Behav Ecol Sociobiol DOI 10.1007/s00265-007-0542-8

ORIGINAL PAPER

To court or not to court: reproductive decisions by male fiddler crabs in response to fluctuating food availability Tae Won Kim & Kotaro Sakamoto & Yasuhisa Henmi & Jae C. Choe

Received: 3 August 2007 / Revised: 24 November 2007 / Accepted: 26 November 2007 # Springer-Verlag 2007

Abstract For males, courting and foraging are often behavioral alternatives, which take time and consume energy. When males have a possibility of mating with receptive females, there may be a behavioral trade-off between courtship and feeding; the outcome of which may be affected by male physiological condition and food availability. Although many mathematical models and empirical studies suggest that the expression of male courtship signals are condition-dependent, decisions about courtship and mating strategies in relation to food availability have not attracted much attention. In this study, we tested whether daily changes in food availability affect males’ decisions about whether to court. We conducted experiments with the fiddler crab Uca lactea by providing males with additional food every other day. In foodCommunicated by P. Backwell T. W. Kim School of Biological Sciences, Seoul National University, Seoul 151-747, South Korea e-mail: [email protected] T. W. Kim : J. C. Choe (*) Division of EcoScience, Ewha Womans University, Seoul 120-750, South Korea e-mail: [email protected] K. Sakamoto Graduate School of Science and Technology, Kumamoto University, Kusrokami, Kumamoto 860-8555, Japan Y. Henmi Aitsu Marine Station, Center for Marine Environment Studies, Kumamoto University, Matsushima Kami-Amakusa, Kumamoto 861-6102, Japan e-mail: [email protected]

supplemented enclosures, males did not increase courtship activity on the days when food was supplemented. However, they built more courtship structures (semidomes) and waved more on the days when they were not given additional food. Male size had a strong influence on the number of days the males courted. We also tested whether the frequency of surface mating, as an alternative reproductive tactic, decreased when food was supplemented. Contrary to our expectation, the number of males that exhibited the surface-mating tactic increased when food was supplemented whereas the number of mate-searching females did not change. Our findings in this field study suggest that reproductive decisions by male fiddler crabs are affected by fluctuating food availability and present body condition, and the alternative mating tactic of this species may be more frequently used by males under good condition. Keywords Condition dependence . Courtship . Feeding . Fiddler crab . Food availability . Reproductive decision . Surface mating . Trade-off

Introduction A male animal that acquires a mate via female choice must spend substantial quantities of time and energy for courtship to attract females (Andersson 1994; Choe and Crespi 1997). However, males must also forage to obtain energy for courtship and survival. Thus, there may be tradeoffs between reproductive activity and foraging in environments with varying food availability (Stearns 1992; Abrahams 1993). The optimal behavior will depend on the fitness returns of each behavior. When animals are presented with the choice of whether to forage or court,

Behav Ecol Sociobiol

males may give up a mating opportunity if the benefits gained from foraging exceed those gained from courtship (Morell 2004). When males are confronted with unpredictable environments with varying food availability, how do they decide whether to court or feed? Observing behavioral responses to changing environments may be a key to understanding the decision-making processes in animals (Gass 1985). Only a few empirical studies have dealt with the decision whether to court or feed, despite the fact that feeding is a significant behavioral alternative to courting (Abrahams 1993; Griffiths 1996). Males may choose their mating tactics according to food availability, following one mating tactic in good condition, but choosing an alternative in poor condition (Thornhill 1981; Kolluru and Grether 2005). A growing body of studies provide evidence of the condition dependence of sexual advertisements (for review, see Johnstone 1995; Cotton et al. 2004). Few studies, however, have investigated the changes in male reproductive strategies in response to the changes in food availability. Previous studies in fiddler crabs suggested that male courtship signals are condition-dependent (Backwell et al. 1995; Jennions and Backwell 1998; Kim and Choe 2003). In Uca lactea, Kim and Choe (2003) found that food supplementation not only increases male courtship intensity but also makes males initiate courtship earlier. In this study, we explore how males decide whether to court and how they change their reproductive tactics in response to fluctuations in food availability. The semiterrestrial fiddler crab U. lactea is an excellent species for studying the behavioral trade-off between feeding and courtship and the variation in reproductive strategies according to food availability. Males have a single exaggerated white claw, which is used for courting females and fighting against neighboring males. On the other side is a small cheliped, which is used only for feeding. They generally feed on organic particles in the sediments around their burrows after the tides recede. Although they can feed and court concurrently (wave–feed) during the mating season, their feeding and courtship activities usually occur at discrete times during a semilunar cycle (Kim et al. 2004a). They allot most of their time to courtship when females are ready to breed, but devote more time to feeding in preparation for the next reproductive periods when females are not ready to breed (Kim et al. 2004a). This species has two mating tactics that correspond to the differences in mating location. In burrow mating, males build semidomes as sexual signals (Kim et al. 2004b) and wave their claws to attract females (Murai et al. 1987; Yamaguchi 2001) and females visit several males’ burrows before choosing a mating partner. Males who successfully attract females into their burrows generally plug the burrow entrance with mud, keeping themselves and the female in

