Fig 1.
Secretion from the ventral eversible gland (VEG) is released near the mouthparts.
(A) Head and prothoracic segment of a final instar T. zethus showing the location of the orifice (O) for the VEG and of the spinneret (S), which releases saliva from the labial salivary gland. (B) VEG dissected from a final instar T. zethus.
Fig 2.
Theroa zethus larvae on Chamaesyce maculata.
(A) First instar disabling latex canals by biting repeatedly into the leaf midrib before feeding distal to the cuts. (B) Second instar feeding after constricting the petiole. (C) Third instar compressing the petiole with its mandibles. (D) Final instar deactivating the laticifers in the stem by repeatedly pinching the stem with its mandibles; ~10 dark spots along the stem indicate the location of the constrictions. (E) Necrotic zone developing in a C. maculata leaf adjacent to a petiole constriction produced by a second instar T. zethus.
Fig 3.
Volume of latex exuding from C. maculata leaves severed 5 mm from the tip.
For each instar, leaves pretreated by T. zethus larvae with vein cuts or petiole/stem constrictions released significantly less latex than control leaves. N = 10 for each instar, except N = 7 for the final instar; data are presented as means ± 1 SEM.
Fig 4.
Time T. zethus larvae spent cutting veins or constricting petioles/stems before initiating feeding.
All larvae were tested on C. maculata. N = 10 for first and third instars, N = 9 for the second instar, N = 7 for the final instar; data are presented as means ±1 SEM.
Fig 5.
Depth of furrows created by T. zethus larvae in poinsettia midribs.
The depth of the midrib in the center of a furrow was divided by the original midrib depth, which was estimated by averaging depths at X and Y on each side of the furrow. The top photograph shows a furrow created by an intact T. zethus larva. Caterpillars with a blocked VEG did not decrease midrib depth at all, in contrast to larvae with a functional VEG (intact larvae and larvae with their spinneret blocked). Data are presented as means ±1 SEM; bars with different letters differ significantly at P < 0.05 using Steel-Dwass tests.
Fig 6.
Behaviors of final instar T. zethus larvae on poinsettia.
Control larvae pressing the VEG opening against the midrib surface (A), constricting the midrib (B) and cutting small leaf veins (C). Spinneret-cauterized larva hanging motionless from a leaf (D).
Fig 7.
Wet weight of latex exuding from poinsettia leaves.
Leaves were cut 1 cm (top) and 6 cm (bottom) from the leaf tip. All larvae were able to compress the midrib near the leaf tip, and thereby reduce latex emission at 1 cm; only larvae with a functional VEG were successful in significantly reducing latex levels 6 cm from the leaf tip. Data are presented as means ± 1 SEM; bars with different letters differ significantly at P < 0.05 using Steel-Dwass tests.
Fig 8.
Wet weight of latex exuding from poinsettia leaves.
Leaves were cut 6 cm from the leaf tip after secretions were applied to the leaf midrib 24 hours previously. VEG secretions reduced latex exudation whether tested alone or with saliva (ground salivary glands). Data are presented as means± 1 SEM; bars with different letters differ significantly (P < 0.05 Steel-Dwass tests).
Fig 9.
Depth of furrows created in poinsettia midribs by T. zethus secretions.
The depth in the center of a furrow was divided by the original midrib depth, which was estimated by averaging depths at X and Y on each side of the furrow. Only the VEG secretion alone or together with ground salivary glands produced a detectable furrow. Data are presented as means ± 1 SEM; bars with different letters differ significantly (P < 0.05 Steel-Dwass tests). The photograph at the top shows a furrow created by 5 μl of VEG secretion, which resembles furrows created by intact T. zethus larvae as illustrated in Fig 5.
Fig 10.
Growth of final instar larvae of T. zethus on poinsettia over 16 hours.
Control and VEG-blocked larvae differed significantly in weight change on intact plants (* P < 0.005, one way ANOVA), but not on excised leaves (n.s. = not significant). Larvae grew more slowly on the intact plants, especially when their VEG opening was blocked. Treatments that differed significantly (P < 0.001, t-tests) have different letters above the bars (lowercase letters comparing control larvae, upper case comparing VEG-blocked larvae). Data are presented as means ±1 SEM; N = 10 larvae/treatment.
Fig 11.
Final instar T. zethus and Praeschausia zapata photographed in the field in Arizona.
T. zethus feeding beyond girdles in (A) Euphorbia cyathophora, and (B) Chamaesyce serpyllifolia. (C) P. zapata feeding on C. hyssopifolia after constricting the petiole.