Periodic regression in long-term agricultural experiments

Publisher Website
Google Scholar
Abstract
Evans & Roberts (1979) considered that loworder polynomials such as mean, linear and quadratic were adequate to describe the trends that occur in treatment contrasts in long-term agricultural experiments. This conclusion was based on experience with experiments on levels of fertilizers and rates of stocking on similar pastures. In such experiments, the marked seasonal pattern in the data disappears when comparisons between treatments are followed. However, further experience with experiments comparing cultivars has shown that some comparisons still show a seasonal pattern. Such patterns may be accounted for by periodic regression (Bliss, 1970).