Journal of Chemical Ecology, Vol. 12, No. 1, 1986
BEHAVIORAL RESPONSES OF MALE AND FEMALE MEXICAN FRUIT FLIES, Anastrepha ludens, TO MALE-PRODUCED CHEMICALS IN LABORATORY EXPERIMENTS
D.C. ROBACKER
and W . G .
HART
Subtropical Crop Insects Research, USDA-ARS, Weslaco, Texas 78596 (Received March 12, 1985; accepted May 21, 1985) Abstract--The behavioral responses of male and female Mexican fruit flies elicited by male abdominal extracts were measured in laboratory cages where pheromone was applied to the undersides of some leaves on a treated tree but to none of the leaves on a control tree. After arrival to the treated tree, females came directly to pheromone sources. Females on the treated tree visited leaves and fought other females at higher rates than on the control tree. Females stayed on treated leaves and trees longer than on control leaves and trees. In separate experiments, the number of males on pheromone-treated trees and leaves was higher than on controls, but other behavior was unchanged. The results indicate that the pheromone stimulates a complex of behavior involved in the mating ecology of the species. Key Words--Sex pheromones, Mexican fruit flies, Diptera, Tephritidae, Anastrepha ludens.
INTRODUCTION M a l e - p r o d u c e d sex p h e r o m o n e s are k n o w n in several species o f tephritid fruit flies i n c l u d i n g Ceratitis capitata (Ohinata et al., 1977), Dacus tryoni (Fletcher, 1968), D. dorsalis and D. cucurbitae (Ohinata et al., 1982), D. oleae (De M a r z o et al., 1978), Rhagoletis pomonella ( P r o k o p y , 1975), R. cerasi (Kats o y a n n o s , 1982), Anastrepha suspensa (Nation, 1972) and A. ludens ( E s p o n d a G a x i o l a , 1977). M o s t studies h a v e f o c u s e d on attraction o f sexually active females to a p h e r o m o n e source. A few studies h a v e also d e m o n s t r a t e d that m a l e p r o d u c e d c h e m i c a l s attract conspecific m a l e s in the field (Fletcher, 1968; Per39
40
ROBACKER AND HART
domo et al., 1976; Ohinata et al., 1977). Other than attraction, little has been published on the behavior of either males or females elicited by the pheromones of their species. The purpose of this paper is to investigate behavior of both male and female Mexican fruit flies (Anastrepha ludens Loew) elicited by the male-produced pheromone in laboratory bioassays. METHODS AND MATERIALS
General. All flies were from a laboratory culture maintained for ca. 60 generations with no wild fly introductions. The original stock came from mango (Mangifera indica L.) fruit collected near Mexico City, Mexico. All experiments were conducted in the laboratory. Temperature varied between 20 and 30~ and relative humidity between 50 and 70%. Photoperiod was shifted so that lights came on at 0230 hr and went off at 1630 hr CST or CDT depending on the time of year, Pheromone extract was prepared by grinding abdomens of sexually mature, virgin males in hexane and allowing them to soak overnight. The extract was filtered through glass wool and concentrated to one male equivalent (ME) per 5/zl of solvent under a stream of nitrogen. Extracts were stored at 0~ Experiments. The experiments were intended to investigate the following aspects of behavior: attraction to the vicinity of the pheromone, leaf visitation rate in the vicinity of pheromone, searching strategy near pheromone sources, attraction directly to pheromone point sources, searching strategy after arrival to point sources, and agonistic behavior and sexual displays by males near and at pheromone sources. Attraction to the vicinity of the pheromone was measured by the number of arrivals to the trees. Leaf visitation rate in the vicinity of pheromone was investigated by analysis of the number of flights (walking from leaf to leaf was negligible) originating and ending within the trees, per female on the trees. The number of females was estimated from the number observed during counts. The null hypothesis was that the number of leaves visited per female on each tree should be the same if the pheromone did not stimulate leaf visitation. Searching strategy near pheromone sources was evaluated by analysis of the time spent by individual females on the treated (T) and control (C) trees. Attraction directly to pheromone point sources was determined from the number of arrivals to T and C leaves on the T tree. Searching strategy after arrival to point sources was investigated by analysis of the time spent by females on T and C leaves. Tendencies for agonistic behavior and male displays (calling) were determined from the number of fights and calling displays, respectively, which occurred in various locations, per fly present: Bioassays were conducted in wood-framed cages (0.7 • !.6 • 1.0 m) with aluminum screening and a Plexiglas door (0.7 • 0.9 m) with two round service openings (15 cm diam). The cages contained a 2-cm layer of sand and
B E H A V I O R A L RESPONSES O F M E X I C A N F R U I T FLIES
4t
