Journal of Chemical Ecology. Vol. 2I, No. 4, 1995
REJECTION CABBAGE
OF HOST
SENSITIVITY
J.A.A.
PLANT
BUTTERFLY: TO AN
RENWlCK*
and
BY LARVAE
OF
DIET-DEPENDENT ANTIFEEDANT
XIN
PEI
HUANG
Boyce Thompson Institute, Tower Road Ithaca, New York 14853 (Received November 4, 1994; accepted January 3, 1995)
Abstract--Garden nasturtium, Tropaeolum majus (Tropaeolaceae), is an acceptable host plant for the cabbage butterfly, Pieris rapae. Eggs are readily laid on the plant and hatching larvae feed and develop into normal pupae and adults. However, when second- to fifth-instar larvae were transferred from cabbage to nasturtium, they refused to feed and starved to death. Similar results were obtained when larvae were transferred from other host plants to nasturtium. However, larvae that were reared on nasturtium readily accepted cabbage as a new host plant. We have demonstrated the presence of strong antifeedants in nasturtium foliage and identified the most prominent active compound as chlorogenic acid. However, larvae reared on nasturtium had limited sensitivity, and larvae reared on a wheat germ diet were completely insensitive to the antifeedants. Larvae apparently develop sensitivity to the deterrent as a result of feeding on other host plants, whereas continuous exposure to the deterrent causes habituation or suppression of sensitivity development. The results demonstrate that dietary experience can dramatically affect the response of an insect to a potentially antifeedant compound in a plant. Key Words--Sensitization, habituation, dietary experience, feeding deterrents, Pieris rapae. INTRODUCTION
Behavioral responses of phytophagous insects to chemical constituents of their host plants are known to be affected by previous learning experience, which may involve habituation, conditioning, or imprinting (Szentesi and Jermy, 1989). The most common examples of learning appear to be the result of an associative *To whom correspondence should be addressed. 465 0098~331/95/0400-0465507.5010 O 1995 Plenum Publishing CoqxJration
466
RENWICK AND HUANG
process, whereas habituation and sensitization are generally considered to be nonassociative. Therefore, quite different neural mechanisms are thought to be involved. Most previous studies of altered feeding behavior have focused on induced food preferences as a result of initial experience. When late-instar larvae are transferred from one acceptable host plant to another, they may initially refuse or lower their consumption rate of the new food (Hanson, 1976; Schoonhoven and Meerman, 1978; Scriber, 1982; Grabstein and Scriber, 1982). However, in a no-choice situation, the new host is generally accepted after a period of temporary starvation. Similar changes in behavior have been observed for insects that were transferred from plants to artificial diet. Some insects will refuse to feed on diet after feeding on a plant (Schoonhoven, 1977; St/idler and Hanson, 1978), whereas others may actually perform better (Hajek, 1989). On the other hand, larvae reared on artificial diet may accept a wider range of potential host plants, and, in the case of Manduca sexta, this change in behavior can be explained by changes in the sensitivity of chemoreceptors (Schoonhoven, 1967, 1969). When late-instar larvae of Pieris brassicae were given a choice of food plants, they generally preferred the plant on which they were reared ( Johansson, 1951). Larvae reared to fifth instar on Tropaeolum majus L. (nasturtium) also developed a preference for this plant, so induced preference was clearly indicated. However, when P. brassicae larvae were transferred from Brassica oleracea to T. majus, they refused to feed (Ma, 1972). Larvae of P. brassicae also refused to feed on artificial diet after feeding on cabbage, even when a feeding stimulant was added (David and Gardiner, 1966). However, no explanation for the rejection behavior was provided. Recent studies on both P. brassicae and P. rapae indicate that caterpillars reared on artificial diet have reduced sensitivity to deterrents compared to those reared on a host plant, and this could be related to changes in chemosensory activity (van Loon, 1990). However, the mechanisms responsible for induced preference or changes in sensitivity to plant constituents have not been studied. Here we report a phenomenon of diet-induced changes in the acceptance of a host plant by larvae of the cabbage butterfly, Pieris rapae L., similar to that reported for P. brassicae (Ma, 1972) and further show that this can be explained by differential development of sensitivity to antifeedants in that plant. Larvae that were transferred from cabbage to nasturtium refused to feed and consequently starved to death. We have investigated the temporal and dietary requirements for this extreme rejection behavior and have isolated the primary deterrent constituent from nasturtium foliage. METHODS AND MATERIALS
Insects. Pieris rapae larvae were obtained from a colony that is continuously maintained in the laboratory but is renewed annually by the addition of
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
467
field-collected butterflies. Experimental larvae were reared from eggs laid on the plant that served as the initial food. Oviposition occurred in greenhouse cages with supplemental lighting, and as the larvae developed, additional plants were provided to ensure an adequate supply of the specific food. Neonate larvae were obtained from eggs laid on a 4-cm strip of Parafilm wrapped around beakers that were covered with a cabbage leaf, so that tarsal contact provided the necessary stimulant for oviposition (Webb and Shelton, 1988; Renwick et al., 1992). The naive neonates were then transferred to either 2- to 3-week-old nasturtium or 4- to 6-week-old cabbage plants, using a small paint brush. Survival rates for neonates transferred from Parafilm to either cabbage or nasturtium were generally ca. 70%. For rearing on wheat germ diet, larvae were allowed to hatch from eggs laid on Parafilm strips that were subsequently placed in the diet cups. Plants and Diet. All plants were grown under the same conditions, in Cornell Mix (Boodley and Sheldrake, 1977), with supplemental lighting in an air-conditioned greenhouse maintained at ca. 25°C. Seeds of nasturtium, Tropaeolum majus var. Double Gleam, were obtained from Thompson & Morgan, Jackson, New Jersey; Cleome spinosa vat. Pink Queen, from The Page Seed Co., Greene, New York; and Reseda luteola (Weld) from Otto Richter & Sons Ltd., Goodwood, Ontario, Canada. Seeds of Sinapis alba were a gift from B. Gabrys and Brassicajuncea was obtained from Agway, Inc., Syracuse, New York. All plants were 4-6 weeks old when used in experiments. Wheat germ diet was prepared according to the recipe of Bell et al. (1979) and poured into 250-ml styrofoam cups. Bioassays. Acceptance of new food plants was tested by allowing butterflies to oviposit on a plant and rearing the hatching larvae to third instars on this plant, before transferring them to their new host