Masked trends in field data: nematode population assessments as an example uri icon

abstract

  • Nematode population and plant yield data from a field experiment were subjected to a range of analyses in an attempt to reveal underlying trends and relationships. Rank correlation proved to be a useful technique for preliminary indication of relationships. Several moving-average techniques were convenient, but introduced significant bias. A robust, locally-weighted regression procedure, although computationally intensive, was the most useful tool for investigating the nature of the relationships among plant-parasitic nematode populations at different densities, and between crop yield and nematode population density. In a field in which cotton and cowpea were grown in rotation, cotton yield was more strongly related to total numbers of plant-parasitic nematodes than it was to numbers of any individual species. There were indications of incompatible or antagonistic associations among the plant-parasitic nematode species, especially at higher population densities. These interactions would not have been revealed by more conventional analyses, and may provide a basis for hypothesis formulation and additional experimentation.

publication date

  • 1993
  • 1993
  • 1993