P1: JLS Journal of Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

Journal of Insect Behavior, Vol. 17, No. 3, May 2004 (C 2004)

Long-Term Stability in the Mating System of the Bot Cuterebra austeni (Cuterebridae)

John Alcock1,3 and Darrell J. Kemp2

Accepted July 16, 2003; revised January 12, 2004

The bot fly Cuterebra austeni exhibits year-to-year consistency in its landmark-based hilltopping mating system. were found on some of the same localized ridgetop perching areas that were known to be mate encounter sites 22 years previously. In both 1980 and 2002, males were drawn preferen- tially to locally elevated, largely vegetation-free sites. In addition, a substantial proportion of individuals exhibited site fidelity in both years, as demonstrated by their willingness to return to the site where marked after being captured and handled. In both years of observation, a smaller proportion returned the second day after marking and only 10–20% of marked males were seen on the third day. Thus males apparently have only a very limited window of oppor- tunity to secure a given perching territory. In both years, a number of males (“residents”) succeeded in monopolizing a perching area over a substantial part of the brief morning flight period. One unexplained difference across the decades: residents appeared on a lower proportion of days of observation in 2002 than in 1980.

KEY WORDS: bot fly; Cuterebra; Cuterebridae; mating system.

INTRODUCTION

The bot fly Cuterebra austeni is a lekking insect whose males perch on and defend flat, gravel patches on hilltops and ridgelines in the Sonoran Desert (see Alcock and Schaefer, 1983, Figs. 1, 2; Baird, 1987). Sites that are higher

1Division of Life Sciences, Arizona State University, Tempe, Arizona 85287-4501. 2Department of Zoology, Stockholm University, SE-10691 Stockholm, Sweden. 3To whom correspondence should be addressed. Fax: 480 965 2519. e-mail: [email protected].

273

0892-7553/04/0500-0273/0 C 2004 Plenum Publishing Corporation P1: JLS Journal of Insect Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

274 Alcock and Kemp

on ascending ridges are more attractive to territorial residents and nonterri- torial visitors than are open patches lower in altitude. Perched males wait on the ground at these places over an approximately 2-h period in the morning. They fly up to chase conspecific females and males as they arrive. Females may be captured and mated after the pair land. Rival males sometimes elicit rapid flights back and forth over the perching area, after which the newcomer often departs and the resident male returns to land. The result is that often only a single male occupies a site that may be as much as 18 m2 (Alcock and Schaefer, 1983), although at other times several males may perch within a few meters of one another, interacting at intervals without any one male displacing all the others. However, even resident defenders rarely exclude all other males from their landmark territory for more than a day; rapid turnover in territorial ownership is the rule. These generalizations are based largely on observations made during two spring flight seasons (1980 and 1981), with attention focused largely on a single hilltop along an ascending ridge in the Usery Mountains, Maricopa County, Arizona (Alcock and Schaefer, 1983). In the spring of 2002, we returned to the Usery Mountains to study the physiological parameters that affect territorial success in C. austeni. This work was conducted on a different ridge located higher in the mountains, approximately a kilometer from the 1980–1981 study site. In the course of this study and another in the fall of 2002, we collected additional data on the site fidelity and selection of perching territories by males of C. austeni. These data permit us to evaluate the robustness of earlier conclusions on the mating system of this species.

MATERIALS AND METHODS

We visited the study site, the highest ridge in the Usery Mountains, on 23 days between 31 March and 5 May 2002 and again on 20 days between 17 September and 20 November 2002. We selected a 175-m-long segment of the undulating ridgeline as the sample transect and measured the altitude of points 10 paces (ca. 7.5 m) apart along the transect using a handheld digital altimeter (Sportline Model EB 833). The entire transect was censused irregularly during spring 2002, with most effort focused on capturing and marking every male fly that arrived at two locations along the ridge, peaks 1 and 2 (at roughly pace 140 and pace 340; see Fig. 1). These high points along the ridge seemed to be favored by the flies in some other years. During censuses, we tried to capture every unmarked fly in an insect net shortly after its arrival. We then placed marks on the dorsum of the fly’s thorax, creating a distinctive combination of dots of Bic Wite-Out Correction Fluid or Liquitex acrylic paints. We released individually marked males near the P1: JLS Journal of Insect Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

