Significant differences in some reproductive parameters of the B and Q biot
ypes of Bemisia tabaci (Gennadius) with regard to four species of winter we
eds were determined in a no-choice assay. The highest fecundity (eggs) and
fertility (pupae and adults) were obtained with Malva parviflora L. as host
, followed by Capsella bursa-pastoris L., Brassica kaber (DC) and Lactuca s
erriola L. The percentage of adult emergence (from egg to adult) was signif
icantly higher (P < 0.001) with M. parviflora, C. bursa-pastoris and B. kab
er than with L. serriola. Except on L. serriola, the mean values of the rep
roductive parameters for the Q-biotype were significantly (P < 0.05) greate
r than those for the B-biotype. On the other hand, in choice experiments wi
th summer weeds, both B- and Q-biotypes preferred Datura stramonium L. and
Solanum nigrum L. to Amaranthus retroflexus L., Chenopodium album L. and Ec
hinochloa crus-galli L. Significantly more males and females, and more pupa
e and empty pupal cases per plant were found on D. stramonium than on S. ni
grum. Adult abundance on A. retroflexus, C. album and E. crus-galli plants
was very low. At a constant temperature of 26 +/- 1 degrees C, the Q-biotyp
e developed significantly faster than the B-biotype on S. nigrum and D. str
amonium. Both biotypes required, for this development, significantly more d
ays on D. stramonium than on S. nigrum. These results suggest that it is im
portant to suppress the growth of M. parviflora, C. bursa-pastoris, B. kabe
r, D. stramonium and S. nigrum in order to avoid the risk of further expans
ion of whitefly populations where these weeds and both B- an Q-biotypes are
present.