- Email: [email protected]
Web placement in sympatric linyphiid spiders (Arachnida, Araneae): Individual foraging decisions reveal inter-specific competition
Acrcr Oecologico
19 (I) (1998)
67-7 I / 0 Elsevier.
Paris
Web placement in sympatric linyphiid spiders (Arachnida, Araneae): individual foraging decisions reveal inter-specific competition Marie Elisabeth Herberstein*
ReceivedJune
10 1997; accepted
September
10 1997.
Abstract - The distribution of two sympatric web spiders, Frontirlrllirlufi-urrtoncln (C. L. Koch) and Nrr?e/ze m&m (Walckenaer) (Araneae: Linyphiidae) was studied on an area of forest regrowth in eastern Austria. Both species utilised significantly different heights on young conifer trees to construct their webs. F ,frutetorurnselected higher vegetation layers, whereas N. rudinrcr constructed its webs closer to the ground. This distribution may either be evidence of competition for web space or it may reflect specific distribution patterns unrelated to spider density. An experiment showed that when spiders of either species were released onto vacant trees they selected similar vegetation heights for web construction. On trees already occupied by a heterospecitic individual however, Efruferonrm placed its webs significantly higher and N. rtrdium significantly lower compared to web placement on vacant trees suggesting that Efrutrforum and N. rcditrtrr compete for web space. 0 Elsevier, Paris
displacement
I competition
/ Linyphiidae
I Frontinellina
frutetorum
/ Neriene
1. INTRODUCTION
Communities of web-building spiders are thought to be shaped by competition for limited space [3, 16, 18, 271 and/or by competition for limited food [ 18, 28, 301. However, conclusive direct evidence supporting the role of competition in forming community patterns is still rare [see 30, 31 for a summary]. Although spatial patterning is thought to reflect competitive interactions among different species of web-building spiders [see 321, most experimental field investigations yield no evidence to support interspecific competition among web spiders [1 1, 16, 19, 291. These studies report no changes in web placement, fecundity or survival following experimental manipulation of competitor densities. The few exceptions demonstrated that the removal of competing spiders resulted in an increase in fecundity and greater vertical stratification of web sites [20, 211 while the addition of competitors lead to an increase in web takeovers [26]. Nevertheless, Wise [32] concluded that interspecific competition between web-building spiders is uncommon, and unlikely to play an important role in shaping community patterns. Ecological patterns, such as species distribution, are the consequence of individual behaviour [ 141. Under:“Current
address:
Dept of Zoology,
University
of Melbourne,
Park\ ille 3052.
radiata
/ web placement/Austria
standing ecological patterns requires a knowledge of interactions between individuals [2, 241. Therefore, manipulating and measuring the response of individuals may provide more insights to understanding population level responses than measuring at the population level [5]. Furthermore, the partial removal or addition of competitors may not necessarily yield a short term result because the biological effect may be small and a design accommodating the biological maximum difference (no competitor vs. competitor) may be necessary to elicit a response. Consequently, rather than manipulating competitor densities and observing the response of the remaining spiders [e.g. 11, 16, 20, 21, 26, 291, the response of spiders that were introduced into either a competitor free-environment, or an environment that was already occupied by the competitor, was observed. Such a procedure may determine if competitive interactions among individuals translate into population level effects. Descriptive observations of the ecology of two sympatric web spiders, Frontinellina frutetorum (C. L. Koch, 1834) and Neriene rudiata (Walckenaer, 1841) in an area of forest regrowth [9] revealed significant differences in web placement, suggesting that the two spider species may be competing for web sites. These Australia
M. E. Herberstein
68
two spiders are approximately the same size (adult total body length c. 0.5 cm) and share similar development times, maturing concurrently in early summer [9]. They construct species specific sheet webs with a centrally located platform. Barrier threads above the sheet intercept flying prey, knocking them to the platform where they are captured by the resident spider. Reported here are the results of a field experiment that incorporates the foraging decisions of these spiders and which offers direct evidence for the effect of competitor presence on web placement. By comparing the web heights of F: frutetorum and N. radiata on vacant trees and trees occupied by a heterospecific individuals the following hypotheses were tested, based on the natural distribution of the spiders in which N. radiata selects web sites significantly lower than those of F: frutetorum ([9] and results from present study). First, N. radiata will place its webs significantly lower on trees occupied by F: frutetorum compared to vacant trees. Second, F frutetorum will place its webs significantly higher on trees occupied by N. radiata compared to vacant trees. 2. MATERIALS
AND METHODS
2.1. Sampling The study site consisted of a patch of forest regrowth (total area: 2 854 m’) within a mature mixed deciduous forest in eastern Austria, near the town of Hartberg (Styria). Most webs (100 % of F:frutetorum webs and 76 % of N. radiata webs) observed in 1994 were built on young cultivated fir trees (Pseudotsuga menziesii), that had been planted in 1990. The height of the webs was sampled from March to October in 1994 to confirm natural distribution [9]. The distance from the ground to the horizontal sheet web was measured at the beginning of each month in 10 randomly selected transects of 10 x 1 m. A competition experiment that investigated the potential driving forces that influence web-site preference in these two spider species was conducted from 10 to 26 May 1994 and from 5 to 12 September 1994. Twenty fir trees were selected in close proximity with similar heights (Mean f SD = 203 I& 9 cm) to reduce the effect of tree variability and habitat differences on web placement. To reduce the effect of the surrounding shrub vegetation on web site selection [IO] shrub vegetation was removed, leaving a clearance of 1 m in diameter around each tree. Any spider and its web found on the trees were manually removed in order to remove any potential inter- or intraspecific interference by other spiders present on the tree. Spiders were collected for the experiment in the morning, marked with a dot of red paint on the dorsal side of their abdomen and released that evening onto the lowest branch of each tree.
