Population structure, the physical arrangement of related and unrelated individuals, can have profound effects on the apparent outcrossing rate ([Formula: see text]) in a plant population. However, detailed experimental investigations of the impact of population structure on [Formula: see text] are few. We compared the apparent outcrossing rates of experimental populations of grain sorghum with seed families spatially arranged in stratified and overdispersed treatments. Using alcohol dehydrogenase allozymes as genetic markers, [Formula: see text] was calculated for each of the treatments at two locations over three years. For all six comparisons, the overdispersed treatment yielded significantly larger apparent outcrossing estimates than the stratified treatment. In one case, the difference was over five-fold. Site-specific and time-specific differences were small compared to treatment differences. Whether natural structuring plays a role in altering the effective outcrossing rates of natural populations has been addressed by only a few descriptive studies; structuring effects appear to have an impact in only about half of such studies. The sample is still too small to make any generalizations.Population structuring can also be of significance in plant breeding programs. Controllable variation in [Formula: see text] values of the magnitude reported herein may be useful in optimizing selection methods for quantitative characters in experimental plant breeding populations. Further work is under way to determine the effects of the variation in apparent outcrossing rates on genetic gains from selection.

Ninety‐three accessions of primitive domesticated diploid wheat (Triticum monococcum L.), five each of two wild tetraploid wheats (T. turgidum L. var. dicoccoides Körn and T. timopheevii Zhuk. var. araraticum Jakubz.), one cultivated durum wheat (T. turgidum L. var. durum Desf. ‘Modoc’) and one cultivated bread wheat (T. aestivum L. var. aestivum ‘Anza’) were compared for plant height, seed weight, flour protein content, flour lysine content, lysine content in protein, spike weight, days to head, and days to flower under irrigated and dryland conditions in 1977. Superior lines were retested in 1979. Variation among lines for each trait in different species was significant except plant height in dicoccoides and araraticum. Over all species, plant height and seed weight were significantly higher, and protein and lysine contents were significantly lower in irrigated than in dryland conditions, although differences within dicoccoides for seed weight and protein and lysine contents were nonsignificant. Two accessions of monococcum were short and headed and flowered as early as Modoc and Anza. Primitive and wild wheats were higher in protein content and lysine content but lower in spike weight and seed weight than the two modern cultivars. Wild tetraploids were higher in flour protein content and in flour lysine content but lower in lysine content in protein than monococcum. Accessions of dicoccoides and araraticum having 30.9 and 30.5% protein, respectively, were identified. Correlations between traits were influenced by species and by moisture conditions. Best lines of monococcum, dicoccoides, and araraticum were identified as sources of genes for different traits.

There is an interest in breeding for deep, extensive root systems to increase crop yields in semiarid environments. Our objective was to develop a field technique with which large numbers of genotypes could be tested for this trait. The experiments were conducted with cowpea (Vigna unguiculata [L.] Walp.) grown on stored soil moisture in the field (coarse‐loamy, mixed, thermic Haplic Durixeralf). The herbicide, metribuzin, was banded into the root zone at sowing at specific depths and lateral distances from the seed rows. In the main field trial, lateral distances of 46, 61, and 76 cm were tested in combination with vertical herbicide bands of 30 and 15 cm at average vertical depths of 53 and 60 cm, respectively, and with herbicide rates of 4.5 and 9.0 kg active ingredient ha−1 in the band. Plant leaves were observed for herbicide symptoms. With this technique, we succeeded in detecting the progress of root growth in the field. Herbicide symptoms consistently developed soonest in plants which were closest to the herbicide band, in either the horizontal or vertical direction. We also succeeded in detecting significant genotypic differences in mean numbers of days to first herbicide symptoms among five cowpea genotypes; California Blackeye #5 and Grant developed symptoms the earliest, 8006 and PI302457 developed symptoms the latest, and PI293579 was intermediate. With four replicates of two‐row plots, mean genotypic differences of 2 to 3 days were usually significant at the 5% level. The ranking of genotypes obtained with this herbicide‐band technique was consistent with estimates of relative depth of effective rooting obtained from soil moisture extraction measurements. Prior to using this technique in different environments, or to screen other crop species, tests should be conducted to determine the most appropriate herbicide, rate of application, and band location in the soil profile. This technique requires fairly uniform soil throughout the root zone, and relatively uniform genetic material. We conclude that this herbicide‐band technique can be used to screen large numbers of genotypes in the field for the presence of rapid root growth. For short season crops, rate of root growth in the first 2 months in the field should be a good predictor of final extensiveness and depth of rooting.