Conference on Applied Statistics in Agriculture

MULTIVARIATE STATISTICAL ANALYSIS OF TERRESTRIAL INVERTEBRATE INDEX OF BIOTIC INTEGRITY

The Index of Biotic Integrity (IBI) is designed to measure the changes in ecological and environmental conditions as affected by human disturbances. In practice, the IBI is used in various ecological applications to detect divergence in biological integrity attributable to human actions. Last year during this conference, methodologies for developing an Avian Index of Biotic Integrity (A-IBI) were presented and discussed. The objective of this paper is to demonstrate the construction and statistical evaluation of a multi-metric terrestrial Invertebrate Index of Biotic Integrity (I-IBI) using the same multivariate statistical techniques. Canonical correlation analyses were utilized to select pertinent invertebrate metrics as impacted by vegetation and hydrology variables. The resulting invertebrate metrics were then ranked, according to a pre-specified scale of human disturbance, and the I-IBI scores were subsequently computed. The multivariate model, as well as the final I-IBI scores were statistically validated using independent temporal data sets. Furthermore, the sensitivity of I-IBI values was determined in order to access their reliability. The techniques are demonstrated using five years of invertebrate survey data collected from the terrestrial environments within the historic fifty-year floodplain of the Kootenai River in Northern Idaho.