plant. Acceptance of nasturtium foliage by larvae was tested by placing one larva on a nasturtium leaf contained in a 250-ml ice cream cup. The petiole of each leaf was passed through a hole in the bottom of the cup and immersed in water, so that the leaf remained fresh for the duration of the experiment. Larval transfers were made with a small paint brush, and 50 replications of each transfer were performed. Nasturtium plants were 2-3 weeks old when used for these assays. Continued feeding and frass production on test leaves within 24 or 48 hr was considered to be acceptance, whereas small nibbles followed by avoidance or a complete lack of biting was considered to be rejection. The feeding deterrent assay (Dimock et al., 1991) consisted of a choice of four leaf discs, 13 mm diameter, cut from fresh cabbage leaves with a cork borer. Two of these were treated with the test extract and the other two with solvent (methanol) alone. Five larvae (early fourth instars) of P. rapae were introduced at the center of each cup and allowed to feed for a period of 6 hr (when ca. 50-70% of the control disk area had been eaten) under controlled conditions of high humidity in an incubator. Leaf area consumed was measured
468
RENWICK AND HUANG
using a Li-Cor 3100 area meter. A feeding deterrent index (FDI) was calculated, based on the areas consumed from all the disks. FDI = 100 (C - T)/ (C + T) where C is the area consumed of control disks and Tthe area consumed of treated disks. The area of treated disks consumed was expressed as a proportion of the total (treated + control) disk area consumed, and the data were subjected to arcsine square-root transformation. Differences between treatments and controls were analyzed using a one-sample t test on the transformed data, under the null hypothesis that the total area consumed was distributed evenly over treated and control disks. Differences among treatments were analyzed using a Waller-Duncan K-ratio t test (K --- 100) on the transformed data. Extraction o f Foliage. Foliage was extracted following a standard procedure for the isolation of deterrents (Renwick and Radke, 1987). Fresh foliage was dropped into boiling ethanol, followed by homogenization and filtration. The ethanolic extract was evaporated under reduced pressure and the residue was defatted with hexane before dissolving in water. The aqueous extract was partitioned 3 x with 1-butanol, and all fractions were tested for deterrent activity. Isolation of Deterrent. Butanol extracts were first fractionated by flash chromatography on a 45 × 2-cm column packed with 30 g 55-105 /,m Ci8 silica, and fractions were eluted with 90 ml each of water, 10%, 20%, 50%, and 100% acetonitrile in water. The most deterrent fraction was subjected to reverse-phase HPLC, using a semipreparative C~8 column, with a gradient of 100% water to 100% acetonitrile to further isolate fractions for bioassays. The most prominent active component was isolated by repeated HPLC and characterized by UV and NMR spectroscopy. Details of the isolation, identification, and deterrent activity will be published elsewhere (Huang and Renwick, unpublished data).
RESULTS
In an initial experiment, P. rapae larvae at each stage from second to fifth instar were removed from cabbage plants on which they had been reared and placed on foliage of nasturtium. Observations of possible feeding were made after 24 and 48 hr. None of these larvae initiated feeding on nasturtium. Most individuals left the leaves and all eventually died, apparently as a result of starvation. However, when neonate (naive) larvae were placed on nasturtium plants, ca. 70% of these fed successfully, and the percentage surviving was similar to that for neonates placed on cabbage plants. The possibility of induced dependence on the first host plant as a general mechanism for refusal behavior was tested by performing the reverse transfer. P. rapae butterflies were allowed to oviposit on nasturtium plants, and when the resulting larvae reached third instar, they were transferred to cabbage. These Larvae readily accepted cabbage as their new food plant (Table 1). Similarly,
469
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
TABLE 1. ACCEPTANCE OR REJECTION OF NEW HOST PLANTS BY eieris rapae LARVAE (N = 50) REARED TO THIRD INSTAR ON VARIOUS PLANTS OR ARTIFICIAL DIETa
Initial diet
New plant
Feeding on new plant (%)
Cabbage (Brassica olevacea)
Brassica juncea Sinapis alba Cleome spinosa Reseda luteola Nasturtium Cabbage
98 100 98 96 0 100
Nasturtium (Tropaeolum majus) Cabbage Brassica juncea Sinapis alba Cleome spinosa Reseda luteola Wheat germ diet
Nasturtium
0 2 6 4 14 94
"Larvae on the initial host plants hatched from eggs laid on these plants. The larvae on wheat germ diet were from eggs laid on a strip of Parafilm so that no previous contact with a plant occurred.
the nasturtium-reared larvae accepted a range of alternate hosts, including members of the Cruciferae (Brassica juncea and Sinapis alba), Resedaceae (Reseda luteola) and Capparidaceae (Cleome spinosa). Cabbage-reared larvae also fed readily when transferred as third instars to these alternate host plants. Thus, nasturtium is unique among this group of experimental plants in its unacceptability to P. rapae from cabbage. The requirements for development of refusal behavior were examined by rearing larvae to third instars on each of the alternate host plants as well as wheat germ diet, and then transfemng these insects to individual leaves of T. majus. In general, almost all larvae that had developed on the test plants refused nasturtium foliage. Only a few larvae that were transferred from Reseda luteola showed a tendency to accept the nasturtium. However, larvae reared on the "artificial" diet readily fed on nasturtium leaves within 24 hr (Table 1). Since neonate larvae accept nasturtium for feeding, the onset of rejection behavior must occur during the first-instar stage of larval development. The time required to trigger this behavioral change was measured by allowing neonate larvae to feed on cabbage foliage for increasing periods of time, from 0 to 24 hr, before transferring them to nasturtium leaves. Since these neonates were initially transferred from Parafilm, ca. 30% of them did not feed successfully on either nasturtium (0 time) or cabbage, probably as a result of handling. No change in behavior was induced after 2 hr o f feeding, but the number of larvae
470
RENWICK AND H U A N G
80 c 60-
II
E u
40-
c to
¢
'u
20-
o
o
.-¢ 0
2
4
6
8
12
16
24
Duration (hr) of feeding on cabbage before transfer
FIG. 1. Change in acceptance of nasturtium by first-instar P. rapae after feeding for an increasing length of time on intact cabbage plants. After each time period, 50 larvae were transferred to individual leaves of Tropaeolum majus, and the number feeding was recorded after 24 and 48 hr. The data are shown for 24 hr, at which time acceptance or rejection generally was clear, based on either distinct consumption of tissue or lack of damage to the leaf.
