Trichomes as dangerous lollipops

Abstract

When attacked by herbivores, plants produce toxic secondary metabolites that function as direct defenses, as well as indirect defenses that attract and reward predators of the offending herbivores. These indirect defenses include both nutritive rewards such as extra floral nectar, as well as informational rewards, such as the production and release of volatile compounds that betray the location of feeding herbivores to predators. Herbivory of Nicotiana attenuata by the tobacco hornworm (Manduca larvae) alters the volatile profiles of both the plant and larval headspace. Herbivory-elicited specific changes in the volatile profiles are detected by arthropod predators of Manduca larvae. The known predators that perceive volatile cues induced by Manduca herbivory of N. attenuata are insects that target Manduca at early developmental stages, when the larvae are still small; large, late-instar larvae may have outgrown these predation risks. However, here we offer evidence that branched chain aliphatic acids derived from the digestion of plant O-acyl sugars from trichomes may betray Manduca larvae to lizard predators during late developmental stages as well.

Acyl Sugars are Part of Plants’ Defensive Arsenal

Recently we showed that volatile aliphatic acids from the headspace of the tobacco hornworm (Manduca sexta, Lepidoptera, Sphingidae) are used by Pogonomyrmex ants to locate their prey.¹ These branched chain aliphatic acids (BCAA) originate from the digestion of plant O-acyl sugars consumed by Manduca larvae. Secreted by trichomes on the phylloplane of the wild tobacco Nicotiana attenuata, O-acyl sugars serve as a first meal for larvae but tag these herbivores with a distinctive body odor that is detected by the rough harvester ant, Pogonomyrmex rugosus, as it forages along the ground. O-acyl sugars, the most abundant metabolite of trichomes, are thus a potent indirect defense against herbivores that combines both nutritional as well as informational components. Ants are not the only threat to emerging Manduca larvae; for example, the big-eyed bug Geocoris spp detects changes in the Z/E ratio of the isomers of green leaf volatiles to locate Manduca eggs and early instar caterpillars.² Additionally, predatory insects use fecal odors, such as the BCAAs from M. sexta larvae, which are emitted from frass, to optimize their prey foraging behavior.^3,4 Previously, the risk of predation was thought to be negligible for large, late-instar Manduca larvae. Here, we provide evidence that desert lizards prey upon large, late-instar, Manduca larvae and suggest that these reptilian predators also use plant-derived larval body and frass odors to enhance their foraging success.

Observation of Lizard Predation in the 2009 Field Season

During field experiments in 2009, we monitored the daily movements of 3rd-instar M. sexta and M. quinquemaculata larvae among N. attenuata plants growing in an irrigated field plot as well as among plants growing in a native population. In the field plot, Manduca larvae (M. sexta, n = 30; M. quinquemaculata, n = 30) were distributed among pairs of N. attenuata plants (n = 60, 1 larva per plant pair). In the native population, M. quinquemaculata (n = 11) and M. sexta (n = 19) were distributed among 428 plants growing in a roadside wash.

We observed high lizard activity (mostly Aspidoscelis tigris, Uta stansburiana and Sceloporus magister) during the twice-daily censuses. In the field plot, Manduca larvae frequently disappeared between the two observation intervals. The experimental design allowed us to determine that lost larvae were not present on either of the plants in each pair and extended searches of all other plant pairs and other plant species in a 100 m radius of the focal pair were conducted when larvae disappeared. After 3 d, 36 individuals had disappeared. Most disappearances (n = 29) occurred between the two daytime observation intervals, when lizards are active (Fig. 1). Manduca larvae in the native population showed a similar pattern of disappearance over a longer (4 week) observation time.

Figure 1.

Predation of Manduca larvae by diurnally active lizards. Manduca larvae that were unaccounted for during an observation time during field experiments during the 2009 field season were subject to repeated, expanded searches in each subsequent observation time. Most individuals that were lost and never recovered initially disappeared between daytime observation points, correlating with lizard activity. Later baiting experiments indicated predation of Manduca by lizards in both the field plot and a native population.

Initially, we could not rule out loss of larvae due simply to extensive long-distance movement. However during the 2009 field season, we twice witnessed lizards preying upon Manduca in the native population. Furthermore, baiting with exposed M. sexta larvae in the field plot revealed that lizards would readily take Manduca larvae as prey. We videotaped 3rd- and 4th-instar M. sexta (n = 10) placed on N. attenuata plants in the native population. After 2 h, three larvae had disappeared, and the videotape revealed a lizard crawling up the stem of a N. attenuata plant and preying upon the larva (Video S1). After 6 h, half of the larvae were gone.

These observations demonstrated that lizard predation can be a major threat to Manduca larvae, and the characteristic tongue flicking behavior of Aspidoscelis tigris lizards as they forage along the ground before climbing N. attenuata stalks to search for larvae, suggested that they might use the BCAA volatiles from freshly deposited frass to locate prey. Hence we designed experiments for the following field season to evaluate whether lizards might use olfactory cues to assist their clearly visually-guided foraging behavior.

