Bb. Pendleton et al., PREDICTING SORGHUM MIDGE (DIPTERA, CECIDOMYIIDAE) GENERATIONS AND ABUNDANCE, Journal of economic entomology, 87(4), 1994, pp. 993-998
A total of 39,687 assessments of abundance of ovipositing sorghum midg
e, Contarinia sorghicola (Coquillett), and sorghum, Sorghum bicolor (L
.) Moench, spikelets in flower were used to evaluate, refine, and vali
date a sorghum midge population dynamics simulation model. Number of s
pikelets per panicle was related positively to panicle length, with 42
% of the variability in the number of spikelets attributable to length
. Uniformly growing fields of sorghum were in flower less-than-or-equa
l-to 13 d. Most spikelets (388.2) per panicle were in flower the seven
th day after flowering began in a field. The refined model was used to
predict accurately time of sorghum midge emergence in all (mean R2 =
0.83) but one field (R2 = 0.59). However, predictions of sorghum midge
abundance varied in accuracy (observed/predicted range of sorghum mid
ges = 0.77-1.21). Predicted time of development of the F1 to F5 genera
tions of sorghum midges ranged from 21.0 to 15.8 d. Abundance of ovipo
siting sorghum midges increased 2.5-, 3.7-, 24.6-, and 12.6-fold betwe
en the P-F1, F1-F2, F2-F3, and F3-F4 generations, respectively, and th
en declined. Abundance was greatest (22.7 sorghum midges per panicle)
in the F4 generation. The economic-threshold level was exceeded during
the last days of the F3 generation.