Three different experiments were designed to test competitive interactions among individuals. First, the spiders were released onto empty trees (one spider per tree) and the following morning the height of their webs was recorded. Second, the spiders were released onto trees that already contained a competitor of the opposite species that had already been released onto the tree the previous morning and that had established a web. Thus, N. radiata were released onto trees already occupied by E frutetorum and vice versa. The height of webs of the newly released spiders was recorded the following morning. Third, the spiders were released onto trees already occupied by an individual of the same species that had been released the day before and that had established a web. I? ,frutetorum was placed on a tree occupied by another F: frutetorum, whereas N. radiata was released onto a tree occupied by another N. mdiata. The heights selected for web construction by the newly released spiders were measured the following morning. 2.2. Statistical Analysis All statistical analyses were performed using SYSTAT 5.2 [33] and all data were distributed normally (Kolmogorov-Smirnov Goodness of Fit tests). As webs were found on trees of various sizes, differences in web height between the two species collected from March to October were analysed using a 2-way ANCOVAR with species and month as the two factors and tree height as the covariate. The results of the experiment were analysed with t-tests (one tailed where specific hypotheses were formulated) using Bonferroni’s correction (a’=&, where k = the number of non-independent tests; P = 0.05/5 = 0.01) in order to avoid inflation of the type I error probability. 3. RESULTS
Both spider species construct permanent webs once a web site is established, which are not renewed on a daily basis but silk is added to the original web. This permanence was confirmed by remeasuring the height of the webs of the resident spider and searching for deserted webs on the tree. Thus, any differences in web height observed during the experiment were not due to changes in the location of the web sites by the resident spiders, which remained constant, but due to the response of the newly arriving spiders. The analysis of monthly web heights measured in the field revealed that both month (F = 10.3; df = 7; P < 0.0001) and species (F = 160.5; df = 1; P < 0.0001) have significant effects on web height, while the interaction was not significant (F = 1.58; df = 7: P = 0.1392). In particular, the height of webs spun by both species increased from March to October and
Competition
69
in web spiders
resident heterospecific individual. E;: frutetorum moved above the resident E frutetorum, while N. radiata stayed below the resident N. radiata cfigure 2). A larger sample size may have also revealed displacement on an intraspecific level. The mechanism responsible for the displacement in web sites may be interference competition, where site owners actively prevent other spiders from foraging in their vicinity [ 171. Whether or not these new web sites are sub-optimal can, however, only be answered through further investigations assessing the quality of web sites and their consequences on the survival, growth and fecundity of the spiders. The results also imply that newly arriving spiders were able to detect the presence of the resident competitor spider. Spiders detect pheromones of conspecifics [4], other spider species [23] and even ants [I], and it therefore seems plausible that F. frutetorum as well as N. radiata utilise such chemical cues to keep a sufficient distance to other individuals. What is the nature of the resource over which the spiders are competing? Space per se does not seem to be limited because the trees were 2 m high. However, spiders may compete for web sites that provide prey of a particular quality. A survey of potential prey using sticky traps at different vegetation heights showed significant variation in the vertical distribution of insects [8]. Alternatively, spiders may be competing for secure hiding places. Bird predation has a high impact on spider populations living on spruce branches [7]. The density of needles on these spruce branches are the key to avoiding predation. Areas with dense needles provide more and better hiding places than areas with sparse needles [6, 7, 221. It is likely that the fir trees used in
webs built by E frutetorum were always placed higher than those of N. rudiatu (figure I). The web height of both spider species released onto vacant trees during the first part of the experiment was not significantly different (t = 1.2; df= 53; P = 0.24). E frutetorum and N. radiata selected similar web heights when each had access to the entire tree. Web placement differed significantly when the spiders were released onto trees already occupied by a heterospecific spider. E frutetorum placed its web higher on trees occupied by N. radiata compared with vacant trees (t = -2.75; df = 47, P = 0.005, one tailed). Like,wise, N. radiata placed its webs lower on trees occupied by E frutetorum than on vacant trees (t = 3.3; df = 48; P = 0.001, one tailed). In contrast, on trees occupied by a conspecific, E frutetorum (t = -1.7; df = 49; P = 0.1) and N. rudiata (t = 1.92; df = 51; P = 0.06) selected similar heights as they did on vacant trees @gure 2). Finally, all E frutetorum individuals placed their webs above the webs of the resident N. rudiatLl and all but one N. radiata individual constructed their webs below those of the resident E .frutetorum. 4. DISCUSSION
The results of these experiments suggest that and N. radiata compete for web sites. The preferred web site, which is located in the middle layer of the tree is taken up by the spider that arrived first, forcing other, heterospecific spiders to displace further up or down the tree. The response of spiders towards a conspecific individual present in the tree is not quite as clear. While the comparison to web placement on vacant trees revealed no significant difference, the trends are similar to spiders faced with a F: frutetorum
160!z +I
bn 3c n .aJ
s
14012
27
i
+
P P
48
i d 9
11
60 40 - P. I I I I I I I 40 ’ ‘! March April May June July Aug. Sept. Oct. Month
Figure 1. Average 1994. The numbers Vol. 19 (1) 1998
web height (Mean + SE in cm) of E fruretorum above and below the error bars denote sample
(0) and N. rudiatu size.
