SOCIAL DOMINANCE: A BEHAVIORAL MECHANISM FOR RESOURCE ALLOCATION IN CRAYFISH
Kandice Christine Fero
A Dissertation
Submitted to the Graduate College of Bowling Green State University in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
August 2008
Committee:
Paul Moore, Advisor
Verner Bingman Graduate Faculty Representative
Sheryl Coombs
Rex Lowe
Steve Vessey ii
ABSTRACT
Paul Moore, Advisor
Social dominance is often equated with priority of access to resources and higher relative fitness. But the consequences of dominance are not always readily advantageous for an individual and therefore, testing of such assumptions is needed in order to appropriately characterize mechanisms of resource competition in animal systems. This dissertation examined the ecological consequences of dominance in crayfish. Specifically, the following questions were addressed: is resource allocation determined by dominance and how does the structure of resources in an environment affect dominance relationships? By examining the mechanism of how dominance may allocate resources in groups of crayfish, we can begin to answer questions concerning what environmental selective pressures are shaping social behavior in this system.
Shelter acquisition and use was examined in a combination of natural, semi-natural, and laboratory studies in order to observe dominance relationships under ecologically relevant conditions. The work presented here shows that: (1) social status has persisting behavioral consequences with regard to shelter use, which are modulated by social context; (2) dominance relationships influence the spatial distribution of crayfish in natural environments such that dominant individuals possess access to more space; (3) resource use strategies differ depending on social history and these strategies may influence larger scale segregation across habitats; and finally, (4) shelter distribution modulates the extent to which social history and shelter ownership influence the formation of subsequent dominance relationships. Taken together, these results demonstrate that dominance has significant consequences for crayfish resource acquisition and holding. However, a complex picture has been revealed as to the nature of dominance iii establishment and the potential resource benefits associated with dominance. Dominant crayfish posses control over shelter and space but the likelihood of becoming dominant and the subsequent resource use consequences are largely dependent on social and environmental context.
iv
For Dan, Young, Don, and Lee v
ACKNOWLEDGMENTS
I am grateful for this opportunity to express my deepest thanks and love for those who have supported me in this endeavor. I first must thank Dr. Paul Moore. Thank you Paul, for having so much passion and enthusiasm for the pursuit of knowledge and for always asking questions. Thank you for always being so generous with your time and energy. Thank you for your confidence in my abilities. Thanks to my dissertation committee, Dr. Vern Bingman, Dr.
Sheryl Coombs, Dr. Rex Lowe, and Dr. Steve Vessey. Thank you for your suggestions, for stimulating conversations, and for your generous support and encouragement. Thanks to the
Department of Biological Sciences and Bowling Green State University for four years of financial support and specifically, thank you Chris Hess, Linda Treeger, and Steve Queen for helping me countless times. Thank you Jay Skock, Nick Kolderman, Michael Braddock, and Jen
Bergner for assistance with video analysis and Tom Zulandt, Art Martin, Jodie Simon and Jen
Bergner for their assistance with field work. And thank you to the members of the Laboratory for
Sensory Ecology for arguing with me, editing countless papers, collecting crayfish, helping me set up experiments and for being such dear friends. Thanks to the University of Michigan
Biological Station for housing and research support. Funding for this research was provided by the following awards and fellowships to K.F.: the BGSU Katzner fund, Non-Service Fellowship and Outstanding Graduate Student Award, the Crustacean Society Graduate Research Fellowship and Best Student Paper Award (SICB 2006). Funding was also provided by a NSF grant to
P.A.M. (IBN# 0131320).
Now I must thank all of the wonderful people that helped me survive graduate school.
Thank you Stephanie Coray and Katie Crane for your continuing friendship. Thank you Paula
Furey for your friendship and camaraderie. I will never forget how we went through this process together, dragging each other to the finish line! Thank you Grounds for Thought for letting me vi spend hundreds of hours occupying your tables for the mere price of a couple of cups of coffee.
Thank you to the Moore family for your kindness and always making me feel welcome. Thank you Rex and Sheryn Lowe for trusting me and Paula with your home. My time in Bowling Green has been enriched by your friendship and generosity. Thank you to the Chibucos family for welcoming me into your lives. Thank you to Marcus. Thank you for your immeasurable love and patience. Thank you for always reminding me of the joy and beauty in my life. Thank you for helping me trust myself. And finally, thank you my loving family. You’ve always accepted me for who I am and have trusted my decisions. You have given me an education but more importantly, you taught me how to think for myself, which is the reason I have chosen the path that I have. Your unconditional support and love make up the core of my self. I dedicate my accomplishment to you.
vii
TABLE OF CONTENTS
Page
CHAPTER I. GENERAL INTRODUCTION ...... 1
CHAPTER II. CONSEQUENCES OF SOCIAL DOMINANCE ON CRAYFISH
RESOURCE USE ...... 6
Introduction ...... 6
Materials and Methods...... 10
Results ...... 16
Discussion ...... 19
CHAPTER III. SOCIAL SPACING OF CRAYFISH IN NATURAL HABITATS:
WHAT ROLE DOES DOMINANCE PLAY? ...... 25
Introduction ...... 25
Materials and Methods...... 28
Results ...... 33
Discussion ...... 34
CHAPTER IV. RESOURCE DISTRIBUTION AND SOCIAL STATUS INFLUENCE
CRAYFISH HABITAT CHOICE ...... 38
Introduction ...... 38
Materials and Methods...... 41
Results ...... 44
Discussion ...... 47
CHAPTER V. RESOURCE DISTRIBUTION EVENS THE PLAYING FIELD FOR
SOCIAL DOMINANCE IN CRAYFISH...... 53 viii
Introduction ...... 53
Materials and Methods...... 55
Results ...... 60
Discussion ...... 63
CHAPTER VI. SUMMARY AND GENERAL CONCLUSIONS ...... 70
LITERATURE CITED ...... 77
APPENDIX A. FIGURE AND TABLES...... 89 ix
LIST OF FIGURES
Figure Page
1 Mean intensities of agonistic interactions for all-male populations ...... 89
2 Dominance activity indices for three all-male populations ...... 90
3 Proportion of agonistic interactions that males engaged in and their
final rank within the hierarchy...... 91
4 Mean number of hours that crayfish of differing ranks spent in shelter...... 92
5 Mean percent of time that crayfish spent in shelter, out of shelter, and at
the feeding area...... 93
6 Mean shifts in dominance score of individuals in populations...... 94
7 Mean of the total shifts in dominance score for crayfish at three ranks ...... 95
8 Experimental arena with three different shelter habitats ...... 96
9 Percent of time that crayfish spent in or out of shelter in each habitat...... 97
10 Number of agonistic encounter in each habitat...... 98
11 Alpha and omega shelter use ...... 99
12 Experimental arena with resident and intruder shelter at different distances ...... 100
13 Proportion of trials where residents are dominant to intruders...... 101
14 Proportion of trials where dominance reversals occurred...... 102 x
LIST OF TABLES
Table Page
1 Dominance activity index and body size linear regressions ...... 103
2 Nearest neighbor distance and dominance score regressions ...... 104
3 Dominance score and diurnal activity across habitats regressions ...... 105
4 Dominance score and nocturnal activity across habitats regressions ...... 106
5 Crayfish behavioral descriptions...... 107
1
CHAPTER I
GENERAL INTRODUCTION
Social dominance, defined by Drews (1993), is characterized by the consistent victory of one individual over another in repeated agonistic interactions. Agonistic interactions are aggressive contests between individuals and are associated with resource acquisition and defense
(King 1973). Thus, an expected consequence of winning is to possess or control desired resources and thereby increase fitness. Dominance in nature is strongly correlated with priority of access to resources, such as food, shelter, and mates (Wilson 1975). In agonistic contests, resources may be acquired by dominance establishment in dyadic interactions, or they may be allocated with respect to relative dominance rank within a hierarchy. Individuals may acquire rank based on predetermined qualities associated with dominance, such as physical attributes that contribute to fighting ability (RHP, Parker 1974), or ranking may result from “social dynamics,” such as winner-loser-effects or bystander effects (Chase 2002). Dominance hierarchies are ultimately predicted to decrease agonism among conspecifics. Agonistic interactions incur high energetic costs and potential injury (Case & Gilpin 1974); as subordinate ranks recognize the fighting ability of higher-ranking individuals, subordinates will subsequently reduce challenges for rank (Clutton-Brock & Harvey 1976).
Dynamics of agonistic interactions are also affected by resources themselves. When resources are clumped or patchy, they are more easily defended and less available to other individuals than when resources are randomly distributed (Noel et al. 2005). A strong relationship exists between resource availability and animal social organization (Johnson et al.
2002). As resources become less available, increased aggression may be expected (Robb &
Grant 1998), thus changing dynamics of agonistic interactions and dominance relationships. 2
Increased resource control due to dominance has been documented in many taxa (Wilson
1975); however, dominance does not always confer higher fitness. In some animal systems in which dominance interactions are prevalent, no consequences are observed on direct fitness due to dominance, for example in female lizards (Langkilde et al. 2005). Dominant male jackdaws were actually found to have decreased reproductive success relative to subordinate males
(Verhulst & Salomons 2004). Studies such as these suggest that the relationships among dominance, resources, and fitness are not as straightforward as previously thought. Acquired resources are an assumed benefit of dominance that out weigh the costs of agonism. Trade-offs of dominance status or rank ultimately dictate levels of agonism present in groups and a question arises concerning why dominance behavior is exhibited when there is no apparent benefit to offset incurred costs. What are the ecological consequences of dominance; is resource allocation determined by dominance relationships, and if so, how may the structure of resources in a habitat be imposing selective pressures on agonism and aggression?
Decapod crustaceans, such as crayfish, serve as an excellent model for examining agonistic interactions (Dingle 1983); a particularly promising opportunity exists to answer fundamental questions concerning the relationship between dominance and resource allocation due to an extensive literature characterizing dominance in this system. Crayfish readily establish dominance during dyadic interactions and form linear hierarchies in the lab that exhibit reduced aggression (Guiasu & Dunham 1997). Crayfish exhibit stereotyped aggression with distinct phases of intensity where opponents will engage each other even in the absence of any potential resources (Bovbjerg 1956; Bruski & Dunham 1987). When resources are not present, reliable predictors of dominance include intrinsic factors such as size (Pavey & Fielder 1996), chelae length (Rutherford, et al. 1995), and aggressive state (Panksepp, et al. 2003). Extrinsic factors 3 such as winner-loser-effects (Daws et al. 2002; Bergman et al. 2003) and perception of chemical cues from opponents (Zulandt-Schneider, et al. 2001) can also predict dominance status.
Agonistic interactions occurring in the presence of resources, such as shelter, indicate that resource ownership reliably predicts positive fight outcomes, even when other physical asymmetries favor the challenging opponent (Peeke, et al. 1995). Shelter is likely a valued resource for crayfish, is often limiting, and is highly contested (Gherardi 2002).
Competition for shelter resources has largely been examined in interspecific studies that focus on mechanisms of competitive exclusion and species displacement. Species such as
Orconectes rusticus and Procambarus clarkii are often characterized as highly aggressive species (Tierney 2000; Klocker & Strayer 2004). This aggressiveness is thought to be a key factor in the success of these two species as invaders that displace native crayfish. In agonistic interactions, O. rusticus and P. clarkii, as well as Orconectes virilis, will commonly evict heterospecifics from shelters (Klocker & Strayer 2004; Gherardi & Daniels 2004, Bovbjerg
1970). Gherardi & Daniels (2004) found that after evicting P. acutus , P. clarkii did not utilize shelter resource, but instead, defended the shelter from the open. This behavior is indicative of
“spiteful behavior” in which individuals may improve their relative fitness by excluding others
(Davies 1978). Species displacements across habitats have also been attributed to interspecific differences in chelae size (Garvey and Stein 1993). Additionally, distributions of O. rusticus, O. virilis , and O. propinquus with respect to habitat type may change depending on risk of predation. During periods of high predation, O. rusticus will monopolize preferred cobble habitats (Hill & Lodge 1994). Exclusion of species from shelter or shelter rich habitat may be directly attributable to agonistic interactions that result in one species being dominant over another. 4
Intraspecific studies have yielded similar findings where dominance is correlated with access to preferred habitat. Levenbach & Hazlett (1996) found that substrate type was correlated with relative health of crayfish, O. virilis , with injured, diseased, or molting crayfish being found in areas with little shelter. Statzner et al. (2000) found that subordinate crayfish of Orconectes limosus were displaced from rocky, higher flow areas to sediment dominated areas in an artificial stream. Intraspecific competition for resources is predicted to operate under the same mechanisms, but should be more intense than interspecific competition (Wilson 1975).