the burrow chamber for 1–5 days as a form of mate guarding (Goshima and Murai 1998). After the males emerge from the burrows, females continue to incubate their eggs in the burrow chambers for 2 weeks before releasing the larvae (Yamaguchi 2001). In the case of surface mating, however, a male directly approaches a neighboring female’s burrow and taps on the burrow opening until she comes out, then they mate on the surface. Some females do not respond to these solicitations. Surface-mating females occasionally mate with several males (T. W. Kim, personal observation). Surface mating may be an alternative male mating tactic when the number of mate-searching females is low (Murai et al. 1987; Kim et al. 2006) or predation risk is high (Koga et al. 1998). But it could also be a mating tactic used by males that are unlikely to successfully attract females to their burrows (Kim et al. 2006). Males in poor condition may try to copulate with females without courting because they do not have enough energy to win the competition among courting males. The objective of this study was to determine whether daily fluctuations in food availability affect male courtship decisions and mating strategies. If food availability is an important determinant of male behavioral decisions, then males should adjust their reproductive tactics when food availability changes. We predict that males in foodsupplemented enclosures will spend less time courting on days when they are given additional food than on days when they are not, but that food-supplemented males should court more on the days when they are not given additional food than males never given supplemented food. This is because males should benefit more by investing their time in feeding on days when food is supplemented. We also predict that food-supplemented males will perform less surface mating than males without additional food because males in good condition should court females for burrow mating.

Materials and methods Study site and species The study was carried out from June 12 to July 30, 2003 at the Nagaura intertidal sandy mudflat in Amakusa-Matsusima, Kumamoto, Japan. The maximum tidal amplitude during the study period was approximately 4 m (from 0.2 to 4.2 m in tidal height). A population of U. lactea inhabits the upper intertidal mudflat and the area is usually covered twice by high tide each day. Crabs emerge from their burrows and become active as soon as the tide recedes from their habitat. The duration of habitat exposure is approximately 4–7 h, depending on the tidal period. Male courtship begins in late June and ends in early August at the field site.

Behav Ecol Sociobiol

Enclosure setup In June 2003, we set up four 1.5 m square enclosures of 0.3 m height in the middle of the habitat of U. lactea. Each enclosure was made from soft plastic netting used to make garden fences. Thirty centimeter plastic bars were inserted into the netting at 50-cm intervals to provide structural support. A 5-cm strip of plastic tape was sewn to the upper edge of the mesh to prevent crabs from climbing out of the enclosures. The enclosures were inserted 15 cm into the sediment and withstood the tides every day. We captured all males inside the enclosures using two methods that do not disturb the crabs’ activities. In the first method, we placed a 1.5-m long plastic tube in front of a crab’s burrow entrance and waited for the crab to emerge from the burrow opening. Then, we quickly moved the tube over the burrow opening and the bewildered crabs were caught easily by hand. A second, newly developed method using a device named the “Sakamoto drop” was also employed (Fig. 1). The device, consisting of a frame supporting a movable wooden stick, was placed around the burrow and was activated so that the wooden stick inside the device plugged the burrow opening after the crabs exited their burrows, permitting easy capture of the crabs. After capturing the crabs, we measured their carapace width, length, and claw length to the nearest 0.1 mm using vernier calipers. We marked each crab on the carapace using paint marker pens (Mitsubishi Paint Marker). Each individual was marked with a unique combination of four different colored dots on each of four regions of the carapace. Instant sticky glue was coated on the painted dots after marking to prevent the dots from being erased by abrasion. Forty males in each enclosure were marked and the remaining unmarked males were removed from the enclosures. The marked crabs were released to their original burrows. Mean male size did not significantly differ among the enclosures (carapace width: F3,155 =2.286, P=0.081; major claw length: F3,152 =1.945, P=0.124). We did not capture all of the females but counted the number of females active on the surface. There were approximately 40

