- Email: [email protected]
Evidence of circadian rhythm in antipredator behaviour in the orb-weaving spider Larinioides cornutus
Animal Behaviour 82 (2011) 549e555
Contents lists available at ScienceDirect
Animal Behaviour journal homepage: www.elsevier.com/locate/anbehav
Evidence of circadian rhythm in antipredator behaviour in the orb-weaving spider Larinioides cornutus Thomas C. Jones*, Tamer S. Akoury, Christopher K. Hauser, Darrell Moore Department of Biological Sciences, East Tennessee State University
a r t i c l e i n f o Article history: Received 26 April 2011 Initial acceptance 11 May 2011 Final acceptance 31 May 2011 Available online 20 July 2011 MS. number: A11-00339R Keywords: aggression circadian rhythm huddle response Larinioides cornutus orb-weaving spider thanatosis
Ecologically, spiders are both predators and prey. Therefore, they must balance being aggressive enough to forage successfully, but not so aggressive that they become overly exposed to predation. Some species of spiders actively forage during clearly defined periods of the day, leading to the hypothesis that they should be less aggressive (or more defensive) during periods when they are not foraging, predicting that antipredator behaviour should be more pronounced during inactive foraging times. We tested the antipredator ‘huddle response’ in a nocturnal foraging orb-weaver, Larinioides cornutus, and found that, as predicted, the spiders huddled longer in the day than at night. We then conducted tests to determine whether the cycling of the response was regulated by an internal clock (circadian), and we found that huddle duration continued to cycle under constant dark (with periodicity significantly less than 24 h) as well as under constant light (periodicity nonsignificantly longer than 24 h). This work adds a novel behaviour to the list of behaviours under circadian control and also to the surprisingly short list of studies demonstrating circadian rhythm in spiders. Ó 2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
It is commonly assumed that circadian rhythms are adaptive, directing the organism to perform each of its various functions, including behaviour, at the most appropriate time of day. Circadian rhythms typically are entrained by environmental cycles, most notably the daily alternation of light and dark. Entrainment ensures that the rhythm maintains a fixed phase relationship with the environmental cycle. Often, the entrained rhythm makes it possible for the organism to anticipate (i.e. perform a function ahead of) environmental changes such as sunrise or sunset. The entrained circadian rhythm thus may allow the organism to be proactive rather than reactive with respect to daily periodic environmental stimuli. Experiments conducted in the laboratory with abrupt changes in light levels are able to differentiate circadian-based anticipatory behaviour from behaviours that respond directly to (gradually) changing light levels. For example, the entrained circadian rhythm of activity in bats (DeCoursey & DeCoursey 1964) and flying squirrels (DeCoursey 1989) allows these nocturnal animals to anticipate sunset, thereby enabling them to begin their night-time activities at the appropriate light intensity level. In another laboratory study, performance of the calling song is primarily a nocturnal behaviour in male crickets (Teleogryllus commodus): it occurs throughout the night but actually begins * Correspondence: T. C. Jones, Department of Biological Sciences, East Tennessee State University, Johnson City, TN, 37614, U.S.A. E-mail address: [email protected] (T. C. Jones).
1e3 h before sunset (Loher 1972). In the field, honeybees, using their circadian clock-driven time-memory, anticipate the time of day at which they encounter profitable food sources on previous days by scheduling reconnaissance flights to that particular location in the environment, with the amount of anticipation depending on the amount of previous experience with that source (Moore & Doherty 2009). In an ecological sense, spiders are both predators and prey (Wise 1993). In particular, orb-weaving spiders aggressively attack insects caught in their webs, but are themselves subject to attack from wasps and birds (Foelix 1996). Most orbweaving species are either nocturnal or diurnal, actively foraging during only one part of the daily cycle, and remaining secluded during the other (Carico 1986; Stowe 1986). In a sense, these spiders must behaviourally ‘switch’ from being predators to avoiding being prey on a daily basis. One would therefore expect there to be a selective advantage for spiders to anticipate dawn and dusk, particularly if prey levels are higher during these environmental transitions. However, very little information exists concerning the circadian control of behaviour in spiders. Most of the studies to date are concerned with locomotor activity (e.g. Seyfarth 1980; Schmitt et al. 1990; Suter 1993) or changes in sensitivity of the eyes (e.g. Yamashita & Nakamura 1999). Here, we present evidence for circadian regulation of an antipredator behaviour in the orb-weaving spider Larinioides cornutus. An antipredator behaviour commonly observed in orb-weaving spiders is ‘bailing out’, in which the spider drops from the web and
0003-3472/$38.00 Ó 2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.anbehav.2011.06.009
550
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
remains motionless with its legs flexed tightly to its ventral midline for a period of time (Rayor 1996). This behaviour is illustrated for the species used in the present study in Jones et al. (2011). Uetz et al. (2002) found that a colonial orb-weaving spider would express this behaviour in response to continuous vibrations applied to a support silk line. That study also found that the likelihood of exhibiting the response increased in those spiders whose neighbours had already been coaxed to bail out. The physical posture assumed by spiders when they bail out of their webs can be elicited in the laboratory, even if the spiders are removed from their web. Studies by Pruitt et al. (2008, 2010) on social cobweb-building spiders suggested that this ‘huddle response’ (also known as thanatosis, or death feigning) is part of an overall behavioural syndrome (sensu Sih et al. 2004) related to aggression: spiders that are more aggressive towards prey and conspecifics do not hold the huddle posture as long as less aggressive spiders. A recent study of the effect of biogenic amines on the huddle response in the orb-weaving spider Larinioides cornutus found that experimentally elevating octopamine significantly shortened the duration of the huddle response, while elevating serotonin significantly prolonged it (Jones et al. 2011). That study found that the effect of serotonin decreased over the first 6 h after treatment, but then increased and was most pronounced 24 h after dosing, leading to the speculation that these spiders may show rhythmicity in antipredator behaviour. In the present study we explored whether the huddle response of the nocturnal orb-weaver L. cornutus (Aranaeidae) shows diel periodicity, and if so, whether periodicity is regulated by an internal (i.e. circadian) clock. We hypothesized that these nocturnal spiders would huddle longer during the day than at night, reflecting the fact that they forage at night but not during daylight hours. Our rationale was that there is a significant cost to bailing out of the web at night, in terms of missed prey capture opportunities, that does not exist during the day. Alternatively, the predation risk for the spider (particularly from wasps and birds) may be higher for the spider during daylight hours than at night. METHODS Study Species, Collection and Housing Larinioides cornutus (Clerck 1757) is an orb-weaver in the family Araneidae, which is common to Europe and North America (Levi & Levi 1990). It is a nocturnal spider, constructing its web at dusk and