Although crayfish dominance, both between and within species, has been correlated with preferred habitat, the consequences of agonistic interactions on individual resource access and use have not been quantified. Agonistic interactions have rarely been quantified in nature
(Bergman & Moore 2003) and it is unclear whether the degree of agonism and hierarchy formation seen in the laboratory exists in natural populations. Thus, the degree to which dominance directly impacts resource allocation remains undetermined. Exclusion of subordinates from shelter resources may create patterns of distribution based on status or hierarchy rank.
Shelter or burrow use has been found to increase during agonistic competition (Gherardi & Cioni
2004; Herberholz et al. 2003), but it remains unclear how individual resource use affects group dynamics and what specific factors are motivating the increased use. Are resources accessed by winning individual agonistic interactions, or does dominance status dictate how resources are allocated among crayfish in groups? The general objective of my dissertation is to elucidate the relationship between dominance and resource allocation in crayfish. I will specifically examine the following questions: 1) does dominance status affect resource use in different social contexts
2) is access to shelter resources and individual spacing correlated with dominance in nature, 3) to what extent do dominant individuals control the access to resource rich habitat, and 4) how does 5 resource distribution and ownership alter dominance relationships. By examining the mechanism of how dominance may allocate resources in groups of crayfish, we can begin to answer questions concerning what environmental selective pressures are shaping social behavior in this system. 6
CHAPTER II
CONSEQUENCES OF SOCIAL DOMINANCE ON CRAYFSIH RESOURCE USE 1
Introduction
Social dominance, defined by Drews (1993), is characterized by the consistent victory of one individual over another in repeated agonistic interactions. Agonistic interactions are aggressive contests between conspecifics and are associated with resource acquisition and defense (King 1973). Thus, an expected consequence of winning is to possess or control limited resources and thereby increase fitness. Increased resource control due to dominance has been documented in many taxa (Wilson 1975); however, dominance does not always confer higher fitness. Dominant female lizards, Eulampris heatwolei , achieve no increase in fitness as a result of their higher social status (Langkilde et al. 2005). Dominant male jackdaws, Corvus monedula , have decreased reproductive success relative to subordinate males (Verhulst & Salomons 2004).
Studies such as these suggest that the relationships among dominance, resources, and fitness are not as straightforward as previously thought (Wilson 1975).
Resources may be acquired by dominance status established in agonistic contests, or resources may be allocated with respect to relative dominance rank within a hierarchy (Drews
1993). Individuals may acquire rank based on predetermined qualities associated with dominance, such as physical attributes that contribute to fighting ability (resource holding potential (RHP), Parker 1974), or ranking may result from “social dynamics,” such as winner- loser-effects or bystander effects (Dugatkin 1997; Chase et al. 2002). Asymmetries in RHP and resource value have conventionally been examined with respect to their affects on agonistic behavior (Enquist & Leimar 1990; Stocker & Huber 2001). In addition, social dynamics are
1 Published as: Fero, K., Simon, J.L., Jourdie, V., Moore, P.A. 2007. Consequences of social dominance on crayfish resource use. Behaviour 144: 61-82.
7 dependent upon the context in which they take place. Factors such as group composition, sex ratio, and density can alter both the structure of agonism and the fitness benefits of status (Emlen
& Oring 1977; Clutton-Brock et al. 1979; Usio et al. 2001). Yet, the impact of social context on the ecological benefits associated with dominance is rarely tested (Eggleston & Lipcius 1992).
This is particularly true for the well-studied aggressive systems of crayfish.
Decapod crustaceans, such as crayfish, serve as an excellent model for examining agonistic interactions (Dingle 1983). Crayfish readily establish dominance during dyadic interactions and in laboratory settings form linear hierarchies that exhibit reduced aggression
(Guiasu & Dunham 1997). Crayfish exhibit stereotyped aggression with distinct phases of intensity where opponents engage each other even in the absence of resources (Bovbjerg 1956;
Bruski & Dunham 1987). Because of these characteristics, crayfish aggression lends itself to experimental manipulation and as a result, the neural substrates of aggression (Panksepp et al.
2003), the communication of aggressive state (Zulandt-Schneider et al. 2001), and the factors influencing dominance establishment have all been thoroughly described (Rutherford et al. 1995;
Pavey & Fielder 1996; Daws et al. 2002; Bergman et al. 2003). The outcome and structure of crayfish agonistic interactions are also shaped by the presence of resources, such as shelters, food, and mates (Peeke et al. 1995; Stocker & Huber 2001; Figler et al. 2005) these resources greatly influence crayfish survival. Although studies have explored proximate mechanisms of dominance in crayfish, the behavioral and fitness consequences of agonism and the function of aggression and dominance remain elusive topics. One measure of fitness consequences is resource acquisition and use. By examining how dominance affects individual resource use, we can gain a firmer grasp on answers to the ultimate mechanisms that give rise to social hierarchies in crayfish. 8
Many species that exhibit dominance systems show priority of access to a limited food supply (Wilson 1975). Crayfish are omnivorous detritivores thus food is not likely to be a limited resource. However, foraging behavior may be affected by such factors as the presence of predators and alarm cues (Pecor & Hazlett 2003). Motivation to forage may also shift over the course of the year, such as during spring breeding seasons when male crayfish allocate more time toward agonism and copulation with females. When females extrude their eggs later in the spring and subsequently retreat to shelters, males spend more time foraging (Berrill & Arsenault 1982).
Food may be limiting in habitats where food is clumped or patchily distributed. Bergman &
Moore (2003) found that agonistic encounters occurred more frequently on detrital patches, indicating these patches were valued resources.
Mates are also a valuable resource for crayfish. Studies examining crayfish mating behavior have not found conclusive evidence of male or female mate choice, possibly due to male-biased sex ratios (Berrill & Arsenault 1982). Microsatellite data have shown that
Orconectes placidus have multi-male broods with skewed genetic contributions from brood fathers (Walker et al. 2002). As a resource, females may not be economically defendable. Male crayfish attempt to interrupt copulating pairs (Berrill & Arsenault 1982) but the success rates of such behavior are undocumented. The extent to which dominant crayfish may monopolize mating opportunities remains unknown.
Considerably more literature has been devoted to examining shelter use in association with agonism. Studies tend to focus on mechanisms of competitive exclusion and species displacement. Several crayfish species commonly evict heterospecifics from shelters in agonistic interactions (Bovbjerg 1970; Gherardi & Daniels 2004; Klocker & Strayer 2004). Crayfish that successfully acquire or defend shelters do not necessarily use these shelters (Gherardi & Daniels 9
2004). This behavior is indicative of “spiteful behavior” in which individuals may improve their fitness by excluding others from resources (Davies 1978). Species displacements across habitats have also been attributed to interspecific differences in chelae size (Garvey & Stein 1993).
Additionally, species distributions with respect to habitat type may change depending on risk of predation where more aggressive species monopolize shelter-rich habitats (Bovbjerg 1970; Hill
& Lodge 1994). Intraspecific studies have yielded similar findings where dominance is correlated with access to preferred shelter-rich habitat (Levenbach & Hazlett 1996; Statzner et al.
2000). The mechanisms by which these distributions occur are attributed to exclusion by dominant ranking crayfish; however, there is evidence to suggest that behaviors associated with shelter use may change depending on dominance status (Herberholz et al. 2003). Thus, persisting behavioral effects following agonism may be another factor determining shelter use and crayfish distributions. Differential shelter use with respect to dominance status/rank in crayfish has not been fully explored.
The purpose of this study is to elucidate how resource use is impacted by social status and how this influence changes with varying social context. We examined dominance affects on resource use with 3 experiments. Experiment 1 tests the effects of dominance rank on shelter occupancy. Since shelter eviction is a commonly documented phenomenon in crayfish (Bovbjerg
1956; Gherardi & Daniels 2004; Figler et al 2005), we predict that higher ranking individuals will exhibit increased shelter use and exclude lower ranking individuals. Experiment 2 tests the effects of social history on shelter occupancy and feeding. If crayfish of differing social status are presented with resources (food and shelter) in the absence of conspecifics, any differences in resource use must be dependent upon social history rather than direct resource competition.
Finally, experiment 3 tests the effects of dominance rank of both males and females on shelter 10 occupancy and mating behavior. We predict that high dominance rank will correlate with increased mating opportunities.
Materials and Methods
Animals
Crayfish, Orconectes rusticus , were collected from the Portage River, Wood County,
Ohio, in the fall of 2001 and 2004. Individual form I (reproductive) males and reproductive females were placed in 10.0 × 10.0 cm ventilated plastic containers in a flow-through holding tank (154.0 × 48.0 × 31.0 cm). Crayfish were fed commercial rabbit food pellets three times a week and maintained on a 12h:12h light-dark cycle. All crayfish were isolated for a minimum of seven days to remove any effects of prior social interaction (Zulandt-Schneider et al. 2001).
Crayfish were used only once during these experiments.
Experiment 1: Shelter use in the presence of male conspecifics
Test arena
All trials were run in opaque fiberglass tanks (78.0 × 78.0 × 35.0 cm) filled with gravel to simulate the natural substrate in crayfish habitats. The tank was filled with de-chlorinated water to a depth of 25.0 cm and was oxygenated with four air stones placed in the corners of the tank.
The tank was placed under a wooden frame (180.0 × 123.0 × 82.0 cm) from which a
Remington security camera (model #00807) was mounted to record behavioral interactions.
Video was recorded on a time-lapse VCR (Samsung SSC-960) at a rate of 1 image per 3 s. Four lamps were clipped to the frame to illuminate the tanks: white lights were used during diurnal periods and red lights were used during nocturnal periods. Five shelters consisting of a variety of
PVC pipe, Plexiglas, and clay flower pots were placed in the tank to simulate natural shelter variability. 11
Experimental protocol
To test the effects of dominance rank on resource use, agonistic interactions and shelter occupancy were monitored in populations of male crayfish (mean ± SE; 3.5 ± 0.3 cm carapace length). Previous studies have shown that fight outcomes are predictable by relative size of opponents (Pavey & Fielder 1996). Therefore, crayfish within a given population were size- matched with no greater than a 10% difference in carapace and chelae length. Reflective tape was placed in different locations on the body in order to distinguish individual crayfish from one another for subsequent analysis. Five male crayfish were placed in the arena at 1900 (the beginning of the nocturnal period) and video taped continuously for 96 hrs. Ten replicates were performed (50 male crayfish).
Data Analysis
Video tapes were analyzed for frequency and intensity of crayfish agonistic interactions and for time spent occupying shelter. An interaction begins when the distance between two crayfish is equal to or less than one body length. Conversely, an interaction ends when the distance between two crayfish exceeds one body length and no interaction occurs for 10 s.
Intensity of interactions and winner/loser identification was determined using an ethogram adapted from Bruski & Dunham (1987). Comparisons between days for number and intensity of interactions were made using a one-way MANOVA with a Fisher-LSD post hoc test (Statsoft©
Statistica, ver 6).
Dominance activity index (DAI) is a measure of dominance status that equals the proportion of fights an individual wins (Bartos 1986). DAI was calculated per hour for each crayfish within a population for the first 10 hrs of a trial. Subsequent calculations were made at 12
24 hr intervals. DAI values were then used to assign hierarchy ranks with ‘1’ indicating the most dominant crayfish.
Finally, the amount of time spent occupying shelter was recorded for each crayfish.
Crayfish were considered to be in a shelter if the end of the cephalothorax was not visible past the shelter entrance. Time spent occupying shelter and hierarchy status were compared using a one-way ANOVA.
Experiment 2: Shelter use and feeding in the absence of conspecifics
Test arenas
Test arenas consisted of 37 liter (31.1 x 51.1 x 25.7 cm) aquaria, each containing gravel substrate, one shelter, and a marked feeding area where food was placed (hereafter referred to as
‘home tanks’). Black felt was taped around the sides of each home tank in order to visually isolate crayfish that were inside. PVC pipe, cut in half longitudinally (diameter = 8.7 cm, height
= 3.9 cm, length = 12.7 cm), was used as shelter with one end affixed to the tank glass, forming a single opening for crayfish to enter or exit. The feeding area consisted of a 6.4 x 6.4 cm piece of
Plexiglas with a darkened perimeter and was placed at the opposite end of the tank from the shelter, approximately 33.0 cm away. Crayfish activity was continuously recorded via a mounted security camera (Model# SG2281UQ-A), and a time-lapse video recorder as previously described. Tank illumination and light-dark cycle were the same as previously described.