Fig. 1 Schematic feature of the crab-capturing device “Sakamoto drop.” The device is composed of a small pole fitting inside a 15-cmlong tube, which is supported by three legs. A researcher holds the pole about 10 cm above the burrow by pulling a string tied to the pole.

females in each enclosure, but the number of individuals active on the ground varied from day to day. Food supplementation We started the experiments 1 day after the day when low tides fell at the time of sunset or sunrise. As soon as the tides exposed the habitat, 10 g of sardine flake (from a fishing market) mixed with 3 l of seawater was added to two of the enclosures and 3 l of seawater was added to the other two enclosures. The food treatment was repeated every other day for 10 days (treatment for 5 days and nontreatment for 5 days) to determine the effects of daily fluctuations in the food supply. The treatment was finished before the day when low tides fell again on sunset or sunrise. Each enclosure was divided into four sectors. Two researchers collected behavioral data simultaneously from different enclosures. Each researcher observed one of two enclosures (one supplemented with food and the other with seawater only) for approximately 30 min and then switched to the other enclosure for 30 min every hour. We repeated this procedure three times (for a total of 3 h). During the first observation session of each day, we plotted the location of the burrow of each male in the enclosure on a map using the male’s identifying back marking. The following aspects of each male’s activity were recorded every hour: whether the crab was active on the surface, whether he built a semidome, and whether he waved. If the male displayed at least one bout of waving (waved his claw continuously for a few seconds), we regarded the male as a “waver.” Our decision to divide each enclosure into four sectors may have led to a pseudoreplication problem. This method was nonetheless chosen for practical reasons. We could not set up more than four experimental enclosures because it would be impossible to monitor all the males’ activity in a larger number of enclosures of the same size. An alternative would be to decrease the size of the enclosures and increase the number of replicates. However, females in smaller enclosures do not behave naturally and search for male partners.

When a crab comes out of the burrow and moves a sufficient distance (①) the researcher releases the string to drop the pole, which plugs the entrance of the burrow (②). The crab can then be easily caught by hand

Behav Ecol Sociobiol

equal between-group correlations and group variances (“sphericity”) was violated, we applied the Greenhouse– Geisser correction of P values (Zar 1999, p. 259). In the food-supplemented enclosures, rates of male courtship fluctuated depending on the daily food treatment (Fig. 2). On food-treatment days, there was no difference in the number of waving males (Table 1) or semidomes built (Table 2) between the control and food-supplemented conditions, but there was a significant effect of the interaction between day and food treatment on the number of semidomes built. This indicates that males increased semidome-building activity on the days of food supplementation over the course of the experiment. On nontreatment days, males in the food-supplemented enclosures courted (waved and built semidomes) more than males in the control enclosures (Tables 3 and 4). Changes in individual courtship activity in response to variation in food availability To examine the influence of food supplementation on individual activities, we compared individual activities (activity days, waving days, and the number of semidomes built) between the same treatment enclosures. For each individual, we defined “activity days” as the number of days that the crab was active on the surface of the mudflat and “waving days” as the number of days during which the