2 potted sour orange (Citrus aurantium L.) host trees, ca. 1 m tall. Filter paper squares (1.5 • 1.5 cm) were attached to the undersides of 20 leaves of each tree with double stick cellophane tape. Pheromone extract was applied to 10 T papers on the T tree (T = pheromone treated). The other papers on both the T and C trees were treated with hexane only and served as controls (C). Undersides of leaves were chosen as treatment sites since male sexual displays (calling) and mating frequently occurs there in A. ludens (Robacker and Hart 1985). For each replication, 50 sexually mature, virgin females (experiment 1) or males (experiment 2) were released into a cage containing water but not food between 0900 and 1100 hr on the test day. A. ludens is not sexually active at this time (Robacker and Hart 1985). Flies were between the ages of 7 and 20 days posteclosion. Tests began at 1400 hr when A. ludens is sexually active. At the beginning of each test, the 10 T leaves were treated with 1.0 ME of abdominal extract in 5 tA of hexane and the 30 C leaves with 5 ~1 of hexane. One observer then recorded the number of flies to arrive on the T tree during a 2-min period, followed by 2-min observations of arrivals to the C tree, the T tree again, and the C tree again. The number of arrivals on T and C leaves was also recorded at this time. Next, the number of flights originating and ending in each tree was recorded for 2 min per tree. At this point, the numbers of flies on the T and C leaves and on the T and C trees were counted. Finally, arrivals to trees and leaves were monitored for another 2 min. per tree. Agonistic interactions (physical fighting only) were recorded whenever observed throughout the experiment. During experiment 2, the number of sexually displaying (abdomen puffing, rectal pouch eversion, rapid wing vibration; Robacker and Hart, 1985) males was also recorded during the fly-location counts. In both experiments, a second observer recorded the amount of time individual flies spent on various locations during the experiment. For this purpose, a fly was considered to be on a tree until it flew away from the tree and landed somewhere else. A fly was considered to be on a leaf until it arrived at another location. Thus, flights which ended where they began were not departures from the original location. As part of the same replication, the positions of the two trees were switched and the papers on the leaves were treated again with 1.0 ME of abdominal extract or hexane. The same series of observations was recorded again. After each test the trees were cleaned with soap and water. The two trees alternated as T and C trees and the positions of the T and C leaves were rerandomized for each replication. Twenty replications of experiment 1 and 30 of experiment 2 were conducted. Each replication took about 1 hr to conduct. Statistical Analyses. Responses of flies on T vs. C trees were compared using paired t tests. Responses on T leaves, C leaves on T trees, and C leaves on C trees were compared with t tests using the error mean square obtained from a randomized complete block analysis of variance. Tendencies for certain
42
ROBACKER AND HART
behaviors to vary either with the number of flies present at various locations or with other behaviors was measured by regression. RESULTS AND DISCUSSION
Experiment 1. Responses of Females. The average number of arrivals to T trees was not significantly greater (P = 0.14) than to C trees (Table 1). Therefore, no conclusions concerning attraction to the vicinity of the pheromone source from distances of at least 0.3 m, the distance from the cage walls to the T tree canopy, can be inferred from the experiment. The variable attraction response may have been due to lack of consistent air currents in the cage which could have enabled orientation to the T tree. The total number of within-tree flights on the T tree was significantly greater (P < 0.01) than on the C tree (Table 1). The number of within-tree flights per female on the T trees was also significantly greater (P < 0.01) than the corresponding values for C trees. Regressions of the number of within-tree flights per female on the number of females on trees were not significant (C tree: b = 0.035, P = 0.5; T tree: b = 0.24, P = 0.21), thus discounting the possibility that within-tree activity per female was related to the number of females. Regression of within-tree flights on the number of agonistic interactions was also not significant for the C trees (b = 1.9, P = 0.5), but was significant for the T tree (b = 2.3, P = 0.05). These results demonstrate that females visited leaves (made "within-tree" flights) at higher rates in the vicinity of the pheromone, but they do not show whether the increase in activity was caused direcOy by pheromone or indirectly through agonistic encounters with other females on the T trees. Another question investigated was whether females modified their searching strategy so as to remain for longer periods near the pheromone source. The amount of time spent searching (flying, walking, and sitting) T trees was significantly greater (P = 0.01) than C trees (Table 1). Combining this result with the effects on leaf visitation rate, we conclude that females not only searched more leaves per unit time but also spent more time searching on the T trees. These pheromone-induced modifications in searching should cause higher encounter rates with males. The number of arrivals to T leaves was significantly greater (P = 0.001) than to C leaves (Table 1), demonstrating that females are probably attracted directly onto male territories after arrival to the vicinity ( < 0.3 m) of the pheromone source. This confirms and expands the finding of Esponda-Gaxiola (1977), who showed that male abdominal extract acts as an attractant to sexually active females. Females spent significantly greater (P = 0.001) amounts of time on T than on C leaves of the T trees (Table 1). Thus, searching strategy after arrival to
3.4y 2.7y
0,25y
0.088y
5.6x 3.7x
2.5x
0.28x
0.076y
0.098y 0.024y
0.25x 0.25x
0.64x
1.4y
0.46y 0.088x
C leaves
2.8x
1.0x 0.13x
T leaves
0.062y
0,10y 0,012y
1.9xy
0.21z 0.042y
C leaves
C tree
~Data entries within the same row, corresponding to comparisons between trees or among leaves, followed by the same letter are not significantly different at the 5 % level. T = pheromone treated; C = untreated. bNumber of females to arrive, per 12 min of observation, per tree or leaf. Number of females present, per count, per tree or leaf. aNumber of flights originating and ending in tree, per 4 rain of observation, per tree. e Per observation. :Number of agonistic interactions, per tree or leaf.