Experimental Protocol
To test the effect of social status on shelter use and feeding behavior, socially conditioned male crayfish (mean ± SE; 3.8 ± 0.1 cm mean carapace length) were monitored for feeding behavior and shelter occupancy in the absence of other individuals. For each status treatment 13
(dominant, subordinate, and control), social conditioning was applied during each day and resource use behavior was recorded during each night over the course of a three day period.
Social Conditioning
Crayfish to be used in behavioral trials were repeatedly paired with other crayfish in predictable agonistic interactions. This procedure has been demonstrated to sustain social status in crayfish (Daws et al. 2002).
‘Dominant’ treatment crayfish were repeatedly paired with a smaller (between 10% and
30% difference in size) crayfish while ‘subordinate’ treatment crayfish were paired with a larger
(between 10% and 30% difference in size) crayfish. ‘Control’ crayfish never interacted with another one. Crayfish were always paired with the same individual to decrease the likelihood of dominance reversals.
Each pair was placed in a fight tank (37 liter aquaria) on opposite sides of an opaque divider and allowed to acclimate to the fight arena for 15 min. Fight tanks are separate from the home tanks described above. After the divider was removed, crayfish were allowed to interact for 15 min. and were visually monitored to confirm predicted dominance establishment and/or reinforcement. All crayfish interactions were also video taped (Canon XL1) for further confirmation. ‘Control’ treatment crayfish were placed in an empty fight tank for 15 min. to control for possible effects from handling. After the 15 min. interaction, all treatment crayfish were placed in their respective home tanks and the other crayfish were re-isolated. Treatment crayfish were removed from the experiment when pairings did not result in predicted outcomes.
As a result, six crayfish (two replicates of each treatment) were removed from the study. Social conditioning was performed three times per day (for three consecutive days), between 1000 and
1800 hrs with no less than one hour between consecutive sessions. 14
Behavioral Trials
Raw halibut fillet (2.0 g) was tied with twine onto the feeding area of each home tank.
Three size-matched crayfish, one of each status treatment (dominant, subordinate, and control) were then placed in the home tanks after one day of social conditioning. Crayfish were then video taped for an average of 16.3 ± 0.1 hrs during the night to record feeding behavior and shelter use. Start times for recording periods were variable between 1530 and 1830 hrs and all were ended at 0900 hrs. At the end of the recording period, any remaining fish was removed from the tank. This procedure was repeated for two more nights (total three nights of recording).
20 replicates of each status treatment were performed.
Data Analysis
Video tapes of shelter/feeding trials were analyzed for proportion of time crayfish spent performing specific behaviors each night. Video tape underwent blind analysis by research assistants that were unaware of status treatments. Video was analyzed for proportion of time each crayfish spent in shelter, out of shelter, and feeding for each night. A crayfish was considered ‘in shelter’ as previously described. If the tip of the rostrum was within the marked feeding area, the crayfish was considered to be ‘feeding.’ Otherwise, crayfish were considered
‘out of shelter.’ Mean proportion of time spent exhibiting these behaviors over three nights was calculated for each crayfish and then square-root arcsine transformed (Zar 1999) from proportions to degrees for statistical analysis. Comparison of time spent performing specific behaviors across social status was analyzed using a two-way ANOVA with a Fisher-LSD post- hoc test with social status and trial day as factors (Statsoft© Statistica ver 6). Size varied between 15 treatment replicates (range = 2.8 - 4.8 cm) therefore linear regressions were performed to test for effects of size on observed behavioral differences (Microcal™ Origin ver 6.0).
Experiment 3: Shelter use and reproductive opportunity
Test arena
Test arenas, video recording equipment, and lighting were identical to those described in experiment 1. Trials were run in an environmental chamber with an ambient temperature of 10ºC and a 14h:10h light-dark to simulate natural conditions during crayfish breeding season.
Experimental protocol
To test the effect of dominance on mating and mate choice, mixed-sex populations of crayfish (mean ± SE; males: 3.9 ± 0.1 cm carapace length; females: 3.4 ± 0.1 cm carapace length) were monitored with respect to agonistic interactions, shelter use, and mating behavior.
Three size-matched males and eight randomly chosen females were placed in a tank with 12
PVC shelters and video taped continuously for 96 hrs. Crayfish were marked with correction fluid to distinguish between individuals. Ten replications were performed (total 30 males and 80 females).
Data Analysis
Videos were analyzed for three classes of behaviors including male-male agonistic interactions, shelter occupancy, and mating behavior. Analysis of agonistic interactions was performed as previously described. Shelter occupancy was only examined for males. Crayfish were considered ‘in shelter’ when chelae markings were not visible. Mating behavior between males and females, as described in Mason (1970), was recorded and timed for duration. Mating events began upon seizure of the female by the male and ended upon release of the female
(Mason 1970). Although mating was quantified, sperm transfer could not be determined and 16 therefore individual reproductive success cannot be inferred. Because crayfish were not visible when they were occupying shelters, only mating events that occurred outside of shelters were recorded.
Final DAI values and rank assignments were calculated for each male crayfish in a population (refer to experiment 1). Linear regressions were used to compare male DAI, male size
(carapace and chelae length), and female-male size ratio to number of mating events, mean time per mating event, total time spent mating, percent time spent mating, time spent in shelter, and number of mates per hour . All regression analyses were performed using Statsoft© Statistica ver
6.
Results
Experiment 1: Shelter use in the presence of male conspecifics
Hierarchy formation
As trial days progressed, overall aggressive behavior in populations decreased. Nightly agonistic interactions decreased significantly each day until day 3 of the trial period (ANOVA,
Fisher LSD, F (2, 10) = 29.51, p < 0.01; N = 10). Average hourly interactions per day were as follows: (mean ± SE) 23.0 ± 0.7 (day 1), 11.3 ± 0.4 (day 2), and 7.4 ± 0.4 (day 3). The intensity of interactions also decreased significantly each day until day 3 (p < 0.05), (Figure 1). Agonistic interactions were both higher in frequency and in intensity during nocturnal periods (ANOVA,
Fisher LSD, F (2, 18) = 83.50, p< 0.01) (Figure 1).
DAI values indicate the rapid establishment of rank 1 dominant crayfish within the first two hours of being in population (Figure 2). Dominant crayfish maintained rank throughout the trial period with little fluctuation. In contrast, lower ranking crayfish displayed relatively high fluctuations in individual rank until day 3, at which point resulting hierarchies tended to stabilize 17
(Figure 2). Rank 1 dominant crayfish also participated in significantly more agonistic interactions than all lower ranking crayfish (ANOVA, Fisher LSD, F (9, 36) = 35.08, p < 0.001;
Figure 3).
Shelter use
Only top and bottom ranked crayfish differed significantly in shelter use in comparison with other crayfish. Rank 1 dominant crayfish spent significantly less time in shelter, 11.0 ± 1.8 hrs, than lower ranking crayfish (ANOVA, Fisher LSD, F (9, 36) = 21.91, p < 0.05; Figure 4).
Intermediate ranked crayfish spent a mean 14.3 ± 1.1 hrs to 15.2 ± 0.8 hrs in shelter and did not differ significantly from each other (ANOVA, Fisher LSD, F (9, 36) = 2.35, p = 0.83). Rank 5 subordinate crayfish spent significantly more time in shelter, 18.0 ± 0.4 hrs, than all higher ranking crayfish (ANOVA, Fisher LSD, F (9, 36) = 18.56, p < 0.05; Figure 4).
Experiment 2: Shelter use and feeding in the absence of conspecifics
Crayfish across treatments exhibited significant differences in the amount of time spent in shelter, out of shelter, and feeding within home tanks (ANOVA, F (2, 38) = 225.0, p < 0.005; N =
20). Crayfish generally spent more than 50% of the time out of shelter and under 5% at the feeding area (Figure 5). Behavioral differences were observed between status treatments over three nights of observation; however, dominant shelter use appears to contradict behavior observed in experiment 1. Dominant crayfish spent more time in shelter (47.7%) than either control (28.5%) or subordinate crayfish (25.8%), (ANOVA, Fisher LSD; p < 0.005), (Figure 5).
Control and subordinate crayfish did not differ. Comparing treatments for proportion of time spent out of shelter showed dominant crayfish spent less time out (57.4%) than subordinate crayfish (72.9%), (p < 0.05), while control crayfish did not differ significantly compared with either treatment (70.0%), (p = 0.13 compared with dominant; p = 0.64 compared with 18 subordinate). Proportion of time at the feeding area did not differ across status treatments (p >
0.05). Neither trial day (ANOVA, Fisher LSD, F (8, 92) = 1.251, p = 0.22) or crayfish size (out of shelter: p = 0.38, r 2 = 0.0121; in shelter: p = 0.77, r 2 = 0.0013; no regression was run for food) was correlated with the behavior of status treatments.
Experiment 3: Shelter use and reproductive opportunity
Male shelter use
Males across trials spent an average of 23.19 ± 3.66 hr occupying shelter over the 96 hr period ( N = 30). Overall, time in shelter was not correlated with male DAI (p = 0.7547, r 2 =
0.0035; Table 1) but was positively correlated with male size (carapace: p < 0.005, r 2 = 0.2924; chelae: p < 0.005, r 2 = 0.2790; Table 1). DAI was not correlated with the small variations in size between males within the same trial (carapace: p = 0.3405, chelae: p = 0.2619) demonstrating the efficiency of the size-matching.
Male reproductive opportunities
Male DAI was not correlated with the number of mating events that a male participated in (4.57 ± 0.44), the total amount of time a male spent mating (5.35 ± 0.66 hr), the percent time spent mating (0.08 ± 0.01 %), the mean time spent per mating event (1.22 ± 0.21 hr), and mating frequency (matings per hour) (0.07 ± 0.01), ( N = 30). Results are summarized in Table 1. Male size, across trials, was negatively correlated with the percent time that males spent mating
(carapace: p < 0.05, r 2 = 0.1704; chelae: p < 0.05, r 2 = 0.1645; Table 1). Male size was not correlated with other observed mating behaviors (results summarized in Table 1).
Female reproductive opportunities
Females across trials participated in an average 1.71 ± 0.19 mating events, each lasting
1.22 ± 0.21 hr, and spent a total 2.00 ± 0.31 hr mating ( N = 80). Female size (carapace length) 19 was not correlated with time spent mating (p = 0.0547, r 2 = 0.0465), number of mating events (p
= 0.2765, r 2 = 0.0152), or mean time spent per mating event (p = 0.1943, r 2 = 0.0215). However, female-male size ratio turned out to be the most predictive of whether mating would occur. The closer in size a female was to the size-matched males in her trial, the more mating events she participated in (p < 0.01, r 2 = 0.0885; Table 1) and the more total time she spent mating (p <
0.01, r 2 = 0.0837; Table 1).
Discussion
Overall, our study demonstrates that dominance in crayfish correlates with shelter use but not with feeding or mating. This is a surprising result given current views on the role of dominance and aggression in many animal systems (Wilson 1975). Both feeding and mating are behaviors that are heavily associated with the health and fitness of individuals, yet we find no indication that dominance confers an advantage with respect to these behaviors. Likelihood of mating in crayfish may instead be more dependent upon biomechanical factors involved in mating. Shelter use, however, is affected by dominance and in addition, we find variable behavioral responses of dominants and subordinates depending on the social context. Such responses may represent particular behavioral strategies or motivations that arise in different situations.
Feeding
When conspecifics are not present, crayfish spend the same amount of time feeding independent of dominance status (Figure 5). Thus, we find no indication that hunger state or some other motivating factor involved in feeding is altered by previous social interactions.
Differential feeding success in crayfish would therefore be more attributable to direct competition for food resources. Bergman and Moore (2003) found that under natural conditions 20 when resources are patchily distributed, agonistic interactions occur more frequently on food patches. Whether dominant crayfish obtained increased access to food patches was undetermined. In our experiment, food was abundant and there was no competition; dominance may only impact feeding behavior when food is limited and defendable.
Mating opportunities
Dominance in many animal systems confers increased reproductive opportunity (Clutton-
Brock et al. 1979) yet we found no correlation between mating and dominance in mixed sex populations of O. rusticus (Table 1). Much like food resources, dominance may only confer a mating advantage when females are limiting. In our experiment, females were abundant and were all in reproductive state. Although dominant males did not acquire increased access to females, we cannot conclude that dominance does not correlate with differential reproductive success as sperm deposition and fertilization were unobservable.
