Below is a Ph.D. dissertation abstract describing research conducted at the University of Minnesota. Publications on this research have appeared in Ecology of Freshwater Fish and Transactions of the American Fisheries Society, and others are in preparation (see publications). Future research will examine trout production dynamics in tailwater rivers.

Dissertation Reference

Kwak, T.J. 1993. Influence of physical and biotic factors on trout production dynamics in southeastern Minnesota streams. Doctoral dissertation. University of Minnesota, St. Paul.

Objectives

The production dynamics of brook trout (Salvelinus fontinalis ), brown trout (Salmo trutta ), and rainbow trout (Oncorhynchus mykiss ) were studied in southeastern Minnesota streams during 1988-1990 to identify and describe physical and biotic influences. Primary objectives were to: (1) identify physical or chemical factors that best explain the variation in salmonid production dynamics and population parameters of other fishes, (2) develop empirical models to predict fish density, biomass, and production, (3) examine density-dependent effects on growth and mortality and other biotic interactions, and (4) evaluate stream habitat development projects. A secondary objective was to develop microcomputer software to estimate fish population parameters, production rates, and associated variances.

Methods

Fish populations were estimated using a mark-recapture (salmonids) or removal method (nonsalmonids), and production was estimated by the instantaneous rate of growth method. All estimates and variances were computed using Pop/Pro Modular Statistical Software, which was developed specifically for this research and is available to interested colleagues.

Results and Implications

Fish assemblages sampled from 15 stream reaches were abundant, but low in diversity. Thirteen species were collected, and mean parameters (and range) among reaches were species richness: 4.1 (1-8); density: 29,490 (1,247-110,602) fish/ha; and biomass: 253.5 (49.6-568.6) kg/ha. Means for salmonids among reaches were annual mean density: 2,279 (343-8,096) fish/ha; annual mean biomass: 162.0 (32.5-355.5) kg/ha; and annual production: 155.6 (36.7-279.6) kg/ha. The salmonid annual production to mean biomass (P/B) ratio averaged 1.06 (0.64-1.42).

Fish population parameters were similar between the two years of study (P > 0.05), but age-0 salmonid density in fall estimates was significantly lower during the second year (P = 0.02), presumably due to spring flooding. This finding suggests that segments of a population may be variably impacted by stochastic events (e.g., flooding and air temperature), but overall population parameters are less likely to be impacted. Water temperature and chemical composition were monitored monthly, and instream and riparian habitat was surveyed during summer. Stream habitat and water quality were variable depending on local geology, geomorphology, and land use. All measures of fish populations (density, biomass, and production) were generally much greater in reaches where habitat had been artificially developed, compared to reference reaches on the same streams.

Significant variation in salmonid density, biomass, and production (P < 0.05) was explained by models incorporating attributes of water depth, substrate composition, and stream bank condition with reduced populations associated with shallow water, fine substrates, and eroded out-sloped banks. Percent eroded bank was the best single predictor of salmonid production. Water temperature was negatively correlated with salmonid density and biomass and positively related to P/B ratio, and results suggested that annual maximum water temperatures of 19-20 C were optimum for salmonid production with an upper thermal limit of about 24 C. Nonsalmonid density and biomass were related to instream habitat and positively to water temperature, and biomass was positively correlated with pH. Models describing variation in salmonid biomass and production and nonsalmonid biomass were also developed using principal components derived from physical variables.

Three habitat indices developed by other investigators as predictors of fish populations were evaluated in this study. The WRRI trout cover rating (WCR) and the qualitative habitat evaluation index (QHEI) yielded significant models of salmonid production, and the QHEI also significantly predicted salmonid density and biomass. Model II of the habitat quality index (HQI) failed to correlate with any fish population parameter.

In general, multiple regression models and those incorporating composite variables accounted for more variance and were more significant than simple regression models. Measures of water temperature and bank condition most commonly described variability in fish populations among streams, but the trends related to temperature were converse for salmonids and nonsalmonids, indicating differences in fish ecology. Cluster analyses also suggested ecological differences among species and showed geographic variation in water temperature and quality within the region

Salmonid biomass was positively associated with amphipod relative abundance. Salmonid P/B ratio differed among species and could be estimated with a linear model based on the number of age classes present. A strong negative correlation between salmonid biomass and P/B ratio suggested that growth may be related to density. Further exploration revealed that growth through fall was density dependent for age-0 brook trout and brown trout (growth negatively related to density), and growth of older fish was significantly different between years. Mortality of age-2 brown trout was negatively related to density and may be owing to angling impacts.

The physical and biotic influences on salmonid production dynamics were summarized in conceptual growth-survivorship curves, where the scale and bounds of biotic processes are set by the physical environment. Spatial and temporal scale should be considered in future application and development of models. Management efforts for these streams should focus on improvement in riparian and watershed land use in combination with artificial habitat development techniques to achieve the short- and long-term goals of mitigation, protection, and enhancement of these productive and dynamic ecosystems.

 Return to Tom Kwak's home page.

 Return to North Carolina Cooperative Fish and Wildlife Research Unit home page.

 Return to Department of Biology home page.