Houghton, D.C. 2002. Biodiversity of Minnesota caddisflies (Insecta: Trichoptera). Ph.D. Dissertation, University of Minnesota. 194 pp.
No Abstract for Chapter 1: INTRODUCTION
Abstract for Chapter 2: DELINEATATION AND CHARACTERIZATION OF MINNESOTA CADDISFLY REGIONS
Over 300,000 adult caddisfly specimens representing 224 species were collected and analyzed from samples of 248 Minnesota aquatic habitats within 58 watersheds. Detrended Correspondence Analysis and a flexible unweighted pair group method using arithmetic averages (UPGMA) dendogram of caddisfly relative abundance data determined that five regions of caddisfly exist within the state. Species richness and diversity were significantly highest in the Lake Superior and Northern regions, lowest in the Northwestern and Southern regions, and intermediate in the Southeastern region. Canonical Correspondence Analysis determined that caddisfly species composition was related to temperature, percentage of disturbed habitat, and stream gradient. A strong correlation between temperature and percentage of disturbed habitat made determination of the relative importance of those variables difficult. Caddisfly species richness in small and medium sized streams correlated negatively with percentage of disturbed upstream habitat in the Northern, Northwestern, and Southern regions. Change in the composition of trophic functional groups based on habitat type generally followed a pattern predicted by the river continuum concept in the Lake Superior, Northern, and Southeastern regions, whereas fine particle filtering collectors dominated all sizes of lakes and streams in the Northwestern and Southern regions. Although the absence of historical data makes it difficult to separate the relative importance of natural and anthropogenic factors, loss of caddisfly biodiversity and homogenization of feeding ecology has probably occurred in at least the Northwestern and Southern regions due to human disturbance.
Abstract for Chapter 3: EVALUATION OF MINNESOTA GEOGRAPHIC CLASSIFICATIONS BASED ON CADDISFLY DATA
The ability to partition the variation of faunal assemblages into homogenous units is referred to as classification strength (CS). In this study, the CSs of three types of geographic classifications: watershed basin, ecological region, and caddisfly region, were compared based on 254 light trap samples of adult caddisflies collected in Minnesota during 1999 2001. The effect on CS of three different levels of taxonomic resolution: family, genus, and species, was also assessed. Primary (broadest possible) a priori classifications by watershed basin and ecological region had a lower CS than did secondary classifications by these regions. Caddisfly region, an a posteriori classification based directly on caddisfly distribution data, had nearly twice the CS of the a priori classifications. CS decreased approximately 20% with a decrease in taxonomic resolution from species to genus, and from genus to family. These results suggest that geographic classification, spatial scale, and taxonomic resolution are all important factors to consider when sampling aquatic insects.
Abstract for Chapter 4: PHENOLOGY AND HABITAT AFFINITIES OF THE MINNESOTA CADDISFLIES (INSECTA: TRICHOPTERA)
Over 300,000 caddisfly specimens were examined from 317 light trap samples collected during 1999 - 2001, and from museum and literature records. Two hundred eighty four caddisfly species were determined to occur in Minnesota, representing 20 families and 74 genera. The distribution and relative occurrence in different regions, habitat types, and levels of upstream disturbance was documented for each species. The majority of Minnesota species appeared to be univoltine, with adult emergence and flight occurring in June and July. There were, however, other species that were detected in the spring, fall, winter, or during multiple seasons. Fifty three percent of all specimens were represented by eight species, whereas 81 species were known from <10 specimens. Seven of the 13 species with official protective status in Minnesota were located during 1999 2001. Two were more widespread than previously thought. Binomial regression determined that 16 species were statistical indicators of particular habitats and site conditions in Minnesota. Presence of three species appeared to indicate habitat disturbance, particularly the disturbance of lakes and small streams. Presence of the remaining 13 species was related to both latitude and habitat disturbance. These species may be valuable as indicators of undisturbed habitats, but their response to latitude must first be separated from their response to habitat disturbance.