(0)
on fir trees sampled
monthly
from
March
to October
in
70
M. E. Herberstein
22 24 4 i
80-
d 28
60 40
P 25
P I vacant
22
I I t heterospecific t conspecific
Tree Figure 2. Average web heterospecific individual
height (Mean f SE in cm) of I? frutetorum (0) and N. rudiutu (\3) when placed on vacant trees, trees occupied and trees occupied by a conspecific individual. The numbers above and below the error bars denote sample size.
the present experiment were not homogenous in terms of needle density. The very top as well as the bottom layers of the tree may be more sparse and thus not as attractive to spiders than the more dense middle layers. Thus, the spiders may be competing for secure hiding places from bird predators. Previous studies on other web spiders have not provided data supporting evidence of competition theory [ll, 16, 19, 29, 301. In these studies, a competitive effect revealed by changed microhabitat use was predicted following density manipulations of the competitor species. However, spiders should only relocate their web sites when the expected gain from moving is higher than the expected gain from remaining plus the cost of moving [12]. The potentially high risk of predation during relocation, together with little information about the quality of the new web site may favour spiders remaining at the current web site [ 151, even if new web sites are suddenly available following the removal of a competitor. A further cost for these spiders is the considerable nutritional investment in producing silk for their relatively permanent webs [ 12, 13, 251. Considering these potential costs and constraints, more conclusive results may be obtained by comparing the behaviour of a spider placed in a competitor-free habitat with one placed in a habitat already occupied by a competitor. The present study focused on the distribution of two spider species in young conifers, which is probably not the only microhabitat occupied by these two species. Consequently, no general conclusions regarding web site selection by the spiders and interactions in other habitats can be drawn. However, this systems allowed an easy and efficient manipulation of spiders for field experiments and the trees provided manage-
by a
able ‘experimental units’ in contrast to fully grown pine trees. Additionally, spiders may have chosen web sites differently had they been released at random heights, rather than always at the bottom of the trees. The results of the present study show that, using appropriate methods, the presence of a competitor can influence individual decisions on web-site preference and thus affect the distribution patterns of webbuilding spiders on a population level. Acknowledgments I thank Y. Lubin, and K. K. Thaler stein for support.
