Application of Longitudinal Analysis on Repeated Measures in a Nutrient Runoff and Leaching Experiment.

Experiments involving repeated measures in agricultural and environmental research are becoming common. One essential feature of this kind of data is that measurements of the same individuals are taken repeatedly through time; hence they may be positively correlated, which invalidates the assumption of independence among the measurements. New developments in the analysis of longitudinal data allow the analyses of within-individual differences (changes over time) and cross-sectional differences (changes among treatments) to be performed simultaneously in a parsimonious way using a mixed linear model. This presentation will illustrate the benefit of longitudinal analysis on a dataset from an experiment to minimize nutrient runoff and leaching from stacks of manure. This dataset has measurements of the runoff volume and concentration of nutrients for 12 stacks after each major rainfall event in four treatments. The longitudinal analysis could let us test the changes of nutrients in leachate among the four treatments and within stacks (the experiment unit) over time. Similar analysis can be used for data from long-term field studies with repeated measures of plots during different seasons.