Application of statistical techniques to modeling crop growth.

摘要

Crop models that incorporated genetics, climate, and management operations may be useful tools to make better decisions throughout the growing season for turnip (Brassica rapa L.), mustard (Brassica juncea L.), sweet and field corn (Zea mays L.). Experiments were conducted near Gainesville, Florida, to test plant populations and N fertilizer on turnip and mustard; to determine the effect of cowpea (Vigna unguiculata (L.) Walp.) mulch on sweet corn yield, plant nutrition, and nematode populations, as well as the effect of nematodes on yield; and to develop dry matter growth models over time for field corn as affected by genotypes and planting dates. Variance, correlation, linear and non-linear regression, surface response, ridge, and longitudinal analysis were used to optimize management decisions.; Ridge analysis predicted that a combination of 168 kg N ha-1 and more than 6 plants m-2 will produce maximum mustard and turnip shoot as well as total biomass yields. Sweet corn yield responded to increasing cowpea mulch, peaking at rates corresponding to at least 201 kg N ha-1. Quadratic and linear-plateau models agreed with the critical N levels reported by others and could be used as tools to select N fertilizer requirements. Diagnostic leaf area and weight as well as N, K, Zn, and Na concentrations of the leaf increased as total and fancy ears increased and all could be good predictors of sweet corn yield. Fancy ear yield reached maximum levels around 63 days after planting (DAP) and between 4.4 and 6.6 kg of cowpea mulch m-2. Nematode population also peaked at 63 DAP but there was little impact of nematodes on ear yield. Analysis showed that the N-sufficient corn plants were able to sustain greater nematode densities compared to N-deficient plants.; For field corn, genotype and planting date affected total dry matter accumulation over the life cycle of the field corn. At early planting (March) 'Pioneer 3320', a temperate hybrid, had greater yield than 'Pioneer X304C' and FLOPUP. Delaying planting until May favored 'Pioneer X304C', a tropical hybrid. When planted in August FLOPUP, an open pollinated variety developed under Florida conditions, performed better than the other corn hybrids. FLOPUP had more resistance to late season diseases, especially compared to Pioneer 3320.