R. Alatalo, CCraig, M. Elgar, R. Graham, C. Kampichler, N. Milasowszky, W. Pie], M. Predavec, K. SLnger, G. Spitter Peter Zulka for help and comments on the manuscript; for identifying the spiders; J. A. Herberstein and L. Herbertheir generous help and the University of Vienna for financial
REFERENCES [ 11 Allan R. A., Elgar M. A. & Capon R. T., Exploitation of an ant chemical alarm signal by the zodariid spider Hubmnesre.s brudleyi Walckenaer, Proc. R. Sot. Lond., Ser. B 263 (1996) 69-73. [2] Bernstein C., Kacelnic A. & Krebs J. R., Individual decisions and the distribution of predators in a patchy environment, J. Anim. Ecol. 57 (1988) 1007-1026. [3] Bradley R. A., The influence of prey availability and habitat on activity patterns and abundance of Argiope keyserlingi (Araneae: Araneidae), J. Arachnol. 2 I (1993) 9l- 106. [4] Foelix R. F, Biologie der Spinnen, Georg Thieme Verlag, Stuttgart, 1992, 331~. [Sl Goss-Custard J. D., Caldow R. W. G., Clarke R. T., le. V. dit Durell S. E. A., Sutherland W. J. Deriving population parameters from individual variations in foraging behaviour. I. EmpirActu
Oecologicrr
Competition in web spiders
[6]
[7]
[8]
[9]
[IO]
[I I]
[12] [ 131 [ 141 [ 151
[ 161
[ 171
ical game theory distribution model of oystercatchers Haematopus ostralegus feeding on mussels Mytilus edulis, J. Anim. Ecol. 64 (1995) 265-276. Gunnarsson B., Vegetation structure and the abundance and size distribution of spruce-living spiders, .I. Anim. Ecol. 59 (1990) 743-752. Gunnarsson B., Bird predation and vegetation structure affecting spruce-living arthropods in a temperate forest, J. Anim. Ecol. 65 (I 996) 389-397. Herberstein M. E., The prey captured by web building spiders (Araneae: Linyphiidae) in comparison with feeding experiments, Rev. Suisse Zool. suppl. (1996) 279-288. Herberstein M. E., Niche partitioning in three sympatric web building spiders (Araneae: Linyphiidae), Bull. Br. Arachnol. Sot. 10, 7 (1997a) 233-238. Herberstein, M. E., The effect of habitat structure on web height preference in three sympatric web building spiders (Araneae: Linyphiidae) J. Arachnol. 25 (1997b) 93-96. Horton C. C., Wise D. H., The experimental analysis of competition between two syntopic species of orb-web spiders (Araneae: Araneidae) Ecology, 64.4 ( 1983) 929-944. Janetos A. C., Active foragers vs. sit-and-wait predators: a simple model, J. Theor. Biol. 95, 2 (1982a) 381-385. Janetos A. C., Foraging tactics of two guilds of web-spinning spiders, Behav. Ecol. Sociobiol. 10, (1982b) I 19-27. Krebs J. R., Davies N. B., Behavioural Ecology, Blackwell Scientific Publications, Oxford, 1991,482~. Lubin Y., Ellner S., Kotzman M., Web relocation and habitat selection in a desert widow spider, Ecology 74, 7 (1993) l9151928. Riechert S. E., Cady A. B., Patterns of resource use and tests for competitive release in a spider community. Ecology 64, 4 (1983) 899-913. Riechert S. E., Gillespie R. G., Habitat choice and utilization in web-building spiders, in: Shear W. A. (ed.), Spiders - Webs, behavior, and evolution, Stanford University Press, Stanford, California, 1986, pp.23.48.
1181 Rypstra A. L., The web-spider denisties: (1983) 312-316.
Vol. 19 (I)
1998
importance of food and space in limiting a test using field enclosures, Oecologia 59
71 [19] Schaefer M., Some experiments on regulation of population density in the spider Floronia bucculentu (Araneidae: Linyphiidae), Sym. 2001. Sot. Lond. 42 (1978) 203-210. [20] Spiller D. A., Competition between two spider species: experimental field study, Ecology 65, 3 (1984a) 909-919. [2l] Spiller D. A., Seasonal reversal of competitive advantage between two spider species, Oecologia 64 (1984b) 322-33 I [22] Sundberg I., Gunnarsson B., Spider abundance in relation to needle density in spruce, J. Arachnol. 22 (I 994) 190- 194. [23] Suter R. B., Shane C. M., Hirscheimer A. J., Spider vs spider: Frontinella pyramitela detects Argyrodes trigonum via cuticular chemicals, J. Arachnol. I7 (1989) 23’7-240. [24] Sutherland W. J., Dolman P. M., Combining behaviour and population dynamics with applications for predicting consequences of habitat loss, Proc. R. Sot. Lond. Ser. B 255 (1994) 133-138. [25] Tanaka K., Energetic cost of web construction and its effect on web relocation in the web-building spider Agelena limbata, Oecologia 81 (I 989) 459-464. [26] Toft S., Interactions among two coexisting Linyphiu spiders, Acta 2001. Fenn. 190 (1990) 367-372. [27] Ward D., Lubin Y., Temporal and spatial segregation of webbuilding in a community of orb-weaving spiders, J. Arachnol. 20 (1992) 73-87. [28] Wise D. H., Effects of an experimental increase in prey abundance upon the reproductive rates of two orb-weaving spider species (Araneae: Araneidae), Oecologia 41 (1979) 289-300. [29] Wise D. H., Inter- and intra specific effects of density manipulations upon females of two orb-weaving spiders (Araneae: Araneidae), Oecologia 48 (I 981) 252-256. [30] Wise D. H., Competitive mechanisms in a food-limited species: relative importance of interference and exploitive interactions among labyrinth spiders (Araneae, Araneidae), Oecologia 58 (1983) 1-9. [3l] Wise D. H., The role of competiton in spider communities: insights from field experiments with a model organism, in: Strong D. R., Simberloff D., Abele L. G., Thistle A. B. (eds), Ecological communities: conceptual issues and the evidence, Princeton University Press, Princeton, 1984, pp. 42-53. [32] Wise D. H., Spiders in Ecological Webs, Cambridge University Press, Cambridge, 1993,328 p. [33] Wilkinson L., SYSTAT: statistics, version 5.2 edn, SYSTAT, Evanston, 1992.