accepting the new host declined steadily thereafter with increasing time spent on cabbage (Figure 1). The possible presence of antifeedants in nasturtium was investigated by extraction of foliage with hot ethanol, fractionation by solvent partitioning, and bioassay of the fractions. When aqueous fractions of the nasturtium extracts were partitioned with n-butanol, the butanol fraction proved to be highly active in standard feeding deterrent assays. At concentrations of 0.1 g leaf equivalents (gle), fourth instars that had been reared on cabbage were completely deterred from feeding on treated cabbage leaf disks. Deterrent activity was obtained at concentrations as low as 0.004 gle/disk (between treatment and control t = 2.25, P < 0.05), but disappeared at 0.001 gle/disk (t = 0.16, P > 0.05) (Figure 2). A significant difference was found between FDIs for these two treatments. Reverse-phase HPLC of the active fraction on a C~ s column with a diode array detector revealed the presence of compounds that gave UV spectra typical of phenolic derivatives. The most prominent compound was isolated and found to be strongly deterrent on its own. This compound was subsequently
CABBAGE
BUTTERFLY
LARVAE
D x 0 "0 c
AND
HOST
471
PLANTS
100-
75-
50-
c "o o ¢D I,k
25-
O-
~,
z,
~,
'
"o
"O
"o
w "o
"=.
g , o
~ ~
g ~.
0
o
FIG. 2. Feeding deterrent activity of the butanol-soluble components of nasturtium on P. rapae larvae reared from hatching to early fourth instar on cabbage, and then given a choice of feeding on two cabbage disks treated with butanol extract of nasturtium at doses of 0. i, 0,01, 0.004, or 0,001 gle/disk, or feeding on two control cabbage disks each treated with solvent only, Concentrations are expressed as gram leaf equivalents (gle) based upon fresh weight of extracted foliage. The feeding deterrent index (FDI) is based on areas of discs consumed. FDI = 100 x (control - treated)/(treated + control). Means (+SE) with the same letters are not significantly different according to a WallerDuncan K-ratio t test (K = 100) on the transformed data; n = number of replications of the bioassay.
identified as chlorogenic acid, on the basis o f spectral analyses, acid and base hydrolysis, and c o m p a r i s o n with a sample o f the authentic chemical. Feeding deterrent assays o f the butanol extract of nasturtium were performed to c o m p a r e the sensitivities o f fourth-instar P. rapae that were reared on c a b b a g e , nasturtium, o r w h e a t g e r m diet. At a deterrent concentration o f 0.01 gle/disk, the c a b b a g e - r e a r e d larvae were strongly deterred. H o w e v e r the nasturtium-reared larvae were m u c h less responsive to the deterrent (between treatment and control, t = 3.00, P < 0.05), and the diet-reared larvae were completely unaffected (t = 0.11, P > 0.05) (Figure 3). Significant differences were found b e t w e e n FDIs for c a b b a g e - r e a r e d larvae and larvae reared on nasturtium or diet.
472
RENWlCK
n=lOa
AND HUANG
0.01 gle/dlsc I
100 tl.
n
8060-
c
P i. ¢D
40-
¢D
01 mC "0 ¢D it
20O-20
i
i
i
--
a
Z
FIG. 3. Response of P. rapae larvae reared from hatching to early fourth instar on cabbage, nasturtium, or artificial diet, and then given a choice of feeding on two cabbage disks each treated with butanol extract of nasturtium at doses of 0.01 gle/disk or two control cabbage disks each treated with solvent only. The feeding index was calculated as in Figure 2. Means (+ SE) with the same letters are not significantly different according to a Waller-Duncan K-ratio t test (K = 100) on the transformed data; n = number of replications of the bioassay.
DISCUSSION Absolute rejection of a normally acceptable host plant by an insect appears to be relatively unique. This behavior differs from the more typical induced preference behavior that occurs as a result of an insect becoming conditioned on a plant (Szentesi and Jermy, 1989). However, the refusal of artificial diet after feeding on a plant has been observed for some insects (Schoonhoven, 1977; St~dler and Hanson, 1978), as has the reverse situation where conditioning on artificial diet actually increases the host acceptance range of Manduca sexta larvae (Schoonhoven, 1967). Rejection of nasturtium by cabbage-reared larvae of P. rapae can now be explained by the presence of antifeedants, the most prominent of which is chlorogenic acid, in the new food plant. However, sensitivity to the deterrents is dependent on the previous experience of these larvae. Newly hatched first-instar larvae appear to be insensitive to the negative stimuli,
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
473
but, by the time they have molted to second instar on cabbage or other host plants, sensitivity to deterrents has developed. This could mean that an inducer present in these plants, but absent in nasturtium, is responsible for triggering the development of sensitivity. However, relative insensitivity of larvae reared on nasturtium might also suggest that habituation to the deterrents occurs. The lack of sensitivity of diet-reared larvae would then indicate the presence in diet of compounds that cause habituation, or suppression of development of sensitivity, to the nasturtium semiochemicals. A similar development of refusal behavior in P. brassicae was reported by Schoonhoven (1977), who transferred larvae to artificial diet after various periods of time feeding on cabbage. After three days, the proportion of larvae accepting the artificial diet was negligible. Schoonhoven suggested that the larvae become " f i x e d " to cabbage in this time, thus making the artificial diet unacceptable. The development of deterrent sensitivity in first instars may occur gradually, but the timing appears to be highly variable. The majority of larvae required about 12 hr of feeding on cabbage, but 4 hr was sufficient for some individuals. Actual consumption by individual larvae is highly variable, so the physiological trigger responsible for the change in behavior may depend on the quantity of food consumed. However, individual differences in dose dependency may also occur, and, if changes in sensory peception are involved, individual variation in thresholds for response is likely. Another source of variation is the time taken for an individual larva to initiate feeding. More precise measurement of feeding times would require recording the time at which each larva actually begins to ingest leaf material. The concept of sensitization to a negative stimulus may suggest new approaches to understanding nonassociative learning in insects. Previous studies of sensitization have focused on increased response to a positive cue after repeated presentation, a reaction referred to as "pseudoconditioning" (McGuire, 1984). M a n d u c a sexta larvae were deterred by constituents of wheat germ diet after feeding on a host plant (St/idler and Hanson, 1978). However, when reared on wheat germ diet, they were less sensitive to the deterrents. This has been interpreted as meaning that the larvae became habituated to compounds in the diet, resulting in sensitization of these larvae to feeding stimulants in plant leaves (Szentesi and Jermy, 1989). Our results would suggest an alternative explanation, i.e., that the larvae became sensitized to the deterrents in the diet as a result of their feeding on a host plant. Possible habituation or adaptation of insects to feeding deterrents is a major concern that could limit the practical application of such natural compounds for insect pest control. The physiological mechanisms of habituation and sensitization in vertebrates as well as invertebrates have yet to be elucidated in detail. In insects, sensory adaptation or increased activity of deterrent receptors may occur, or more basic effects on the central nervous system could be involved
474
RENWICK AND HUANG
(Szentesi and B e r n a y s , 1984). P i e r i s r a p a e m a y well s e r v e as a m o d e l f o r future study o f the p h y s i o l o g i c a l p r o c e s s e s i n v o l v e d in the c h a n g i n g r e s p o n s e s o f org a n i s m s to c h e m i c a l stimuli.