3.8x 1.3y 6.0y
6.2x 2.3 x 15.7x
Arrivals~ Number present c Within-tree flightsa Within-tree flights/number present on tree c' d Time spent (min)" Arrivals/number on tree b' " Agonistic interactionsf Agonistic interactions/ number present on tree or leaff':
C tree
T tree
Measurement
T tree
TABLE 1. RESPONSES OF VIRGIN FEMALE MEXICAN FRUIT FLIES TO MALE ABDOMINAL EXTRACT APPLIED TO UNDERSIDES OF LEAVES ON CITRUS TREES IN INDOOR CAGES a
q
7~
x
ROBACKER AND HART
pheromone point sources was also modified relative to strategy on untreated leaves. It was not possible to determine from our data whether locomotor arrest or a modified searching strategy in which females engaged in short, round-trip flights was responsible for the increase in time spent on T leaves since both behaviors were regarded as time spent on the leaf. However, the conclusion that females modified their searching strategy so as to remain near the pheromone point source is justified. This translates into a tendency for females to remain on male territories if pheromone is released by the male. At this point it is appropriate to return to the subject of leaf-visitation rates. The number of arrivals to C leaves was almost identical on the T and C trees, per female on each tree (Table 1). This result suggests similar searching rates on the two trees, thus contradicting the previous conclusion that the pheromone stimulated activity on the T tree. However, the number of females actually available to respond to C leaves on the T tree was probably less than the number observed on the tree since females spent so much time on or near T leaves as a result of modified searching strategy. If the number of females available to respond to C leaves could be determined, the number of arrivals to C leaves on the T tree per available female would probably be greater than the arrival rate to leaves on the C tree. Much more agonistic behavior occurred on T vs. C trees and T vs. C leaves of the T tree both before and after the appropriate conversions to the number of agonistic interactions per female present on the trees or leaves (P < 0.01). Regression of the number of agonistic interactions per female on either the number of females present on trees (C tree: b = 0.04, P = 0.07; T tree: b = 0.08, P = 0.05) or the number of females on leaves of the T trees (C leaves: b = 0.05, P = 0.10; T leaves: b = 0.16, P = 0.08) showed that, with more replications, significant positive relationships may exist. However, in each case, the regression coefficient is too small to account for the dramatic differences in the number of agonistic interactions on T vs. C treatments. This discounts the possibility that the greater number of females on T trees and leaves was primarily responsible for the increase in agonistic behavior. Rather, the increase in agonistic behavior was probably caused by the presence of pheromone which altered the tendency of individual females to fight. Experiment 2. Responses of Males. Behavior at pheromone-treated vs. untreated trees and leaves was not quantitatively different, with a few exceptions. The numbers of males on the T trees during counts was significantly higher (P < 0.05) than on the C trees (Table 2). Also, the number of males on T leaves was significantly greater (P = 0.01) than on C leaves of the C tree: These results suggest that pheromone either attracted or arrested flight behavior of males. However, there were no significant differences in arrivals to T vs. C trees and leaves or in the time spent at T vs. C trees and leaves. The contradictory nature of these results compels us to accept the null hypothesis that male distribution was not affected by pheromone.
2.9x 2.7x
3.9x 0.97x 2.8x
0.72x
2.4x 2.6x
4.3x 0.90x 3.8x
0.78x
1.4x
1.1 x
1.0x 0.15xy
0.095x 0.14xy
0.12x 0.15x 0.96x 0.23x
2.5x
0.61x 0.12xy
C leaves
2.7x
0.72x 0.15x
T leaves
1.3x
0.97x 0.14y
0.15x 0.1 ly
2.6x
0.72x 0.10y
C leaves
C tree
~Data entries within the same row, corresponding to comparisons between trees or among leaves, followed by the same letter are not significantly different at the 5 % level. T = pheromone treated; C = untreated. bNumber of males to arrive, per 12 min of observation, per tree or leaf. r Number of males present, per count, per tree or leaf. dNumber of flights originating and ending in tree, per 4 rain of observation, per tree. e Per observation. fNumber of males calling, per count, per tree or leaf. s' Number of agonistic interactions, per tree or leaf.