19 (I) (1998)
67-7 I / 0 Elsevier.
Paris
Web placement in sympatric linyphiid spiders (Arachnida, Araneae): individual foraging decisions reveal inter-specific competition Marie Elisabeth Herberstein*
ReceivedJune
10 1997; accepted
September
10 1997.
Abstract - The distribution of two sympatric web spiders, Frontirlrllirlufi-urrtoncln (C. L. Koch) and Nrr?e/ze m&m (Walckenaer) (Araneae: Linyphiidae) was studied on an area of forest regrowth in eastern Austria. Both species utilised significantly different heights on young conifer trees to construct their webs. F ,frutetorurnselected higher vegetation layers, whereas N. rudinrcr constructed its webs closer to the ground. This distribution may either be evidence of competition for web space or it may reflect specific distribution patterns unrelated to spider density. An experiment showed that when spiders of either species were released onto vacant trees they selected similar vegetation heights for web construction. On trees already occupied by a heterospecitic individual however, Efruferonrm placed its webs significantly higher and N. rtrdium significantly lower compared to web placement on vacant trees suggesting that Efrutrforum and N. rcditrtrr compete for web space. 0 Elsevier, Paris
displacement
I competition
/ Linyphiidae
I Frontinellina
frutetorum
/ Neriene
1. INTRODUCTION
Communities of web-building spiders are thought to be shaped by competition for limited space [3, 16, 18, 271 and/or by competition for limited food [ 18, 28, 301. However, conclusive direct evidence supporting the role of competition in forming community patterns is still rare [see 30, 31 for a summary]. Although spatial patterning is thought to reflect competitive interactions among different species of web-building spiders [see 321, most experimental field investigations yield no evidence to support interspecific competition among web spiders [1 1, 16, 19, 291. These studies report no changes in web placement, fecundity or survival following experimental manipulation of competitor densities. The few exceptions demonstrated that the removal of competing spiders resulted in an increase in fecundity and greater vertical stratification of web sites [20, 211 while the addition of competitors lead to an increase in web takeovers [26]. Nevertheless, Wise [32] concluded that interspecific competition between web-building spiders is uncommon, and unlikely to play an important role in shaping community patterns. Ecological patterns, such as species distribution, are the consequence of individual behaviour [ 141. Under:“Current
address:
Dept of Zoology,
University
of Melbourne,
Park\ ille 3052.
radiata
/ web placement/Austria
standing ecological patterns requires a knowledge of interactions between individuals [2, 241. Therefore, manipulating and measuring the response of individuals may provide more insights to understanding population level responses than measuring at the population level [5]. Furthermore, the partial removal or addition of competitors may not necessarily yield a short term result because the biological effect may be small and a design accommodating the biological maximum difference (no competitor vs. competitor) may be necessary to elicit a response. Consequently, rather than manipulating competitor densities and observing the response of the remaining spiders [e.g. 11, 16, 20, 21, 26, 291, the response of spiders that were introduced into either a competitor free-environment, or an environment that was already occupied by the competitor, was observed. Such a procedure may determine if competitive interactions among individuals translate into population level effects. Descriptive observations of the ecology of two sympatric web spiders, Frontinellina frutetorum (C. L. Koch, 1834) and Neriene rudiata (Walckenaer, 1841) in an area of forest regrowth [9] revealed significant differences in web placement, suggesting that the two spider species may be competing for web sites. These Australia
M. E. Herberstein
68
two spiders are approximately the same size (adult total body length c. 0.5 cm) and share similar development times, maturing concurrently in early summer [9]. They construct species specific sheet webs with a centrally located platform. Barrier threads above the sheet intercept flying prey, knocking them to the platform where they are captured by the resident spider. Reported here are the results of a field experiment that incorporates the foraging decisions of these spiders and which offers direct evidence for the effect of competitor presence on web placement. By comparing the web heights of F: frutetorum and N. radiata on vacant trees and trees occupied by a heterospecific individuals the following hypotheses were tested, based on the natural distribution of the spiders in which N. radiata selects web sites significantly lower than those of F: frutetorum ([9] and results from present study). First, N. radiata will place its webs significantly lower on trees occupied by F: frutetorum compared to vacant trees. Second, F frutetorum will place its webs significantly higher on trees occupied by N. radiata compared to vacant trees. 2. MATERIALS
AND METHODS
2.1. Sampling The study site consisted of a patch of forest regrowth (total area: 2 854 m’) within a mature mixed deciduous forest in eastern Austria, near the town of Hartberg (Styria). Most webs (100 % of F:frutetorum webs and 76 % of N. radiata webs) observed in 1994 were built on young cultivated fir trees (Pseudotsuga menziesii), that had been planted in 1990. The height of the webs was sampled from March to October in 1994 to confirm natural distribution [9]. The distance from the ground to the horizontal sheet web was measured at the beginning of each month in 10 randomly selected transects of 10 x 1 m. A competition experiment that investigated the potential driving forces that influence web-site preference in these two spider species was conducted from 10 to 26 May 1994 and from 5 to 12 September 1994. Twenty fir trees were selected in close proximity with similar heights (Mean f SD = 203 I& 9 cm) to reduce the effect of tree variability and habitat differences on web placement. To reduce the effect of the surrounding shrub vegetation on web site selection [IO] shrub vegetation was removed, leaving a clearance of 1 m in diameter around each tree. Any spider and its web found on the trees were manually removed in order to remove any potential inter- or intraspecific interference by other spiders present on the tree. Spiders were collected for the experiment in the morning, marked with a dot of red paint on the dorsal side of their abdomen and released that evening onto the lowest branch of each tree.
