An investigation of the impact of urban riparian forest composition and diversity on carabid assemblages

Urbanization impacts native ecosystems directly through replacement of native communitieswith urban infrastructure and indirectly through habitat fragmentation and degradation. Urban forests develop from the survival of remnant trees, shrubs, and herbaceous species, and the intentional planting and unmanaged reproduction and regeneration of native and non-native species. These forests represent novel ecosystems comprised of new species compositions and structures new abiotic conditions compared to historic vegetation. Because humans have suppressed fire and eliminated most grazing, the two disturbances that limited shrub growth and maintained herbaceous understory growth, many such understories are now dominated by woody vegetation. One of the biggest threats to urban forests is invasion by non-native plant species, which can alter decomposition rates, nutrient cycling, and regeneration. Urban forests generally have lower integrity and resilience than more natural or rural counterparts. Understanding how and to what extent urbanization has impacted forest composition and structure is vital because these forests continue to provide a diverse array of ecosystem services. This is especially true for small, urban headwater streams because of their strong influence on downstream water quality and ecosystem processes. Our study area included the upland urban riparian forests of the cities of Rock Island and Moline, IL. We sought to answer the following research questions: (1) What is the variation in overstory and understory plant community composition and diversity, and especially levels of invasive species dominance? and(2) What is the relative importance of the fine-scale land use and other factors driving observed patterns of community composition and diversity? (3) How are these patterns in community composition reflected in the abundance and diversity of carabid abundance? In 2015, we sampled 19 sites across an urbanization gradient. Aerial photographs, soil surveys, and other remote sensing data were used to stratify the urban forests and capture as much variation in the factors thought to be important drivers of forest composition as possible. Each of the 19 study sites consisted of a 25m radius circular plot. Composition of the mature and sapling forest overstory was measured using the point quarter and understory using the line-point intercept methods, respectively. Carabid species abundance and diversity was measured within these study sites from June-August using 10 pitfall traps 20m apart. We present our preliminary findings on the forest understory community. We identified a total of 182 species consisting of 121 native and 61 non-native species. Floristic Quality Index (FQI) ranged from 6.36 to 22.98, and invasive cover ranged from 0 to 69%. Hierarchical agglomerative cluster analysis identified four groups of forest sites that differed in the composition of the understory community (MRPP, A = 0.14, p < 0.0001) and combined abiotic (patch size, nearby levels of impervious surface, bare soil, heat loads, and soil depth) and biotic conditions (shade tolerant tree density) (MRPP, A= .19, p < .0001). first group consisted of 7 sites within the largest forest patches, with least nearby impervious surface (<5%), but highest levels of bare soil (>20%) and densities of shade tolerant species.They were characterized by the lowest levels of invasive species (7.7%) but also the lowest FQI (13.1). The second and third groups consisted of 11 sites within the smallest patches characterized by the most nearby impervious surface (>50%) but lowest levels of bare soil (<5%). They were characterized by the highest levels of invasive species, 49% and 26% cover, and intermediate FQIs (17.3, 16.8). They differed primarily in Loniceraspp. (honeysuckle) dominance. The fourth group consisted of 4 sites spanning a range of patch sizes with intermediate levels of impervious surface. They were characterized by intermediate levels of invasive species, 21% cover, but the highest FQI (19.0). When compaired with these groups we found a significan correlation to changes in carabid beetle assemblages suggesting that patterns in fine-scale land use are impacting multiple trophic levels.