We did observe a correlation between mating and crayfish size (both carapace and chelae length), independent of dominance. Overall, larger males spent a smaller proportion of time mating and the most mating occurred between males and females of similar size (Table 1). A possible interpretation is that considerable size differences between males and females impose constraints on successful handling of females. Larger chelae are thought to confer an advantage to males in controlling females during mounting (Stein 1976) and larger females are often more fecund (Rahman et al. 2004). Relative fecundity may then produce a preference in males to approach females that are as large as possible that the male can successfully handle. The trend we observed, in which smaller males spent a higher percent of the time mating (Table 1), may have resulted from smaller average female size (refer to exp. 3; experimental protocol). It would 21 be necessary to determine if males had less success seizing or mounting females that differed considerably in size.
Shelter
As an ecological resource for crayfish, shelter plays an important role in providing protection from predators and conspecifics during different phases of their life cycle (Gherardi
2002). As predicted, shelter occupancy was found to correlate with dominance but only in experiments one and two. No effects of dominance were observed in experiment three. An additional finding is that status specific shelter use appears to vary depending on the social context presented in each experiment. In male populations, dominant crayfish spent less time in shelter (Figure 4), whereas when dominant males were alone, shelter use increased relative to other statuses (Figure 5). When females are present in a population, no differential shelter use exists between ranks (Table 1). By altering group composition and presence of direct versus indirect social interactions, we altered the social context under which shelters were used.
Behavioral choices regarding shelter use were subsequently altered. We hypothesize that this variable affect of dominance on shelter use is due to changes in the underlying behavioral motivations that arise in different social contexts.
Motivation to use shelter may be influenced by causal factors such as sensory information concerning shelter quality, perception of a threat (predator), reproductive state, degree of shelter competition, etc. (Eggleston & Lipcius 1992; Alberstadt et al. 1995).
Interactions between these factors determine how shelter use behavior manifests itself. In these three experiments, reproductive state and resource quality (e.g. shelter type, resource abundance, etc.) were the same for all crayfish, whereas individual social status and the social interactions crayfish were exposed to (e.g. in population versus alone) varied. Underlying causal factors 22 motivating shelter use appeared to change due to the interaction between dominance status and the social context of the experiment. These behavioral differences across status/rank may be attributed to changes in crayfish physiological state that are apparent at the onset of dominance establishment (neurology: Edwards & Kravitz 1997; excretion: Zulandt-Schneider et al. 2001; metabolism: Schapker et al. 2002), or they may be due to the use of different strategies in different situations.
The males in experiment one appear to invest more time towards status reinforcement and may consequently exhibit reduced time in shelter (Figures 3). Males that are alone have no need to reinforce status thus motivating factors to use shelter may have been altered. Since conspecifics were absent and differential shelter use is still observed, these differences must be due to causal factors that were affected by previous social interactions. Dominant crayfish may have increased relative shelter use when conspecifics were absent as a result of lack of competition for shelter or decreased need to reinforce status. Other evidence of shifting motivation to possess shelter has been found in female crayfish where ovigerous females will defend shelter more intensely than females without young (Peeke et al. 1995). Males of all ranks were similarly motivated to use shelter when females were present. While shelter is necessary for mate acquisition and defense in some decapod crustaceans (Rahman et al. 2004), this is not the case for crayfish species. Females acquire shelters on their own when extruding eggs and have been found to be able to defend shelter from male and female intruders (Peeke et al. 1995).
Shelter possession by males does not appear to apply an advantage towards acquiring mating opportunities. Worth noting is a correlation between body size and shelter occupancy when data were examined across trials in experiment three. Larger males may have spent more time in 23 shelter due to different preferences for shelter size. Explanations for the observed size affect on shelter use remain speculative.
Hierarchy formation
The hierarchy formation observed in experiments 1 and 3 was congruent with typical formation in laboratory crayfish populations (Goessman et al. 2000), being characterized by high frequency and high intensity agonistic interactions when individuals were first introduced, followed by declining agonism as ranks were established (Figures 1 & 2). It is worth noting that the process of hierarchy formation itself may certainly lend insight into how differential resource use arises. As agonism changes during rank establishment, individual resource use may also change. Examining differential resource use at different points of time during the course of hierarchy formation should be examined in future studies.
Conclusions
Our results characterize an animal system that does not adhere to conventional predictions concerning the consequences of social dominance. We find no apparent consequences of social interactions on feeding or mating, two primary behaviors impacting fitness. In contrast, social interactions do affect shelter use, thus shelter may be closely associated with aggression and individual fitness. These findings highlight the importance of social history and social context as factors in behavioral strategies and motivation in crayfish. This should be taken into account in future studies that attempt to extrapolate fitness consequences in nature. Crayfish live in a variety of habitats, are highly mobile, and experience seasonal changes in food and mate availability
(Berrill & Arsenault 1982; Gherardi 2001; Light 2003). We expect to see resource use manifest itself in nature as in our experiments given similar conditions. We predict that the addition of predation pressures and low resource availability or patchiness would alter crayfish resource use 24 but our experiments indicate that crayfish with differing social histories will react to these changes in different ways. Future studies should examine under which conditions dominance may impact feeding and mating and whether differential resource use results in differential reproductive success. 25
CHAPTER III
SOCIAL SPACING OF CRAYFISH IN NATURAL HABITATS: WHAT ROLE DOES
DOMINANCE PLAY? 2
Introduction
The manner in which animals are distributed relative to conspecifics often represents a tradeoff between the costs and benefits of proximity. The benefits of closer spacing may include increased protection from predators (‘selfish herd:’ Hamilton 1971; Treisman 1975) and increased foraging efficiency (Beauchamp 1998). Conversely, these benefits may be offset by the costs incurred through increased competition for resources (Amano et al. 2006) and the risks of conspecific aggression (Alexander 1974). These costs and benefits arise from and are modulated by various environmental and behavioral factors. Specifically, environmental factors such as resource availability, dictate habitat choice and where animals will aggregate. Resource availability can vary temporally (Weir and Grant 2004), spatially (Grant 1993), with competitor- to-resource ratio (Noel et al. 2005), or change in response to population density (Mares et al.
1982). Behavioral factors, such as species specific individual distances (Hediger 1955), social cues (Webster & Hart 2006), conspecific aggression including distances imposed by territorial behavior (Maher & Lott 1995) and social dominance (Hemelrijk 2000), may further impose constraints on how animals are spaced relative to one another. The interaction of all of these factors ultimately defines the consequences that spacing has for individual fitness. Identifying the factors and interactions that have the most influence in determining social spacing in a given animal system may lend insight into the selection pressures on different suites of behavior.
2 Publushed as: Fero, K. and Moore, P.A. 2008. Social spacing of crayfish in natural habitats: what role does dominance play? Behav. Ecol. Sociobiol . 62: 1119-1125.
26
When resources are limited or economically defensible, competitive interactions play a significant role in spacing (Grant 1993). Habitat complexity and the degree to which food resources are clumped have a significant impact on the frequency and intensity of aggressive interactions (Jensen et al. 2005). Aggressive animals that group with regard to resource distribution, consequently experience higher encounter rates and thus increased levels of aggression. At intermediate levels of competitor to resource ratio, aggression is highest at that point when resource patches are the most economically defensible (Noel et al. 2005). The presence of territories may modulate aggression within these groups. Maher and Lott (1995) define a territory as “a fixed space from which an individual or group of mutually tolerant individuals, actively excludes competitors for a specific resource or resources.” Actively defended territory boundaries and the factors that affect territory size and shape, such as resource availability and population density, may create observed spatial distributions of conspecifics
(Adams 2001).
In animal systems that are characterized social dominance, dominant individuals (i.e. those that are consistently successful in agonistic contests) are predicted to control preferred territories and/or spatial positions. Dominants have been shown to inhabit preferred spatial positions within a group (Hall & Fedigan 1997; Herrera & Macdonald 1993); however, this spatial pattern may result from avoidance by other group members rather than interference competition (Hall & Fedigan 1997). When food resources are abundant, subordinate animals may need to travel further from dominant ones as a result of socially mediated foraging (Rands et al. 2006). In convict cichlids, dominance impacts territory size and the use of defended refuges
(Hamilton 2004). Additionally, different factors affecting the decision to engage in or retreat from agonistic interactions may produce unique spatial patterns with regard to social status 27
(Hemelrijk 2000). Examining spatial distributions that result from dominance interactions may reveal whether dominance functions to confer control over resources or to decrease aggression within a population (Drews 1993; Hemelrijk 2000).
Crayfish are aggressive animals that exhibit ritualized agonistic behavior and form sustained dominance relationships (Dingle 1985). Extensive research has elucidated many aspects of the neural basis of aggression, as well as the behavioral ecology of aggression in crayfish (Bergman & Moore 2003; Bovbjerg 1970; Edwards and Kravitz 1997; Edwards et al.
2003; Hazlett et al. 1992). However, the degree to which crayfish aggression and social dominance function in the acquisition and control over resources and ultimately impact individual fitness in natural contexts remains unclear. In crayfish, dominance has been shown to increase access to shelter resources in the laboratory (Gherardi & Daniels 2004; Klocker &
Strayer 2004) and resource holding potential (RHP: Parker 1974) has been correlated with preferred habitat (Garvey & Stein 1993; Statzner et al. 2000). Bergman and Moore (2003) demonstrated that the intensity of crayfish agonistic interactions was influenced by perceived resource quality. Resource quality, and the context in which resources are presented, may also impact whether dominant crayfish monopolize resources (Fero et al. 2007; Hill & Lodge 1994).
Exclusion of subordinates from shelter resources may create patterns of distribution based on status or hierarchy rank. Shelter or burrow use has been found to increase during agonistic competition (Gherardi & Cioni 2004; Herberholz et al. 2003), but it remains unclear how individual resource use affects group dynamics and what specific factors motivate the increased use. In addition to shelter, space may be a valuable resource for crayfish as ready access to multiple shelters may facilitate foraging in the presence of predators, may grant exclusive access to valuable food resources, and may attract potential mates. 28
We investigated social dominance as an underlying mechanism of social spacing of crayfish in natural habitats. We also assessed the likelihood of territoriality as a mechanism for social spacing. The spatial location of individual crayfish in situ was recorded and then dominance was quantified by reconstructing social hierarchies in the lab. Dominant crayfish were predicted to exhibit increased control over space, evidenced by larger nearest neighbor distances. Dominant crayfish, in particular, were predicted to enforce increased individual distance by territorial shelter defense, based on shelter eviction data from the literature (Gherardi
& Daniels 2004; Klocker & Strayer 2004). If crayfish distributions within sample sites are correlated with reconstructed hierarchy rank, then social history can affect individual spacing within habitats.
Methods
Study site
Site observations and population sampling were conducted in an area approximately 40 m offshore of Grapevine Point, Douglas Lake at the University of Michigan Biological Station
(UMBS), Pellston, MI (45° 33’ N, 84° 40’ W) from July to August, 2005 (site number: sample date; 1: 7/14; 2: 7/24; 3: 7/28; 4: 7/30; 5: 8/02). Douglas Lake contains three known species of crayfish: native Orconectes propinquus , native Orconectes virilis , and invasive Orconectes rusticus ; only O. propinquus was observed in this section of the lake. Water depth never exceeded 1.5 m. The lake substrate in this area was characterized by sand covered with a mat of blue-green algae and diatoms (approximately 1 cm thick), which was also interspersed with outcroppings of cobble and small boulders and patches of small macrophytes. Crayfish were observed using cobble and eroded substrate for shelter. Five sample sites were then selected based on the presence of available shelter and visual confirmation of crayfish in the area. Due to 29 the nature of the shelter resources at the sampling area, quantifying resource distribution was infeasible as discrete shelters could not be identified. Thus, only sites that possessed abundant potential shelter were selected for sampling. This was also done in order to minimize the effect of resource distribution on crayfish spacing.