Acknowledgments--This research was supported in part by NSF grant BSR-9107322 and USDA grant 91-37302-6198. We thank Ann E. Hajek, V. Macko, J.J.A. van Loon, L.M. Schoonhoven, and three anonymous reviewers for their critical review of the manuscript.
REFERENCES BELL, R.A., OWENS, C.D., SHAPIRO, M., and TARDIF, J.R. 1979. Development of mass rearing technology, pp. 599-633, in C.C. Doane and M.L. McManus (eds.). The Gypsy Moth: Research Toward Integrated Pest Management. USDA-FS Technical Bulletin 1584. BOODLEY, J.W., and SHELDRAKE,R. 1977. Comell peat-lite mixes for commercial plant growing. Cornell lnfo. Bull. 42:1-8. DAVID, W.A.L., and GARDINER, O.B.C. 1966. The effect of sinigrin on the feeding of Pieris brassicae L. larvae transferred from various diets. Entomol. Exp. Appl. 9:95-98. DIMOCK, M.B., RENWICK,J.A.A., RADKE, C.D., and SACHDEV-GUPTA,K. 1991. Chemical constituents of an unacceptable cmcifer, Erysimum cheiranthoides, deter feeding by Pieris rapae. J. Chem. Ecol. 17:525-533. GRABSTEIN, E.M., and SCRIBER, J.M. 1982. Host-plant utilization by Hvalaphora cecropia as affected by prior feeding experience. EntomoL Exp. Appl. 32:262-268. HAIEK, A.E. 1989. Effects of transferring gypsy moth, Lymantria dispar, larvae between artificial diet and Quercus rubra foliage. Entomol. Erp. Appl. 51 : 14 I- 148. HANSON, F.E. 1976. Comparative studies in induction of food choice preferences in lepidopterous larvae. Syrup. Biol. Hung. 16:71. JOHANSSON, A.S. 1951. The food plant preference of the larvae of Pieris brassicae L. (Lepid., Pieridae). Norsk Entomol. Tidsskr. 8:187-195. MA, W.C. 1972. Dynamics of feeding responses in Pieris brassicae Linn. as a function of chemosensory input: A behavioral, ultrastructural and electrophysiological study. Meded. Landbouwhogesch. Wageningen 72( 11): 162 pp. McGUIRE, T. 1984. Learning in three species of Diptera: The blowfly Phormia regina, the fruit fly Drosophila melanogaster, and the house fly Musca domestica. Behav. Genet. 14:479. RENWlCK,J.A.A., and RADKE,C.D. 1987. Chemical stimulants and deterrents regulating acceptance or rejection of crucifers by cabbage butterflies. J. Chem. Ecol. 13: 1771-1776. RENWlCK, J.A.A., RADKE, C.D., SACHOEV-GUPTA,K., and SThDLER, E, 1992. Leaf surface chemicals stimulating oviposition by Pieris rapae (Lepidoptera: Pieridae) on cabbage. Chemoecology 3:33-38. SCHOONHOVEN, L.M. 1967. Loss of hostplant specificity by Manduca sexta after rearing on an artificial diet. EntomoL Exp. Appl. 10:270-272. SCHOONHOVEN,L.M. 1969. Sensitivity changes in some insect chemoreceptors and their effect on food selection behavior. Proc. K. Ned. Akad. Wet. Ser. C 72:491-498. SCHOONHOVEN,L.M. 1977. On the individuality of insect feeding behavior. Proc. K. Ned. Akad. Wet. Ser. C 80:341-350. SCHOONHOVEN,L.M,, and MEERMAN,J. 1978. Metabolic cost of changes in diet and neutralization of allelochemics. Entomol. Exp. Appl. 24:689-693. SCRIBER, J+M. t982. The behavior and nutritional physiology of southern armyworm larvae as a function of plant species consumed in earlier instals. Entomol. Erp. Appl. 31:359-369,
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
475
ST,~DLER, E.. and HANSON, F.E. 1978. Food discrimination and induction of preference for artificial diets in the tobacco homworm, Manduca sexta. PhysioL Entomol. 3:121-133. SZENTESJ, A., and BERNAYS, E.A. 1984. A study of behavioral habituation to a feeding deterrent in nymphs of Schistocerca gregaria, Physiol. Entomol. 9:329-340. SZENTESl, A., and JERMY, T. 1989. The role of experience in host plant choice by phytophagous insects, pp. 39-74, in E.A. Bemays (ed.). Insect-Plant Interactions, Vol. II. CRC Press, Boca Raton, Florida. VAN LOON, J.J.A, 1990. Chemoreception of phenolic acids and flavonoids in larvae of two species of Pieris. J. Comp. Physiol. A. 166:889-899. WEBB, S.E., and SHEL'rON, A.M. 1988, Laboratory rearing of the imported cabbagewonn. Bull. N.Y. Food Life S('i. 122: 1-6.
