An Algorithm for Fitting Seinhorst Curves to the Relationship between Plant Growth and Preplant Nematode Densities.

A data base representing nematode counts and soil weight from 1,936 individual soil cores taken from a 7-ha alfalfa field was used to investigate sample optimization for five plant-parasitic nematodes: Meloidogyne arenaria, Pratylenchus minyus, Merlinius brevidens, Helicotylenchus digonicus, and Paratrichodorus minor. Sample plans were evaluated by the accuracy and reliability of their estimation of the population and by the cost of collecting, processing, and counting the samples. Interactive FORTRAN programs were constructed to simulate four collecting patterns: raudom; division of the field into square sub-units (cells); and division of the field into rectangular sub-traits (strips) running in two directions. Depending on the pattern, sample numbers varied from 1 to 25 with each sample representing from I to 50 cores. Each pattern, sample, and core combination was replicated 50 times. Strip stratification north/south was the most optimal sampling pattern in this field because it isolated a streak of fine-textured soil. The mathematical optimmn was not found because of data range limitations. When practical economic time constraints (5 hr to collect, process, and count nematode samples) are placed on the optimization process, all species estimates deviate no more than 25 % from the true mean. If accuracy constraints are placed on the process (no more than 15% deviation from true field mean), all species except Merlinius required less than 5 hr to complete the sample process.