Crayfish Sampling
Sampling took place during daylight hours, between 1300 and 1900 hrs, when crayfish activity is relatively low (personal observations; Martin & Moore 2008). Mean daily movements of crayfish ( Austropotamobius pallipes ) have been shown to be less than 5 m per day (Robinson et al. 2000) and movements of Orconectes virilis have been shown to be highly variable but most frequently range from 0 – 5 m between capture events (Hazlett et al. 1974). Taken together with observations of abundant shelter and crayfish (crayfish density noted further in this section) at the sample site, we predicted that a 4 x 4 m area would be of sufficient size to contain individuals that repeatedly interact. Thus, crayfish were collected from 4 x 4 m square plots in order to sample individuals that likely had pre-existing dominance relationships. After a site was selected, a fenced enclosure was used to delineate the site boundaries and avoid the loss of any individuals while collecting. Researchers, in positively buoyant dive suits and snorkels, floated above the site and placed numbered flags by shelters where crayfish were visible. All crayfish were found inhabiting shelters and were collected either by hand or by suction pump, and then immediately placed into individually labeled containers that corresponded with flag numbers. Crayfish spatial distributions within the site were quantified by measuring marked flag locations on an X,Y coordinate system using the NW corner of the site as the origin (0,0). The NW corner was also designated as a GPS reference point for each site. Following collection and spatial measurements, researchers then dug into the substrate, turned over cobble, and pumped deep 30 holes ensuring that sites were exhaustively sampled. No crayfish were missed at the five sample sites. All crayfish were immediately transported to the UMBS research facility.
Animals
Upon transport to UMBS, crayfish species ( Orconectes propinquus ) was confirmed and sex, reproductive form and bodily injuries were recorded. All crayfish possessed intact chelae, walking legs, and sensory appendages. Carapace (from the tip of the rostrum to the end of the cephalothorax) and chelae length were also measured (mean ± SE; 2.36 ± 0.04 cm carapace; 1.76
± 0.07 cm chelae). A total of 53 O. propinquus individuals were collected (10.6 ± 1.0 crayfish per site) with a mean density of 0.7 crayfish / m 2 per site. Overall sex ratio was 2:1 (males: females) but varied within the five sites: 1:6, 8:1, 3:1, 1:0, 1:1, respectively. All males were in reproductive form I and females were in non-reproductive (non-glair) form. Of the 53 collected crayfish, five were juveniles (carapace < 1.0 cm) and were subsequently excluded from this study, leaving 48 total crayfish.
Hierarchy Reconstruction
Pre-existing in situ dominance relationships were reproduced by transferring each sampled population to an artificial pond, housed at UMBS, to allow for hierarchy formation.
Each pond consisted of a 2 m diameter plastic-lined wading pool containing lake water, sand, detritus, and 11 halved terra cotta pots, 9 cm in diameter, were used for shelter. The number and distribution of shelters was kept constant across populations in order to control for potential effects of resource availability and distribution on aggressive interactions. Shelter resources were not limited in either the lab or field conditions. Because of the limitations of available space and resolving power of the video equipment used to record crayfish behavior, artificial ponds were decreased in area as compared to the 4 x 4 m sites from which crayfish were sampled. Crayfish 31 were marked with correction fluid on the back of the carapace or chelae in order to differentiate individuals and were left in the pond for three days to allow for hierarchy formation (Fero et al.
2007; Goessmann et al. 2000). All agonistic interactions were video recorded using a mounted security camera (Model# SG2281UQ-A) and a time lapse video recorder (Samsung SSC-960) set at one frame per three seconds. Red lights (25 Watt bulbs) were mounted around the perimeter and lit continuously in order to illuminate the pond at night. Ponds were exposed to natural light and temperature regimes. The light-dark cycle for housed populations did not deviate from that of sample sites being approximately 15 h: 9 h light-dark. At the end of the recording period, crayfish were returned to Grapevine Point approximately 500 m away from the sample area to prohibit recollection of any individuals as populations were sampled sequentially over the month of July.
Data Analysis
Crayfish dominance and hierarchy stability were determined through analysis of video recordings. Agonistic bouts were analyzed using a standard ethogram adapted for our lab (see
Bergman & Moore 2003). Bouts began when crayfish were within one body length of each other and concluded when a retreat was followed by no interaction for 10 s and with more than one body length between opponents (Bergman & Moore 2003). Dominance scores, ranging from zero (low dominance) to one (high dominance), were calculated based on the percentage of agonistic bouts an individual won, both per 5 hour intervals and over the total course of analysis.
Five hour interval dominance scores were used to plot hierarchy stability (Fero et al. 2007) and total dominance score was used for all other analyses (Poisbleau et al. 2006). Dominance hierarchies may be considered stable at the point where the change in individual dominance scores decreases appreciably throughout continued interactions (Fero et al. 2007). For this study, 32 we defined stability as the point when mean shifts in dominance score over time decreased to below ± 0.015 ∆ dominance score / 5 hr (Figure 6) and mean DS shifts at each time interval were compared using one-way ANOVA with a Tukey HSD post hoc. Additionally, ordinal rank (only alpha, intermediate, and omega ranks) and average total shifts in dominance score over 15 hrs for each population were compared using a one-way ANOVA with a Tukey HSD post hoc.
Intermediate rank was determined by either taking the median DS of a population or by taking the average when two median ranks were present (e.g. populations with an even number of individuals). These three ranks were examined in order to analyze the timing of hierarchy formation in the present study with previous work. Once we had established a criterion for hierarchy stability, all crayfish within a population were used in the subsequent analysis involving nearest neighbor distance. When hierarchies reached stability, video analysis of agonistic interactions was concluded. To determine whether dominant crayfish actively defend shelters and surrounding areas, the number of shelter evictions that crayfish performed was also quantified during hierarchy formation. Shelter evictions occurred when the approach of a crayfish into a shelter was immediately followed by the retreat of the shelter resident.
From the spatial data obtained at collection sites, nearest neighbor distances were calculated for each individual and subsequently used as our measure of social spacing. Nearest neighbors were identified for each individual by the shortest distance between one crayfish and the next (hence forth referred to as ‘nearest neighbor distance’). The difference in dominance scores between nearest neighbors was also calculated ( │DS Individual – DS Neighbor │). Linear regression was used to examine the relationships between dominance score, chelae length, shelter eviction, and nearest neighbor distance. Neither sample site nor crayfish sex were significant predicting factors in the spatial analysis (NND × site × DS: F2,45 = 5.51, P = 0.007; 33
site beta = 0.01, P = 0.94; DS beta = 0.44, P = 0.002), (NND × sex × DS: F2,45 = 5.96, P = 0.005; sex beta = 0.13, P = 0.39; DS beta = 0.49, P = 0.002); therefore, data from all five populations were pooled for statistical analysis ( n = 48). All statistical tests were performed using Statistica ver 6.0.
Results
Hierarchy Stabilization
Hierarchies of sampled populations stabilized when mean shifts in dominance score decreased to below ± 0.015 ∆ DS/5 hr (Figure 6). Dominance shifts at 5 hrs were significantly higher than shifts at 15 hrs (ANOVA; n = 48, F2, 137 = 5.27, P = 0.004), at which point behavioral analysis was concluded. Ordinally ranked alpha (mean DS ± SE: 0.99 ± 0.002), intermediate
(0.37 ± 0.05) and omega individuals (0.09 ± 0.03) from each of the five populations significantly differed in the sum of dominance shifts each individual experienced (ANOVA; n = 5, F2, 12 =
0.69, P = 0.01; Figure 7). Alpha individuals experienced significantly smaller magnitudes of dominance shifts, 0.09 ± 0.05, than did intermediately ranked individuals, 0.80 ± 0.21 ( P =
0.008). Omega dominance shifts (0.37 ± 0.10) did not differ from alpha or intermediate ranks ( P
= 0.34 and P = 0.11, respectively). Dominance hierarchies in this study stabilized rapidly relative to previous crayfish studies (Fero et al. 2007; Goessmann et al. 2000); therefore video was not analyzed for the entire three day recording period.
Spatial Distribution and Dominance
Nearest neighbor distances in field sites (mean ± SE; 68.7 ± 2.0 cm) were significantly greater for crayfish that acquired higher dominance scores (DS) in reconstructed hierarchies ( R2
= 0.19, F1, 4 = 11.25, P = 0.002). Thus, we found that dominance is correlated with increased space between individuals. Additionally, the difference between dominance scores of nearest 34 neighbor pairs was positively correlated with the distance between them ( R2 = 0.08, P = 0.05).
Not only is dominance correlated with increased space, but the relative dominance between two individuals is also correlated with space; the larger the difference in dominance, the greater the distance between individuals. Chelae size accurately predicted dominance ( R2 = 0.50, P <
0.0001) but size did not predict nearest neighbor distance ( R2 = 0.01, P = 0.59). Thus, the effect of dominance on nearest neighbor distance is not attributable to size alone. These results are summarized in Table 2.
Shelter Eviction
Shelter eviction behavior varied considerably across individual crayfish. The total number of evictions an individual performed ranged from 0 to 17 (2.6 ± 0.5) during the 15 hr observation period. Fifteen crayfish did not perform any evictions, while only 3 crayfish performed 10 or higher. Overall, we found that the number of evictions that crayfish performed increased with higher dominance scores ( R2 = 0.48, P < 0.0001) as well as with nearest neighbor distance ( R2 = 0.19, P = 0.002), (Table 2). As dominance increases, crayfish perform more evictions and evictions are also independently correlated with the amount of space between nearest neighbor pairs.
Discussion
This study demonstrates that crayfish social status significantly impacts the spatial distribution of individuals in nature. Crayfish that were dominant had greater nearest neighbor distances within the field sites (Table 2). In addition, as the difference in dominance score between nearest neighbors increased, the distance between neighbors also increased, further indicating a correlation between dominance and social spacing. Dominant crayfish also performed significantly more shelter evictions during hierarchy formation and eviction rate 35 predicted nearest neighbor distance (Table 2). Crayfish size did not significantly impact spatial distribution, even though size significantly predicted individual dominance score (Table 2).
Social status appears to have a significant impact on crayfish spatial distribution and shelter acquisition such that dominant crayfish may possess increased control over space and shelter.
Finally, the observed correlations between data from reconstructed hierarchies and field spatial data, suggest the possibility that stable crayfish dominance hierarchies exist in nature.
These results indicate that dominant crayfish are better able to exclude other individuals from shelter resources. In turn, less dominant crayfish may disperse away following eviction shelters in order to avoid continued aggression from dominants or to find accessible resources.
Computer modeling of such behavioral strategies has produced spatial patterns that resemble the pattern measured in this study. Hemelrijk’s (2000) ‘risk-sensitive’ and ‘obligate attack’ strategies both outline patterns of behavior where aggression in a population subsequently decreases as a result of subordinate dispersal away from dominants. The ‘obligate attack’ strategy, in particular, is one where the goal of the action is to acquire dominance rank. Strategies that aim to reduce aggression, such as the ‘ambiguity reducing’ strategy, produce spatial structures where individuals of disparate social rank are closer together than those that are of similar rank. Our results suggest that crayfish use behavioral strategies to increase dominance rank, in addition to gaining knowledge of the fighting abilities of others. Additionally, the correlation between shorter neighbor distance and smaller difference in neighbor social status may reflect that these individuals are less able to defend space as the dominance relationships are more tenuous; in other words, the probability of a dominance reversal increases when individuals are close in hierarchy rank (Pagel & Dawkins 1997). 36
In crayfish, size (carapace and chelae length) is one of many intrinsic and extrinsic factors that determines or reinforces crayfish social status (Moore & Bergman 2005; Pavey &
Fielder 1996; Ranta & Lindstrom 1992). The observed effect of dominance on nearest neighbor distance may be attributable to increased space requirements for large crayfish or size-dependent spacing, rather than size-dependent dominance. Size-dependent spatial distribution has been documented as a reflection of RHP based decisions to engage in agonistic interactions
(Hemelrijk & Kunz 2004) and of the ability of larger males to monopolize females, along with their home ranges (Haenel et al. 2003). Even though size was significantly correlated with dominance in reconstructed hierarchies, behavioral strategies such as size-dependent ‘risk avoidance’ (Hemelrijk & Kunz 2004) may be excluded as crayfish size did not correlate with social spacing.
Shelter ownership is another factor that may contribute to social status in crayfish
(Gherardi & Daniels 2004). Shelter eviction by more dominant crayfish may be indicative of territorial behavior or some other mode of competition for space which enforces nearest neighbor distance. We are unable to derive any definitive conclusions concerning crayfish territoriality in this study as resource distribution (shelter and food) was not quantified.
The fact that dominance correlated with spacing, but size did not, indicates that social dynamics (e.g. winner-loser effects and bystander effects; Chase et al. 2002; Dugatkin 2001) may have greatly contributed to the outcome of dominance interactions in the field. In contrast with many studies examining crayfish hierarchy formation (Fero et al. 2007; Goessmann et al.