REJECTION CABBAGE
OF HOST
SENSITIVITY
J.A.A.
PLANT
BUTTERFLY: TO AN
RENWlCK*
and
BY LARVAE
OF
DIET-DEPENDENT ANTIFEEDANT
XIN
PEI
HUANG
Boyce Thompson Institute, Tower Road Ithaca, New York 14853 (Received November 4, 1994; accepted January 3, 1995)
Abstract--Garden nasturtium, Tropaeolum majus (Tropaeolaceae), is an acceptable host plant for the cabbage butterfly, Pieris rapae. Eggs are readily laid on the plant and hatching larvae feed and develop into normal pupae and adults. However, when second- to fifth-instar larvae were transferred from cabbage to nasturtium, they refused to feed and starved to death. Similar results were obtained when larvae were transferred from other host plants to nasturtium. However, larvae that were reared on nasturtium readily accepted cabbage as a new host plant. We have demonstrated the presence of strong antifeedants in nasturtium foliage and identified the most prominent active compound as chlorogenic acid. However, larvae reared on nasturtium had limited sensitivity, and larvae reared on a wheat germ diet were completely insensitive to the antifeedants. Larvae apparently develop sensitivity to the deterrent as a result of feeding on other host plants, whereas continuous exposure to the deterrent causes habituation or suppression of sensitivity development. The results demonstrate that dietary experience can dramatically affect the response of an insect to a potentially antifeedant compound in a plant. Key Words--Sensitization, habituation, dietary experience, feeding deterrents, Pieris rapae. INTRODUCTION
Behavioral responses of phytophagous insects to chemical constituents of their host plants are known to be affected by previous learning experience, which may involve habituation, conditioning, or imprinting (Szentesi and Jermy, 1989). The most common examples of learning appear to be the result of an associative *To whom correspondence should be addressed. 465 0098~331/95/0400-0465507.5010 O 1995 Plenum Publishing CoqxJration
466
RENWICK AND HUANG
process, whereas habituation and sensitization are generally considered to be nonassociative. Therefore, quite different neural mechanisms are thought to be involved. Most previous studies of altered feeding behavior have focused on induced food preferences as a result of initial experience. When late-instar larvae are transferred from one acceptable host plant to another, they may initially refuse or lower their consumption rate of the new food (Hanson, 1976; Schoonhoven and Meerman, 1978; Scriber, 1982; Grabstein and Scriber, 1982). However, in a no-choice situation, the new host is generally accepted after a period of temporary starvation. Similar changes in behavior have been observed for insects that were transferred from plants to artificial diet. Some insects will refuse to feed on diet after feeding on a plant (Schoonhoven, 1977; St/idler and Hanson, 1978), whereas others may actually perform better (Hajek, 1989). On the other hand, larvae reared on artificial diet may accept a wider range of potential host plants, and, in the case of Manduca sexta, this change in behavior can be explained by changes in the sensitivity of chemoreceptors (Schoonhoven, 1967, 1969). When late-instar larvae of Pieris brassicae were given a choice of food plants, they generally preferred the plant on which they were reared ( Johansson, 1951). Larvae reared to fifth instar on Tropaeolum majus L. (nasturtium) also developed a preference for this plant, so induced preference was clearly indicated. However, when P. brassicae larvae were transferred from Brassica oleracea to T. majus, they refused to feed (Ma, 1972). Larvae of P. brassicae also refused to feed on artificial diet after feeding on cabbage, even when a feeding stimulant was added (David and Gardiner, 1966). However, no explanation for the rejection behavior was provided. Recent studies on both P. brassicae and P. rapae indicate that caterpillars reared on artificial diet have reduced sensitivity to deterrents compared to those reared on a host plant, and this could be related to changes in chemosensory activity (van Loon, 1990). However, the mechanisms responsible for induced preference or changes in sensitivity to plant constituents have not been studied. Here we report a phenomenon of diet-induced changes in the acceptance of a host plant by larvae of the cabbage butterfly, Pieris rapae L., similar to that reported for P. brassicae (Ma, 1972) and further show that this can be explained by differential development of sensitivity to antifeedants in that plant. Larvae that were transferred from cabbage to nasturtium refused to feed and consequently starved to death. We have investigated the temporal and dietary requirements for this extreme rejection behavior and have isolated the primary deterrent constituent from nasturtium foliage. METHODS AND MATERIALS
Insects. Pieris rapae larvae were obtained from a colony that is continuously maintained in the laboratory but is renewed annually by the addition of
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
467
field-collected butterflies. Experimental larvae were reared from eggs laid on the plant that served as the initial food. Oviposition occurred in greenhouse cages with supplemental lighting, and as the larvae developed, additional plants were provided to ensure an adequate supply of the specific food. Neonate larvae were obtained from eggs laid on a 4-cm strip of Parafilm wrapped around beakers that were covered with a cabbage leaf, so that tarsal contact provided the necessary stimulant for oviposition (Webb and Shelton, 1988; Renwick et al., 1992). The naive neonates were then transferred to either 2- to 3-week-old nasturtium or 4- to 6-week-old cabbage plants, using a small paint brush. Survival rates for neonates transferred from Parafilm to either cabbage or nasturtium were generally ca. 70%. For rearing on wheat germ diet, larvae were allowed to hatch from eggs laid on Parafilm strips that were subsequently placed in the diet cups. Plants and Diet. All plants were grown under the same conditions, in Cornell Mix (Boodley and Sheldrake, 1977), with supplemental lighting in an air-conditioned greenhouse maintained at ca. 25°C. Seeds of nasturtium, Tropaeolum majus var. Double Gleam, were obtained from Thompson & Morgan, Jackson, New Jersey; Cleome spinosa vat. Pink Queen, from The Page Seed Co., Greene, New York; and Reseda luteola (Weld) from Otto Richter & Sons Ltd., Goodwood, Ontario, Canada. Seeds of Sinapis alba were a gift from B. Gabrys and Brassicajuncea was obtained from Agway, Inc., Syracuse, New York. All plants were 4-6 weeks old when used in experiments. Wheat germ diet was prepared according to the recipe of Bell et al. (1979) and poured into 250-ml styrofoam cups. Bioassays. Acceptance of new food plants was tested by allowing butterflies to oviposit on a plant and rearing the hatching larvae to third instars on this plant, before transferring them to their new host plant. Acceptance of nasturtium