a

8 7

No. of semidomes/sector

To determine whether food supplementation influences the surface-mating tactic, we recorded the frequency of surface-mating attempts by each male and the outcome of those attempts (rejection or acceptance of mating by the female) in each enclosure during each 30-min observation period. Because changes in the number of mate-searching females or females active on the ground may influence male mating tactics, we also counted the number of active females and mate-searching females. When a female visited a male burrow following a waving male, we designated her as a mate-searching female. We plotted the routes of matesearching females on the map. We counted the number of females active on the surface during each 30-min observation period and used the maximum number recorded during each interval as the number of active females. In addition to those periods, we continued to observe female matesearching and surface-mating attempts (and the outcomes of those attempts) by males until the tides covered their habitat. To see the effect of food on the frequency of surface mating, one cycle did not provide sufficient data. Therefore, to increase the statistical power, we repeated the experiment for one more semilunar cycle after a 3-day pause. Treatments were exchanged between the enclosures (i.e., previously food-supplemented enclosures became controls, and previous controls were food-supplemented), and food treatment was conducted in the same way for 8 days. We observed that they did not mate on the other 2 days in the first 10-day cycle. Thus, we did not include these days in our analysis, but duplicated the data collection procedure for just the eight mating days in the second cycle. Each statistical method applied to analyze the data is explained at the head of each section of the results. All tests were two-tailed; the significance level was set at 0.05. All means are presented ±1 SD and the analyses were conducted using SPSS 11.0 and Statview 5.01 for Windows (SAS Institute).

6 5 Control Food-supplemented

4 3 2 1 0 8

9

10

11

12

13

14

15

16

17

18

July

Results

b 10

Repeated-measures ANOVA was used to determine if food supplementation influenced the number of wavers and the number of semidome builders in each sector. Data were divided into two groups based on whether the enclosure was food-treated that day. We applied separate repeatedmeasures ANOVAs for treatment days and nontreatment (post treatment) days in the food-treated and control sectors, thus producing two repeated-measures analyses for each behavioral measurement. When the assumption of

9

No. of wavers/sector

Changes in courtship activity per sector in response to variation in food availability

8 7 6

Control

5

Food-supplemented

4 3 2 1 0 8

9

10

11

12

13

14

15

16

17

18

July

Fig. 2 Distribution of the mean (±SD) number of semidomes (a) and the number of wavers (b) per sector (0.75×0.75 m). Circled days represent days when food was supplemented

Behav Ecol Sociobiol Table 1 Repeated-measures ANOVA results for the experimental treatment effect on the number of wavers (on treatment days)

Table 3 Repeated-measures ANOVA results for the experimental treatment effect on the number of wavers (on nontreatment days)

Source

Source

Within-subject effect Day Day×food treatment Error Between-subject effect Food treatment Error

Mean square

df

F

P

47.291 7.265 4.692

2.398 2.398 33.578

10.079 1.548

<0.001 0.225

1.013 2.755

1 14

0.367

0.554

Within-subject effect Day Day×food treatment Error Between-subject effect Food treatment Error

Mean square

df

F

P

21.478 9.654 3.288

2.960 2.960 41.44

6.532 2.936

0.001 0.045

74.113 3.705

1 14

20.001

0.001

“Day” represents the day from the starting day of food treatment. “Food treatment” represents food supplementation or control treatment.

male waved his claw to females. We pooled data from both enclosures within the same treatment because there was no significant difference in the number of activity days (foodsupplemented: t=0.800, df=78, P=0.129; control: t=−0.250, df=78, P=0.698), the number of semidomes built (foodsupplemented: t=1. 194, df=78, P=0.236; control: t=1.434, df=78, P=0.156), or the number of waving days (foodsupplemented: t=0.448, df=78, P=0.655, control: t=−0.800, df=78, P=0.129) in enclosures with the same treatment. Data were divided into two groups of 5 days (treatment days vs nontreatment days) based on whether food was added on that day for evaluation of the daily effects of food treatment. The numbers of semidomes and waving days for each male on the five treated days and five untreated days were summed separately, and then we used an unpaired t test to test for differences in the mean number of semidomebuilding and waving days between treatments. There was no difference in the mean number of activity days per individual between treatments (food-supplemented: 6.4±2.9, control: 6.3±2.4; t=0.242, df=158, P= 0.809). On food-treated days, the total number of semidomes built per individual was not significantly different for food-supplemented and nonsupplemented (t=−0.913, df=158, P=0.363) (Fig. 3a). The mean number of waving days per individual was also not significantly different (t=−0.836, df=158, P=0.404) (Fig. 3b). On untreated days, however, food-supplemented males built more semidomes (t=−3.379, df=158, P<0.001) (Fig. 3a) and waved for