Three different experiments were designed to test competitive interactions among individuals. First, the spiders were released onto empty trees (one spider per tree) and the following morning the height of their webs was recorded. Second, the spiders were released onto trees that already contained a competitor of the opposite species that had already been released onto the tree the previous morning and that had established a web. Thus, N. radiata were released onto trees already occupied by E frutetorum and vice versa. The height of webs of the newly released spiders was recorded the following morning. Third, the spiders were released onto trees already occupied by an individual of the same species that had been released the day before and that had established a web. I? ,frutetorum was placed on a tree occupied by another F: frutetorum, whereas N. radiata was released onto a tree occupied by another N. mdiata. The heights selected for web construction by the newly released spiders were measured the following morning. 2.2. Statistical Analysis All statistical analyses were performed using SYSTAT 5.2 [33] and all data were distributed normally (Kolmogorov-Smirnov Goodness of Fit tests). As webs were found on trees of various sizes, differences in web height between the two species collected from March to October were analysed using a 2-way ANCOVAR with species and month as the two factors and tree height as the covariate. The results of the experiment were analysed with t-tests (one tailed where specific hypotheses were formulated) using Bonferroni’s correction (a’=&, where k = the number of non-independent tests; P = 0.05/5 = 0.01) in order to avoid inflation of the type I error probability. 3. RESULTS
Both spider species construct permanent webs once a web site is established, which are not renewed on a daily basis but silk is added to the original web. This permanence was confirmed by remeasuring the height of the webs of the resident spider and searching for deserted webs on the tree. Thus, any differences in web height observed during the experiment were not due to changes in the location of the web sites by the resident spiders, which remained constant, but due to the response of the newly arriving spiders. The analysis of monthly web heights measured in the field revealed that both month (F = 10.3; df = 7; P < 0.0001) and species (F = 160.5; df = 1; P < 0.0001) have significant effects on web height, while the interaction was not significant (F = 1.58; df = 7: P = 0.1392). In particular, the height of webs spun by both species increased from March to October and
Competition
69
in web spiders
resident heterospecific individual. E;: frutetorum moved above the resident E frutetorum, while N. radiata stayed below the resident N. radiata cfigure 2). A larger sample size may have also revealed displacement on an intraspecific level. The mechanism responsible for the displacement in web sites may be interference competition, where site owners actively prevent other spiders from foraging in their vicinity [ 171. Whether or not these new web sites are sub-optimal can, however, only be answered through further investigations assessing the quality of web sites and their consequences on the survival, growth and fecundity of the spiders. The results also imply that newly arriving spiders were able to detect the presence of the resident competitor spider. Spiders detect pheromones of conspecifics [4], other spider species [23] and even ants [I], and it therefore seems plausible that F. frutetorum as well as N. radiata utilise such chemical cues to keep a sufficient distance to other individuals. What is the nature of the resource over which the spiders are competing? Space per se does not seem to be limited because the trees were 2 m high. However, spiders may compete for web sites that provide prey of a particular quality. A survey of potential prey using sticky traps at different vegetation heights showed significant variation in the vertical distribution of insects [8]. Alternatively, spiders may be competing for secure hiding places. Bird predation has a high impact on spider populations living on spruce branches [7]. The density of needles on these spruce branches are the key to avoiding predation. Areas with dense needles provide more and better hiding places than areas with sparse needles [6, 7, 221. It is likely that the fir trees used in
webs built by E frutetorum were always placed higher than those of N. rudiatu (figure I). The web height of both spider species released onto vacant trees during the first part of the experiment was not significantly different (t = 1.2; df= 53; P = 0.24). E frutetorum and N. radiata selected similar web heights when each had access to the entire tree. Web placement differed significantly when the spiders were released onto trees already occupied by a heterospecific spider. E frutetorum placed its web higher on trees occupied by N. radiata compared with vacant trees (t = -2.75; df = 47, P = 0.005, one tailed). Like,wise, N. radiata placed its webs lower on trees occupied by E frutetorum than on vacant trees (t = 3.3; df = 48; P = 0.001, one tailed). In contrast, on trees occupied by a conspecific, E frutetorum (t = -1.7; df = 49; P = 0.1) and N. rudiata (t = 1.92; df = 51; P = 0.06) selected similar heights as they did on vacant trees @gure 2). Finally, all E frutetorum individuals placed their webs above the webs of the resident N. rudiatLl and all but one N. radiata individual constructed their webs below those of the resident E .frutetorum. 4. DISCUSSION
The results of these experiments suggest that and N. radiata compete for web sites. The preferred web site, which is located in the middle layer of the tree is taken up by the spider that arrived first, forcing other, heterospecific spiders to displace further up or down the tree. The response of spiders towards a conspecific individual present in the tree is not quite as clear. While the comparison to web placement on vacant trees revealed no significant difference, the trends are similar to spiders faced with a F: frutetorum
160!z +I
bn 3c n .aJ
s
14012
27
i
+
P P
48
i d 9
11
60 40 - P. I I I I I I I 40 ’ ‘! March April May June July Aug. Sept. Oct. Month
Figure 1. Average 1994. The numbers Vol. 19 (1) 1998
web height (Mean + SE in cm) of E fruretorum above and below the error bars denote sample
(0) and N. rudiatu size.