6. l x 3.9y 1 1.8x
5.9x 4.9x 11.4x
Arrivalsb Number presentc Within-tree flightsd Within-tree flights/number present on tree c'a Time spent (min) ~ Arrivals/number on tree b'" Sexually displaying males f Sexually displaying males/ number on tree or leaf C'.r Agonistic interactionsg Agonistic interactions/ number present on tree or leaf c' g
C tree
T tree
Measurement
T tree
OF LEAVES ON CITRUS TREES IN INDOOR CAGES a
TABLE 2. RESPONSES OF VIRGIN MALE MEXICAN FRUIT FLIES TO MALE ABDOMINAL EXTRACT APPLIED TO UNDERSIDES
'q r"
Z~ -n
m o
46
ROBACKER AND HART
More sexual displays and agonistic interactions were also observed on T leaves (T tree) than on C leaves of the C tree (P = 0.05) (Table 2). Dividing by the number of males on T and C leaves showed that these results reflect the greater number of males on T leaves rather than any stimulatory effect of the pheromone on sexual displays or agonistic behavior. Compared to female responses in Table 1, male responses to pheromone treatments were weak or absent for most behaviors examined. To date, no published reports have demonstrated that tephritid males respond in laboratory bioassays specifically to pheromone from conspecific males, even though conspecific females do respond in laboratory bioassays (Nation, 1977). Males of at least three species apparently are attracted to male-produced pheromones in field assays (Fletcher, 1968; Perdomo et al., 1976; Ohinata et al., 1977). Also, the lek system used by most tropical Tephritidae requires that males have an aggregating mechanism, and possible roles of pheromones have been proposed (Perdomo et al., 1976; Prokopy, 1980; Burk and Calkins, 1983). It seems likely that experimental designs are at fault for the lack of success at proving responses of males to pheromone in the lab. One possible problem is that males themselves act as treatments wherever they produce pheromone in the bioassay chamber. These unwanted pheromone sources may then compete with the experimental treatments and also convert controls to treatments. Field tests avoid this problem since males are less likely to produce pheromones at sites where control traps are installed. Better designed laboratory and field experiments are needed to elucidate the roles of male-produced pheromones in the mating ecology of both male and female tephritid fruit flies. Acknowledgments--We thank Ernesto Solis for technical assistance and Sammy Ingle, Claudio Garcia, and Ruben Garcia for supplying the test insects. We also thank Dr. Theodore Burk of Creighton University at Omaha, Nebraska; Dr. Donovan E. Hendricks of USDA-ARS at Brownsville, Texas; and Dr. James L. Nation of the University of Florida at Gainesville for critical review of the manuscript.
REFERENCES
BURK,T., and CALKINS,C.O. 1983. Medfly mating behavior and control strategies. Fla. Entomot. 66: 3-18. DE MARZO, L., NUZZACI, G., and NALINAS,M. 1978. Studio anatomico, istologico, ultrastmtturale e fisiologico del retto e osservazioni etologiche in relazione alla possibile produzione di ferormoni sessuali nel maschio di Dacus oleae Gmel. Entomologica 14: 203-266. ESPONDA-GAXIOLA,R.E. 1977. Contribucion al estudio quimico del atrayente sexual de la mosca Mexicana de la fmta, Anastrepha ludens (Loew). Unpublished thesis, Instituto Technologico y de Estudios Superiores de Monterrey, Monterrey, N.L. Mexico (cited in Nation, 1977). FLETCHER, B.S. 1968. Storage and release of a sex pheromone by the Queensland fruit fly, Dacus tryoni (Diptera: Trypetidae). Nature 219:631-632. KATSOYANNOS, B.I. 1982. Male sex pheromone of Rhagoletis cerasi L. (Diptera: Tephritidae):
BEHAVIORAL RESPONSES OF MEXICAN FRUIT FLIES
47
Factors affecting release and response and its role in the mating behavior. Z. Angew. Entomol. 94: 187-198. NAT~ON, J.L. 1972. Courtship behavior and evidence for a sex attractant in the male Caribbean fruit fly, Anastrepha suspensa. Ann. Entomol. Soc. Am. 65: 1364-1367. NATION, J.L. 1977. Pheromone research in tephritid fruit flies (Diptera: Tephritidae). Proc. Int. Soc. Citriculture 2: 481-485. OrnNATA, K., JACOBSON, M., NADACAWA,S., FUJIMOTO, M., and HI6A, H. 1977. Mediterranean fruit fly: Laboratory and field evaluation of synthetic sex pheromone. J. Environ. Sci. Health A12:67-78 OHINATA, K., JACOBSON, M., KOBAYASHI, R.M., CHAMBERS,D.L., FUJIMOTO, M.S., and HIGA, H.H. 1982. Oriental fruit fly and melon fly: Biological and chemical studies of smoke produced by males. J. Environ. Sci. Health A17: 197-216. PERDOMO, A.J., NATION, J.L., and BARANOWSrd,R.M. 1976. Attraction of female and male Caribbean fruit flies to food-baited and male-baited traps under field conditions. Environ. Entomol. 5: 1208-1210. PROKOPY, R.J. 1975. Mating behavior in Rhagoletis pomonella (Diptera: Tephritidae) V. Virgin female attraction to male odor. Can. Entomol. 107: 905-908. PROKOPY,R.J. 1980. Mating behavior of frugivorous Tephritidae in nature, pp. 37-46, in J. Koyarea (ed.). Fruit Fly Problems. Proc. Symp. Natl. Inst. Agric. Sci. ROBACKER, D.C., and HART, W.G. 1985. Courtship and territoriality of laboratory-reared Mexican fruit flies, Anastrepha ludens (Diptera: Tephritidae), in cages containing host and nonhost trees. Ann. Entomol. Soc. Am. 78:488-494.
