How Systematic Age Underestimation Can Impede Understanding of Fish Population Dynamics: Lessons Learned from a Lake Superior Cisco Stock

Abstract

Systematic underestimation of fish age can impede understanding of recruitment variability and adaptive strategies (like longevity) and can bias estimates of survivorship. We suspected that previous estimates of annual survival (S; range = 0.20‐0.44) for Lake Superior ciscoes Coregonus artedi developed from scale ages were biased low. To test this hypothesis, we estimated the total instantaneous mortality rate of adult ciscoes from the Thunder Bay, Ontario, stock by use of cohort‐based catch curves developed from commercial gill‐net catches and otolith‐aged fish. Mean S based on otolith ages was greater for adult females (0.80) than for adult males (0.75), but these differences were not significant. Applying the results of a study of agreement between scale and otolith ages, we modeled a scale age for each otolith‐aged fish to reconstruct catch curves. Using modeled scale ages, estimates of S (0.42 for females, 0.36 for males) were comparable with those reported in past studies. We conducted a November 2005 acoustic and midwater trawl survey to estimate the abundance of ciscoes when the fish were being harvested for roe. Estimated exploitation rates were 0.085 for females and 0.025 for males, and the instantaneous rates of fishing mortality were 0.089 for females and 0.025 for males. The instantaneous rates of natural mortality were 0.131 and 0.265 for females and males, respectively. Using otolith ages, we found that strong year‐classes at large during November 2005 were caught in high numbers as age‐1 fish in previous annual bottom trawl surveys, whereas weak or absent year‐classes were not. For decades, large‐scale fisheries on the Great Lakes were allowed to operate because ciscoes were assumed to be short lived and to have regular recruitment. We postulate that the collapse of these fisheries was linked in part to a misunderstanding of cisco biology driven by scale‐ageing error.

Can pelagic forage fish and spawning cisco (Coregonus artedi) biomass in the western arm of Lake Superior be assessed with a single summer survey?

“…This range of biomass levels was similar to that obtained for the same regions by Mason et al (2005) in 1997 (7.9-24.4 kg ha −1 ). Obtaining comparable estimates to Mason et al (2005) is not overly surprising given (1) the dominant fauna by weight (cisco) can live up to age-21 in Lake Superior, and (2) strong year-classes of this species recruited in Lake Superior during 1998 and 2003 (Yule et al, 2008). estimated biomass of omul in Lake Baikal over differ- Rudstam et al (1987) also had higher cisco biomass levels than we measured (Trout Lake 51 kg ha −1 and Lake Mendota 223 kg ha −1 ), while one was lower (Big Muskellunge 2.4 kg ha −1 ).…”

“…Biomass of spawningsize cisco, however, is best estimated during November when pre-spawning aggregations form. Yule et al (2008) demonstrated that November acoustic and midwater-trawl surveys could be combined with commercial catch data to estimate exploitation levels at a given locale, provided that the fishery-independent survey work occurs after the bulk of spawners have returned to that site. However, given the time needed to process the survey results, the data cannot be used to set harvest quotas until the next fishing season.…”

Using pectoral fin rays as a non-lethal aging structure for smallmouth bass: precision with otolith age estimates and the importance of reader experience

“…Accurate age data coupled with other fisheries metrics can be used to estimate growth, mortality rates, population structure, and change in species abundance, biomass, and yield. However, inaccurate age estimation can lead to gross errors in stock metrics (e.g., recruitment, growth, and mortality) and adversely affect management strategies (Mills and Beamish 1980;Beamish and McFarlane 1983;Lai and Gunderson 1987;Leaman and Nagtegaal 1987;Tyler et al 1989;Reeves 2003;Yule et al 2008). Therefore, it is crucial that fisheries managers and biologists use the best available methods to estimate age.…”

The Global Status of Freshwater Fish Age Validation Studies and a Prioritization Framework for Further Research

“…If age information is unreliable, population models used for prediction of population dynamics may result in the implementation of liberal catch limits and the potential for overharvest. For instance, Yule et al (2008) suggested non-reliable ageing structures resulted in faulty age information and inaccurate harvest models with the subsequent over-harvest and depletion of cicso (Coregonus artedi, Salmonidae) population abundances and a collapse of the fishery in Lake Superior, USA. As such, fisheries professionals need reliable information on the true ages of organisms of interest.…”