more days (t=−3.057, df=158, P<0.01) (Fig. 3b) than did nonsupplemented males. When the data from food-treated days and untreated days are combined, food-supplemented males waved for more days than nonsupplemented males did (t=−2.154, df=158, P<0.05). However, the number of semidomes built per male did not differ significantly between the food-supplemented and nonsupplemented enclosures (t=−1.757, df=158, P=0.08). Does size affect reproductive tactics? To test whether male size influenced the number of days of courtship and surface-mating attempts, we used ANCOVA with male carapace width as a covariate, food treatment (food supplementation and nonsupplementation) as a fixed factor, and the number of semidome-building and waving days as dependent variables. For this test, we used data from both treatment and nontreatment days. Analysis by ANCOVA indicated that carapace width affected the number of semidomes built and waving days, but that neither food supplementation nor the interaction between food supplementation and carapace width affected either variable (Tables 5 and 6). Male carapace width and the number of waving days were positively correlated in both food-supplemented (r = 0.455, F 1,78 = 20.316, P<0.0001) and control enclosures (r=0.374, F1,77 =12.484, P=0.0007) (Fig. 4a). The total number of semidomes built was also positively correlated with male carapace width in both food-supplemented (r = 0.371, F1,78 = 12.457, P=

Table 2 Repeated-measures ANOVA results for the experimental treatment effect on the number of semidomes (on treatment days)

Table 4 Repeated-measures ANOVA results for the experimental treatment effect on the number of semidomes (on nontreatment days)

Source

Source

Within-subject effect Day Day×food treatment Error Between-subject effect Food treatment Error

Mean square

df

F

P

18.437 4.325 1.388

4 4 56

13.28 3.115

<0.001 0.022

1.513 3.063

1 14

0.494

0.494

Within-subject effect Day Day×food treatment Error Between-subject effect Food treatment Error

Mean square

df

F

P

7.575 2.675 2.796

4 4 56

2.709 0.957

0.039 0.439

76.050 8.511

1 14

8.936

0.010

Behav Ecol Sociobiol Table 6 Analysis of covariance for the effects of food treatment and carapace width on the number of waving days

4 3.5 3 2.5 2

Control Supplemented

1.5 1

Mean square

F

P

Food Carapace width Food×carapace width Error

1 1 1 155

0.520 148.802 2.384 4.605

0.113 32.31 0.518

0.737 <0.0001 0.472

On untreated days

4 3.5 3 2.5 Control

2

Supplemented

1.5 1 .5 0 On treated days

On untreated days

Fig. 3 Mean (±SD) number of semidomes built (a) and number of waving days (b) per male on treated and untreated days

0.0007) and control enclosures (r=0.363, F1,77 =11.657, P= 0.001) (Fig. 4b). There was no effect of crab size or the interaction between size and food treatment on the frequency of surface mating (Table 7). The effect of food availability on surface mating For the statistical analyses of the surface-mating frequency, we used 8-day data sets for both semilunar cycles because mating occurred on only 8 days. To test whether food availability influenced the frequency of surface mating, we designated food treatment (food supplementation and nonsupplementation), period (first and second semilunar period), and treatment (treatment and nontreatment days) as fixed factors for the analysis by ANOVA. We used the same method to test for effects of food availability on the number of females active on the surface and the number of mate-searching females. Contrary to our prediction, food-supplemented males attempted surface mating more frequently than nonsupple-

mented males did (Table 8; 2.5±2.0 vs 0.9±1.1; F1,62 = 14.189, P=0.0004). There was no significant difference in the frequency of surface-mating attempts between foodtreatment and nontreatment days (Table 8). The female rejection rate of surface-mating attempts was not significantly different between treatments (food-supplemented: 23.9± 30.1%, control: 10.4±22.7%, F1,40 =2.378, P=0.131). Consequently, there were more surface-mating pairs in the foodsupplemented enclosure than in the control enclosures in both semilunar periods (Fig. 5; F1,62 =9.589, P=0.003). The numbers of mate-searching females (F1,62 =2.004, P=0.161) and of females active on the surface (F1,62 =0.333, P=0.565) did not significantly differ between food treatments.