BEHAVIORAL RESPONSES OF MALE AND FEMALE MEXICAN FRUIT FLIES, Anastrepha ludens, TO MALE-PRODUCED CHEMICALS IN LABORATORY EXPERIMENTS
D.C. ROBACKER
and W . G .
HART
Subtropical Crop Insects Research, USDA-ARS, Weslaco, Texas 78596 (Received March 12, 1985; accepted May 21, 1985) Abstract--The behavioral responses of male and female Mexican fruit flies elicited by male abdominal extracts were measured in laboratory cages where pheromone was applied to the undersides of some leaves on a treated tree but to none of the leaves on a control tree. After arrival to the treated tree, females came directly to pheromone sources. Females on the treated tree visited leaves and fought other females at higher rates than on the control tree. Females stayed on treated leaves and trees longer than on control leaves and trees. In separate experiments, the number of males on pheromone-treated trees and leaves was higher than on controls, but other behavior was unchanged. The results indicate that the pheromone stimulates a complex of behavior involved in the mating ecology of the species. Key Words--Sex pheromones, Mexican fruit flies, Diptera, Tephritidae, Anastrepha ludens.
INTRODUCTION M a l e - p r o d u c e d sex p h e r o m o n e s are k n o w n in several species o f tephritid fruit flies i n c l u d i n g Ceratitis capitata (Ohinata et al., 1977), Dacus tryoni (Fletcher, 1968), D. dorsalis and D. cucurbitae (Ohinata et al., 1982), D. oleae (De M a r z o et al., 1978), Rhagoletis pomonella ( P r o k o p y , 1975), R. cerasi (Kats o y a n n o s , 1982), Anastrepha suspensa (Nation, 1972) and A. ludens ( E s p o n d a G a x i o l a , 1977). M o s t studies h a v e f o c u s e d on attraction o f sexually active females to a p h e r o m o n e source. A few studies h a v e also d e m o n s t r a t e d that m a l e p r o d u c e d c h e m i c a l s attract conspecific m a l e s in the field (Fletcher, 1968; Per39
40
ROBACKER AND HART
domo et al., 1976; Ohinata et al., 1977). Other than attraction, little has been published on the behavior of either males or females elicited by the pheromones of their species. The purpose of this paper is to investigate behavior of both male and female Mexican fruit flies (Anastrepha ludens Loew) elicited by the male-produced pheromone in laboratory bioassays. METHODS AND MATERIALS
General. All flies were from a laboratory culture maintained for ca. 60 generations with no wild fly introductions. The original stock came from mango (Mangifera indica L.) fruit collected near Mexico City, Mexico. All experiments were conducted in the laboratory. Temperature varied between 20 and 30~ and relative humidity between 50 and 70%. Photoperiod was shifted so that lights came on at 0230 hr and went off at 1630 hr CST or CDT depending on the time of year, Pheromone extract was prepared by grinding abdomens of sexually mature, virgin males in hexane and allowing them to soak overnight. The extract was filtered through glass wool and concentrated to one male equivalent (ME) per 5/zl of solvent under a stream of nitrogen. Extracts were stored at 0~ Experiments. The experiments were intended to investigate the following aspects of behavior: attraction to the vicinity of the pheromone, leaf visitation rate in the vicinity of pheromone, searching strategy near pheromone sources, attraction directly to pheromone point sources, searching strategy after arrival to point sources, and agonistic behavior and sexual displays by males near and at pheromone sources. Attraction to the vicinity of the pheromone was measured by the number of arrivals to the trees. Leaf visitation rate in the vicinity of pheromone was investigated by analysis of the number of flights (walking from leaf to leaf was negligible) originating and ending within the trees, per female on the trees. The number of females was estimated from the number observed during counts. The null hypothesis was that the number of leaves visited per female on each tree should be the same if the pheromone did not stimulate leaf visitation. Searching strategy near pheromone sources was evaluated by analysis of the time spent by individual females on the treated (T) and control (C) trees. Attraction directly to pheromone point sources was determined from the number of arrivals to T and C leaves on the T tree. Searching strategy after arrival to point sources was investigated by analysis of the time spent by females on T and C leaves. Tendencies for agonistic behavior and male displays (calling) were determined from the number of fights and calling displays, respectively, which occurred in various locations, per fly present: Bioassays were conducted in wood-framed cages (0.7 • !.6 • 1.0 m) with aluminum screening and a Plexiglas door (0.7 • 0.9 m) with two round service openings (15 cm diam). The cages contained a 2-cm layer of sand and
B E H A V I O R A L RESPONSES O F M E X I C A N F R U I T FLIES
4t
2 potted sour orange (Citrus aurantium L.) host trees, ca. 1 m tall. Filter paper squares (1.5 • 1.5 cm) were attached to the undersides of 20 leaves of each tree with double stick cellophane tape. Pheromone extract was applied to 10 T papers on the T tree (T = pheromone treated). The other papers on both the T and C trees were treated with hexane only and served as controls (C). Undersides of leaves were chosen as treatment sites since male sexual displays (calling) and mating frequently occurs there in A. ludens (Robacker and Hart 1985). For each replication, 50 sexually mature, virgin females (experiment 1) or males (experiment 2) were released into a cage containing water but not food between 0900 and 1100 hr on the test day. A. ludens is not sexually active at this time (Robacker and Hart 1985). Flies were between the ages of 7 and 20 days posteclosion. Tests began at 