Journal of lnsect Behavior, Vol. 7, No. 3, 1994

Size-Related Mating and Mate Guarding in the Orb-Web Spider Nephila clavata (Araneae,

Araneidae) Tadashi Miyashita l Accepted October 11, 1993; revised November 18, 1993

The orb-web spider Nephila clavata satisfies three conditions for assortative mating proposed by Ridley (The Explanation of Organic Diversity. The Comparative Method and Adaptations for Mating, Clarendon, Oxford, 1983); (1) a large male advantage in male-male competition, (2) a correlation between female size and fecundity, and (3) a long pairing duration. To test Ridley's hypothesis, size assortative mating and guarding were examined in the field. When data were pooled over time, assortative mating was found but this was due to temporal covariation of body sizes of males and receptive females. After controlling for the effect of time, size assortative guarding was not detected, although females guarded by males were larger than those not guarded in the early breeding season. Possible reasons for the absence of size assortative guarding were discussed. KEY WORDS: mating; mate guarding; male choice; spider.

INTRODUCTION Although size assortative mating has been reported in many arthropods in nature (e.g., McCauley and Wade, 1978; Ridley, 1983; McLain and Boromisa, 1987; Fairbaim, 1988; Crespi, 1989), the mechanisms are not fully understood. Ridley (1983) proposed that assortative mating is expected when the following three conditions are satisfied: (1) a large male advantage in male-male competition, (2) a correlation between female size and fecundity, and (3) a long pairing duration. Other conditions that could cause assortative mating have also been proposed (e.g., Adams and Greenwood, 1987; Crespi, 1989). ~Laboratoryof Forest Zoology,Facultyof Agriculture, Universityof Tokyo,Tokyo 113, Japan. 289 0892-7553/94/0500~0289507.00/0 © 1994 Plenum Publishing Coq~oration

290

Miyashita

Adult males of the orb-web spider Nephila clavata show intense male-male competition on female webs, and mating success increases with male size (Miyashita, 1993). Also, there is a strong positive correlation between female body size and fecundity, and fecundity varies greatly among individuals (Miyashita, 1986). Furthermore, mate guarding lasts for several days (Miyashita, 1993). Thus, this spider satisfies Ridley's three conditions for size assortative mating. In this paper, first, I compare the body size of females with and without males on female webs to detect the possibility of male choice. Next, I examine the correlations of body size between males and females during mating and during guarding. MATERIALS AND METHODS

N. clavata is an univoltine species in Japan. After overwintering in the egg stage, eggs hatch in May and spiderlings emerge from egg cocoons in early June in central Japan. Spiders reach sexual maturity from late August to early October. Oviposition takes place from mid-October to November, and only one egg sac is produced by a female. Immature spiders spin their own webs, while adult males abandon webs after maturation and cohabit with females on their webs. Field observations were conducted at the Tokyo University Forest Experimental Station in Tanashi in Tokyo, from late August to October. I established one site (site A) in 1989 and two sites (sites A and C) in 1990. Site A (14 x 3 m in area) was a young pine plantation (Pinus densiflora) with a tree height of about 3 m. Site C (18 × 10 m) was a garden with rather sparse vegetation and consisted of maple (Acer spp.) and pine trees. The population density at this site was less than one-tenth of that at site A. One to six censuses were taken every day between 1000 and 1700 h except on rainy days, which could detect all matings of males with just-molted females since such matings lasted for about 2 days (Miyashita, 1993). All females and most males were marked on the dorsal abdomen and/or legs with paint, and their body length (cephalothoraxabdomen length) was measured. The position of males on the female web, i.e., whether they were near the hub position (hub male) or at the periphery (peripheral male), was recorded because only hub males mated with newly molted females. If males touched the female epigynum with their palps, I checked whether hematodochal bulb contraction occurred. If this was observed, it was regarded as a copulation. RESULTS Since mating occurred infrequently on a given day in this species, I used guarding pairs to assess the possibility of male choice. Table I compares the body lengths of females with and without males on their webs. Overall differ-

13.5 + 1,5 (9) 15.2 -+ 3.2 (5) 1 3 . 0 + 1.6(6) 14.5 + 2.6 (10) 17.3 5 : 3 . 2 (11) 19.4 _+ 3,1 (10)

19.75:4,0(6) 20.55:4,2(9)

20.1+3.3(11) 20.4_+2.5(8)

-

14.7 _+ 2.3 (9) 16.0 + 2.5 (11) 17.6_+2.3(10) 17.5 -+ 2.3 (I I) 18.3 + 3.1 (9) 18.5 _+ 3.2 (9)

+

ns ns

ns as <0.01 <0.02 ns as

P

6 10 14 19 25 Oct. 2 9

Sept.

Date

_+ 2.4 (5) -+2.3(9) _+ 1.9 (8) ± 1.8 (15) _+ 2.0(13)

15.3-+2.8(13) 15.95:2.9(11)

14.3 13.5 14.4 14,6 14.9

+

-

1.9 (16) 1.7(12) 1.9 (15) 1.6 (10) 1,7(7)

14.5_+2.5(10) 16.5_+3,5(10)

11.0 + 10.6-+ 11.3 _+ II.0 + 11.5 +_

90A

ns ns

<0.03 <0.01 <0.01 <0.005 <0.005

P

10 14 19 25 Oct. 2 9

Sept.