a 12 10

No. of Semidomes

On treated days

No. of waving days/male

df

.5 0

b

Factor

8 6

Control

4

Supplemented

2 0 -2 9

10

11

12

13

14

15

16

17

Carapace width(mm)

b 10 9 8 7

Waving days

No. of Semidomes/male

a

6 5

Control

4

Supplemented

3 2 1

Table 5 Analysis of covariance for the effects of food level and carapace width on the number of semidome-building days

0 -1 9

Factor

df

Mean square

F

P

Food Carapace width Food×carapace width Error

1 1 1 155

0.258 110.786 0.023 4.606

0.056 24.05 0.005

0.813 <0.0001 0.944

10

11

12

13

14

15

16

17

Carapace width (mm)

Fig. 4 The relationship between carapace width and number of semidomes (a) and waving days (b). Number of semidomes: supplemented (bold line), Y=−3.456+0.537×carapace width; control (thin line), Y=−4.123+0.522×carapace width. Waving days: supplemented (bold line), Y=−5.039+0.691×carapace width; control (thin line), Y=−4.094+0.536×carapace width

Behav Ecol Sociobiol Table 7 Analysis of covariance for the effects of food treatment and carapace width on the frequencies of surface-mating attempts by males

a

10

Food Carapace width Food×Carapace width Error

df 1 1 1 155

Mean square 0.016 1.211 0.172 0.455

F 0.035 2.663 0.379

P 0.851 0.105 0.539

Frequency of SM

9

Factor

8 7 6

Supplemented

5

Control

4 3 2 1 0 9

Discussion

10

11

12

13

14

15

16

17

18

July

Table 8 ANOVA results for the effects of food level, period, and treatment on the number of surface-mating attempts by males Factor

df

Mean square

F

P

Food Period Treatment Food×period Food×treatment Period×treatment Food×period×treatment Error

1 1 1 1 1 1 1 56

39.063 9.000 0.063 3.063 0.250 7.563 1.000

14.608 3.366 0.023 1.145 0.093 2.828 0.374

0.003 0.071 0.879 0.289 0.760 0.098 0.543

b 10 9

Frequency of SM

Our results show that courtship decisions of male fiddler crabs vary in response to daily fluctuations in food availability. On days when males were provided with additional food, they did not show a significant change in courtship behavior compared with males in the control enclosures. On the days after food supplementation, however, males exhibited a significant increase in courtship behaviors such as claw-waving and semidome-building. The results suggest that males respond to changes in food quality and/or quantity by deciding whether to court based on the trade-off in costs and benefits between courting and feeding. Our results did not conform perfectly to our prediction that males would invest more time in feeding (and less in courting) on days when they were presented with additional food. While there was no effect of food supplementation on semidome-building activity overall, there was a significant interaction between food treatment and day. It seems that males track short-term fluctuations in food availability until they obtain enough energy for courtship. After they obtain enough energy for courtship, time investment for feeding may decrease. Previous studies of various animals have shown that sexual advertisement by males is condition-dependent and increases linearly with food quality and/or quantity (e.g., Kotiaho et al. 1998; Griffith et al. 1999; David et al. 2000). However, these studies did not examine changes in courtship decisions in response to short-term variation in the food supply. Our study indirectly showed that males