(0)
on fir trees sampled
monthly
from
March
to October
in
70
M. E. Herberstein
22 24 4 i
80-
d 28
60 40
P 25
P I vacant
22
I I t heterospecific t conspecific
Tree Figure 2. Average web heterospecific individual
height (Mean f SE in cm) of I? frutetorum (0) and N. rudiutu (\3) when placed on vacant trees, trees occupied and trees occupied by a conspecific individual. The numbers above and below the error bars denote sample size.
the present experiment were not homogenous in terms of needle density. The very top as well as the bottom layers of the tree may be more sparse and thus not as attractive to spiders than the more dense middle layers. Thus, the spiders may be competing for secure hiding places from bird predators. Previous studies on other web spiders have not provided data supporting evidence of competition theory [ll, 16, 19, 29, 301. In these studies, a competitive effect revealed by changed microhabitat use was predicted following density manipulations of the competitor species. However, spiders should only relocate their web sites when the expected gain from moving is higher than the expected gain from remaining plus the cost of moving [12]. The potentially high risk of predation during relocation, together with little information about the quality of the new web site may favour spiders remaining at the current web site [ 151, even if new web sites are suddenly available following the removal of a competitor. A further cost for these spiders is the considerable nutritional investment in producing silk for their relatively permanent webs [ 12, 13, 251. Considering these potential costs and constraints, more conclusive results may be obtained by comparing the behaviour of a spider placed in a competitor-free habitat with one placed in a habitat already occupied by a competitor. The present study focused on the distribution of two spider species in young conifers, which is probably not the only microhabitat occupied by these two species. Consequently, no general conclusions regarding web site selection by the spiders and interactions in other habitats can be drawn. However, this systems allowed an easy and efficient manipulation of spiders for field experiments and the trees provided manage-
by a
able ‘experimental units’ in contrast to fully grown pine trees. Additionally, spiders may have chosen web sites differently had they been released at random heights, rather than always at the bottom of the trees. The results of the present study show that, using appropriate methods, the presence of a competitor can influence individual decisions on web-site preference and thus affect the distribution patterns of webbuilding spiders on a population level. Acknowledgments I thank Y. Lubin, and K. K. Thaler stein for support.
R. Alatalo, CCraig, M. Elgar, R. Graham, C. Kampichler, N. Milasowszky, W. Pie], M. Predavec, K. SLnger, G. Spitter Peter Zulka for help and comments on the manuscript; for identifying the spiders; J. A. Herberstein and L. Herbertheir generous help and the University of Vienna for financial
REFERENCES [ 11 Allan R. A., Elgar M. A. & Capon R. T., Exploitation of an ant chemical alarm signal by the zodariid spider Hubmnesre.s brudleyi Walckenaer, Proc. R. Sot. Lond., Ser. B 263 (1996) 69-73. [2] Bernstein C., Kacelnic A. & Krebs J. R., Individual decisions and the distribution of predators in a patchy environment, J. Anim. Ecol. 57 (1988) 1007-1026. [3] Bradley R. A., The influence of prey availability and habitat on activity patterns and abundance of Argiope keyserlingi (Araneae: Araneidae), J. Arachnol. 2 I (1993) 9l- 106. [4] Foelix R. F, Biologie der Spinnen, Georg Thieme Verlag, Stuttgart, 1992, 331~. [Sl Goss-Custard J. D., Caldow R. W. G., Clarke R. T., le. V. dit Durell S. E. A., Sutherland W. J. Deriving population parameters from individual variations in foraging behaviour. I. EmpirActu
Oecologicrr
Competition in web spiders
[6]
[7]
[8]
[9]
[IO]
[I I]
[12] [ 131 [ 141 [ 151
[ 161
[ 171