actively foraging throughout the night, then removing its web at dawn, and remaining in a hidden retreat during the day (Bellmann 1997). Female adult L. cornutus were collected in September and October 2009 from buildings and shrubbery in northeastern Tennessee (U.S.A.). The spiders were extracted from their webs and individually housed in 250 ml plastic deli containers. Spiders that were housed for more than a few days were given water twice a week, fed crickets every other week, and kept at approximately 23 C under 12:12 h light:dark cycle. Quantifying the Huddle Response Web-building spiders typically show an antipredator huddle response in which they draw their legs in tight to their body and remain motionless. If left undisturbed, the spiders will break out of the huddle after a period of time. This huddling appears to be an ‘allor-nothing’ response, and there is no known effect of stimulus strength on huddle duration in L. cornutus (T. C. Jones, unpublished data) or in other spiders (Pruitt et al. 2008, 2010). There is also no evidence of habituation to the stimulus (Jones et al. 2011). We used the method of Jones et al. (2011) to quantify the response. Individual spiders were removed from their containers and placed in a glass bowl (15 cm diameter, 6 cm high), and were given 30 s to acclimate to the dish. To trigger the huddle response we delivered a puff of air from approximately 10 cm away. The air movement deflects fine sensory hairs (trichobothria) on the spider’s legs, stimulating the stereotypical huddle response (Foelix 1996). We determined the duration of the huddling response to the nearest second with a stopwatch from the moment the spiders initiated the huddle to the moment that they broke out of it. Individuals that did not huddle after three attempts to initiate the response received a score of 0 s; individuals that initiated a response but immediately broke out of it, received a score of 1 s. We cleaned each dish with ethanol between trials. Measuring Diel and Circadian Rhythmicity in the Huddle Response For this part of the study, adult L. cornutus were housed in a light-controlled room under an LD 12:12 h for at least 5 days prior to testing. The room was carefully sealed against light leaks, and the door was double-screened with black-out curtains to prevent light from leaking in when opened. All dark observations were conducted under low light filtered with two layers of GamColorÒ Medium Red XT gel. Ten spiders were tested for 72 h under LD
Table 1 Chi-square and F periodogram analyses of huddle response duration for individual spiders under LD 12:12 h, DD and LL LD 12:12 h
DD
ID
c2 a¼0.001
a¼0.01
Lc-22 Lc-14 Lc-1 Lc-12 Lc-CCC Lc-QQQQ Lc-JJJ Lc-HHH Lc-11 Lc-26
23.9 23.9 28 28 23.9 23.9 23.9 23.9 23.9 23.9
23.9 23.9 28 28 23.9 23.9 23.9 23.9 23.9 23.9
F
LL
ID
c2 a¼0.001
a¼0.01
Lc-C Lc-P Lc-12 Lc-D1 Lc-JJJ Lc-D2 Lc-CCC Lc-1 Lc-11 Lc-22 Lc-18 Lc-MM Lc-LL Lc-HHH Lc-26
22 19.9 19.9 20 24 23.9 23.9 25 18.7 22 23.9 22.7 21.3 19.9 NS
22 19.9 21.3 20 24 23.9 23.9 25 18.7 21.4 23.9 22.7 21.3 19.9 NS
F
ID
c2 a¼0.001
a¼0.01
Lc-I Lc-XI Lc-HHH Lc-CCC Lc-D2 Lc-X2 Lc-NNN Lc-X4 Lc-II Lc-JJJ Lc-12 Lc-M Lc-C Lc-X5 Lc-X3
26.7 26.7 26.7 22 NS 29.5 23.9 NS 27 28.7 20.6 28 28.7 20 23.9
26.7 26.7 26.7 22 NS NS 23.9 NS 27 28.7 20.6 28 28.7 18 23.9
F
Given are the main periods (s), in hours, that were significant at a ¼ 0.001 for chi-square periodograms and that were significant at a ¼ 0.01 for F periodograms (individuals with nonsignificant periodograms are indicated in bold).
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
350
1200 (b)
(a)
300
1000
250 Amplitude
Huddle duration (s)
551
200 150
800 600
100
400
50
200
08
00 12 00 16 00 20 0 00 0 00 04 00 08 00 12 00 16 00 20 00 00 00 04 00 08 00 12 00 16 00 20 00 00 00 04 00
0
0 18 19 20 21 22 23 24 25 26 27 28 29 30 Period
Time of day (hours) Figure 1. Diel rhythms in huddle responses of L. cornutus under LD 12:12 h. (a) Means and standard errors of huddle response durations of 10 spiders tested at six times per day over the course of 3 days. Open bars: spiders tested with lights on; solid bars: spiders tested with lights off. (b) Chi-square periodogram of the mean huddle durations. The solid line indicates the amplitude of detected periodicities; a significant periodicity (P < 0.01) is indicated by a peak above the diagonal line.
12:12 h conditions. In this group, observations began at 0800 hours (0.5 h after lights on), and were conducted every 4 h thereafter. We also tested huddle responses of spiders under constant dark (DD) and constant light (LL) conditions, at 23 C. In both cases, the spiders were tested every 4 h beginning at 0700 hours (0.5 h prior to lights on). Fifteen spiders were tested for 24 h under normal LD, and then for 84 h in constant dark (DD). Fifteen spiders also were tested under constant light (LL). In this case, the 24 h LD pretrial was followed by 108 h of continuous light. Prior to both the DD and LL trials, the spiders were entrained to LD 12:12 h for 5 days. Statistical Analyses The data were compiled and figures were created with Excel 2007, and statistical analyses were run in Minitab 15. We analysed periodicity and created periodograms with ClockLab (Actimetrics, Wilmette, IL, U.S.A.). Periodicities were determined objectively by employing chi-square periodograms at an alpha level of 0.001 for periods between 18 and 30 h. Only the highest peaks above the significance line were used. These periodicities were not accepted
Light
80
as significant unless additionally confirmed by F periodograms at an alpha level of 0.01 (Table 1). RESULTS Diel Rhythmicity Spiders under LD 12:12 h conditions clearly showed diel rhythmicity in the huddle response, holding the posture longer in the light than in the dark (Fig. 1a). Chi-square periodogram analysis of mean huddle duration found significant periodicity (P < 0.001), with a dominant period of 23.9 h (Fig. 1b). Periodogram analyses of individual spiders found significant periodicity (P < 0.001) of 23.9 h for eight individuals and 28 h for two individuals. Data pooled for all individuals showed that the distributions of huddle response durations in both the light and dark phases were skewed to the right, indicating that there were relatively few very long huddle durations (Fig. 2). However, huddle durations in the light phase (median ¼ 130.1) tended to be longer than in the dark (median ¼ 39.2), with nine of the 10 individuals huddling
Dark
70
Frequency
60 50 40 30 20 10 0
0
100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 Huddle duration (s)
Figure 2. Huddle durations of L. cornutus under LD 12:12 h. Observations under light and dark conditions are shown separately.
552
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
significantly longer in the light phase than in the dark (Manne Whitney U tests comparing each individual’s responses, P range 0.019 to <0.001). There also were differences between individuals in huddle duration (KruskaleWallis test: H9 ¼ 40.54, P < 0.001). While all individuals displayed some short huddle durations, a few individuals consistently huddled longer than others (Fig. 3).
800
Huddle duration (s)
700 600 500 400 300
Rhythmicity under Constant Conditions
200 100 0 S−01 S−02 S−03 S−04 S−05 S−06 S−07 S−08 S−09 S−10 Individual spiders
Figure 3. Huddle durations of individual spiders under LD 12:12 h. Boxes represent the interquartile ranges, split with the median lines. Whiskers represent the range of the data excluding outliers (>2 SD), which are represented by open circles.
