Assessing fish biodiversity in northeastern Ontario drainage basins: methodological and landscape effects

Abstract

Inland waters and their fish communities constitute a valuable resource for Canada. Long-term monitoring of fish communities in Ontario supports both fisheries management and biodiversity assessment. Traditionally, fish monitoring programs have relied on gill net surveys of sportfish-dominated larger lakes (≥ 50 ha) which may not adequately sample some aquatic habitats or fish species, and thus underestimate biodiversity. The issue is important in northeastern Ontario, where historical acidification led to significant loss of fish resources. A more thorough and robust assessment of the recovery of fish communities in this region may require a multi-gear and multi-habitat approach. My research objectives were to examine the efficiency of current gill net community surveys as a fish biodiversity assessment tool, and to explore patterns in fish biodiversity in relation to landscape characteristics at multiple scales in the historic acid deposition zone of northeastern Ontario. My first chapter compares fish species richness estimates from Broad-scale Monitoring (BsM) surveys with those derived from alternate protocols, and a multi-gear approach (BsM surveys plus alternate sampling protocols) designed to sample habitats and species that are believed to be poorly sampled by gill nets. Sampling was carried out on 30 lakes in the historical acid-damaged zone of northeastern Ontario. Species detected increased asymptotically with sampling effort. Maximum species detection (Smax) differed greatly among gear types, but no individual gear was able to effectively sample the full fish community. In contrast, the effort required to achieve 80% of Smax was quite similar among individual gears. Integrating all individual gears (Combo protocol) detected significantly more species than the BsM protocol at higher sampling effort (≥18 efforts). BsM surveys detected 73.8 ± 15.4% (mean ± 1 SD) of all fish species detected in my study lakes, often missing species that use extreme nearshore and tributary habitats. BsM detection efficiency was negatively related to abundance of smallmouth bass. My second chapter examines patterns in fish biodiversity within and among six

drainage basins covering a gradient of historical acidification damage (three acid- recovering, three reference) in northeastern Ontario. Fish biodiversity was higher in

reference drainage basins than acid-recovering drainage basins; this trend was most evident when biodiversity was measured as species richness and Shannon’s diversity, and less evident when measured as Simpson’s diversity. Differences in fish biodiversity between reference and acid-recovering drainage basins were also more pronounced when analyzed with BsM survey data than with nearshore and tributary Gee trap data. Biodiversity estimates tended to be higher in large sentinel lakes located at lower elevations in the drainage basin. Fish biodiversity in lakes was most strongly linked to contemporary pH, but latitude, relative elevation and lake size were also significant factors in some analyses. My study provides a critical evaluation of the role of the BsM program in fish biodiversity assessment, provides an updated and more comprehensive approach to examining status and recovery of fish communities, and will inform targeted conservation efforts in north temperate landscapes.