Biodiversity Analysis of Selected Riparian Ecosystems within a Fragmented Landscape Prepared by: Reuben R Goforth, PhD, David Stagliano, Yu Man Lee, Joshua Cohen, Michael Penskar Michigan Natural Features Inventory PO Box 30444 Lansing, MI 48909-7944 For: Michigan Great Lakes Protection Fund and Michigan Department of Environmental Quality, Office of the Great Lakes December 2002 Report Number 2002-26 Biodiversity Analysis of Selected Riparian Ecosystems within a Fragmented Landscape Prepared by: Reuben R Goforth, PhD, David Stagliano, Yu Man Lee, Joshua Cohen, Michael Penskar Michigan Natural Features Inventory PO Box 30444 Lansing, MI 48909-7944 For: Michigan Great Lakes Protection Fund and Michigan Department of Environmental Quality, Office of the Great Lakes December 2002 Report Number 2002-26 ABSTRACT Riparian ecosystems of the Great Lakes Basin influence the quality of the Great Lakes and provide habitat for many characteristic elements of biodiversity within the region Extensive human landscape modifications have dramatically changed the character of terrestrial and aquatic ecosystems in Michigan, especially in Lower Michigan, where riparian ecosystems are among the only remaining contiguously forested areas within highly fragmented landscapes The significance of these isolated riparian ecosystems for maintaining regional biodiversity in a highly fragmented landscape is not fully understood Historically, these areas have been poorly inventoried, and only a few elements of biodiversity are locally well known This study was initiated to gain a better understanding of the biodiversity refuge potential of riparian corridors within fragmented landscapes Our approach was unique in that we surveyed multiple elements of both terrestrial and aquatic communities, including plants, natural communities, breeding birds, amphibians and reptiles, and multiple aquatic taxa We used multivariate statistics to determine whether these community parameters were patterned among riparian corridors with varied levels of riparian forest width and connectivity Overall, the results of this study provided some support for the idea that biodiversity refuge potential of riparian corridors within fragmented landscapes can be predicted based solely on corridor width and contiguity, primairly with respect to terrestrial flora and some vertebrate groups However, aquatic community parameters were much more responsive to varied channel types than to riparian corridor widths Spatial analysis of land cover properties of local and upstream riparian buffer areas provided an additional level of correlation analysis for riparian community components and multi-scale environmental properties of landscapes These multi-spatial analyses identified some strong associations between community measures and upstream properties, suggesting that riparian biodiversity modeling and management may need to be conducted at larger spatial scales in order to be effective While the overall results of this study did not wholly support the sole use of riparian corridor width and contiguity as guiding factors for identifying riparian biodiversity potential in fragmented landscapes of southern Lower Michigan, further study that includes appropriate criteria for determining the integrity of streams with varied channel characteristics may lead to more definitive models of riparian biodiversity that do provide greater evidence for the use of riparian corridors as broad scale models for prioritizing conservation targets within landscapes Contents INTRODUCTION 1 METHODS 2 Study Areas 2 Aquatic Community Surveys 2 Vegetation and Floristic Surveys 5 Terrestrial Vertebrate Surveys 7 Spatial Analysis 8 Statistical Analysis 9 RESULTS 9 Overall Results 9 Aquatic Community Results 13 Vegetation and Floristic Results 26 Overall Vegetation and Floristic Sampling Summary 26 Natural Community and Rare Species Occurrences 26 Adventive Plant Species 35 Vegetation and Floristic Parameters 35 Vegetation and Floristic Responses to Varied Riparian Forest Buffer Width Classes 36 Vegetation and Floristic Responses to Varied Channel Types 36 Vegetation and Floristic Sampling Results by Zone 46 Terrestrial Vertebrate Results 47 Spatial Analysis Results 54 Aquatic Community Spatial Analysis Results 54 Terrestrial Community Spatial Analysis Results 64 DISCUSSION 64 Summary 64 Aquatic Community Discussion 67 Terrestrial Vertebrate Discussion 70 Spatial Analysis Discussion 72 ACKNOWLEDGEMENTS 73 LITERATURE CITED 74 APPENDICES 80 Riparian Ecosystems Phase II - i Riparian Ecosystems Phase II - ii List of Tables Table 1 Riparian class and channel type designations for riparian study areas sampled in 2000 and 2001 Channel types include shallow (A), moderately incised (B) and deeply incised (C) 4 Table 2 Natural community (C), animal (A), invertebrate (I) and rare plant (P) occurences documented during 2000 and 2001 riparian ecosystem surveys Riparian width classes (<125m, 125-250m and 250-500m) and channel types (A, B and C) are indicated for each study site Rivers include the Grand (GR), Kalamazoo (KZ), Raisin (RR), St Joseph (SJ), Pine (PR), Maple (MR), Looking Glass (LG), Red Cedar (RC), Shiawassee (SH), Thornapple (TR) and Sycamore Creek (SC) 11 Table 3 Summary of habitat, fish, mussel and benthic macroinvertebrate community indices for 27 riparian sites in southern Lower Michigan sampled in 2000 and 2001 Indices include Habitat Quality Index (HQI), total aquatic species richness (TASR), fish species richness (FSR), relative abundance of intolerant fish (RAIF), fish catch per unit effort (FCPUE), fish IBI (FIBI), benthic invertebrate species richness (BNSR), Ephemeroptera, Plecoptera and Trichoptera Index (EPT), benthic invertebrate biotic index (INBI), relative abundance of intolerant benthos (RAIB), mussel species richness (MSR), relative abundance of intolerant unionids (RAIU), relative abundance of tolerant unionids (RATU) and Mussel Biotic Tolerance Index (MBTI) Increasing values for HBI, FIBI, INBI and EPT reflect greater biological integrity, while larger MBTI scores reflect greater community tolerance to degraded environmental conditions RAIF, RAIB and RAIU are expected to increase with increasing site ecological integrity, while RATU values are expected to increase with increasing levels of environmental degradation at a site Channel types include shallow (A), moderately incised (B) and deeply incised (C) 15 Table 4 Spearmans Rank correlation coefficients (R) and two-tailed statistical significance values (p) for correlations between aquatic community descriptors of the 27 riparian survey sites Correlations with p<0005 are highlighted in light gray Correlations for autocorrelated data are indicated in dark gray Community descriptors include Habitat Quality Index (HQI, Barbour et al 1999), fish species richness (FSR), fish Index of Biotic Integrity (FIBI, Karr 1981) relative abundance of intolerant fish (RAIF), fish catch per unit effort (FCPUE), mussel species richness (MSR), relative abundance of intolerant unionids (RAIU), relative abundance of tolerant unionids (RATU), mussel catch per unit effort (MCPUE), mussel biotic tolerance index (MBTI), benthic species richness (BNSR), Benthic Invertebrate Biotic Index (INBI), relative abundance of intolerant benthos (RAIB), Ephemerotpera, Plecoptera and Trichoptera Index (EPT), and total aquatic species richness (TASR) 27 Table 5 Floristic and ecological variables measured at riparian study sites, including total number of plant species (TSP), total number of native plant species (TNPS), total number of adventive plant species (TAPS), percent of all species as native species (%Native), percent of all species as adventive species (%Adventive), Floristic Quality Index (FQI), Coefficient of Conservatism (COC), number of ecological zones (#Zones) and coefficient of topo graphic variation (CTV) River basins include the Grand, (GR), Kalamazoo (KZ), Raisin (RR), St Joseph (SJ), Pine (PR), Shiawassee (SH), Looking Glass (LG), Red Cedar (RC), Maple (MR), and Thornapple (TR) Rivers and Sycamore Creek (SC) Riparian forest buffer width classed include <125m, 125-250m and 250-500m 31 Table 6 Weighted means for ecological variables measured at 27 riparian forest sites River basins sampled include the Grand, (GR), Kalamazoo (KZ), Raisin (RR), St Joseph (SJ), Pine (PR), Shiawassee (SH), Looking Glass (LG), Red Cedar (RC), Maple (MR), and Thornapple (TR) Rivers and Sycamore Creek (SC) Riparian forest buffer width classes include <125m, 125-250m and 250-500m 37 Table 7 Summary of floristic parameters by buffer width class and channel type Parameters include the total number of plant species/site (TPS), total number of native plant species (TNPS), total number of adventive plant species (TAPS), Floristic Quality Index (FQI), and Coefficient of Conservatism (COC) 38 Table 8 Summary of vegetation measures by buffer width class and channel type Parameters include basal area, number of tree species/plot (NTS), diameter at breast height by prism plot (DBH), number of woody stems/plot (USSt), number of understory species/plot (USSp), number of ground cover species/plot (GCS) and the percentage of ground cover/plot (%GC) 38 Table 9 Summary of vegetation measures according to channel type Parameters include basal area (BA), number of tree species/plot (NTS), diameter at breast height by prism plot (DBH), number of woody stems/plot (USSt), number of understory species/plot (USSp), number of ground cover species/plot (GCS) and the percentage of ground