8 7 6

Supplemented

5

Control

4 3 2 1 0 23

24

25

26

27

28

29

30

31

July

Fig. 5 The distribution of frequencies of surface mating in the first semilunar cycle (a) and in the second semilunar cycle (b)

adjust their courtship and feeding behavior in response to changes in food quantity. Deposit-feeding crustaceans like fiddler crabs could be considered to be simple foragers that follow feeding and reproductive schedules to be dictated by their usual environments. However, our results suggest that these animals have evolved a rather complex decisionmaking mechanism allowing them to respond flexibly to daily changes in food abundance. Our results confirm the results of previous studies suggesting that male courtship signals in fiddler crabs are influenced by food availability (Backwell et al. 1995; Jennions and Backwell 1998; Kim and Choe 2003). However, the ANCOVA results suggest that the major factor which influences courtship activity is not food availability but body size. Although food-supplemented males courted more intensely, the size dependence of courtship behavior was not affected by food availability. In a previous study, male size did not influence male reproductive activity (Kim and Choe 2003). Why was there a significant correlation between male size and the number of courtship days in this study? The different results of the two studies might be due to a difference in the sample size. In this study, we obtained higher statistical power by increasing the sample size. Larger males can deter other smaller males from courting by threatening them with their large claw (T. W. Kim, unpublished data; Zucker 1984). Accordingly, male size appears to influence neighboring males’ reproductive activity and to determine the males’ own reproductive tactics.

Behav Ecol Sociobiol

Contrary to our prediction, male fiddler crabs did not reduce the frequency of surface-mating attempts when they were given additional food. We had predicted that surface mating might be an alternative mating tactic used in conditions of food scarcity when males do not have enough energy to court. In our test, however, males in foodsupplemented enclosures attempted surface mating more than males in control enclosures. It is possible that the frequency of surface mating is not controlled by males but by females. However, we did not find any supporting evidence of behavioral changes in females in food-supplemented enclosures that might have influenced male mating tactics. First, neither the number of mate-searching females nor the number of females on the surface differed between treatments in our study. Second, there was no difference in the female rejection rate for surface mating between treatments. Given these results, it would be hard to conclude that female behavior alone determines male mating tactics although it is possible that both males and females play a role in determining the rate of surface mating. Surface-mating females have a high probability of mating with multiple males. Therefore, paternity assurance is lower for surface-mating males than for males who mate in their burrows. However, if the benefit from obtaining more mates via surface mating exceeds the cost of losing assured paternity, surface mating may be favored. Males in good condition can increase their mating opportunities by increasing both courtship and mate-searching activity. deRivera and Vehrencamp (2001) found that the mating system of Uca is also related to sediment grain size. Surface mating is more common in sediments with small grain size. When the grain size is smaller, food is more abundant (e.g., Stamhuis et al. 1998). These results are consistent with our results in suggesting that the evolution of mating systems in fiddler crabs might be highly related to food availability of their habitat. In particular, food availability appears to play an important role in determining male reproductive decisions. In conclusion, we found that small animals like crabs can choose their reproductive and feeding tactics based on current food availability and their energetic needs. We believe that this research will open a new avenue of exploration for studies of animal decisions in variable environments. Acknowledgements T. W. Kim is grateful to the Aitsu Marine Station of Kumamoto University for the supporting facilities and convenience during the study. We are very thankful to Mark Abrahams, Susan Lappan, Virginia Weaver, Dennise Pope, and five anonymous reviewers for the valuable comments on the manuscript. This research was supported by Shilla Chemical and Brain Korea 21 research fellowship from the Korean Ministry of Education and Human Resource Development. Amore Pacific Research and Cultural Foundation provided a special research fund during the publication of this study. Jae C. Choe was funded by Ewha Womans University.