Long-Term Mating System Stability 275

Fig. 1. An altitudinal cross section of the study transect on Usery Mountain with the number of census records totaled for each 30- pace segment of the transect. Bot flies appeared most regularly at the peaks along the ridgeline.

point of capture and noted the identity of any released flies that returned after first flying away. Marked males that perched at a given spot for at least 30 min were considered site-faithful residents. During the fall study period, we modified the procedure slightly by censusing the entire transect five times per morning (at 15-min intervals). During each census we recorded the location of every bot fly and again tried to net, mark, and release any new arrivals. The data collected enable us to determine the proportion of marked flies returning to the study area on the day they were marked and on subse- quent days. These measures were then compared with similar data gathered on 25 days scattered between 14 April and 20 May 1980. In addition, the census results identify the sites that the male bot flies considered most at- tractive, permitting us to test the proposition that higher altitude landmarks are indeed more popular with the flies than those lower in altitude.

RESULTS

Site Fidelity of Male C. austeni in Different Flight Seasons

TableI presents mark-resighting data for bot flies marked in 2002 during two flight seasons, spring and fall, for comparison with similar data collected from a nearby location in 1980. The differences in same-day site fidelity of the three populations of males marked in the spring and fall of 2002 are not statistically significant ( 2 = 1.54, df = 2, P > 0.40). Therefore, the data from all three 2002 populations were combined for comparison with the P1: JLS Journal of Insect Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

276 Alcock and Kemp

Table I. Mark–Resighting Data for Populations of Male C. austeni That Were Observed at Two Locations (Sites A and B) Over Three Flight Seasons Separated by 22 Years

No. of flies returning to the site where they were marked Site Marked flies on the same day

1980, lower site (spring) 89 49 (55%) 2002, upper site (spring) Peak 1 83 23 (28%) Peak 2 100 34 (31%) 2002, upper site (fall) 84 22 (26%) No. marked on day 1 of No. seen again on day 2 2 consecutive days of observation 1980, lower site (spring) 34 20 (59%) 2002, upper site (spring) 100 21 (21%) 2002, upper site (fall) 61 24 (39%) No. marked on day 1 with No. seen again on day 3 observations made 2 days later 1980, lower site (spring) 42 8 (19%) 2002, upper site (spring) 89 8 (9%) 2002, upper site (fall) 52 6 (12%)

spring 1980 figure. Flies were far less likely to return on the day of marking in 2002 than in 1980 ( 2 = 18.8, df = 1, P < 0.001). Not only were marked males less likely to return on the day of marking in 2002, but also they were less likely to appear again the day after they were marked (multiple-day site fidelity) than flies had been in 1980. This conclusion is based on data from males that were marked on the first of 2 consecutive days of observations in the different years. For these subsamples, the proportions of males known to return to the transect study site for 1980, spring 2002, and fall 2002 are statistically different ( 2 = 17.8, df = 2, P < 0.001). However, the differences in the proportion of males from the three populations that were known to have returned 2daysafter marking are not significant ( 2 = 2.75, df = 2, P > 0.05) based on data collected from just those males that were marked 2 days prior to a return survey of the study transects. Another way to compare male site fidelity across seasons is to exam- ine the proportion of census days in which one male was a clear territorial resident at the most popular sites. (Note that no more than one male was identified as a resident at any given site on any given day.) In 1980, the study centered on a single territory on peak X. This site attracted a resident male (one that occupied the site for at least 30 min) on 23 (92%) of the 25 days that peak X was visited by an observer. In contrast, during the spring of P1: JLS Journal of Insect Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

Long-Term Mating System Stability 277

2002, peak 1 was occupied by a resident on 10 (59%) of the 17 days when this site was thoroughly checked, while peak 2 produced a resident male on 13 (68%) of 19 days of close observation. Likewise, in the fall, peak 1 had a resident male on only 8 (40%) of 20 census days, while peak 2 had a resident on only 10 (50%) of these 20 days. If we combine data from the spring of 2002 and compare it with data from the fall of 2002, the difference is not significant ( 2 = 2.72, df = 1, P = 0.10). Lumping all data from 2002 and then comparing the combined data with the 1980 results, we find a highly significant difference between years ( 2 = 16.2, df = 1, P < 0.001).