Date

14.4_+1.8(6) 15.6_+1.6(4)

12.0-+0.8(5) 12.0 + 1.2 (8) 13.3 5 : 1 . 2 (8) 12.2 5:1.7 (6)

+

+ + -+ 5:

-

1.2(9) 1.4 (9) 1.1 (5) 1.2(6) 14.2_+1.0(7) 14.5_+1.8(5)

I1.1 10.9 11.3 13.5

90C

ns ns

ns ns <0.02 ns

P

Males on T h e i r W e b ( M a n n - W h i t n e y U test, two-tailed)"

" V a l u e s in parentheses represent s a m p l e sizes. 8 9 A , site A in 1989; 9 0 A , site A in 1990; 9 0 C , site C in 1990.

Sept, I 5 II 15 19 26 Oct. 2 6

Date

89A

III

T a b l e I, C o m p a r i s o n o f B o d y L e n g t h s o f F e m a l e s ( M e a n + S D m m ) With ( + ) a n d Without ( - )

292

Miyashita

ences in size between female status assessed by two-way ANOVA (i.e., adjusted for differences among dates) were significant on three occasions [89A, F(1,128) = 3.97, P < 0.05; 90A, F(1,140) = 35.42, P < 0.001; and 90C, F(1.65) = 4.02, P < 0.05], which suggests a kind of male choice. The size difference seemed to be more pronounced in the early part of the breeding season. Relationships between body lengths of males and females observed in copula are presented in Table II. Only matings of newly molted females were included, because most matings were observed at this time and first-male sperm precedence is known in the congener N. clavipes (Christenson and Cohn, 1988). A significant correlation was found at 89A. Although the correlation at 90C was as large as that at 89A, it was not significant due to the small sample size. Overall trends were assessed using covariance analysis (Nie et al., 1975). The overall correlation, or adjusted r, was significant. Body size of females at maturation decreased significantly with the progress of the season (Fig. 1). Also, the mean body size of adult males present at the site decreased. So early in the season both males and receptive females were larger, while later they were smaller. Table III shows the correlation between body length of spiders in copula and date of copulation. The signs of the correlations were all negative, though statistical significance was detected in only one instance. Overall correlations estimated by covariance analysis were close to significant, which suggests that mating pairs tend to be larger early in the season. Since the sample size of mating was too small to control the effects of the seasonal trend of body size, I estimated the possibility of size assortative guarding instead of mating (Table IV). Although guarding males in Table IV did not always succeed in mating with the partner [54.5% (n = 22) in 1989 and 50.0% (N = 26) in 1990) due to the limited chance of mating in this species, guarding should be intended for copulation. The correlations between hub male and female sizes on a particular day was not significant except for one instance. Moreover, the r values were mostly small. Thus, size assortative guarding on a given day is not a general phenomenon in this species.

Table H. Correlation Coefficients(r) Between Male and Female Lengths Observed in Copula~ i

Covariance analysis Site

r

n

P

Adjusted r ~

P

89A 90A 90C

0.647 -0.010 0.644

13 13 7

< 0.02 >0.8 >0.1

0.539

< 0.005

~The correlationbetween male and female length adjusted for differencesamong sites.

Size-Related Mating a n d Mate Guarding in the Spider

293

A 20

IIII

•

•

r=-0.522

•

p<0.05

16

J _¢

I

12

I

I

I

I

10-

B

~1

t F (2,28)=8.27

8

p<0.003 9

6 I

Sep.5

I

I

I

I

10

15

20

25

Date Fig. 1. Seasonal changes in the body length of females just after the final molt (A) and mean body length of males present on a particular date (B) at site 89A. Vertical bars and numerals in B represent 95 % confidence limits and sample size, respectively.

Table HI. Correlation Coefficients (r) Between Body Length of Males or Females in Copula and Date of Copulation Covariance analysis Site Male 89A 90A 90C Female 89A 90A 90C

r

n

P

-0.573 -0.045 -0.578

13 13 7

<0.05 >0.8 > 0.1

-0.281 - 0.465 -0.310

13 13 7

>0.3 > 0. I >0.4

aThe correlation adjusted for the differences among sites.