foliage by larvae was tested by placing one larva on a nasturtium leaf contained in a 250-ml ice cream cup. The petiole of each leaf was passed through a hole in the bottom of the cup and immersed in water, so that the leaf remained fresh for the duration of the experiment. Larval transfers were made with a small paint brush, and 50 replications of each transfer were performed. Nasturtium plants were 2-3 weeks old when used for these assays. Continued feeding and frass production on test leaves within 24 or 48 hr was considered to be acceptance, whereas small nibbles followed by avoidance or a complete lack of biting was considered to be rejection. The feeding deterrent assay (Dimock et al., 1991) consisted of a choice of four leaf discs, 13 mm diameter, cut from fresh cabbage leaves with a cork borer. Two of these were treated with the test extract and the other two with solvent (methanol) alone. Five larvae (early fourth instars) of P. rapae were introduced at the center of each cup and allowed to feed for a period of 6 hr (when ca. 50-70% of the control disk area had been eaten) under controlled conditions of high humidity in an incubator. Leaf area consumed was measured
468
RENWICK AND HUANG
using a Li-Cor 3100 area meter. A feeding deterrent index (FDI) was calculated, based on the areas consumed from all the disks. FDI = 100 (C - T)/ (C + T) where C is the area consumed of control disks and Tthe area consumed of treated disks. The area of treated disks consumed was expressed as a proportion of the total (treated + control) disk area consumed, and the data were subjected to arcsine square-root transformation. Differences between treatments and controls were analyzed using a one-sample t test on the transformed data, under the null hypothesis that the total area consumed was distributed evenly over treated and control disks. Differences among treatments were analyzed using a Waller-Duncan K-ratio t test (K --- 100) on the transformed data. Extraction o f Foliage. Foliage was extracted following a standard procedure for the isolation of deterrents (Renwick and Radke, 1987). Fresh foliage was dropped into boiling ethanol, followed by homogenization and filtration. The ethanolic extract was evaporated under reduced pressure and the residue was defatted with hexane before dissolving in water. The aqueous extract was partitioned 3 x with 1-butanol, and all fractions were tested for deterrent activity. Isolation of Deterrent. Butanol extracts were first fractionated by flash chromatography on a 45 × 2-cm column packed with 30 g 55-105 /,m Ci8 silica, and fractions were eluted with 90 ml each of water, 10%, 20%, 50%, and 100% acetonitrile in water. The most deterrent fraction was subjected to reverse-phase HPLC, using a semipreparative C~8 column, with a gradient of 100% water to 100% acetonitrile to further isolate fractions for bioassays. The most prominent active component was isolated by repeated HPLC and characterized by UV and NMR spectroscopy. Details of the isolation, identification, and deterrent activity will be published elsewhere (Huang and Renwick, unpublished data).
RESULTS
In an initial experiment, P. rapae larvae at each stage from second to fifth instar were removed from cabbage plants on which they had been reared and placed on foliage of nasturtium. Observations of possible feeding were made after 24 and 48 hr. None of these larvae initiated feeding on nasturtium. Most individuals left the leaves and all eventually died, apparently as a result of starvation. However, when neonate (naive) larvae were placed on nasturtium plants, ca. 70% of these fed successfully, and the percentage surviving was similar to that for neonates placed on cabbage plants. The possibility of induced dependence on the first host plant as a general mechanism for refusal behavior was tested by performing the reverse transfer. P. rapae butterflies were allowed to oviposit on nasturtium plants, and when the resulting larvae reached third instar, they were transferred to cabbage. These Larvae readily accepted cabbage as their new food plant (Table 1). Similarly,
469
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
TABLE 1. ACCEPTANCE OR REJECTION OF NEW HOST PLANTS BY eieris rapae LARVAE (N = 50) REARED TO THIRD INSTAR ON VARIOUS PLANTS OR ARTIFICIAL DIETa
Initial diet
New plant
Feeding on new plant (%)
Cabbage (Brassica olevacea)
Brassica juncea Sinapis alba Cleome spinosa Reseda luteola Nasturtium Cabbage
98 100 98 96 0 100
Nasturtium (Tropaeolum majus) Cabbage Brassica juncea Sinapis alba Cleome spinosa Reseda luteola Wheat germ diet
Nasturtium
0 2 6 4 14 94
"Larvae on the initial host plants hatched from eggs laid on these plants. The larvae on wheat germ diet were from eggs laid on a strip of Parafilm so that no previous contact with a plant occurred.
the nasturtium-reared larvae accepted a range of alternate hosts, including members of the Cruciferae (Brassica juncea and Sinapis alba), Resedaceae (Reseda luteola) and Capparidaceae (Cleome spinosa). Cabbage-reared larvae also fed readily when transferred as third instars to these alternate host plants. Thus, nasturtium is unique among this group of experimental plants in its unacceptability to P. rapae from cabbage. The requirements for development of refusal behavior were examined by rearing larvae to third instars on each of the alternate host plants as well as wheat germ diet, and then transfemng these insects to individual leaves of T. majus. In general, almost all larvae that had developed on the test plants refused nasturtium foliage. Only a few larvae that were transferred from Reseda luteola showed a tendency to accept the nasturtium. However, larvae reared on the "artificial" diet readily fed on nasturtium leaves within 24 hr (Table 1). Since neonate larvae accept nasturtium for feeding, the onset of rejection behavior must occur during the first-instar stage of larval development. The time required to trigger this behavioral change was measured by allowing neonate larvae to feed on cabbage foliage for increasing periods of time, from 0 to 24 hr, before transferring them to nasturtium leaves. Since these neonates were initially transferred from Parafilm, ca. 30% of them did not feed successfully on either nasturtium (0 time) or cabbage, probably as a result of handling. No change in behavior was induced after 2 hr o f feeding, but the number of larvae
470
RENWICK AND H U A N G
80 c 60-
II
E u
40-
c to
¢
'u
20-
o
o
.-¢ 0
2
4
6
8
12
16
24
Duration (hr) of feeding on cabbage before transfer
FIG. 1. Change in acceptance of nasturtium by first-instar P. rapae after feeding for an increasing length of time on intact cabbage plants. After each time period, 50 larvae were transferred to individual leaves of Tropaeolum majus, and the number feeding was recorded after 24 and 48 hr. The data are shown for 24 hr, at which time acceptance or rejection generally was clear, based on either distinct consumption of tissue or lack of damage to the leaf.