2000; Issa et al. 1999), sample site populations in this study exhibited strong asymmetries in resource holding potential (RHP; Parker 1974), mainly in terms of size and sex of individuals
(refer to methods). Hierarchies, where all individuals possess similar RHP arise primarily by 37 social dynamics that are stochastic in nature and are not reproducible (Chase et al. 2002). In nature, RHP differs greatly between individuals and resulting social dynamics may build upon these RHP differences in a predictable manner. For example, early winning experiences by larger crayfish may instill strong winner and loser effects, causing hierarchies to differentiate rapidly
(Hock & Huber 2006; Hsu et al. 2006). The results from the present study suggest that crayfish,
Orconectes propinquus , maintain stable and reproducible hierarchies in nature.
Ultimately, dominance in crayfish appears to confer increased control over space as demonstrated by the correlations found between dominance, territorial behavior and social spacing. Consequently, dominant crayfish may acquire increased access to shelter and food resources. Examining crayfish social dominance has yielded many insights into the neural basis of aggression (Edwards et al. 2003). This study provides an evolutionary context for such insights by revealing potential selection pressures that shape the formation and maintenance of dominance relationships. Future studies should examine factors such as social context, space use and resource distribution in concert to further elucidate what resource advantage dominance yields for crayfish. 38
CHAPTER IV
RESOURCE DISTRIBUTION AND SOCIAL STATUS INFLUENCE CRAYFISH HABITAT
CHOICE
Introduction
Investigations of how various factors interact to produce habitat selection and segregation tend to be dominated by studies examining local distributions of different species (Ruckstuhl &
Neuhaus 2002). Far fewer studies have tackled the driving factors behind intraspecific habitat segregation (Conradt et al. 1999). In interspecific studies, observed distributions often represent trade-offs between key ecological and behavioral factors that are the most relevant for the species in question (Martin 2000; Grabowski & Isely, 2007; Macpherson 1998). Observed distributions in species that have overlapping geographical ranges and ecological niches may be determined by, for example, trade-offs between aggressive superiority and environmental tolerance (crayfish: Bovbjerg 1970, coral reef fish: Bay et al. 2001), and predation versus relative levels of intraspecific competition (marine isopods: Franke et al. 2007). But as individuals within the same population grow, reproduce and senesce, habitat segregation may also be influenced by changing resource requirements and intraspecific competition.
Within species, various abiotic and biotic factors determine the suitability of a local habitat and constrain the distribution of individuals across different habitat types such as the presence and distribution of resources, microclimate, and habitat structure (abiotic), versus habitat preference (based on intrinsic factors of the individual), predation, competition, social affinity, and dominance relationships (biotic) (Martin 2001). Identifying the relative impact and importance of such factors in shaping habitat segregation and resource partioning within populations is a necessary step in understanding mechanisms underlying social structure and 39 social behavioral evolution. In particular, testing whether biotic factors such as preference and competition are driving observed patterns of segregation behavior has largely been restricted to a few key animal systems (e.g. social segregation in ungulates; Ciuti et al. 2004).
Habitat selection may be subject to influence by biotic factors such as subjective resource value, activity budgets, behavioral responses to competition and predation. For example, the relative value of shelter fluctuates as requirements for shelter differ significantly for lobsters at different stages of molt cycle (Karnofsky et al. 1989). Hence, the motivation to use certain resources and habitats shifts depending on the context in which an animal finds itself. Resource partitioning may thus arise, not necessarily as a direct consequence of RHP (resource holding potential; Parker 1975), but as a result of differences in resource use motivation. In certain taxa, size segregation may arise as a direct product of different habitat preferences based on resource requirements; smaller individuals compete for food resources due to an increased investment in body growth, whereas larger individuals compete for mating opportunities (Frank et al 2006;
Dennenmoser & Thiel 2007). Habitat segregation may also arise as a byproduct of social interactions that produce differences in resource use (David et al. 2007; Nakano 1995; Marra et al. 2001). In taxa that compete via interference competition, fighting ability or resource holding potential may largely determine resource partitioning and habitat segregation among individuals
(Wasserberg et al. 2006). The ability of competitively superior individuals to control resource/habitat access can also largely depend upon resource dispersion or distribution.
Resource distribution and habitat structure can greatly influence resource value (e.g. defensibility) and habitat quality. Distributions of conspecifics across habitats of varying quality may be driven by resource availability, where resource rich habitats support a larger number of individuals (Fretwell & Lucas 1970), or by competitive interactions. In contexts where resource 40 monopolization occurs, individual distributions should reflect an interaction between resource dispersion (e.g. defensibility) and interference competition (Weir & Grant 2004). In this case, resource density is far more important than resource value in determining how defensible the resource is, independent of how individuals compete for resources (Flaxman & Roos 2007). The extent to which habitat selection and individual dispersion are ecologically and behaviorally motivated are crucial elements needed for understanding the selection pressures on both habitat choice and grouping behavior.
Studies focusing on taxa that exhibit dominance hierarchies based on body size show concurrent differences in resource use and activity patterns. Socially dominant individuals can not only restrict access to valued resource patches, but can exclude lower ranking individuals from habitats and even influence subordinate diel activity in much the same manner as predation
(Nakano 1995; David et al. 2007; Marra et al. 2001; Bruinzeel et al. 2006; van Oort et al. 2007).
Identifying whether distributions of conspecifics indeed result from dominant control or from status specific differences in resource use (i.e. subjective value) is a crucial element in understanding the role that dominance relationships play in resource partitioning and the ultimate causation for habitat selection in a given taxa.
The aim of this study is to examine how aggressive competition and resource distribution interact to produce patterns of conspecific distribution. Both of these factors have been shown to influence crayfish distribution (Hill & Lodge 1994; Englund & Krupa 2000; Levenbach &
Hazlett 1996), but have not been examined in such a way as to test differences in resource use, habitat use, and preference as a consequence of social status. Status specific shelter use has been observed in Orconectes rusticus , which is also contingent upon social context (Fero et al. 2007;
Martin & Moore 2007). Dominance has also been shown to afford individuals with increased 41 access or control over space surrounding inhabited shelters (Fero & Moore 2008). This semi- natural study was conducted in order to quantify agonism in a controlled setting while providing a large enough scale that habitat choice may be observed. We predict that dominant O. virilis will monopolize shelter rich habitat, from which lower ranking individuals will be excluded and consequently be found more frequently in less desirable habitat.
Methods
Animals
Crayfish, Orconectes virilis , were collected from the shore of Maple Bay, Burt Lake, MI
(lat. 45˚28 ΄ N, long. 84˚40 ΄ W). Following collection, crayfish were immediately transported to the University of Michigan Biological Station Stream Research Facility, Pellston, MI.
Individuals were sexed, inspected for bodily injuries, and carapace (tip of the rostrum to the end of the cephalothorax) and chelae length were measured. Carapace and chelae lengths for males were (mean ± SE) 4.3 ± 0.0 cm and 3.6 ± 0.1 cm respectively; females measured at 4.3 ± 0.1 cm for carapace and 3.3 ± 0.1 cm for chelae lengths. Forty intact, non-reproductive males and females (Form II and non-glair respectively), were isolated in 10.0 x 10.0 cm ventilated plastic containers and placed in an outdoor flow through metal trough. Crayfish fed off of the available detritus that flowed in with pumped stream water and were kept under the natural light-dark cycle (15 h: 9 h) over the course of the experiment. All crayfish were isolated for a minimum of seven days in order to eliminate the effects of prior social history (Zulandt-Schneider et al. 2001) and to acclimate to ambient conditions. Individuals were used only once during the experiment.
Experimental Arena
The test arena consisted of a 4.0 x 4.0 x 0.4 m square pool with a sand/pebble substrate and an inflow and outflow on opposite ends. Three horizontal sections of equal size were 42 demarcated on the surface of the substrate using weighted down flagging tape. Each section contained a different abundance of crayfish shelter (halved 7.5 cm diameter, 13 cm long PVC pipe) with 5 shelters in the ‘high’ section, 2 shelters in the ‘low’ section, and with no shelters present in the ‘shelter absent’ section (FIGURE 8). The shelter-absent section was in the center of the pool for all trials. However, the positions of the high and low shelter sections, relative to the pool inflow and outflow, were randomly determined for each trial to serve as an environmental control. Water flow velocity in the pool was negligible during experimental trials
(below 0.01 m/s; Marsh-McBirney flow meter).
Four surveillance cameras (model#CVC-6993CL) were mounted on a 2.45 m x 1.44 m wooden frame erected over the pool. Each camera view spanned a 2.0 x 2.0 m quadrant of the pool in order to continuously record crayfish activity. Video was captured at one frame per three seconds on four time lapse video recorders (Samsung SSC-960 and Sylvania SY96R). In order to reduce environmental interference and protect video equipment, several tarps were draped over the frame and anchored down, completely covering the entire structure. Red lights were also mounted on the frame and remained on continuously to aid in visualizing crayfish in the pools during periods of low light.
Experimental design
For each experimental trial, three ‘large’ males, three ‘small’ males (possessed a 10-30% difference in both carapace (88.7 ± 0.9%) and chelae length (83.5 ± 3.8%) compared to ‘large’ males), and two randomly chosen females (carapace: 4.3 ± 0.1 cm; chelae: 3.3 ± 0.1 cm) were marked with reflective tape on the back of the carapace or chelae in order to differentiate between individuals and to facilitate tracking on video recordings. Trials began with all eight individuals being simultaneously released into the ‘shelter absent’ section of the pool. Crayfish 43 activity was recorded continuously over three days. At the end of the trial, final crayfish positions within the pool were recorded and all individuals were removed. The pool water was drained and then refilled for the next trial. A total of five trials was conducted (40 total crayfish).
Data Analysis
Crayfish agonistic encounters, habitat use, shelter occupation, and shelter evictions were analyzed from the first 20 hrs of video recordings from behavioral trials. Dominance was measured as the proportion of agonistic encounters that an individual won over the course of analysis (Poisbleau 2006). Agonistic encounters began when crayfish approached within one body length of another individual and ended with a retreat that was followed by no interaction for at least 10 s with at least two body lengths separating combatants (for ethogram see Bergman
& Moore 2003). Resulting dominance scores for each individual ranged from zero to one with one being the most dominant. Alpha (highest dominance score in a given trial) and omega
(lowest dominance score) crayfish were also identified.
Habitat use was defined by how much time an individual spent in each section of the pool
(high, low, and shelter absent). An individual was considered inside a habitat when the posterior end of the carapace passed over the flagging tape that marked the habitat boundaries. An individual was considered a shelter inhabitant if any part of the body was obscured by the shelter. Shelter evictions occurred when an inhabitant vacated in response to an individual approaching within one body length. Evictions were included in win percentage calculations.
Alpha and omega shelter use was examined by calculating the difference in the proportion of time each spent in specific shelters (e.g. alpha % time in shelter A – omega % time in shelter A).
Shelters from each trial were then ranked (1-7) by the magnitude of difference with the highest difference in shelter occupation ranked as 1. 44
Linear regressions were performed to test whether dominance was significantly correlated with agonism, and habitat and shelter use. Factorial ANOVAs (trial number × photoperiod × habitat × behavior) with Tukey HSD post hoc analyses were performed to examine differences in overall crayfish behavior depending on habitat and photoperiod (11 h day and 9 hr night). Trial number was included as a factor in all analyses in order to ensure that trial populations did not differ significantly and thus, data from the five populations were lumped together (N = 40). A one-way ANOVA (shelter rank × adjusted % time spent in shelter) was performed with a Tukey posthoc to examine differences in alpha and omega shelter use (N = 5).
All statistical tests were performed using Statistica ver 6.0.
Results
Habitat use
Overall
Crayfish spent significantly different amounts of time in each habitat section (trial x habitat x proportion time spent in habitat). The largest proportion of time was spent in the high shelter habitat, with the smallest proportion of time in the section where no shelters were present
(Tukey HSD; F (1, 105) = 61.12, P << 0.001), (note that in this comparison there was a significant interaction between trial and habitat; F (8, 105) = 3.57, P < 0.001).
Between day and night
When crayfish were not occupying shelters, they spent an equal proportion of time between habitats, regardless of whether shelters were present and regardless of photoperiod
(FIGURE 9).
Shelter use within habitats
Overall 45
Each shelter in the arena was only occupied for approximately 20% of the trial on average (19.85 ± 2.14) and high habitat shelters were occupied as often as low habitat shelters
(F (1,25) = 0.21, P = 0.65).