1400 hr when A. ludens is sexually active. At the beginning of each test, the 10 T leaves were treated with 1.0 ME of abdominal extract in 5 tA of hexane and the 30 C leaves with 5 ~1 of hexane. One observer then recorded the number of flies to arrive on the T tree during a 2-min period, followed by 2-min observations of arrivals to the C tree, the T tree again, and the C tree again. The number of arrivals on T and C leaves was also recorded at this time. Next, the number of flights originating and ending in each tree was recorded for 2 min per tree. At this point, the numbers of flies on the T and C leaves and on the T and C trees were counted. Finally, arrivals to trees and leaves were monitored for another 2 min. per tree. Agonistic interactions (physical fighting only) were recorded whenever observed throughout the experiment. During experiment 2, the number of sexually displaying (abdomen puffing, rectal pouch eversion, rapid wing vibration; Robacker and Hart, 1985) males was also recorded during the fly-location counts. In both experiments, a second observer recorded the amount of time individual flies spent on various locations during the experiment. For this purpose, a fly was considered to be on a tree until it flew away from the tree and landed somewhere else. A fly was considered to be on a leaf until it arrived at another location. Thus, flights which ended where they began were not departures from the original location. As part of the same replication, the positions of the two trees were switched and the papers on the leaves were treated again with 1.0 ME of abdominal extract or hexane. The same series of observations was recorded again. After each test the trees were cleaned with soap and water. The two trees alternated as T and C trees and the positions of the T and C leaves were rerandomized for each replication. Twenty replications of experiment 1 and 30 of experiment 2 were conducted. Each replication took about 1 hr to conduct. Statistical Analyses. Responses of flies on T vs. C trees were compared using paired t tests. Responses on T leaves, C leaves on T trees, and C leaves on C trees were compared with t tests using the error mean square obtained from a randomized complete block analysis of variance. Tendencies for certain
42
ROBACKER AND HART
behaviors to vary either with the number of flies present at various locations or with other behaviors was measured by regression. RESULTS AND DISCUSSION
Experiment 1. Responses of Females. The average number of arrivals to T trees was not significantly greater (P = 0.14) than to C trees (Table 1). Therefore, no conclusions concerning attraction to the vicinity of the pheromone source from distances of at least 0.3 m, the distance from the cage walls to the T tree canopy, can be inferred from the experiment. The variable attraction response may have been due to lack of consistent air currents in the cage which could have enabled orientation to the T tree. The total number of within-tree flights on the T tree was significantly greater (P < 0.01) than on the C tree (Table 1). The number of within-tree flights per female on the T trees was also significantly greater (P < 0.01) than the corresponding values for C trees. Regressions of the number of within-tree flights per female on the number of females on trees were not significant (C tree: b = 0.035, P = 0.5; T tree: b = 0.24, P = 0.21), thus discounting the possibility that within-tree activity per female was related to the number of females. Regression of within-tree flights on the number of agonistic interactions was also not significant for the C trees (b = 1.9, P = 0.5), but was significant for the T tree (b = 2.3, P = 0.05). These results demonstrate that females visited leaves (made "within-tree" flights) at higher rates in the vicinity of the pheromone, but they do not show whether the increase in activity was caused direcOy by pheromone or indirectly through agonistic encounters with other females on the T trees. Another question investigated was whether females modified their searching strategy so as to remain for longer periods near the pheromone source. The amount of time spent searching (flying, walking, and sitting) T trees was significantly greater (P = 0.01) than C trees (Table 1). Combining this result with the effects on leaf visitation rate, we conclude that females not only searched more leaves per unit time but also spent more time searching on the T trees. These pheromone-induced modifications in searching should cause higher encounter rates with males. The number of arrivals to T leaves was significantly greater (P = 0.001) than to C leaves (Table 1), demonstrating that females are probably attracted directly onto male territories after arrival to the vicinity ( < 0.3 m) of the pheromone source. This confirms and expands the finding of Esponda-Gaxiola (1977), who showed that male abdominal extract acts as an attractant to sexually active females. Females spent significantly greater (P = 0.001) amounts of time on T than on C leaves of the T trees (Table 1). Thus, searching strategy after arrival to
3.4y 2.7y
0,25y
0.088y
5.6x 3.7x
2.5x
0.28x
0.076y
0.098y 0.024y
0.25x 0.25x
0.64x
1.4y
0.46y 0.088x
C leaves
2.8x
1.0x 0.13x
T leaves
0.062y
0,10y 0,012y
1.9xy
0.21z 0.042y
C leaves
C tree
~Data entries within the same row, corresponding to comparisons between trees or among leaves, followed by the same letter are not significantly different at the 5 % level. T = pheromone treated; C = untreated. bNumber of females to arrive, per 12 min of observation, per tree or leaf. Number of females present, per count, per tree or leaf. aNumber of flights originating and ending in tree, per 4 rain of observation, per tree. e Per observation. :Number of agonistic interactions, per tree or leaf.