References Abrahams MV (1993) The trade-off between foraging and courting in male guppies. Anim Behav 45:673–681 Andersson M (1994) Sexual selection. Princeton University Press, London Backwell PRY, Jennions MD, Christy JH, Schober U (1995) Pillar building in the fiddler crab Uca beebei: evidence for a conditiondependent ornament. Behav Ecol Sociobiol 36:185–192 Choe JC, Crespi BJ (eds) (1997) The evolution of mating systems in insects and arachnids. Cambridge University Press, Cambridge Cotton S, Fowler K, Pomiankowski A (2004) Do sexual ornaments demonstrate heightened condition-dependent expression as predicted by handicap hypothesis? Proc R Soc Lond B Biol Sci 271:771–783 David P, Bjorksten T, Fowler K, Pomiankowski A (2000) Conditiondependent signaling of genetic variation in stalk-eyed flies. Nature 406:186–188 deRivera CE, Vehrencamp SL (2001) Male versus female mate searching in fiddler crabs: a comparative analysis. Behav Ecol 12:182–191 Gass CL (1985) Behavioral foundations of adaptation. In: Bateson PPG, Klopfer P (eds) Perspectives in ethology, vol. 6: mechanisms. Plenum, New York Goshima S, Murai M (1998) Mating investment of male fiddler crabs, Uca lactea. Anim Behav 36:1249–1251 Griffith SC, Owens IPF, Burke T (1999) Environmental determination of a sexually selected trait. Nature 400:358–360 Griffiths SW (1996) Sex differences in the trade-off between feeding and mating in the guppy. J Fish Biol 48:891–898 Jennions MD, Backwell PRY (1998) Variation in courtship rate in the fiddler crab Uca annulipes: is it related to male attractiveness? Behav Ecol 6:605–611 Johnstone RA (1995) Sexual selection, honest advertisement and the handicap principle: reviewing the evidence. Biol Rev 70:1–65 Kim TW, Choe JC (2003) The effect of food availability on the semilunar courtship rhythm in the fiddler crab Uca lactea (Brachyura:Ocypodidae). Behav Ecol Sociobiol 55:210–217 Kim TW, Kim KW, Srygley RB, Choe JC (2004a) Semilunar courtship rhythm of the fiddler crab Uca lactea in a habitat with great tidal variation. J Ethol 22:63–68 Kim TW, Christy JH, Choe JC (2004b) Semidome building as sexual signaling in the fiddler crab Uca lactea (De Haan) (Brachyura: Ocypodidae). J Crustac Biol 24:673–679 Kim TW, Kim TK, Hong SK, Choe JC (2006) Surface mating as an alternative mating strategy in the fiddler crab Uca lactea. J Ecol Field Biol 29:49–53 Koga T, Backwell PRY, Jennions MD, Christy JH (1998) Elevated predation risk changes mating behaviour and courtship in a fiddler crab. Proc R Soc Lond B Biol Sci 265:1385–1390 Kolluru GR, Grether GF (2005) The effects of resource availability on alternative mating tactics in guppies (Poecilia reticulata). Behav Ecol 16:294–300 Kotiaho JS, Alatalo RV, Mappes J, Nielsen MG, Parri S, Rivero A (1998) Energetic costs of size and sexual signaling in a wolf spider. Proc R Soc Lond B Biol Sci 265:2203–2209 Morrell LJ (2004) Are behavioral trade-offs all they seem? Counterintuitive resolution of the conflict between two behaviours. Behav Ecol Sociobiol 56:539–545 Murai M, Goshima S, Henmi Y (1987) Analysis of the mating system of the fiddler crab, Uca lactea. Anim Behav 35:1334–1342 Stamhuis EJ, Videler JJ, de Wilde PAWJ (1998) Optimal foraging in the thalassinidean shrimp Callianassa subterranean—improving food quality by grain size selection. J Exp Mar Biol Ecol 228:197–208

Behav Ecol Sociobiol Stearns SC (1992) The evolution of life histories. Oxford University Press, New York Thornhill R (1981) Panorpa (Mecoptera: Panorpidae) scorpionflies: Systems for understanding recourse defense polygyny and alternative male reproductive effort. Ann Rev Ecolog Syst 12:355–386

Yamaguchi T (2001) The mating system of the fiddler crab, Uca lactea (Decapoda, Brachyura, Ocypodidae). Crustaceana 74:389–399 Zar JH (1999) Biostatistical analysis. 4th ed. Prentice Hall, New Jersey Zucker N (1984) Delayed courtship in the fiddler crab Uca musica terpsichores. Anim Behav 32:735–742