Between day and night
Of the total time that crayfish spent in shelter, a significantly higher proportion of occupancy occurred during the day and in the high shelter area (F (2, 468) = 41.05, P << 0.001;
FIGURE 9). Crayfish did not occupy shelters for longer periods during a given episode between day and night (F (1,78) = 1.34, P = 0.25), but entered shelters more frequently during the day (F (1,78)
= 58.59, P << 0.001).
Overall agonism
The number of agonistic encounters significantly differed across habitats (F (2,24) = 10.84,
P < 0.001, FIGURE 10). Crayfish fought equally between high and low shelter habitats, but agonism was significantly decreased in the shelter absent area (FIGURE 10). Crayfish also fought significantly more during the night (63.7 ± 4.3 encounters) than during the day (37.4 ± 2.2 encounters); (F (1,78) = 30.23, P < 0.0005).
Status-specific habitat and shelter use
Habitat use
Dominance was correlated with diurnal habitat use (results summarized in TABLE 3 &
2). Individuals with higher dominance scores both spent significantly less time (R 2 = 0.11, P =
0.03) and made fewer entries into the shelter absent area (R 2 = 0.13, P < 0.05) (TABLE 3).
Shelter use
Dominance was positively correlated with the proportion of overall time spent occupying shelter (P < 0.05). Crayfish with higher dominance also had longer average bouts of shelter 46 occupation during the day (R 2 = 0.12, P < 0.05, TABLE 3), which is when shelter use is significantly increased for all crayfish. Additionally, individuals with higher dominance made more entries into shelters than those with lower dominance, but only at night (R 2 = 0.1, P < 0.05,
TABLE 4).
Status specific agonism
Between day and night
Significant correlations between dominance and agonistic behavior were dependent upon photoperiod. During the day, when crayfish fight less overall, the number of agonistic encounters that a given individual engaged in increased for those with higher dominance scores (R 2 = 0.17,
P < 0.01, TABLE 3); whereas no significant dominance trend was observed during the night (R 2
= 0.08, P = 0.07, TABLE 4). Shelter evictions, which occurred almost entirely during the day
(F (1,70) = 17.73, P << 0.001), were also performed significantly more by individuals with higher dominance scores (R 2 = 0.69, P << 0.001, TABLE 3).
Across habitats
Dominance was not correlated with how much an individual fought in any of the three habitats. Shelter evictions did not differ per shelter between high and low shelter habitats (F (1,8) =
0.01, P = 0.92). High area shelters, which are more abundant and spaced closer together, do not appear to be more or less defendable than low area shelters.
Alpha and omega shelter use
When the shelter use of the top (alpha) and bottom (omega) dominance ranks are compared, interesting patterns emerge with regard to shelter preference. Across the five populations, alpha and omega crayfish exhibited a significant difference in the occupation of only one of the seven available shelters. Alphas demonstrated increased preference for a single 47
shelter (F (6, 28) = 9.23, P << 0.001), whereas omega occupation was distributed across the other shelters without a distinct preference (FIGURE 11). We use ‘preference’ here as a term describing the proportion of time spent occupying a shelter that is increased when common occupation by all individuals (e.g. alpha and omega) is subtracted out.
Discussion
Results summary
The presence of shelter was associated with a significant increase in crayfish habitat use.
Overall agonism also increased in habitats where shelter was present. Social segregation by dominance score was not observed across habitats when crayfish were not using shelters; however, lower dominance was correlated with decreased shelter occupation and defense. These status-specific differences in behavior were only observed during the day when shelter-related behaviors were increased overall. Dominant (alpha) individuals also exhibited distinct shelter preference providing evidence for resource partitioning, that is in part mediated by social status.
Taken together, these results suggest that perceived shelter value is driving crayfish habitat segregation, but that value is contingent upon environmental context. It is unclear whether the observed habitat segregation and resource partioning is primarily due to more effective resource holding by dominant individuals or by differences in shelter use that arise as a consequence of social history.
Differences in crayfish habitat use are correlated with shelter availability
Habitat use in this study was strongly influenced by the presence of shelter. Crayfish spent more time in areas that had more shelter (FIGURE 9). This effect is likely attributable to increased shelter availability as crayfish did not differ in the time spent out of shelter in high and low habitats (FIGURE 9). As both shelter occupation and eviction were largely diurnal 48 phenomena (TABLE 3), a concurrent lack of preference for high or low habitat was observed at night (TABLE 4). When individuals were in the high shelter area, a larger proportion of time was spent occupying shelter than when individuals were in the low area (FIGURE 9), most likely due to differences in resource availability between the two areas. Interestingly, crayfish did not stay in shelters for longer periods during the day but instead entered shelters more frequently. The observed increase in diurnal shelter occupation is thus a result of more frequent visits to shelter, not longer episodes of occupation. These results are consistent with crayfish diel activity described in previous research (Hill & Lodge 1994; Martin & Moore 2007) and may be reflective of diurnal predation pressure in natural habitats.
Agonism is correlated with the presence of shelter
Shelter-related agonism in the context of shelter eviction is a well documented phenomenon in crayfish literature (Klocker & Strayer 2004; Gherardi & Daniels 2004, Bovbjerg
1970). But this study provides evidence for a broader association between shelter resources and crayfish agonism in general by the observed increase in agonistic encounters with the presence of shelter (FIGURE 10). Competition for resources is expected to increase at low resource density or availability (competitor to resource ratio; Grant et al. 2000; Noel et al. 2005). In this study however, the scale of the overall habitat may not have been large enough for increased local shelter competition to occur as individuals may simply move on to wherever shelter was available (Fretwell & Lucas 1970). Similar occurrences of agonism between high and low habitats may also be attributed to similar crayfish densities in the open at any given time since the time spent out of shelter was the same for both habitats. Concurrently, crayfish spent the same amount of time in the open in the shelter absent habitat; thus, the decrease in agonism is not attributable to decreased relative densities of individuals. It remains unclear however, 49 whether the increased agonism observed in areas with shelter is due to competition for shelter or other related factors such as increased habitat complexity (Jensen et al. 2003) or resource density
(Grant 1993).
Evidence of dominant resource control and access to preferred habitat
If habitat use is a byproduct of shelter use in crayfish, we would expect to see habitat segregation arise due to the presence of status specific resource use (Ranta & Lindstrom 1992;
Peeke et al. 1995; Fero et al. 2007). Individuals with higher dominance scores evicted shelter more and had longer episodes of occupation. While crayfish overall entered shelters more during the day, crayfish with higher dominance scores made significantly more nightly entries (TABLE
3 & 2), thus exhibiting a higher propensity for shelter use and shelter-related agonism in general.
Dominant individuals also appear to engage in more encounters during the day than lower ranking individuals, irrespective of habitat type, whereas nightly encounter rates are not significantly predicted by social status (TABLE 3 & 2). Individuals with higher dominance are both fighting more and occupying shelter more than subordinate individuals during times when crayfish are more motivated to use shelter resources. Fero (et al. 2007) found that differences in shelter occupation among crayfish of differing dominance rank were not apparent in mixed sex populations of reproductive Orconectes rusticus . As crayfish in this study were non- reproductive, our results suggest that reproductive status may largely impact the resource use consequences that arise due to social history. In addition, individuals that possessed lower dominance scores spent more time in the shelter absent area during the day (TABLE 3 & 2) when shelter-related behavior overall is higher.
Although segregation by social status between high and low shelter habitats was not observed, the correlation between increased shelter-related behavior by dominants and increased 50 presence in ‘lower value’ habitat by subordinates, suggests that some dominance-mediated segregation is indeed occurring. Studies examining local crayfish distributions have observed similar trends with regard to physical asymmetries where unhealthy or molting individuals are more often found in shelter poor habitat (Levenbach & Hazlett 1996). Rare studies have demonstrated that dominant crayfish will actively exclude subordinates from preferred habitat;
Statzner et al. (2000) found subordinate crayfish of Orconectes limosus were displaced from rocky, higher flow areas to sediment dominated areas in an artificial stream. The differences in resource and habitat use observed in this study may arise due to interference competition by dominants, as evidenced by the observed increase in shelter-related behavior. Additionally, this resource partioning may arise as a product of other behavioral consequences of social status formation. Activity budgets of individuals may differ depending on status as activity has been shown to change as a result of dominance establishment (Gherardi & Cioni 2004; Herberholz et al. 2003; Lawton 1987). Lower ranking crayfish may also actively avoid dominant individuals, resulting in partial exclusion from preferred habitat (Hemelrijk 2000; Statzner et al. 2000).
Dominant crayfish exhibit a different pattern of shelter use
Crayfish habitat segregation may alternately arise from smaller scale segregation that is a direct result of differences in shelter and space use by dominant individuals. Alpha dominants exhibited a strong preference for a single shelter, whereas omega (subordinate) individuals shifted their use towards other available shelters with no clear preference (FIGURE 11).
Dominant individuals may be more effective at resource holding or they may be responding to some cue (visual, chemical, etc.) that is motivating shelter fidelity. Additionally, alpha and omega shelter use may represent different resource use strategies where subordinates attempt to avoid shelter eviction by dominants. Given that social status has been correlated with individual 51 spacing in natural contexts (Fero & Moore 2008), dominant individuals exhibit behaviors that indicate a relatively heightened sense of shelter ownership. Shelter has not been shown to be an important component of reproductive behavior in crayfish, as it is lobsters for example (Lawton
1987; Debuse et al. 2003), thus, it remains unclear what the potential consequences of shelter ownership may be. Even in lobsters where shelters maintain high value as breeding sites, shelter possession does not significantly impact sexual selection and affects agonism in the opposite direction from expected (Debuse et al. 2003). The value of shelter for crayfish may be highly context dependent and the benefits of ownership or access may thus be highly variable.
Implications
Animals select habitat based on a complex interplay between various environmental and biological factors. Ontogenetic differences in resource requirements and environmental tolerances are often described as the mechanisms by which conspecifics distribute themselves across habitats (Macpherson 1998; Brotons et al. 2000; Koczaja et al. 2005). Segregation across habitats due to social interactions is largely attributed to sex differences (Ciuti et al. 2004) and interference competition (Hemelrijk & Kunz 2004). Although dominance-mediated habitat segregation has been examined in some taxa such as migratory birds (Marra 2000), the relationship between dominance, resource use and habitat selection has been less well-studied.
Ultimately, this study suggests that habitat selection may be a product of status-specific differences in resource use. Subordinate individuals that exhibit a more transient pattern of shelter occupation may consequently be more vulnerable to predation and environmental disturbance. Alternatively, as size sorting in some fish has also been shown to be a strategy for avoiding larger, more dominant individuals (Hemelrijk & Kunz 2004), ‘shelter-hopping’ may also be a strategy that subordinates employ to cope with pervasive dominants. These findings 52 contribute to elucidating mechanisms of social segregation across habitats. Further studies should address how spatial scale and microhabitat distribution impact social structure and social dynamics. 53
CHAPTER V
RESOURCE DISTRIBUTION EVENS THE PLAYING FIELD FOR SOCIAL DOMINANCE
IN CRAYFISH
Introduction
Individuals that are engaged in conflict over a resource may value the resources differently (payoff asymmetry), possess differences in fighting ability (RHP), or they may differ in a way that does not impact payoff or RHP (uncorrelated asymmetry), e.g. bourgeois strategy
(Maynard Smith & Parker 1976). Studies attempting to tease apart the weight that each of these asymmetries brings to bear on contest outcomes typically focus on interactions between resource quality (i.e. value), body size, and residency status (Elwood et al. 1998; Lindstrom & Pampoulie
2004; Takahashi et al. 2001). Fewer studies have examined how behavioral asymmetries that result from the formation of dominance relationships may interact with other factors that influence fight outcome and ultimately determine resource acquisition.