3.8x 1.3y 6.0y
6.2x 2.3 x 15.7x
Arrivals~ Number present c Within-tree flightsa Within-tree flights/number present on tree c' d Time spent (min)" Arrivals/number on tree b' " Agonistic interactionsf Agonistic interactions/ number present on tree or leaff':
C tree
T tree
Measurement
T tree
TABLE 1. RESPONSES OF VIRGIN FEMALE MEXICAN FRUIT FLIES TO MALE ABDOMINAL EXTRACT APPLIED TO UNDERSIDES OF LEAVES ON CITRUS TREES IN INDOOR CAGES a
q
7~
x
ROBACKER AND HART
pheromone point sources was also modified relative to strategy on untreated leaves. It was not possible to determine from our data whether locomotor arrest or a modified searching strategy in which females engaged in short, round-trip flights was responsible for the increase in time spent on T leaves since both behaviors were regarded as time spent on the leaf. However, the conclusion that females modified their searching strategy so as to remain near the pheromone point source is justified. This translates into a tendency for females to remain on male territories if pheromone is released by the male. At this point it is appropriate to return to the subject of leaf-visitation rates. The number of arrivals to C leaves was almost identical on the T and C trees, per female on each tree (Table 1). This result suggests similar searching rates on the two trees, thus contradicting the previous conclusion that the pheromone stimulated activity on the T tree. However, the number of females actually available to respond to C leaves on the T tree was probably less than the number observed on the tree since females spent so much time on or near T leaves as a result of modified searching strategy. If the number of females available to respond to C leaves could be determined, the number of arrivals to C leaves on the T tree per available female would probably be greater than the arrival rate to leaves on the C tree. Much more agonistic behavior occurred on T vs. C trees and T vs. C leaves of the T tree both before and after the appropriate conversions to the number of agonistic interactions per female present on the trees or leaves (P < 0.01). Regression of the number of agonistic interactions per female on either the number of females present on trees (C tree: b = 0.04, P = 0.07; T tree: b = 0.08, P = 0.05) or the number of females on leaves of the T trees (C leaves: b = 0.05, P = 0.10; T leaves: b = 0.16, P = 0.08) showed that, with more replications, significant positive relationships may exist. However, in each case, the regression coefficient is too small to account for the dramatic differences in the number of agonistic interactions on T vs. C treatments. This discounts the possibility that the greater number of females on T trees and leaves was primarily responsible for the increase in agonistic behavior. Rather, the increase in agonistic behavior was probably caused by the presence of pheromone which altered the tendency of individual females to fight. Experiment 2. Responses of Males. Behavior at pheromone-treated vs. untreated trees and leaves was not quantitatively different, with a few exceptions. The numbers of males on the T trees during counts was significantly higher (P < 0.05) than on the C trees (Table 2). Also, the number of males on T leaves was significantly greater (P = 0.01) than on C leaves of the C tree: These results suggest that pheromone either attracted or arrested flight behavior of males. However, there were no significant differences in arrivals to T vs. C trees and leaves or in the time spent at T vs. C trees and leaves. The contradictory nature of these results compels us to accept the null hypothesis that male distribution was not affected by pheromone.
2.9x 2.7x
3.9x 0.97x 2.8x
0.72x
2.4x 2.6x
4.3x 0.90x 3.8x
0.78x
1.4x
1.1 x
1.0x 0.15xy
0.095x 0.14xy
0.12x 0.15x 0.96x 0.23x
2.5x
0.61x 0.12xy
C leaves
2.7x
0.72x 0.15x
T leaves
1.3x
0.97x 0.14y
0.15x 0.1 ly
2.6x
0.72x 0.10y
C leaves
C tree
~Data entries within the same row, corresponding to comparisons between trees or among leaves, followed by the same letter are not significantly different at the 5 % level. T = pheromone treated; C = untreated. bNumber of males to arrive, per 12 min of observation, per tree or leaf. r Number of males present, per count, per tree or leaf. dNumber of flights originating and ending in tree, per 4 rain of observation, per tree. e Per observation. fNumber of males calling, per count, per tree or leaf. s' Number of agonistic interactions, per tree or leaf.