600
(a)
After entrainment to LD 12:12 h, spiders displayed free-running rhythms in huddle duration when placed in constant darkness (Fig. 4). All but one of the 15 individuals showed significant periodicity under DD (chi-square periodograms: P < 0.001 in all cases), with a mean SE period among individuals of 21.94 0.54 h (range 18.7e25.0 h). This period was significantly shorter than 24 h (one-sample t test: t14 ¼ 3.83, P ¼ 0.002). Similarly, after entrainment to LD 12:12 h, spiders displayed free-runs in huddle duration when placed in constant light (Fig. 5). In this case, all but three of the 15 individuals showed significant periodicity under LL
700
Spider '12' under DD
600
500
400
300
300
200
200
100
900
100 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900
0 18 19 20 21 22 23 24 25 26 27 28 29 30
(c)
Spider 'C' under DD
700 600
Amplitude
Huddle duration (s)
800
500 400 300 200
300
2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900
100 0
(e)
Spider '26' under DD
250 200 150 100
2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900
50 0
19.0
500
400
0
(b)
500 (d) 22.0 450 400 350 300 250 200 150 100 50 0 18 19 20 21 22 23 24 25 26 27 28 29 30
500 450 (f) 400 350 300 250 200 150 100 50 0 18 19 20 21 22 23 24 25 26 27 28 29 30 Period (h)
Time of day (hours) Figure 4. (a, c, e) Diel rhythms in huddle responses of three individual L. cornutus under DD. Black bars: subjective night; grey bars: subjective day. (b, d, f) Chi-square periodograms of huddle durations for each individual. Solid line indicates the amplitude of detected periodicities; a significant circadian periodicity (P < 0.001) is indicated by a peak above the dashed line. The first two individuals showed significant periodicity (aed), while the third individual did not (eef).
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
250
(a)
553
800 Spider 'CCC' under LL
(b)
700
200
21.5
600 500
150
400 100
300 200
50
100 0 18 19 20 21 22 23 24 25 26 27 28 29 30
2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700
0
300
(c)
700 Spider 'NNN' under LL
Amplitude
Huddle duration (s)
24.0
500
200 150 100
400 300 200
50
100 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700
0
1200
(d)
600
250
(e)
0 18 19 20 21 22 23 24 25 26 27 28 29 30
450 Spider 'D2' under LL
400
1000
(f)
350 300
800
250
600
200 150
400
100
200
50 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700
0
0 18 19 20 21 22 23 24 25 26 27 28 29 30 Period (h)
Time of day (hours) Figure 5. (a, c, e) Diel rhythms in huddle responses of three individual L. cornutus under LL. White bars: subjective day; grey bars: subjective night. (b, d, f) Chi-square periodograms of huddle durations for each individual. Solid line indicates the amplitude of detected periodicities; a significant circadian periodicity (P < 0.001) is indicated by a peak above the dashed line. The first two individuals showed significant periodicity (aed), while the third individual did not (eef).
(chi-square periodograms: P ¼ 0.001), with a mean SE period among individuals of 25.24 0.89 h (range 20.0e28.7 h). This period was not significantly longer than 24 h (one-sample t test: t14 ¼ 1.78, P ¼ 0.10). However, the difference in period length between spiders under DD and LL was significant (two-sample t test: t14 ¼ 3.52, two-tailed P ¼ 0.002). DISCUSSION It is clear from our results that there is diel rhythmicity in the antipredator huddle response in L. cornutus. This nocturnal spider huddles for relatively long durations in the photophase, when it does not actively forage, and for shorter durations in the scotophase, when
it is typically foraging in its web. These results support the hypothesis that a cost is associated with long-duration huddle responses at night (when there is the possibility of prey capture) but not during the day. Alternatively, this pattern also would be consistent with there being a greater antipredator advantage from the behaviour during the day (i.e. if there are more diurnal predators). Many of the main predators of spiders (birds and wasps) are diurnal, which is widely presumed to explain the fact that most orb-weavers are nocturnal (Foelix 1996). Future comparative studies of antipredator behaviour in diurnal and continuously foraging orb-weavers may discern between these alternative hypotheses. We found strong evidence that this antipredator behaviour is influenced by a circadian clock, with the majority of the spiders
554
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
tested showing significant free-runs in both DD and LL. The circadian influence on this antipredator behaviour is in sharp contrast with the timing of web building. Observations of nocturnal orbweavers under laboratory conditions (Ramousse & Davis 1976) and a diurnal orb-weaver during a total solar eclipse (Uetz et al. 1994) suggest that web building (and web removal) responds directly to light conditions, rather than being controlled by an endogenous, light-entrainable circadian oscillator. Because of the relative ease of recording, the vast majority of studies on the circadian control of behaviour in animals have been concerned with rhythms of general locomotor activity. Nevertheless, circadian rhythmicity has been confirmed for a diversity of more specific, well-defined behaviours such as feeding, mating, oviposition, predation, territorial defence and migration (Brady 1981; Rusak 1981; Drickamer et al. 2002). Most relevant to the present study are circadian rhythms of aggressive behaviours, given that antipredator behaviour is part of an aggression-related syndrome in other spiders (Pruitt et al. 2008, 2010). Aggression is known to be under circadian control in nocturnal rodents (Landau 1975). Under light:dark cycles, peak aggression occurs just after the light-to-dark transition. Phase control of the behaviour may be accomplished by an increase in plasma corticosterone that just anticipates scotophase (Haller et al. 2000). Similarly, aggression in crayfish (which also are nocturnally active) occurs most often at the light-to-dark transition and is driven by a circadian pacemaker (Farca Luna et al. 2009). In the present study, we have identified a circadian rhythm in a novel, aggression-related behaviour, the huddle response. The behaviour itself is not aggressive; it is an antipredator defence. However, by varying huddling duration, so that it is significantly longer during the day than at night, the spider is able to be more aggressive at the appropriate phase of the daily cycle without sacrificing use of the behaviour. Likewise, one might predict a lower sensory threshold for releasing the huddle response during the day. We speculate that the daily oscillation in antipredator behaviour may be part of a larger oscillation in overall aggression levels. We know that huddle response duration is affected by experimental manipulation via the biogenic amines octopamine and serotonin (Jones et al. 2011), and that these compounds have broad hormonal-like effects on aggression levels in a wide variety of arthropods (Roeder 1999; Kravitz & Huber 2003). We also know that, in another family of spiders, the huddle response duration is negatively correlated with aggressiveness towards prey