Site Preferences of C. austeni

Male bot flies were found at peak X in the lower site on the 3 days in 2002 when this ridge was surveyed (on 5 and 25 April as well as 1 May), demonstrating that individuals continue to occupy the same landmark site that was known to be popular more than two decades previously. In the fall of 2002, data on the perch site preferences of male flies were documented by repeat censuses of the different perching territories at the upper site. Figure 1 shows the relationship between the local topography of the ridgeline study site and the total number of times during the fall censuses (n = 100) that each 30-pace segment was occupied by at least one bot fly. The flies were distributed in a highly nonrandom manner, with the vast majority concentrated on the three most elevated portions of the ridgeline transect. (If we compare the frequency histogram of the observed number of flies for each 10-pace segment with that expected from a Poisson-generated frequency, specifying seven groupings to maintain expected frequencies in excess of five, the observed distribution deviated significantly from random; 2, goodness-of-fit test = 84.3, df = 6, P < 0.0001.) Overall, the correlation between the altitude of transect segments and the number of censuses in which the segments were occupied by perching flies was statistically significant (r = 0.55, n = 13, P = 0.05). The same relationship held for number of males marked in the different segment with 60 (72%) of 83 bot flies captured from the three highest rises along the ridgeline. The pattern also applies for the number of resident males (30 min occupants) found in the different segments, with 18 (82%) of 22 known residents concentrated in the three elevated sections of the transect.

DISCUSSION

This study confirms that males of C. austeni exhibit same-day site fi- delity in that a substantial proportion of males that were captured, marked, P1: JLS Journal of Insect Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

278 Alcock and Kemp

and released, quickly returned to the point of capture. Moreover, some demonstrated multiple-day site fidelity by coming back to the study area on subsequent days, although the dropoff in returns over several days was steep. There was, however, a considerable difference in apparent site fidelity between the two years. In 1980 the proportion of males that returned on the same day to the marking site, as well as the proportion that came back the next morning, was substantially higher than in 2002. That this differ- ence was real is supported by the finding that at the upper site in 2002 resident males rather often failed to claim the two prime perching areas, whereas residents almost always materialized at peak X in the lower site in 1980. These differences in apparent site fidelity may have arisen be- cause of differences in climatic conditions between the two years (2002 was an extreme drought year) or topographic differences between loca- tions (no equivalent or higher peak occurs within hundreds of meters of peak X, whereas peaks 1 and 2 at the upper site are at essentially the same altitude and are separated by only 135 m of transect, less as the bot fly flies). Although the site fidelity of males was evidently not the same across years, most other aspects of male behavior were similar in the different study populations. Turnover rates were very high throughout, such that in both studies only 10–20% of marked males were seen just 2 days after mark- ing. Given that males cannot feed as adults and live at most 2 weeks, judging from mark–recapture studies of C. austeni (Alcock and Schaefer, 1983) and the congener C. latifrons (Catts, 1967), one imagines that males would ex- perience severe selection to maximize time spent at productive landmarks. The fact that even successful territory holders rarely spend more than a sin- gle day at the study site suggests that males (1) rapidly deplete their energy reserves, (2) experience a very high rate of mortality, and/or (3) regularly disperse to different landmark sites. Little information relevant to these alternatives exists, although we do know that energy availability is highly correlated with success in achieving and maintaining territorial residency in C. austeni (Kemp and Alcock, 2003). The 2002 study at the upper site also showed that males were drawn to locally elevated perching spots, just as observations at the ridge lower in the mountains demonstrated that elevated perching areas (relative to other nearby sites) were most consistently occupied (Alcock, 1984). One assumes that receptive females are far more likely to arrive at the portions favored by males, but unfortunately visits by females to the landmark perching ar- eas have been recorded only very infrequently. In spring 2002, one female appeared at Peak 1, while in the fall, the only record of an incoming female was made at Peak 2. P1: JLS Journal of Insect Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

