Surface-sterilized and nonsterilized seeds from eight commercial cultivars of
cotton (Gossyp < /em>ium barbadense L.) were examined for qualitative and quantitative
estimates of seedborne fungi. The observed fungi were Altemaria altemata.
Asp < /em>ergillus f1avus, Aspergillus niger, Aspergillus spp., Cephalosporium sp.,
Cladosp < /em>orium sp., Drechs/era spp., Fusarium moniliforme, Fusarium oxysporum,
Fusarium semitectum, Fusarium solani, Fusarium sp., Nigrospora oryzae, Penicillium
spp., Rhizoctonia solani, Rhizopus stolonifer, Trichoderma spp. and Trichothecium
roseum. The quantitative estimates of the fungi showed that A. niger (71%),
Penicillium sp. (34%) and Cladosp < /em>orium sp. (25.63%) were the most dominant fungi
isolated from the nonsterilized seeds. Other fungi occurred at frequencies ranged from
0.13 to 22.50%. The isolation frequencies of A. niger, Cephalosp < /em>orium sp.,
Cladosporium sp., and T. roseum were significantly decreased by surface sterilization,
while the isolation frequencies of the other fungi were not affected. These results may
suggest that A. niger, Cephalosporium sp., Cladosporium sp., and T. roseum tended
to colonize the outer seed coat more than embryos, while the other fungi tended to
colonize the internal parts of the seeds. Cultivar and cultivar x treatment interactions
were all very highly significant or significant sources of variation in frequencies of the
isolated fungi except F. oxysporum. Cultivar was the first in importance as a source of
variation in frequencies of 6 (40%) of the isolated fungi, while cultivar x treatment
interaction was the first in importance as a source of variation in frequencies of
5(33.33%) of the isolated fungi. No single cultivar yielded all the 18 fungi. Giza 70
yielded the highest number of fungi (14 fungi), while Giza 85 yielded the lowest
number (9 fungi). The other cultivars yielded a number of fungi ranged from 10 to 13.
A. altemata, A. niger, F. moni/iforme and Penicillium sp. were the only fungi, which
were isolated from all the tested cultivars. The present study showed that the role of
seedborne fungi of cotton, as seedling disease incitants, was more evident in the
preemergence stage compared with the postemergence stage. Pearson correlation
coefficient was calculated to evaluate the degree of association among 153 pairs of
the isolated fungi. Eleven (7.19%) of the fungal pairs were significantly associated. Of
the 11 pairs, 9 were positively associated, and 2 were negatively associated. No
significant associations were found in the remainder fungal pairs. Cluster analysis
divided the isolated fungi into distinct groups. One group consisting of Drechs/era
spp., F. solani, Cephalosporium sp., F. semitectum, A. altemata, F. moniliforme, A.
niger, Cladosporium sp., A. f1avus, Fusarium sp., Penicillium spp. and Trichoderma
spp., and a second group consisting of R. solani, T. roseum, Aspergillus spp., N.
oryzae, R. stolonifer, and F. oxysp < /em>orum. Within each group, fungi were associated
strongly and positively, whereas between groups, fungi were associated weakly or
negatively. This result implies the potential existence of cultivar related groups of
fungi. Four regression models, derived from stepwise multiple regression analysis,
were constructed to describe the effect of the isolated fungi (independent variables)
on seedling disease variables (dependent variables). These models showed that
differences in seedling disease variables were due largely to the effects of N. oryzae,
F. semitectum, R. stolonifer, R. so/ani, and Trichoderma spp. It is worth noting that no
regression model was constructed to predict postemergence damping-off, which
reconfirms that soilborne fungi of cotton are more important, as seedling disease
incitants, in the preemergence stage compared with the postemergence stage.