and conspecifics (Pruitt et al. 2008, 2010). It follows that a physiological daily oscillation in aggression-related behaviours would allow spiders to adaptively ‘switch’ from predator to prey mode, effectively balancing the relative risks of predation and lost foraging. Having such a shift under circadian control would give the added selective advantage of being able to anticipate (i.e. physiologically prepare for) changing conditions. Surprisingly, there is a paucity of direct evidence for circadian adaptiveness under natural conditions. In one of the very few such studies, DeCoursey et al. (1997) showed that antelope ground squirrels with lesions of the SCN (suprachiasmatic nuclei, the tiny portion of the brain containing the primary circadian oscillator in mammals) were less likely to survive predation from a feral cat in an outdoor enclosure in the desert than SCN-intact squirrels. A more extensive study (DeCoursey et al. 2000) showed that SCN-lesioned chipmunks suffered significantly greater predation from weasels in a wilderness habitat in the Allegheny Mountains of Virginia than did unoperated and sham-operated controls. We believe that our discovery of circadian control of antipredator behaviour in the orb-weaving spider, and the potential for quantifying the costs and benefits of the behaviour in orb-weavers, will allow future studies to provide robust insight into the adaptiveness of circadian rhythms. Furthermore, the potential
to neurochemically manipulate antipredator behaviour will also facilitate our understanding of the connections between physiology, behaviour and ecology. Acknowledgments We thank the Department of Biological Sciences, N. Weber, B. Linville, J Price and K. Tipton for logistical support of this project. This work was funded in part by the ETSU Honors College through Student-Faculty Collaborative Grants to T. Akoury and C. Hauser. We are also grateful for the helpful comments of the editor and two anonymous referees. References Bellmann, H. 1997. Kosmos-Atlas Spinnentiere Europas. Stuttgart: Frankh-Kosmos Verlag. Brady, J. 1981. Behavioral rhythms in invertebrates. In: Handbook of Behavioral Neurobiology. Vol. 4: Biological Rhythms (Ed. by J. Aschoff), pp. 125e144. New York: Plenum. Carico, J. E. 1986. Web removal patterns in orb-weaving spiders. In: Spiders: Webs, Behavior, and Evolution (Ed. by W. A. Shear), pp. 306e318. Stanford: Stanford University Press. DeCoursey, G. & DeCoursey, P. J. 1964. Adaptive activity rhythms in bats. Biological Bulletin, 126, 14e27. DeCoursey, P. J. 1989. Photoentrainment of circadian rhythms: an ecologist’s viewpoint. In: Circadian Clocks and Ecology (Ed. by T. Hiroshigi & K.-I. Honma), pp. 187e206. Sapporo: University of Hokkaido Press. DeCoursey, P. J., Krulas, J., Mele, G. & Holley, D. 1997. Circadian performance of suprachiasmatic nuclei (SCN)-lesioned antelope ground squirrels in a desert enclosure. Physiology & Behavior, 62, 1099e1108. DeCoursey, P. J., Walker, J. K. & Smith, S. A. 2000. A circadian pacemaker in freeliving chipmunks. Essential for survival? Journal of Comparative Physiology A, 186, 169e180. Drickamer, L. C., Vessey, S. H. & Jakob, E. M. 2002. Animal Behavior. 5th edn. New York: McGraweHill. Farca Luna, A. J., Hurtado-Zavala, J. I., Reischig, T. & Heinrich, R. 2009. Circadian regulation of agonistic behavior in groups of parthenogenetic marbled crayfish, Procambarus sp. Journal of Biological Rhythms, 24, 64e72. Foelix, R. F. 1996. Biology of Spiders. 2nd edn. New York: Oxford University Press. Haller, J., Millar, S., van de Schraaf, J., de Kloet, R. E. & Kruk, M. R. 2000. The active phase-related increase in corticosterone and aggression are linked. Journal of Neuroendocrinology, 12, 431e436. Jones, T. C., Akoury, T. S., Hauser, C. K., Neblett, M. F., II, Linville, B. J., Edge, A. A. & Weber, N. O. 2011. Octopamine and serotonin have opposite effects on antipredator behavior in the orb-weaving spider, Larinioides cornutus. Journal of Comparative Physiology A, Online First, 11 April 2011. Kravitz, E. A. & Huber, R. 2003. Aggression in invertebrates. Current Opinion in Neurobiology, 13, 726e743. Landau, I. T. 1975. Light-dark rhythms in aggressive behavior of the male golden hamster. Physiology & Behavior, 14, 767e774. Levi, H. W. & Levi, L. R. 1990. Spiders and Their Kin. New York: Golden Press. Loher, W. 1972. Circadian control of locomotion in the cricket Teleogryllus commodus Walker. Journal of Comparative Physiology, 79, 173e190. Moore, D. & Doherty, P. 2009. Acquisition of a time-memory in forager honey bees. Journal of Comparative Physiology A, 195, 741e751. Pruitt, J. N., Riechert, S. E. & Jones, T. C. 2008. Behavioural syndromes and their fitness consequences in a socially polymorphic spider, Anelosimus studiosus. Animal Behaviour, 76, 871e879. Pruitt, J. N., Riechert, S. E., Iturralde, G., Vega, M., Fitzpatrick, B. M. & Avilés, L. 2010. Population differences in behavior are explained by shared withinpopulation trait correlations. Journal of Evolutionary Biology, 23, 748e756. Ramousse, R. & Davis, F., III 1976. Web-building time in a spider: preliminary applications of ultrasonic detection. Physiology & Behavior, 17, 997e1000. Rayor, L. S. 1996. Attack strategies of predatory wasps (Hymenoptera: Pompilidae; Specidae) on colonial orb web-building spiders (Araneidae: Metepeira incrassata). Journal of the Kansas Entomological Society, 69, 67e75. Roeder, T. 1999. Octopamine in invertebrates. Progress in Neurobiology, 59, 533e561. Rusak, B. 1981. Vertebrate behavioral rhythms. In: Handbook of Behavioral Neurobiology. Vol. 4: Biological Rhythms (Ed. by J. Aschoff), pp. 183e213. New York: Plenum. Schmitt, A., Schuster, M. & Barth, F. G. 1990. Daily locomotor patterns in three species of Cupiennius (Araneae, Ctenidae): the males are the wandering spiders. Journal of Arachnology, 18, 248e255. Seyfarth, E.-A. 1980. Daily patterns of locomotor activity in a wandering spider. Physiological Entomology, 5, 199e206. Sih, A., Bell, A. M. & Johnson, J. C. 2004. Behavioral syndromes: an ecological and evolutionary overview. Trends in Ecology & Evolution, 19, 372e378.
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555 Stowe, M. K. 1986. Prey specialization in the Araneidae. In: Spiders: Webs, Behavior, and Evolution (Ed. by W. A. Shear), pp. 101e131. Stanford: Stanford University Press. Suter, R. B. 1993. Circadian rhythmicity and other patterns of spontaneous motor activity in Frontinella pyramitela (Linyphiidae) and Argyrodes trigonum (Theridiidae). Journal of Arachnology, 21, 6e22. Uetz, G. W., Hieber, C. S., Jakob, E. M., Wilcox, S., Kroeger, D., McCrate, A. & Mostrom, A. M. 1994. Behavior of colonial orb-weaving spiders during a solar eclipse. Ethology, 96, 24e32.
555
Uetz, G. W., Boyle, J., Heiber, C. S. & Wilcox, S. 2002. Antipredator benefits of group living in colonial web-building spiders: the ‘early warning effect’. Animal Behaviour, 63, 445e452. Wise, D. H. 1993. Spiders in Ecological Webs. Cambridge: Cambridge University Press. Yamashita, S. & Nakamura, T. 1999. Circadian oscillation of sensitivity of spider eyes: diurnal and nocturnal spiders. Journal of Experimental Biology, 202, 2539e2542.