Observation of Lizard Predation in the 2010 Field Season

During the 2010 field season numerous predation experiments were conducted with 1st- and 2nd-instar M. sexta larvae reared on Nicotiana glauca, a plant that completely lacks trichomes and consequently produced larvae that lack the telltale BCAA-redolent body and frass odors.¹ These larvae were reared on N. glauca foliage to produce 3rd- to 4th-instar larvae with BCAA-free body odor and frass. We selected an experimental arena adjacent to an N. attenuata field plantation, where whiptail (Aspidoscelis tigris) and spiny (Sceloporus magister) lizards were commonly seen foraging among N. attenuata plants. In order to determine if natively foraging lizards also used BCAAs to inform their foraging behavior, we created an experimental arena in which large, scentless N. glauca-reared M. sexta larvae were placed individually on 10 dried Datura branches which were tied to a string to keep them in an upright position. The larvae were placed on these branches approximately 15–20 cm above a cement pad so that self recruited lizards would have to climb the stems to capture the larvae (Fig. 2A). Two types of experiments were conducted in which (1) equal volumes (approx 8 cm³) of either fresh, redolent frass from N. attenuata-fed larvae or oven-dried odorless frass (in alternating order) was placed at the base of the branches that touched the cement pad, or (2) the base of the branches were sprayed with BCAA mixtures in frass-equivalent amounts.¹ While the native lizard community quickly learned to locate the larvae on the branches by vision in subsequent replicates, the first trials of both experimental setups (conducted almost a month apart with two weeks of no experimentation in the arena) revealed a strong preference for lizards to climb branches and consume larvae that had been scented with fresh frass or BCAA sprays (Fig. 2B). These results, which will require further experimental support with experimental designs that minimize the ability of lizards to use vision to locate prey, strongly suggest that lizards use BCAA odors to locate larval prey.

Figure 2.

Lizard predation of M. sexta larvae was highest on branches whose bases were treated with fresh BCAA-redolent frass or scented with BCAA volatiles. Predation of 3rd- to 4th-instar M. sexta larvae: (A) Spiny lizard (SL) predating a M. sexta larvae from one of the 10 replicate Datura branches of the experimental arena. (B) Experimental set up and results of the first trials (conducted on June 6 and July 1, 2010). Ten dried wooden D. wrightii branches were affixed to a string that stretched across a slab of concrete across on which native lizards of at least two species regularly foraged. Fresh (black) or oven-dried (gray) M. sexta frass (8 cm³) was applied to the base of the branches that touched the cement slab (upper panel). Larvae (previously reared on N. glauca leaves) were placed 15–20 cm above the ground in a fork of each branch with an affixed piece of N. glauca leaf. Predation by spiny (SL) and whiptailed (WL) lizards is depicted during the subsequent hour. The data from the first trial on June 6 revealed a strong preference of lizards to climb branches with fresh frass at their base compared with branches tagged with dried frass. In four subsequent replicates, conducted on consecutive days, the lizards climbed the first branch they arrived at and consumed larvae on adjacent branches without returning to the ground, clearly finding larvae by vision. On July 1, after 14 d of no experimentation, the experiment was repeated but instead of placing frass at the base of the branches, branches were either sprayed with 100 µL water + BCAA (0.05% Tween-20) (+) or with a water control (-). Lizard predation was monitored for an hour; BCAA-sprayed branches were preferred over water-sprayed branches. A SL after first predating the larvae of the 4th BCAA-perfumed replicate, crossed over to the adjacent control branch to consume the larvae on this branch.

Conclusion

Late-instar Manduca larvae are clearly vulnerable to lizard predation in their native habitats in SW Utah. In an experimental arena with self-recruited lizards, native lizards appeared to use fresh (BCAA-containing) caterpillar frass as well as BCAA-scented branch bases to locate prey located above them on the branches, before they learned the experimental setup and located prey purely by vision. Since larvae are commonly hidden from visually hunting ground-foraging predators by foliage, we propose that BCAAs in fresh frass deposited on the ground by canopy feeding larvae are used by ground-foraging lizards to select plants to climb and locate prey. Because the fatty acid bouquet vanishes quickly under field conditions,¹ these volatile fatty acids are a strong and rapid signal that betrays the location of larvae to predators. Since the BCAA body and frass odor of Manduca larvae is strongly dependent on the ingestion of trichome-produced O-acyl sugars, this compound class seems to be an important component of N. attenuata’s arsenal of indirect defenses. To more rigorously test these inferences, we are currently attempting to engineer plants with morphologically normal trichomes but lacking in O-acyl sugars.

Supplementary Material

Additional material

Supplementary audio file supplied by authors.

Additional material

Supplementary audio file supplied by authors.

Glossary

Abbreviations:

BCAA

branched chain aliphatic acids

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.