Influence of local and landscape scale factors on the presence, relative abundance and characteristics of brook trout (Salvelinus fontinalis) in beaver ponds

Abstract

Study area : Mackenzie River watershed, northeast of Thunder Bay, Northwestern Ontario (drains into Lake Superior)

Beaver ponds are a characteristic component of northern Ontario stream ecosystem s and display a great deal of variation through their natural cycle of establishment and abandonment. Ponds also provide habitat for brook trout (Salvelinus fontinalis) within small-stream systems. However, the habitat characteristics of ponds used by brook trout, and the characteristics of brook trout in the ponds is poorly understood. I evaluated the local and landscape scale habitat characteristics of beaver ponds that are associated with the presence, relative abundance and physical characteristics of brook trout. Brook trout were captured in 40% o f 50 beaver ponds sampled. Angling proved to be the most reliable method to sample brook trout in ponds, however, catchability appeared to be strongly influenced by temperature. Catch per unit effort (CUE) was significantly higher in ponds with water temperatures within the approximate preferred thermal range of brook trout (11 ° C to 18 ° C) (ANOVA , F 2.48 = 5.259, p = 0.026). Peak CUE occurred between approximately 14 °C to 18 °C . Beaver ponds with brook trout present were generally characterized by greater upstream catchment area (UCA), lower water temperature, higher dissolved oxygen, higher conductivity, higher alkalinity, and greater depth. Brook trout were never captured in ponds with an upstream catchment area (UCA ) less than 2.9 sq. km. In a logistic regression model, UCA correctly predicted brook trout presence and absence in beaver ponds (82.4%). In beaver ponds with a mean water temperature greater than 11 ° C, predicted group membership using UCA was greater (92.9%). Model parsimony and predictive ability increased when beaver ponds with an overall mean water temperature greater than 11 °C were used in the analysis. Brook trout captured in beaver ponds were, on average, 105 mm and 72 g larger than those in adjacent streams (Pi,i32i = 1658.2, P < .001). Brook trout average size in ponds was larger in smaller UCAs, whereas, brook trout in streams were larger in larger UCAs. My research illustrates that beaver ponds provide complimentary and/or supplementary habitat for brook trout living in stream systems. It is clear that pond characteristics at both a landscape and local scale are associated with brook trout presence or absence, likely through the ability of brook trout to colonize the pond and the suitability and stability of habitat within the pond. Further research is required to better understand the linkage between local and landscape scale characteristics influencing brook trout habitat and abundance in small headwater streams and insure the protection of these linkages from disturbance.