Adjusted r °

P

-0.355

< 0.06

-0.349

<0.07

294

Miyashita Table IV. Correlation Coefficients (r) Between Hub Male and Female Body Lengths II

Site 89A Date

Site 90A

r

n

P

0.402 0.181 -0.194 0.054

10 9 11 9

>0.2 >0.6 >0.5 >0.8

Sept. 5 11 15 19

Date

r

n

P

-0.094 0.710 -0.225 -0.137

8 8 15 13

>0.8 <0.05 > 0.4 >0.6

Sept. 10 14 19 25

DISCUSSION The correlation of body size between males and females in copula was significant at one site, and the adjusted correlation over the three occasions was also significant (Table II). Note, however, that samples at each site were taken from multiple dates. Body sizes of males and newly molted adult females decreased with time (Fig. 1). Since mating chances concentrate at female molt (Miyashita, 1993), the size of sexually receptive females decreased with time. Actually, sizes of copulating pairs tended to decrease with time (Table III). Thus, the positive correlation of body size is probably due to temporal covariation of body sizes of both sexes and is not associated with mate choice. We must therefore test Ridley's hypothesis using data controlled for temporal factors. Since the sample size of mating pairs on a given day was very small, size assortative guarding is used to test the hypothesis. Although N. clavata satisfies the three conditions of Ridley's hypothesis, it showed no evidence of size assortative guarding (Table IV). Several explanations for this are possible. First, assortative mating might have actually existed despite the absence of assortative guarding. McLain and Boromisa (1987) reported that mating was highly assortative, but pairing without genital contact was not. Testing this possibility is difficult in this species, because obtaining many mating pairs on a given day in a habitat is virtually impossible due to the low mating frequency. Second, positive correlations were not detected simply because sample sizes were small. This, however, does not seem plausible, since most r values were less than 0.2, which was regarded as an arbitrary threshold for the presence of a positive correlation (Fairbairn, 1988). Third, assortative mating on a given day might actually be absent because the ability of males to choose large females may be imperfect. It is considered that most web spiders do not use visual cues for searching mates (Platnick, 1971), though it has not been proved yet. Moreover, the opportunity of meeting potential mates seems relatively lower compared to flying arthropods. However, Rubenstein (1987) found size assortative mating in the spider Meta segmentata. This is probably because this spider often lives in aggregations, and access to many females may be relatively easy for males. If

Size-Related Mating and Mate Guarding in the Spider

295

the third explanation, i.e., imperfect recognition of female size by males, is true, it does not seem that Ridley's three conditions necessarily lead to size assortative mating. Early in the breeding season, females guarded by males were often larger than those not guarded (Table I). This season corresponds to the period before and during the final female molt (Miyashita, 1993). Thus, it is adaptive for males to guard large females during the period, since mating occurs mostly just after the female molt and males of a related species are known to deplete their sperm after copulation with just-molted females (Christenson, 1989). Whether this is evidence of active male choice is uncertain. Since large females tend to mature earlier (Fig. 1), males may respond to the age of females, not their size. This is plausible if males can discriminate female age to molt by some chemical substances or behaviors (Watson, 1990). Alternatively, the probability that males reach female webs may be high simply because large females construct large webs at open sites. These two mechanisms belong to the "passive attraction" of females (Crespi, 1989).

ACKNOWLEDGMENT I thank B. J. Crespi for comments on the manuscript.

REFERENCES Adams, J., and Greenwood, P. J. (1987). Loading constraints, sexual selection and assortative mating in peracarid Crustacea. J. Zool. Lond. 211: 35-46. Christenson, T. E. (1989). Sperm depletion in the orb-weaving spider Nephila clavipes (Araneae, Araneidae). J. Arachnol. 17:115-118. Christenson, T. E., and Cohn, J. (1988). Male advantage for egg fertilization in the golden orbweaving spider, NephUa clavipes. J. Comp. Psychol. 102: 312-318. Crespi, B. J. (1989). Causes of assortative mating in arthmpods. Anita. Behav. 38: 980-1000. Fairbaim, D. J. (1988). Sexual selection for homogamy in the Gerridae: An extension of Ridley's comparative approach. Evolution 42:1212-1222. McCauley, D. E., and Wade, M. J. (1978). Female choice and the mating structure of a natural population of the soldier beetle, Chauliognathus pennsylvanicus. Evolution 32: 771-775. McLain, D. K., and Bommisa, R. D. (1987). Male choice, fighting ability, assortative mating and the intensity of sexual selection in the milkweed longhorn beetle, Tetraopes tetraophthalmus (Coleoptera, Cerambycidae). Behav. Ecol. Sociobiol. 20: 239-246. Miyashita, T. (1986). Growth, egg production, and population density of the spider, Nephila clavata in relation to food conditions in the field. Res. Popul. Ecol. 28: 135-149. Miyashita, T. (1993). Male-male competition and mating success in the orb-web spider Nephila clavata with reference to temporal factors. EcoL Res. 8: 93-102. Nie, N. H., Hadlai Hull, C., Jenkins, J. G., Steinbrenner, K., and Bent, D. H. (1975). Statistical Package for the Social Sciences, McGraw-Hill, New York. Platnick, N. I. (1979). The evolution of courtship behaviour in spiders. Bull. Br. Arachnol. Soc. 2: 40-47.

296

Miyashita

Ridley, M. (1983). The Explanation of Organic Diversity. The Comparative Method and Adaptations for Mating, Clarendon, Oxford, UK. Rubenstein, D. I. (1987), Alternative reproductive tactics in the spider Meta segmentata. Behav. Ecol. Sociobiol. 20: 229-237. Watson, P. J. (1990). Female-enhanced male competition determines the first mate and principal sire in the spider Linyphia litigiosa (Linyphiidae). Behav. Ecol. Sociobiol. 26: 77-90.