accepting the new host declined steadily thereafter with increasing time spent on cabbage (Figure 1). The possible presence of antifeedants in nasturtium was investigated by extraction of foliage with hot ethanol, fractionation by solvent partitioning, and bioassay of the fractions. When aqueous fractions of the nasturtium extracts were partitioned with n-butanol, the butanol fraction proved to be highly active in standard feeding deterrent assays. At concentrations of 0.1 g leaf equivalents (gle), fourth instars that had been reared on cabbage were completely deterred from feeding on treated cabbage leaf disks. Deterrent activity was obtained at concentrations as low as 0.004 gle/disk (between treatment and control t = 2.25, P < 0.05), but disappeared at 0.001 gle/disk (t = 0.16, P > 0.05) (Figure 2). A significant difference was found between FDIs for these two treatments. Reverse-phase HPLC of the active fraction on a C~ s column with a diode array detector revealed the presence of compounds that gave UV spectra typical of phenolic derivatives. The most prominent compound was isolated and found to be strongly deterrent on its own. This compound was subsequently
CABBAGE
BUTTERFLY
LARVAE
D x 0 "0 c
AND
HOST
471
PLANTS
100-
75-
50-
c "o o ¢D I,k
25-
O-
~,
z,
~,
'
"o
"O
"o
w "o
"=.
g , o
~ ~
g ~.
0
o
FIG. 2. Feeding deterrent activity of the butanol-soluble components of nasturtium on P. rapae larvae reared from hatching to early fourth instar on cabbage, and then given a choice of feeding on two cabbage disks treated with butanol extract of nasturtium at doses of 0. i, 0,01, 0.004, or 0,001 gle/disk, or feeding on two control cabbage disks each treated with solvent only, Concentrations are expressed as gram leaf equivalents (gle) based upon fresh weight of extracted foliage. The feeding deterrent index (FDI) is based on areas of discs consumed. FDI = 100 x (control - treated)/(treated + control). Means (+SE) with the same letters are not significantly different according to a WallerDuncan K-ratio t test (K = 100) on the transformed data; n = number of replications of the bioassay.
identified as chlorogenic acid, on the basis o f spectral analyses, acid and base hydrolysis, and c o m p a r i s o n with a sample o f the authentic chemical. Feeding deterrent assays o f the butanol extract of nasturtium were performed to c o m p a r e the sensitivities o f fourth-instar P. rapae that were reared on c a b b a g e , nasturtium, o r w h e a t g e r m diet. At a deterrent concentration o f 0.01 gle/disk, the c a b b a g e - r e a r e d larvae were strongly deterred. H o w e v e r the nasturtium-reared larvae were m u c h less responsive to the deterrent (between treatment and control, t = 3.00, P < 0.05), and the diet-reared larvae were completely unaffected (t = 0.11, P > 0.05) (Figure 3). Significant differences were found b e t w e e n FDIs for c a b b a g e - r e a r e d larvae and larvae reared on nasturtium or diet.
472
RENWlCK
n=lOa
AND HUANG
0.01 gle/dlsc I
100 tl.
n
8060-
c
P i. ¢D
40-
¢D
01 mC "0 ¢D it
20O-20
i
i
i
--
a
Z
FIG. 3. Response of P. rapae larvae reared from hatching to early fourth instar on cabbage, nasturtium, or artificial diet, and then given a choice of feeding on two cabbage disks each treated with butanol extract of nasturtium at doses of 0.01 gle/disk or two control cabbage disks each treated with solvent only. The feeding index was calculated as in Figure 2. Means (+ SE) with the same letters are not significantly different according to a Waller-Duncan K-ratio t test (K = 100) on the transformed data; n = number of replications of the bioassay.
DISCUSSION Absolute rejection of a normally acceptable host plant by an insect appears to be relatively unique. This behavior differs from the more typical induced preference behavior that occurs as a result of an insect becoming conditioned on a plant (Szentesi and Jermy, 1989). However, the refusal of artificial diet after feeding on a plant has been observed for some insects (Schoonhoven, 1977; St~dler and Hanson, 1978), as has the reverse situation where conditioning on artificial diet actually increases the host acceptance range of Manduca sexta larvae (Schoonhoven, 1967). Rejection of nasturtium by cabbage-reared larvae of P. rapae can now be explained by the presence of antifeedants, the most prominent of which is chlorogenic acid, in the new food plant. However, sensitivity to the deterrents is dependent on the previous experience of these larvae. Newly hatched first-instar larvae appear to be insensitive to the negative stimuli,
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
473
but, by the time they have molted to second instar on cabbage or other host plants, sensitivity to deterrents has developed. This could mean that an inducer present in these plants, but absent in nasturtium, is responsible for triggering the development of sensitivity. However, relative insensitivity of larvae reared on nasturtium might also suggest that habituation to the deterrents occurs. The lack of sensitivity of diet-reared larvae would then indicate the presence in diet of compounds that cause habituation, or suppression of development of sensitivity, to the nasturtium semiochemicals. A similar development of refusal behavior in P. brassicae was reported by Schoonhoven (1977), who transferred larvae to artificial diet after various periods of time feeding on cabbage. After three days, the proportion of larvae accepting the artificial diet was negligible. Schoonhoven suggested that the larvae become " f i x e d " to cabbage in this time, thus making the artificial diet unacceptable. The development of deterrent sensitivity in first instars may occur gradually, but the timing appears to be highly variable. The majority of larvae required about 12 hr of feeding on cabbage, but 4 hr was sufficient for some individuals. Actual consumption by individual larvae is highly variable, so the physiological trigger responsible for the change in behavior may depend on the quantity of food consumed. However, individual differences in dose dependency may also occur, and, if changes in sensory peception are involved, individual variation in thresholds for response is likely. Another source of variation is the time taken for an individual larva to initiate feeding. More precise measurement of feeding times would require recording the time at which each larva actually begins to ingest leaf material. The concept of sensitization to a negative stimulus may suggest new approaches to understanding nonassociative learning in insects. Previous studies of sensitization have focused on increased response to a positive cue after repeated presentation, a reaction referred to as "pseudoconditioning" (McGuire, 1984). M a n d u c a sexta larvae were deterred by constituents of wheat germ diet after feeding on a host plant (St/idler and Hanson, 1978). However, when reared on wheat germ diet, they were less sensitive to the deterrents. This has been interpreted as meaning that the larvae became habituated to compounds in the diet, resulting in sensitization of these larvae to feeding stimulants in plant leaves (Szentesi and Jermy, 1989). Our results would suggest an alternative explanation, i.e., that the larvae became sensitized to the deterrents in the diet as a result of their feeding on a host plant. Possible habituation or adaptation of insects to feeding deterrents is a major concern that could limit the practical application of such natural compounds for insect pest control. The physiological mechanisms of habituation and sensitization in vertebrates as well as invertebrates have yet to be elucidated in detail. In insects, sensory adaptation or increased activity of deterrent receptors may occur, or more basic effects on the central nervous system could be involved
474
RENWICK AND HUANG
(Szentesi and B e r n a y s , 1984). P i e r i s r a p a e m a y well s e r v e as a m o d e l f o r future study o f the p h y s i o l o g i c a l p r o c e s s e s i n v o l v e d in the c h a n g i n g r e s p o n s e s o f org a n i s m s to c h e m i c a l stimuli.