Contents lists available at ScienceDirect
Animal Behaviour journal homepage: www.elsevier.com/locate/anbehav
Evidence of circadian rhythm in antipredator behaviour in the orb-weaving spider Larinioides cornutus Thomas C. Jones*, Tamer S. Akoury, Christopher K. Hauser, Darrell Moore Department of Biological Sciences, East Tennessee State University
a r t i c l e i n f o Article history: Received 26 April 2011 Initial acceptance 11 May 2011 Final acceptance 31 May 2011 Available online 20 July 2011 MS. number: A11-00339R Keywords: aggression circadian rhythm huddle response Larinioides cornutus orb-weaving spider thanatosis
Ecologically, spiders are both predators and prey. Therefore, they must balance being aggressive enough to forage successfully, but not so aggressive that they become overly exposed to predation. Some species of spiders actively forage during clearly defined periods of the day, leading to the hypothesis that they should be less aggressive (or more defensive) during periods when they are not foraging, predicting that antipredator behaviour should be more pronounced during inactive foraging times. We tested the antipredator ‘huddle response’ in a nocturnal foraging orb-weaver, Larinioides cornutus, and found that, as predicted, the spiders huddled longer in the day than at night. We then conducted tests to determine whether the cycling of the response was regulated by an internal clock (circadian), and we found that huddle duration continued to cycle under constant dark (with periodicity significantly less than 24 h) as well as under constant light (periodicity nonsignificantly longer than 24 h). This work adds a novel behaviour to the list of behaviours under circadian control and also to the surprisingly short list of studies demonstrating circadian rhythm in spiders. Ó 2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
It is commonly assumed that circadian rhythms are adaptive, directing the organism to perform each of its various functions, including behaviour, at the most appropriate time of day. Circadian rhythms typically are entrained by environmental cycles, most notably the daily alternation of light and dark. Entrainment ensures that the rhythm maintains a fixed phase relationship with the environmental cycle. Often, the entrained rhythm makes it possible for the organism to anticipate (i.e. perform a function ahead of) environmental changes such as sunrise or sunset. The entrained circadian rhythm thus may allow the organism to be proactive rather than reactive with respect to daily periodic environmental stimuli. Experiments conducted in the laboratory with abrupt changes in light levels are able to differentiate circadian-based anticipatory behaviour from behaviours that respond directly to (gradually) changing light levels. For example, the entrained circadian rhythm of activity in bats (DeCoursey & DeCoursey 1964) and flying squirrels (DeCoursey 1989) allows these nocturnal animals to anticipate sunset, thereby enabling them to begin their night-time activities at the appropriate light intensity level. In another laboratory study, performance of the calling song is primarily a nocturnal behaviour in male crickets (Teleogryllus commodus): it occurs throughout the night but actually begins * Correspondence: T. C. Jones, Department of Biological Sciences, East Tennessee State University, Johnson City, TN, 37614, U.S.A. E-mail address: [email protected] (T. C. Jones).
1e3 h before sunset (Loher 1972). In the field, honeybees, using their circadian clock-driven time-memory, anticipate the time of day at which they encounter profitable food sources on previous days by scheduling reconnaissance flights to that particular location in the environment, with the amount of anticipation depending on the amount of previous experience with that source (Moore & Doherty 2009). In an ecological sense, spiders are both predators and prey (Wise 1993). In particular, orb-weaving spiders aggressively attack insects caught in their webs, but are themselves subject to attack from wasps and birds (Foelix 1996). Most orbweaving species are either nocturnal or diurnal, actively foraging during only one part of the daily cycle, and remaining secluded during the other (Carico 1986; Stowe 1986). In a sense, these spiders must behaviourally ‘switch’ from being predators to avoiding being prey on a daily basis. One would therefore expect there to be a selective advantage for spiders to anticipate dawn and dusk, particularly if prey levels are higher during these environmental transitions. However, very little information exists concerning the circadian control of behaviour in spiders. Most of the studies to date are concerned with locomotor activity (e.g. Seyfarth 1980; Schmitt et al. 1990; Suter 1993) or changes in sensitivity of the eyes (e.g. Yamashita & Nakamura 1999). Here, we present evidence for circadian regulation of an antipredator behaviour in the orb-weaving spider Larinioides cornutus. An antipredator behaviour commonly observed in orb-weaving spiders is ‘bailing out’, in which the spider drops from the web and
0003-3472/$38.00 Ó 2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.anbehav.2011.06.009
550
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
remains motionless with its legs flexed tightly to its ventral midline for a period of time (Rayor 1996). This behaviour is illustrated for the species used in the present study in Jones et al. (2011). Uetz et al. (2002) found that a colonial orb-weaving spider would express this behaviour in response to continuous vibrations applied to a support silk line. That study also found that the likelihood of exhibiting the response increased in those spiders whose neighbours had already been coaxed to bail out. The physical posture assumed by spiders when they bail out of their webs can be elicited in the laboratory, even if the spiders are removed from their web. Studies by Pruitt et al. (2008, 2010) on social cobweb-building spiders suggested that this ‘huddle response’ (also known as thanatosis, or death feigning) is part of an overall behavioural syndrome (sensu Sih et al. 2004) related to aggression: spiders that are more aggressive towards prey and conspecifics do not hold the huddle posture as long as less aggressive spiders. A recent study of the effect of biogenic amines on the huddle response in the orb-weaving spider Larinioides cornutus found that experimentally elevating octopamine significantly shortened the duration of the huddle response, while elevating serotonin significantly prolonged it (Jones et al. 2011). That study found that the effect of serotonin decreased over the first 6 h after treatment, but then increased and was most pronounced 24 h after dosing, leading to the speculation that these spiders may show rhythmicity in antipredator behaviour. In the present study we explored whether the huddle response of the nocturnal orb-weaver L. cornutus (Aranaeidae) shows diel periodicity, and if so, whether periodicity is regulated by an internal (i.e. circadian) clock. We hypothesized that these nocturnal spiders would huddle longer during the day than at night, reflecting the fact that they forage at night but not during daylight hours. Our rationale was that there is a significant cost to bailing out of the web at night, in terms of missed prey capture opportunities, that does not exist during the day. Alternatively, the predation risk for the spider (particularly from wasps and birds) may be higher for the spider during daylight hours than at night. METHODS Study Species, Collection and Housing Larinioides cornutus (Clerck 1757) is an orb-weaver in the family Araneidae, which is common to Europe and North America (Levi & Levi 1990). It is a nocturnal spider, constructing its web at dusk and