ical game theory distribution model of oystercatchers Haematopus ostralegus feeding on mussels Mytilus edulis, J. Anim. Ecol. 64 (1995) 265-276. Gunnarsson B., Vegetation structure and the abundance and size distribution of spruce-living spiders, .I. Anim. Ecol. 59 (1990) 743-752. Gunnarsson B., Bird predation and vegetation structure affecting spruce-living arthropods in a temperate forest, J. Anim. Ecol. 65 (I 996) 389-397. Herberstein M. E., The prey captured by web building spiders (Araneae: Linyphiidae) in comparison with feeding experiments, Rev. Suisse Zool. suppl. (1996) 279-288. Herberstein M. E., Niche partitioning in three sympatric web building spiders (Araneae: Linyphiidae), Bull. Br. Arachnol. Sot. 10, 7 (1997a) 233-238. Herberstein, M. E., The effect of habitat structure on web height preference in three sympatric web building spiders (Araneae: Linyphiidae) J. Arachnol. 25 (1997b) 93-96. Horton C. C., Wise D. H., The experimental analysis of competition between two syntopic species of orb-web spiders (Araneae: Araneidae) Ecology, 64.4 ( 1983) 929-944. Janetos A. C., Active foragers vs. sit-and-wait predators: a simple model, J. Theor. Biol. 95, 2 (1982a) 381-385. Janetos A. C., Foraging tactics of two guilds of web-spinning spiders, Behav. Ecol. Sociobiol. 10, (1982b) I 19-27. Krebs J. R., Davies N. B., Behavioural Ecology, Blackwell Scientific Publications, Oxford, 1991,482~. Lubin Y., Ellner S., Kotzman M., Web relocation and habitat selection in a desert widow spider, Ecology 74, 7 (1993) l9151928. Riechert S. E., Cady A. B., Patterns of resource use and tests for competitive release in a spider community. Ecology 64, 4 (1983) 899-913. Riechert S. E., Gillespie R. G., Habitat choice and utilization in web-building spiders, in: Shear W. A. (ed.), Spiders - Webs, behavior, and evolution, Stanford University Press, Stanford, California, 1986, pp.23.48.
1181 Rypstra A. L., The web-spider denisties: (1983) 312-316.
Vol. 19 (I)
1998
importance of food and space in limiting a test using field enclosures, Oecologia 59
71 [19] Schaefer M., Some experiments on regulation of population density in the spider Floronia bucculentu (Araneidae: Linyphiidae), Sym. 2001. Sot. Lond. 42 (1978) 203-210. [20] Spiller D. A., Competition between two spider species: experimental field study, Ecology 65, 3 (1984a) 909-919. [2l] Spiller D. A., Seasonal reversal of competitive advantage between two spider species, Oecologia 64 (1984b) 322-33 I [22] Sundberg I., Gunnarsson B., Spider abundance in relation to needle density in spruce, J. Arachnol. 22 (I 994) 190- 194. [23] Suter R. B., Shane C. M., Hirscheimer A. J., Spider vs spider: Frontinella pyramitela detects Argyrodes trigonum via cuticular chemicals, J. Arachnol. I7 (1989) 23’7-240. [24] Sutherland W. J., Dolman P. M., Combining behaviour and population dynamics with applications for predicting consequences of habitat loss, Proc. R. Sot. Lond. Ser. B 255 (1994) 133-138. [25] Tanaka K., Energetic cost of web construction and its effect on web relocation in the web-building spider Agelena limbata, Oecologia 81 (I 989) 459-464. [26] Toft S., Interactions among two coexisting Linyphiu spiders, Acta 2001. Fenn. 190 (1990) 367-372. [27] Ward D., Lubin Y., Temporal and spatial segregation of webbuilding in a community of orb-weaving spiders, J. Arachnol. 20 (1992) 73-87. [28] Wise D. H., Effects of an experimental increase in prey abundance upon the reproductive rates of two orb-weaving spider species (Araneae: Araneidae), Oecologia 41 (1979) 289-300. [29] Wise D. H., Inter- and intra specific effects of density manipulations upon females of two orb-weaving spiders (Araneae: Araneidae), Oecologia 48 (I 981) 252-256. [30] Wise D. H., Competitive mechanisms in a food-limited species: relative importance of interference and exploitive interactions among labyrinth spiders (Araneae, Araneidae), Oecologia 58 (1983) 1-9. [3l] Wise D. H., The role of competiton in spider communities: insights from field experiments with a model organism, in: Strong D. R., Simberloff D., Abele L. G., Thistle A. B. (eds), Ecological communities: conceptual issues and the evidence, Princeton University Press, Princeton, 1984, pp. 42-53. [32] Wise D. H., Spiders in Ecological Webs, Cambridge University Press, Cambridge, 1993,328 p. [33] Wilkinson L., SYSTAT: statistics, version 5.2 edn, SYSTAT, Evanston, 1992.