Resistance to the reniform nematode (RN) (Rotylenchulus reniformis Linford and Oliveira) was studied in three population sets involving moderately resistant strains (La RN-910, Aub 612-RNR, M 019-PNR), each of which was crossed to a common susceptible strain, 'Deltapine 41' (Dp 41), of upland cotton (Gossypium hirsutum L.). Each population set consisted of six generations (P(1), P(2), F₁, F₂, BC(1)P(1), BC(1)P(2)), two randomized blocks, and two planting dates.

Plants were grown in the greenhouse in individual pots holding 500 g of sterile soil-sand (50:50) mixture. Plots consisted of 10 plants/block for non-segregating generations (P(1), P(2), F₁), 20 plants/block for backcross generations, and 40 plants/block for F₂ generations. Plants were inoculated at first true-leaf stage with 2000 RN juveniles/pot and incubated for 43 (winter) or 32 (summer) days after inoculation. At termination of each experiment, plants were severed at the soil line, the root ball was thoroughly soaked in water, the soil was carefully washed from the roots, and the roots were blotted dry and weighed. Eggs were extracted from roots by soaking and agitating in a 0.26% solution of sodium hypochlorite. Eggs suspended in the sodium hypochlorite solution were collected onto a 500-mesh screen and backwashed into a beaker standardized to contain a 50-ml suspension. A 10-ml aliquot was extracted from the 50ml suspension, and RN eggs were counted with the aid of a microscope (15X). Number of eggs/gram of root (EPGR) was calculated by dividing eggs/plant by root weight.

The genetics of RN resistance were studied by evaluating frequency distribution of individual plant data by generations and by use of the Generation Mean Analysis (GMA) procedure whereby the genotypic values of RN EPGR for each experiment were partitioned into additive (a), dominance (d), and epistatic (aa, ad, dd) gene effects, following Gamble's method. Population means for each cross were calculated from individual plant data obtained from two blocks in each planting date as well as combining data over dates. Plants within blocks were included in estimates of error variance.

Significant differences among generations, high coefficients of variation, and non-discrete frequency distributions in segregating populations indicated that differences between parents for RN resistance in 2 of 3 crosses (La RN-910 X Dp 41 and Aub 612-RNR X Dp 41) were under genetic control and inherited in a quantitative manner. No pattern was observed for the significance of additive and dominant gene effects, but significant epistatic gene effects occurred in both cases. Estimates of epistatic gene effects plus evidence of transgressive segregation for susceptibility indicated that parents of both crosses differed by two or more pairs of genes governing RN resistance. The direction of d and aa effects in the La RN-910 X Dp 41 cross was negative and that of dd effects was positive, whereas direction of these effects was opposite in the Aub 612-RNR X Dp 41 cross. Differences in direction of these gene effects suggest that La RN-910 and Aub 612-RNR have different genetic mechanisms for RN resistance and that their combination through breeding may lead to increased resistance to this nematode. Environmental influence must be reduced if selection for RN resistance is to be effective. Increasing the level of RN inoculum, the use of growth chambers, or limiting studies in the greenhouse to spring and fall may help. Advancing generations to F₄ or F(5), while maintaining genetic variability prior to selection, may improve selection efficiency for RN resistance by reducing nonadditive gene effects through increased homozygosity. The general lack of additive gene effects in these lines suggest that different sources of RN resistance should be explored in future genetic and breeding studies.