actively foraging throughout the night, then removing its web at dawn, and remaining in a hidden retreat during the day (Bellmann 1997). Female adult L. cornutus were collected in September and October 2009 from buildings and shrubbery in northeastern Tennessee (U.S.A.). The spiders were extracted from their webs and individually housed in 250 ml plastic deli containers. Spiders that were housed for more than a few days were given water twice a week, fed crickets every other week, and kept at approximately 23 C under 12:12 h light:dark cycle. Quantifying the Huddle Response Web-building spiders typically show an antipredator huddle response in which they draw their legs in tight to their body and remain motionless. If left undisturbed, the spiders will break out of the huddle after a period of time. This huddling appears to be an ‘allor-nothing’ response, and there is no known effect of stimulus strength on huddle duration in L. cornutus (T. C. Jones, unpublished data) or in other spiders (Pruitt et al. 2008, 2010). There is also no evidence of habituation to the stimulus (Jones et al. 2011). We used the method of Jones et al. (2011) to quantify the response. Individual spiders were removed from their containers and placed in a glass bowl (15 cm diameter, 6 cm high), and were given 30 s to acclimate to the dish. To trigger the huddle response we delivered a puff of air from approximately 10 cm away. The air movement deflects fine sensory hairs (trichobothria) on the spider’s legs, stimulating the stereotypical huddle response (Foelix 1996). We determined the duration of the huddling response to the nearest second with a stopwatch from the moment the spiders initiated the huddle to the moment that they broke out of it. Individuals that did not huddle after three attempts to initiate the response received a score of 0 s; individuals that initiated a response but immediately broke out of it, received a score of 1 s. We cleaned each dish with ethanol between trials. Measuring Diel and Circadian Rhythmicity in the Huddle Response For this part of the study, adult L. cornutus were housed in a light-controlled room under an LD 12:12 h for at least 5 days prior to testing. The room was carefully sealed against light leaks, and the door was double-screened with black-out curtains to prevent light from leaking in when opened. All dark observations were conducted under low light filtered with two layers of GamColorÒ Medium Red XT gel. Ten spiders were tested for 72 h under LD
Table 1 Chi-square and F periodogram analyses of huddle response duration for individual spiders under LD 12:12 h, DD and LL LD 12:12 h
DD
ID
c2 a¼0.001
a¼0.01
Lc-22 Lc-14 Lc-1 Lc-12 Lc-CCC Lc-QQQQ Lc-JJJ Lc-HHH Lc-11 Lc-26
23.9 23.9 28 28 23.9 23.9 23.9 23.9 23.9 23.9
23.9 23.9 28 28 23.9 23.9 23.9 23.9 23.9 23.9
F
LL
ID
c2 a¼0.001
a¼0.01
Lc-C Lc-P Lc-12 Lc-D1 Lc-JJJ Lc-D2 Lc-CCC Lc-1 Lc-11 Lc-22 Lc-18 Lc-MM Lc-LL Lc-HHH Lc-26
22 19.9 19.9 20 24 23.9 23.9 25 18.7 22 23.9 22.7 21.3 19.9 NS
22 19.9 21.3 20 24 23.9 23.9 25 18.7 21.4 23.9 22.7 21.3 19.9 NS
F
ID
c2 a¼0.001
a¼0.01
Lc-I Lc-XI Lc-HHH Lc-CCC Lc-D2 Lc-X2 Lc-NNN Lc-X4 Lc-II Lc-JJJ Lc-12 Lc-M Lc-C Lc-X5 Lc-X3
26.7 26.7 26.7 22 NS 29.5 23.9 NS 27 28.7 20.6 28 28.7 20 23.9
26.7 26.7 26.7 22 NS NS 23.9 NS 27 28.7 20.6 28 28.7 18 23.9
F
Given are the main periods (s), in hours, that were significant at a ¼ 0.001 for chi-square periodograms and that were significant at a ¼ 0.01 for F periodograms (individuals with nonsignificant periodograms are indicated in bold).
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
350
1200 (b)
(a)
300
1000
250 Amplitude
Huddle duration (s)
551
200 150
800 600
100
400
50
200
08
00 12 00 16 00 20 0 00 0 00 04 00 08 00 12 00 16 00 20 00 00 00 04 00 08 00 12 00 16 00 20 00 00 00 04 00
0
0 18 19 20 21 22 23 24 25 26 27 28 29 30 Period
Time of day (hours) Figure 1. Diel rhythms in huddle responses of L. cornutus under LD 12:12 h. (a) Means and standard errors of huddle response durations of 10 spiders tested at six times per day over the course of 3 days. Open bars: spiders tested with lights on; solid bars: spiders tested with lights off. (b) Chi-square periodogram of the mean huddle durations. The solid line indicates the amplitude of detected periodicities; a significant periodicity (P < 0.01) is indicated by a peak above the diagonal line.
12:12 h conditions. In this group, observations began at 0800 hours (0.5 h after lights on), and were conducted every 4 h thereafter. We also tested huddle responses of spiders under constant dark (DD) and constant light (LL) conditions, at 23 C. In both cases, the spiders were tested every 4 h beginning at 0700 hours (0.5 h prior to lights on). Fifteen spiders were tested for 24 h under normal LD, and then for 84 h in constant dark (DD). Fifteen spiders also were tested under constant light (LL). In this case, the 24 h LD pretrial was followed by 108 h of continuous light. Prior to both the DD and LL trials, the spiders were entrained to LD 12:12 h for 5 days. Statistical Analyses The data were compiled and figures were created with Excel 2007, and statistical analyses were run in Minitab 15. We analysed periodicity and created periodograms with ClockLab (Actimetrics, Wilmette, IL, U.S.A.). Periodicities were determined objectively by employing chi-square periodograms at an alpha level of 0.001 for periods between 18 and 30 h. Only the highest peaks above the significance line were used. These periodicities were not accepted
Light
80
as significant unless additionally confirmed by F periodograms at an alpha level of 0.01 (Table 1). RESULTS Diel Rhythmicity Spiders under LD 12:12 h conditions clearly showed diel rhythmicity in the huddle response, holding the posture longer in the light than in the dark (Fig. 1a). Chi-square periodogram analysis of mean huddle duration found significant periodicity (P < 0.001), with a dominant period of 23.9 h (Fig. 1b). Periodogram analyses of individual spiders found significant periodicity (P < 0.001) of 23.9 h for eight individuals and 28 h for two individuals. Data pooled for all individuals showed that the distributions of huddle response durations in both the light and dark phases were skewed to the right, indicating that there were relatively few very long huddle durations (Fig. 2). However, huddle durations in the light phase (median ¼ 130.1) tended to be longer than in the dark (median ¼ 39.2), with nine of the 10 individuals huddling
Dark
70
Frequency
60 50 40 30 20 10 0
0
100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 Huddle duration (s)
Figure 2. Huddle durations of L. cornutus under LD 12:12 h. Observations under light and dark conditions are shown separately.
552
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
significantly longer in the light phase than in the dark (Manne Whitney U tests comparing each individual’s responses, P range 0.019 to <0.001). There also were differences between individuals in huddle duration (KruskaleWallis test: H9 ¼ 40.54, P < 0.001). While all individuals displayed some short huddle durations, a few individuals consistently huddled longer than others (Fig. 3).
800
Huddle duration (s)
700 600 500 400 300
Rhythmicity under Constant Conditions
200 100 0 S−01 S−02 S−03 S−04 S−05 S−06 S−07 S−08 S−09 S−10 Individual spiders
Figure 3. Huddle durations of individual spiders under LD 12:12 h. Boxes represent the interquartile ranges, split with the median lines. Whiskers represent the range of the data excluding outliers (>2 SD), which are represented by open circles.
