Evaluating the relationship between floristic quality and measures of plant biodiversity along stream bank habitats

Measures used to describe the floristic structure of a habitat can vary in their ability to express trends in plant composition along anthropogenic disturbance gradients. This study was based on a survey of vascular plant biodiversity performed along stream bank habitats within an agricultural landscape in southeastern Ontario, Canada. The accuracy of several measures of plant biodiversity – including those related to a regional floristic quality assessment system – was examined to compare their ability to recognize a gradient of anthropogenic disturbance and associated floristic quality along the stream bank habitats. The floristic quality assessment system is a scheme in which all vascular plants of a region have been assigned a score corresponding to a qualitative conservation value based on habitat fidelity and tolerance of disturbance (native species), and on invasiveness (non-native species). Data were collected from a priori designated disturbed, moderate, and pristine zones along 27 stream sections exhibiting a length-wise disturbance gradient. A detrended correspondence analysis (DCA) was used to isolate the plant compositional gradient present along the stream sections. The measures of plant biodiversity recorded in the different study sites were then ranked by the degree to which they were linearly correlated with the identified compositional gradient of the DCA. The “% non-native plant species” measure was most effective at expressing the gradient, though it incorporated nothing about the fidelity and sensitivity of native plant species present in individual zones. Several measures associated with the floristic quality assessment system – including the mean coefficient of “conservatism” (mCC) – were also effective in identifying the gradient, and had the additional benefit of considering the contribution of each native species in a plot. The simple measure of “total plant species richness” proved to be a poor linear indicator due to a quadratic trend across the whole of the compositional gradient. The floristic quality assessment system proved to be a valuable tool for assessing conservation values of the selected sites. It should be extended to include further regions in Canada and North America in general. Our results further suggest that stream banks associated with open non-crop agricultural property are highly susceptible to colonization by non-native upland plants and species of low conservation interest, and that the presence of wooded areas surrounding these same streams is associated with higher numbers of native and disturbance sensitive plant species present in the bank habitats.