Acknowledgments--This research was supported in part by NSF grant BSR-9107322 and USDA grant 91-37302-6198. We thank Ann E. Hajek, V. Macko, J.J.A. van Loon, L.M. Schoonhoven, and three anonymous reviewers for their critical review of the manuscript.
REFERENCES BELL, R.A., OWENS, C.D., SHAPIRO, M., and TARDIF, J.R. 1979. Development of mass rearing technology, pp. 599-633, in C.C. Doane and M.L. McManus (eds.). The Gypsy Moth: Research Toward Integrated Pest Management. USDA-FS Technical Bulletin 1584. BOODLEY, J.W., and SHELDRAKE,R. 1977. Comell peat-lite mixes for commercial plant growing. Cornell lnfo. Bull. 42:1-8. DAVID, W.A.L., and GARDINER, O.B.C. 1966. The effect of sinigrin on the feeding of Pieris brassicae L. larvae transferred from various diets. Entomol. Exp. Appl. 9:95-98. DIMOCK, M.B., RENWICK,J.A.A., RADKE, C.D., and SACHDEV-GUPTA,K. 1991. Chemical constituents of an unacceptable cmcifer, Erysimum cheiranthoides, deter feeding by Pieris rapae. J. Chem. Ecol. 17:525-533. GRABSTEIN, E.M., and SCRIBER, J.M. 1982. Host-plant utilization by Hvalaphora cecropia as affected by prior feeding experience. EntomoL Exp. Appl. 32:262-268. HAIEK, A.E. 1989. Effects of transferring gypsy moth, Lymantria dispar, larvae between artificial diet and Quercus rubra foliage. Entomol. Erp. Appl. 51 : 14 I- 148. HANSON, F.E. 1976. Comparative studies in induction of food choice preferences in lepidopterous larvae. Syrup. Biol. Hung. 16:71. JOHANSSON, A.S. 1951. The food plant preference of the larvae of Pieris brassicae L. (Lepid., Pieridae). Norsk Entomol. Tidsskr. 8:187-195. MA, W.C. 1972. Dynamics of feeding responses in Pieris brassicae Linn. as a function of chemosensory input: A behavioral, ultrastructural and electrophysiological study. Meded. Landbouwhogesch. Wageningen 72( 11): 162 pp. McGUIRE, T. 1984. Learning in three species of Diptera: The blowfly Phormia regina, the fruit fly Drosophila melanogaster, and the house fly Musca domestica. Behav. Genet. 14:479. RENWlCK,J.A.A., and RADKE,C.D. 1987. Chemical stimulants and deterrents regulating acceptance or rejection of crucifers by cabbage butterflies. J. Chem. Ecol. 13: 1771-1776. RENWlCK, J.A.A., RADKE, C.D., SACHOEV-GUPTA,K., and SThDLER, E, 1992. Leaf surface chemicals stimulating oviposition by Pieris rapae (Lepidoptera: Pieridae) on cabbage. Chemoecology 3:33-38. SCHOONHOVEN, L.M. 1967. Loss of hostplant specificity by Manduca sexta after rearing on an artificial diet. EntomoL Exp. Appl. 10:270-272. SCHOONHOVEN,L.M. 1969. Sensitivity changes in some insect chemoreceptors and their effect on food selection behavior. Proc. K. Ned. Akad. Wet. Ser. C 72:491-498. SCHOONHOVEN,L.M. 1977. On the individuality of insect feeding behavior. Proc. K. Ned. Akad. Wet. Ser. C 80:341-350. SCHOONHOVEN,L.M,, and MEERMAN,J. 1978. Metabolic cost of changes in diet and neutralization of allelochemics. Entomol. Exp. Appl. 24:689-693. SCRIBER, J+M. t982. The behavior and nutritional physiology of southern armyworm larvae as a function of plant species consumed in earlier instals. Entomol. Erp. Appl. 31:359-369,
CABBAGE BUTTERFLY LARVAE AND HOST PLANTS
475
ST,~DLER, E.. and HANSON, F.E. 1978. Food discrimination and induction of preference for artificial diets in the tobacco homworm, Manduca sexta. PhysioL Entomol. 3:121-133. SZENTESJ, A., and BERNAYS, E.A. 1984. A study of behavioral habituation to a feeding deterrent in nymphs of Schistocerca gregaria, Physiol. Entomol. 9:329-340. SZENTESl, A., and JERMY, T. 1989. The role of experience in host plant choice by phytophagous insects, pp. 39-74, in E.A. Bemays (ed.). Insect-Plant Interactions, Vol. II. CRC Press, Boca Raton, Florida. VAN LOON, J.J.A, 1990. Chemoreception of phenolic acids and flavonoids in larvae of two species of Pieris. J. Comp. Physiol. A. 166:889-899. WEBB, S.E., and SHEL'rON, A.M. 1988, Laboratory rearing of the imported cabbagewonn. Bull. N.Y. Food Life S('i. 122: 1-6.