600
(a)
After entrainment to LD 12:12 h, spiders displayed free-running rhythms in huddle duration when placed in constant darkness (Fig. 4). All but one of the 15 individuals showed significant periodicity under DD (chi-square periodograms: P < 0.001 in all cases), with a mean SE period among individuals of 21.94 0.54 h (range 18.7e25.0 h). This period was significantly shorter than 24 h (one-sample t test: t14 ¼ 3.83, P ¼ 0.002). Similarly, after entrainment to LD 12:12 h, spiders displayed free-runs in huddle duration when placed in constant light (Fig. 5). In this case, all but three of the 15 individuals showed significant periodicity under LL
700
Spider '12' under DD
600
500
400
300
300
200
200
100
900
100 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900
0 18 19 20 21 22 23 24 25 26 27 28 29 30
(c)
Spider 'C' under DD
700 600
Amplitude
Huddle duration (s)
800
500 400 300 200
300
2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900
100 0
(e)
Spider '26' under DD
250 200 150 100
2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900
50 0
19.0
500
400
0
(b)
500 (d) 22.0 450 400 350 300 250 200 150 100 50 0 18 19 20 21 22 23 24 25 26 27 28 29 30
500 450 (f) 400 350 300 250 200 150 100 50 0 18 19 20 21 22 23 24 25 26 27 28 29 30 Period (h)
Time of day (hours) Figure 4. (a, c, e) Diel rhythms in huddle responses of three individual L. cornutus under DD. Black bars: subjective night; grey bars: subjective day. (b, d, f) Chi-square periodograms of huddle durations for each individual. Solid line indicates the amplitude of detected periodicities; a significant circadian periodicity (P < 0.001) is indicated by a peak above the dashed line. The first two individuals showed significant periodicity (aed), while the third individual did not (eef).
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
250
(a)
553
800 Spider 'CCC' under LL
(b)
700
200
21.5
600 500
150
400 100
300 200
50
100 0 18 19 20 21 22 23 24 25 26 27 28 29 30
2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700
0
300
(c)
700 Spider 'NNN' under LL
Amplitude
Huddle duration (s)
24.0
500
200 150 100
400 300 200
50
100 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700
0
1200
(d)
600
250
(e)
0 18 19 20 21 22 23 24 25 26 27 28 29 30
450 Spider 'D2' under LL
400
1000
(f)
350 300
800
250
600
200 150
400
100
200
50 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700 1100 1500 1900 2300 0300 0700
0
0 18 19 20 21 22 23 24 25 26 27 28 29 30 Period (h)
Time of day (hours) Figure 5. (a, c, e) Diel rhythms in huddle responses of three individual L. cornutus under LL. White bars: subjective day; grey bars: subjective night. (b, d, f) Chi-square periodograms of huddle durations for each individual. Solid line indicates the amplitude of detected periodicities; a significant circadian periodicity (P < 0.001) is indicated by a peak above the dashed line. The first two individuals showed significant periodicity (aed), while the third individual did not (eef).
(chi-square periodograms: P ¼ 0.001), with a mean SE period among individuals of 25.24 0.89 h (range 20.0e28.7 h). This period was not significantly longer than 24 h (one-sample t test: t14 ¼ 1.78, P ¼ 0.10). However, the difference in period length between spiders under DD and LL was significant (two-sample t test: t14 ¼ 3.52, two-tailed P ¼ 0.002). DISCUSSION It is clear from our results that there is diel rhythmicity in the antipredator huddle response in L. cornutus. This nocturnal spider huddles for relatively long durations in the photophase, when it does not actively forage, and for shorter durations in the scotophase, when
it is typically foraging in its web. These results support the hypothesis that a cost is associated with long-duration huddle responses at night (when there is the possibility of prey capture) but not during the day. Alternatively, this pattern also would be consistent with there being a greater antipredator advantage from the behaviour during the day (i.e. if there are more diurnal predators). Many of the main predators of spiders (birds and wasps) are diurnal, which is widely presumed to explain the fact that most orb-weavers are nocturnal (Foelix 1996). Future comparative studies of antipredator behaviour in diurnal and continuously foraging orb-weavers may discern between these alternative hypotheses. We found strong evidence that this antipredator behaviour is influenced by a circadian clock, with the majority of the spiders
554
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555
tested showing significant free-runs in both DD and LL. The circadian influence on this antipredator behaviour is in sharp contrast with the timing of web building. Observations of nocturnal orbweavers under laboratory conditions (Ramousse & Davis 1976) and a diurnal orb-weaver during a total solar eclipse (Uetz et al. 1994) suggest that web building (and web removal) responds directly to light conditions, rather than being controlled by an endogenous, light-entrainable circadian oscillator. Because of the relative ease of recording, the vast majority of studies on the circadian control of behaviour in animals have been concerned with rhythms of general locomotor activity. Nevertheless, circadian rhythmicity has been confirmed for a diversity of more specific, well-defined behaviours such as feeding, mating, oviposition, predation, territorial defence and migration (Brady 1981; Rusak 1981; Drickamer et al. 2002). Most relevant to the present study are circadian rhythms of aggressive behaviours, given that antipredator behaviour is part of an aggression-related syndrome in other spiders (Pruitt et al. 2008, 2010). Aggression is known to be under circadian control in nocturnal rodents (Landau 1975). Under light:dark cycles, peak aggression occurs just after the light-to-dark transition. Phase control of the behaviour may be accomplished by an increase in plasma corticosterone that just anticipates scotophase (Haller et al. 2000). Similarly, aggression in crayfish (which also are nocturnally active) occurs most often at the light-to-dark transition and is driven by a circadian pacemaker (Farca Luna et al. 2009). In the present study, we have identified a circadian rhythm in a novel, aggression-related behaviour, the huddle response. The behaviour itself is not aggressive; it is an antipredator defence. However, by varying huddling duration, so that it is significantly longer during the day than at night, the spider is able to be more aggressive at the appropriate phase of the daily cycle without sacrificing use of the behaviour. Likewise, one might predict a lower sensory threshold for releasing the huddle response during the day. We speculate that the daily oscillation in antipredator behaviour may be part of a larger oscillation in overall aggression levels. We know that huddle response duration is affected by experimental manipulation via the biogenic amines octopamine and serotonin (Jones et al. 2011), and that these compounds have broad hormonal-like effects on aggression levels in a wide variety of arthropods (Roeder 1999; Kravitz & Huber 2003). We also know that, in another family of spiders, the huddle response duration is negatively correlated with aggressiveness