource competition was eliminated by supplying Artemia spp. (Leach) nauplii at surplus feeding rates based on previous studies (Grear et al 2011). Neonates were obtained from laboratory cultures at AED using the spawning chamber technique described in Lussier et al (1988).…”

“…Using digital imaging techniques described in Grear et al (2011), I obtained complete census counts of mysid cohorts on days 7, 14, 21, and 28. Neonates were present in counts that occurred on days 21 and 28 (i.e., at the ends of weeks three and four) and were easily distinguishable from the study cohorts because of the difference in size.…”

“…Importantly, neonates were removed after counting to ensure they were not misidentified as neonates in the following week's count. This is standard practice in cohort studies of mysids and other similar organisms and has the obvious effect of preventing population growth and reducing inter-age interactions (e.g., cannibalism; Grear et al 2011) and competition.…”

“…The population experiment was initiated in July 2010 using the same setup, census intervals, and observation methods as in the cohort study, except that: (1) each tank was initiated with 25 individuals of mixed ages in proportion to stable age distributions observed in previous mysid studies (Grear et al 2011); (2) there were three replicate population tanks per treatment; (3) populations were censused for a longer period; (4) neonates were not removed; and (5) length measurements were obtained each week from the digital images. Note that in contrast to the cohort experiments, abundance was not controlled in the population experiments, so initiation with 25 individuals and eventual population fluctuations caused differences in spatial density, and possibly competition, between tanks.…”

“…For the cohort experiment, I scaled up to the population level using the common practice of assuming that agespecific fecundity and survival are the same in isolated cohorts as in whole populations (the ''cohort scaling assumption''), and then inserting these rates into a matrix population model. For the population experiment, I used the inverse demographic methods described in Grear et al (2011), which allow estimation of demographic rates within intact populations and eliminate the need for the cohort scaling assumption. The two experiments resulted in what I refer to as a cohort-derived and a population-derived matrix model, which I then subjected to LTRE analysis to explore the population-level consequences that can be inferred from the two experimental approaches.…”