6. l x 3.9y 1 1.8x
5.9x 4.9x 11.4x
Arrivalsb Number presentc Within-tree flightsd Within-tree flights/number present on tree c'a Time spent (min) ~ Arrivals/number on tree b'" Sexually displaying males f Sexually displaying males/ number on tree or leaf C'.r Agonistic interactionsg Agonistic interactions/ number present on tree or leaf c' g
C tree
T tree
Measurement
T tree
OF LEAVES ON CITRUS TREES IN INDOOR CAGES a
TABLE 2. RESPONSES OF VIRGIN MALE MEXICAN FRUIT FLIES TO MALE ABDOMINAL EXTRACT APPLIED TO UNDERSIDES
'q r"
Z~ -n
m o
46
ROBACKER AND HART
More sexual displays and agonistic interactions were also observed on T leaves (T tree) than on C leaves of the C tree (P = 0.05) (Table 2). Dividing by the number of males on T and C leaves showed that these results reflect the greater number of males on T leaves rather than any stimulatory effect of the pheromone on sexual displays or agonistic behavior. Compared to female responses in Table 1, male responses to pheromone treatments were weak or absent for most behaviors examined. To date, no published reports have demonstrated that tephritid males respond in laboratory bioassays specifically to pheromone from conspecific males, even though conspecific females do respond in laboratory bioassays (Nation, 1977). Males of at least three species apparently are attracted to male-produced pheromones in field assays (Fletcher, 1968; Perdomo et al., 1976; Ohinata et al., 1977). Also, the lek system used by most tropical Tephritidae requires that males have an aggregating mechanism, and possible roles of pheromones have been proposed (Perdomo et al., 1976; Prokopy, 1980; Burk and Calkins, 1983). It seems likely that experimental designs are at fault for the lack of success at proving responses of males to pheromone in the lab. One possible problem is that males themselves act as treatments wherever they produce pheromone in the bioassay chamber. These unwanted pheromone sources may then compete with the experimental treatments and also convert controls to treatments. Field tests avoid this problem since males are less likely to produce pheromones at sites where control traps are installed. Better designed laboratory and field experiments are needed to elucidate the roles of male-produced pheromones in the mating ecology of both male and female tephritid fruit flies. Acknowledgments--We thank Ernesto Solis for technical assistance and Sammy Ingle, Claudio Garcia, and Ruben Garcia for supplying the test insects. We also thank Dr. Theodore Burk of Creighton University at Omaha, Nebraska; Dr. Donovan E. Hendricks of USDA-ARS at Brownsville, Texas; and Dr. James L. Nation of the University of Florida at Gainesville for critical review of the manuscript.
REFERENCES
BURK,T., and CALKINS,C.O. 1983. Medfly mating behavior and control strategies. Fla. Entomot. 66: 3-18. DE MARZO, L., NUZZACI, G., and NALINAS,M. 1978. Studio anatomico, istologico, ultrastmtturale e fisiologico del retto e osservazioni etologiche in relazione alla possibile produzione di ferormoni sessuali nel maschio di Dacus oleae Gmel. Entomologica 14: 203-266. ESPONDA-GAXIOLA,R.E. 1977. Contribucion al estudio quimico del atrayente sexual de la mosca Mexicana de la fmta, Anastrepha ludens (Loew). Unpublished thesis, Instituto Technologico y de Estudios Superiores de Monterrey, Monterrey, N.L. Mexico (cited in Nation, 1977). FLETCHER, B.S. 1968. Storage and release of a sex pheromone by the Queensland fruit fly, Dacus tryoni (Diptera: Trypetidae). Nature 219:631-632. KATSOYANNOS, B.I. 1982. Male sex pheromone of Rhagoletis cerasi L. (Diptera: Tephritidae):
BEHAVIORAL RESPONSES OF MEXICAN FRUIT FLIES
47
Factors affecting release and response and its role in the mating behavior. Z. Angew. Entomol. 94: 187-198. NAT~ON, J.L. 1972. Courtship behavior and evidence for a sex attractant in the male Caribbean fruit fly, Anastrepha suspensa. Ann. Entomol. Soc. Am. 65: 1364-1367. NATION, J.L. 1977. Pheromone research in tephritid fruit flies (Diptera: Tephritidae). Proc. Int. Soc. Citriculture 2: 481-485. OrnNATA, K., JACOBSON, M., NADACAWA,S., FUJIMOTO, M., and HI6A, H. 1977. Mediterranean fruit fly: Laboratory and field evaluation of synthetic sex pheromone. J. Environ. Sci. Health A12:67-78 OHINATA, K., JACOBSON, M., KOBAYASHI, R.M., CHAMBERS,D.L., FUJIMOTO, M.S., and HIGA, H.H. 1982. Oriental fruit fly and melon fly: Biological and chemical studies of smoke produced by males. J. Environ. Sci. Health A17: 197-216. PERDOMO, A.J., NATION, J.L., and BARANOWSrd,R.M. 1976. Attraction of female and male Caribbean fruit flies to food-baited and male-baited traps under field conditions. Environ. Entomol. 5: 1208-1210. PROKOPY, R.J. 1975. Mating behavior in Rhagoletis pomonella (Diptera: Tephritidae) V. Virgin female attraction to male odor. Can. Entomol. 107: 905-908. PROKOPY,R.J. 1980. Mating behavior of frugivorous Tephritidae in nature, pp. 37-46, in J. Koyarea (ed.). Fruit Fly Problems. Proc. Symp. Natl. Inst. Agric. Sci. ROBACKER, D.C., and HART, W.G. 1985. Courtship and territoriality of laboratory-reared Mexican fruit flies, Anastrepha ludens (Diptera: Tephritidae), in cages containing host and nonhost trees. Ann. Entomol. Soc. Am. 78:488-494.