towards prey and conspecifics (Pruitt et al. 2008, 2010). It follows that a physiological daily oscillation in aggression-related behaviours would allow spiders to adaptively ‘switch’ from predator to prey mode, effectively balancing the relative risks of predation and lost foraging. Having such a shift under circadian control would give the added selective advantage of being able to anticipate (i.e. physiologically prepare for) changing conditions. Surprisingly, there is a paucity of direct evidence for circadian adaptiveness under natural conditions. In one of the very few such studies, DeCoursey et al. (1997) showed that antelope ground squirrels with lesions of the SCN (suprachiasmatic nuclei, the tiny portion of the brain containing the primary circadian oscillator in mammals) were less likely to survive predation from a feral cat in an outdoor enclosure in the desert than SCN-intact squirrels. A more extensive study (DeCoursey et al. 2000) showed that SCN-lesioned chipmunks suffered significantly greater predation from weasels in a wilderness habitat in the Allegheny Mountains of Virginia than did unoperated and sham-operated controls. We believe that our discovery of circadian control of antipredator behaviour in the orb-weaving spider, and the potential for quantifying the costs and benefits of the behaviour in orb-weavers, will allow future studies to provide robust insight into the adaptiveness of circadian rhythms. Furthermore, the potential
to neurochemically manipulate antipredator behaviour will also facilitate our understanding of the connections between physiology, behaviour and ecology. Acknowledgments We thank the Department of Biological Sciences, N. Weber, B. Linville, J Price and K. Tipton for logistical support of this project. This work was funded in part by the ETSU Honors College through Student-Faculty Collaborative Grants to T. Akoury and C. Hauser. We are also grateful for the helpful comments of the editor and two anonymous referees. References Bellmann, H. 1997. Kosmos-Atlas Spinnentiere Europas. Stuttgart: Frankh-Kosmos Verlag. Brady, J. 1981. Behavioral rhythms in invertebrates. In: Handbook of Behavioral Neurobiology. Vol. 4: Biological Rhythms (Ed. by J. Aschoff), pp. 125e144. New York: Plenum. Carico, J. E. 1986. Web removal patterns in orb-weaving spiders. In: Spiders: Webs, Behavior, and Evolution (Ed. by W. A. Shear), pp. 306e318. Stanford: Stanford University Press. DeCoursey, G. & DeCoursey, P. J. 1964. Adaptive activity rhythms in bats. Biological Bulletin, 126, 14e27. DeCoursey, P. J. 1989. Photoentrainment of circadian rhythms: an ecologist’s viewpoint. In: Circadian Clocks and Ecology (Ed. by T. Hiroshigi & K.-I. Honma), pp. 187e206. Sapporo: University of Hokkaido Press. DeCoursey, P. J., Krulas, J., Mele, G. & Holley, D. 1997. Circadian performance of suprachiasmatic nuclei (SCN)-lesioned antelope ground squirrels in a desert enclosure. Physiology & Behavior, 62, 1099e1108. DeCoursey, P. J., Walker, J. K. & Smith, S. A. 2000. A circadian pacemaker in freeliving chipmunks. Essential for survival? Journal of Comparative Physiology A, 186, 169e180. Drickamer, L. C., Vessey, S. H. & Jakob, E. M. 2002. Animal Behavior. 5th edn. New York: McGraweHill. Farca Luna, A. J., Hurtado-Zavala, J. I., Reischig, T. & Heinrich, R. 2009. Circadian regulation of agonistic behavior in groups of parthenogenetic marbled crayfish, Procambarus sp. Journal of Biological Rhythms, 24, 64e72. Foelix, R. F. 1996. Biology of Spiders. 2nd edn. New York: Oxford University Press. Haller, J., Millar, S., van de Schraaf, J., de Kloet, R. E. & Kruk, M. R. 2000. The active phase-related increase in corticosterone and aggression are linked. Journal of Neuroendocrinology, 12, 431e436. Jones, T. C., Akoury, T. S., Hauser, C. K., Neblett, M. F., II, Linville, B. J., Edge, A. A. & Weber, N. O. 2011. Octopamine and serotonin have opposite effects on antipredator behavior in the orb-weaving spider, Larinioides cornutus. Journal of Comparative Physiology A, Online First, 11 April 2011. Kravitz, E. A. & Huber, R. 2003. Aggression in invertebrates. Current Opinion in Neurobiology, 13, 726e743. Landau, I. T. 1975. Light-dark rhythms in aggressive behavior of the male golden hamster. Physiology & Behavior, 14, 767e774. Levi, H. W. & Levi, L. R. 1990. Spiders and Their Kin. New York: Golden Press. Loher, W. 1972. Circadian control of locomotion in the cricket Teleogryllus commodus Walker. Journal of Comparative Physiology, 79, 173e190. Moore, D. & Doherty, P. 2009. Acquisition of a time-memory in forager honey bees. Journal of Comparative Physiology A, 195, 741e751. Pruitt, J. N., Riechert, S. E. & Jones, T. C. 2008. Behavioural syndromes and their fitness consequences in a socially polymorphic spider, Anelosimus studiosus. Animal Behaviour, 76, 871e879. Pruitt, J. N., Riechert, S. E., Iturralde, G., Vega, M., Fitzpatrick, B. M. & Avilés, L. 2010. Population differences in behavior are explained by shared withinpopulation trait correlations. Journal of Evolutionary Biology, 23, 748e756. Ramousse, R. & Davis, F., III 1976. Web-building time in a spider: preliminary applications of ultrasonic detection. Physiology & Behavior, 17, 997e1000. Rayor, L. S. 1996. Attack strategies of predatory wasps (Hymenoptera: Pompilidae; Specidae) on colonial orb web-building spiders (Araneidae: Metepeira incrassata). Journal of the Kansas Entomological Society, 69, 67e75. Roeder, T. 1999. Octopamine in invertebrates. Progress in Neurobiology, 59, 533e561. Rusak, B. 1981. Vertebrate behavioral rhythms. In: Handbook of Behavioral Neurobiology. Vol. 4: Biological Rhythms (Ed. by J. Aschoff), pp. 183e213. New York: Plenum. Schmitt, A., Schuster, M. & Barth, F. G. 1990. Daily locomotor patterns in three species of Cupiennius (Araneae, Ctenidae): the males are the wandering spiders. Journal of Arachnology, 18, 248e255. Seyfarth, E.-A. 1980. Daily patterns of locomotor activity in a wandering spider. Physiological Entomology, 5, 199e206. Sih, A., Bell, A. M. & Johnson, J. C. 2004. Behavioral syndromes: an ecological and evolutionary overview. Trends in Ecology & Evolution, 19, 372e378.
T. C. Jones et al. / Animal Behaviour 82 (2011) 549e555 Stowe, M. K. 1986. Prey specialization in the Araneidae. In: Spiders: Webs, Behavior, and Evolution (Ed. by W. A. Shear), pp. 101e131. Stanford: Stanford University Press. Suter, R. B. 1993. Circadian rhythmicity and other patterns of spontaneous motor activity in Frontinella pyramitela (Linyphiidae) and Argyrodes trigonum (Theridiidae). Journal of Arachnology, 21, 6e22. Uetz, G. W., Hieber, C. S., Jakob, E. M., Wilcox, S., Kroeger, D., McCrate, A. & Mostrom, A. M. 1994. Behavior of colonial orb-weaving spiders during a solar eclipse. Ethology, 96, 24e32.
555
Uetz, G. W., Boyle, J., Heiber, C. S. & Wilcox, S. 2002. Antipredator benefits of group living in colonial web-building spiders: the ‘early warning effect’. Animal Behaviour, 63, 445e452. Wise, D. H. 1993. Spiders in Ecological Webs. Cambridge: Cambridge University Press. Yamashita, S. & Nakamura, T. 1999. Circadian oscillation of sensitivity of spider eyes: diurnal and nocturnal spiders. Journal of Experimental Biology, 202, 2539e2542.