Long-Term Mating System Stability 279

In sum, the description of the mating system of C. austeni made in 1980 largely applies to the flies studied in 2002 at a different location. Not only has this landmark-defending insect utilized the same perching areas over more than two decades, but also some males in 2002, as well as 1980, exhibited strong site fidelity to a given perch territory, although generally only for 1 day. This species is one of a relatively few whose mating system has been reexamined at an interval of decades (see also Alcock, 1981; Kemp and Alcock, 2003). The value of long-term studies is well recognized in standard evolutionary research (e.g, Grant et al., 1998) and in projects on the lifetime reproductive success of long-lived , such as red deer (Clutton-Brock, 1988). Behavioral studies of short-lived insects across generations are less common. In those studies of this sort that do exist, the typical finding is that little changes from generation to generation. Thus, for example, larger-than- average males of the bees Centris pallida and Amegilla dawsoni consistently experience a strong copulatory advantage in the female-defense mating sys- tems of these species (Alcock, 1984; Simmons et al., 2000). Consistency of this sort also characterizes the mating system of C. austeni, which supports the assumption that short-term studies can be trusted to produce a truly accurate picture of species-specific patterns of behavior.

ACKNOWLEDGMENTS

This research was supported in part by a Maytag Postdoctoral Fellow- ship to D.J.K. during his time at Arizona State University. We also thank Jeff Boettner for his helpful assistance with the literature.

REFERENCES

Alcock, J. (1981). Lek territoriality in a tarantula hawk wasp Hemipepsis ustulata (Hymenoptera: Pompilidae). Behav. Ecol. Sociobiol. 8: 309–317. Alcock, J. (1984). Convergent evolution in perching and patrolling site preferences of some hilltopping insects of the Sonoran Desert. Southw. Nat. 29: 475–480. Alcock, J. (1984). Long-term maintenance of size variation in populations of Centris pallida (Hymenoptera: Anthophoridae). Evolution 38: 220–223. Alcock, J., and Schaefer, J. E. (1983). Hilltop territoriality in a Sonoran desert bot fly (Diptera: Cuterebridae). Anim. Behav. 31: 518–525. Baird, C. R. (1987). Bionomics of Cuterebra austeni (Diptera: Cuterebridae) and its association with Neotoma albigula (Rodentia: Cricetidae) in the southwestern United States. J. Med. Entomol. 34: 690–695. Catts, E. P.(1967). Biology of a California botfly Cuterebra latifrons Coquillett (Diptera: Cuterebridae). J. Med. Entomol. 4: 87–101. Clutton-Brock, T. H. (1988). Reproductive Success, University of Chicago Press, Chicago, IL. P1: JLS Journal of Insect Behavior [joib] pp1253-joir-487714 May 28, 2004 7:15 Style file version Feb 08, 2000

280 Alcock and Kemp

Grant, B. S., Cook, A. D., Clarke, C. A., and Owen, D. F. (1998). Geographic and temporal vari- ation in the incidence in melanism in peppered moth population in America and Britain. J. Hered. 89: 465–471. Kemp, D. J., and Alcock, J. (2003). Lifetime resource utilization, flight physiology and the evolution of contest competition in territorial insects. Am. Nat. 162: 190–201. Simmons, L. W., Tomkins, J. L., and Alcock, J. (2000). Can minor males of Dawson’s burrowing bee, Amegilla dawsoni (Hymenoptera: Anthophorini) compensate for reduced access to virgin females through sperm competition? Behav. Ecol. 11: 319–325.