Communication of dominance status may serve as an effective way to assess the RHP of competitors and avoid costly fighting (Parker 1974; Rutte et al. 2006). Winning agonistic contests can also have the reciprocal effect of increasing an individual’s RHP. Neuroendocrine changes that occur as a result of winning/losing experiences may subsequently affect factors such as aggressive state and self-assessment (Hsu et al. 2006), which greatly alter contest outcome. Other behavioral consequences of repeatedly winning fights may involve resource use strategies or preferences (Huntingford & Leaniz 1997; Dennenmoser & Thiel 2006). Thus, social experience may interact in complex ways with other factors that also affect resource use, such as prior residence. 54
Resource ownership can have a significant impact on an individual’s success in aggressive contests (Fayed et al. 2008). This owner/resident advantage may be a product of a number of different mechanisms. Maynard Smith and Parker (1976) proposed that prior residence effects may arise purely from deference to resource holders (uncorrelated asymmetry), which was supported by Davies (1978) examining territorial behavior in the speckled wood butterfly. Subsequent literature has established, however, that prior residence effects are almost always associated with other asymmetries (cf. Kokko et al. 2006). For example, as was eventually shown in the speckled wood butterfly, more aggressive individuals may be more likely to achieve resident status (Kemp & Wiklund 2004). Prior residence may also alter payoff asymmetries as more time is invested in a resource (Tobias 1997) or may increase RHP by conferring a spatial or positional advantage during a fight (Jennions & Backwell 1996; Fayed et al. 2008). Both previous social experience and prior residence can alter or determine subjective resource value (Morrel & Kokko 2003).
Resource distribution may greatly impact the relative value of a resource, in the sense that resources with the same inherent quality may differ with regard to defensibility and how costly they are to obtain. For example, when resources are clumped or patchily distributed, they are easier to defend and less vulnerable to competing individuals (Noel et al. 2005). The value of a territory may be influenced by the distance to neighbors as time spent engaged in defense or intrusion on neighboring patches is affected by territory density (Hamilton 2004). Relative value may also depend on resource availability; as resources become limited, increased aggression or motivation to fight may be expected (Robb & Grant 1998), which in turn may alter the dynamics of agonistic interactions and dominance relationships. 55
The objective of this study is to investigate how previous social experience, prior residence, and resource distribution interact to influence an individuals’ ability to control resources. Factors impacting dominance establishment in crayfish have been extensively examined making crayfish an excellent subject for this type of investigation. Asymmetries in body and claw size (Garvey & Stein 1993; Pavey & Fielder 1996), hunger state (Stocker &
Huber 2001), neuro-amine levels (e.g. serotonin and octopamine) (Panksepp et al. 2003; Yeh et al. 1997), shelter residency status (Peeke et al. 1995), and ability to perceive chemical cues from conspecifics (Zulandt-Schneider 1999), can all predict the likelihood an individual will be dominant relative to an opponent. Less is known about how factors such as resource value (e.g. quality, availability, and defensibility) impacts crayfish agonism and fight outcome.
Crayfish have been shown to exhibit status-specific differences in shelter use (Herberholz et al. 2001, Gherardi & Daniels 2004; Fero et al. 2007) and preference for shelter type (Martin &
Moore 2008). Many studies have observed segregation by apparent RHP correlated with different habitat types (body size: Englund & Krupa 2000; relative health: Levenbach & Hazlett
1996) with higher RHP individuals inhabiting deeper and more shelter rich habitat. Dominance relationships have also been correlated with spacing between neighboring shelter residents in natural habitats (Fero & Moore 2008). But the extent to which social status interacts with other factors to ultimately determine resource acquisition remains unclear. This interaction will be examined by quantifying resident/intruder agonism and shelter use in contexts with varying resident social experience and inter-shelter distance. We predict that residents with prior winning history will exert more control over shelter resources than residents with a history of losing.
Methods
Animals 56
Crayfish, Orconectes rusticus , were collected from the Portage River, Wood County,
Ohio, in the fall of 2007. Form I (reproductive) males were measured for carapace (beginning of the rostrum to the end of the cephalothorax), (mean ± SEM; 3.2 ± 0.0) and chelae length (2.9 ±
0.0), and were housed in flow-through plastic containers that visually and chemically isolated from one another. Crayfish were maintained on a diet of commercial rabbit food pellets fed three times a week and on a 12h:12h light-dark cycle. This L-D cycle was also maintained throughout the course of experimental trials. All crayfish were isolated for a minimum of seven days in order to eliminate affects of prior social history (Zulandt-Schneider et al. 2001). Only crayfish with intact appendages were used in experiments and were only used once. Prior to the beginning of the experimental protocol, crayfish were marked with reflective tape on the back of the carapace or chelae to differentiate between individuals during trials and to aid in visualization.
Experimental design
The experiment was conducted as a 3 × 3 factorial design (72 total trials, N = 8) with three social status treatments (dominant, subordinate, and control) and three shelter distance treatments (20 cm, 60 cm, and 120 cm). Each social status treatment was comprised of twenty four crayfish and body size did not differ across these groups ((treatment: carapace, chelae) control: 3.1 ± 0.1, 2.9 ± 0.1; dominant: 3.2 ± 0.1, 3.0 ± 0.1; subordinate: 3.1 ± 0.1, 2.8 ± 0.1).
Distance treatments were chosen based on nearest neighbor distances recorded in the field for
Orconectes propinquus (Fero & Moore 2008) in order to create a context of ecological relevance. Prior to experimental trials, crayfish (focal individuals) underwent pre-conditioning for social status and shelter residency over the course of two days. At the end of the second day, conditioned focal crayfish (further referred to as ‘resident’) were introduced to a size-matched 57 naive conspecific (further referred to as ‘intruder’) with an additional shelter and subsequent agonism and shelter use was observed.
Social status conditioning and residency
To produce individuals with differing social status, focal crayfish were subjected to a social status conditioning protocol in which the focal was paired with a conditioning crayfish that possessed a size (carapace and chelae) difference of 10 – 30%. Size asymmetries in crayfish are highly predictive of fight outcome (Bovbjerg 1953; Garvey & Stein 1993) and the range of difference used in this experiment allows for manipulation of outcome while ensuring that crayfish still engage each other in aggressive interactions. Studies have shown that repeated winning or losing experiences can have prolonged effects on future agonistic interactions (Fero et al. 2007; Daws et al. 2002). Focal crayfish in the dominant status treatment were paired with conditioning crayfish that were 10 – 30% smaller (focal wins) and focals in the subordinate treatment were paired with crayfish that were 10 – 30% larger (focal loses).
Focal and conditioning crayfish were placed in a 20 L aquarium with a divider between them to prevent interaction and acclimated for 20 min. The divider was then removed and the pair was allowed to interact for 20 min. During this time, agonistic behavior was recorded using an ethogram adapted from Bruski and Dunham 1987 (cf. Bergman & Moore 2005). Conditioning was performed over the course of two days with three conditioning trials per day, separated by at least 1 hr. This protocol has been shown to yield status-specific differences in behaviors related to shelter (Fero et al. 2007). Pairs in which dominance relationships were not clearly established or reinforced during each conditioning trial were not used in experimental trials. Control treatment individuals were subjected to the same protocol except they were not exposed to fighter conspecifics. 58
When focal crayfish were not engaged in a conditioning trial, they were placed in the experimental arena in order to generate a ‘prior residence effect’. Crayfish that are exposed to a shelter for 24 hrs in the absence of a conspecific are subsequently more likely to win agonistic encounters (e.g. larger in body/chelae size), (Peeke et al. 1995). Prior residence effects in crayfish appear to be sex and species dependent as male Procambarus clarkii do not exhibit this effect (Figler et al. 1999; Figler et al. 2005). Body size asymmetries in male Orconectes rusticus are more predictive of fight outcome (Martin & Moore 2007); however it has not been shown whether prior residence effects agonism among size-matched individuals.
Experimental arena
The experimental arena consists of a 1.8 m (diameter) pool containing one shelter (halved
7.5 diameter PVC pipe, 11.0 cm long) and black pumice substrate to aid in visualization of crayfish (FIGURE 12). White and red lamps were mounted around the perimeter of the pool in order to maintain crayfish on a consistent L-D cycle and red lights were left on continuously so that crayfish were visible during dark periods. A surveillance camera (model#CVC-6993CL) was mounted approximately 1.5 m above the pool in order to record crayfish behavior. Video was captured at one frame per three seconds on a time lapse video recorders (Samsung SSC-960 and
Sylvania SY96R).
Experimental protocol
Following the last conditioning trial, focal crayfish (residents) were placed back in the arena, confined in their shelter (resident shelter) by a weighted cage constructed from egg crating. A size matched naïve crayfish (intruders), (carapace: 3.1 ± 0.0; chelae: 2.9 ± 0.1), was also confined to a shelter (intruder shelter) and placed at 20 cm, 60 cm, or 120 cm away from the resident shelter (FIGURE 12). Naive intruders had no contact with conspecifics or experimenters 59 for at least seven days prior to introduction into the experimental arena. Crayfish remained confined for a 20 min acclimation period, at the end of which, experimental trials commenced by lifting the shelter cages and allowing crayfish to move freely for 24 hrs. At the end of the 24 hr trial, both crayfish were removed from the arena, which was drained, refilled, and then aerated for at least 12 hr prior to beginning the next experiment.
Data analysis
Video recordings underwent blind analysis, in which agonism and shelter use were quantified. The number of agonistic encounters was determined using the same ethogram as in social conditioning trials (previously described). An encounter began with an approach by one individual and ended with a retreat followed by a 10 s absence of interaction and spatial separation by two body lengths. The dominance status that resident crayfish acquired during experimental trials was also recorded and was described as a ‘reversal’ if the resulting ‘trial status’ was different than the resident’s ‘conditioned status.’ The proportion of trials in which reversals occurred was calculated, as well as the proportion of trials where the resident was dominant to the naive intruder.
Different measures were used to determine shelter ownership and control. Social interactions were considered ‘shelter-related’ if they occurred within two body lengths of a shelter and were not associated with a shelter eviction. Crayfish shelter occupancy began when crayfish were completely obscured by the shelter and ended when the individual moved more than one half body length away from the shelter. Evictions occurred when a shelter occupant was displaced from the shelter by the other crayfish. Crayfish approaches within one body length of a shelter without entering it were also recorded and considered a possible indicator of ownership.
Additionally, percent shelter entries were calculated (# shelter entries into shelter X / (# of 60 shelter approaches to shelter X + entries into shelter X)). Descriptions of behavioral measures are summarized in Table 5.
A multiple comparisons for proportions contingency table (Zar 1999) was used to compare the effects of conditioned status and shelter distance on the occurrence of resident status reversals and resident dominance. The proportion of control treatment residents that were dominant in experimental trials was compared across distance treatments and against an expected random value of 50% in order to determine whether a prior residence effect was present.
MANOVAs with Tukey post hoc analyses were used to examine the interaction between conditioned status, shelter distance and trial status and the extent to which these factors predicted agonism, shelter use, and shelter control. All proportions were arcsine transformed for statistical analyses (Zar 1999). MANOVAs were performed with Statistica ver 6.0.
Results
Dominance relationships
Status reversals
Dominant and subordinate-conditioned residents exhibited reversals in dominance status during experimental trials. The probability of a reversal occurring (as defined from expected dominance relationships) was dependent upon the distance between shelters (x 2 = 11.94, P <
0.05; q(0.05) ,∞,6 = 4.03), (FIGURE 14). A significantly higher proportion of reversals was observed for conditioned dominants in the 20 cm treatment versus 60 and 120 cm (60 cm: q =
5.63; 120 cm: q = 5.63; P < 0.05), (FIGURE 14). In other words, the probability of a reversal of dominance status decreased as inter-shelter distances increased. Similarly, the probability of a subordinate reversal increased as inter-shelter distances decreased (60 cm: q = 5.42, P < 0.05; 20 61 cm: q = 2.71, P > 0.05), (FIGURE 14). Thus, social conditioning accurately predicted the status of residents only in experimental trials where shelters were farther apart from each other.
Resident dominance
The likelihood of a resident being dominant to an intruder was also dependent upon the
2 distance between shelters (x = 16.39, P < 0.05; q(0.05),∞,9 = 4.39). Control and dominant- conditioned residents were dominant to intruders more often when shelters were farther apart.
Control residents were dominant to intruders in significantly more of the 120 cm trials than in the
20 cm trials (q = 5.63; P < 0.05), (FIGURE 14). Dominant-conditioned residents were dominant in a significantly higher proportion of both 60 and 120 cm trials compared to 20 cm (60 cm: q =
5.63; 120 cm: q = 5.63; P < 0.05), (FIGURE 13). Subordinate-conditioned residents exhibited different dominance outcomes than those observed in dominant and control treatments; dominance was established by the residents in significantly fewer of the 120 cm trials compared to 60 cm trials and did not differ from 20 cm trials (60 cm: q = 5.42, P < 0.05; 20 cm: q = 2.71, P
> 0.05),(FIGURE 13). As indicated by the observed pattern of status reversal, social conditioning alone did not predict resident dominance; when inter-shelter distance is not considered, residents were dominant to intruders in the same proportion of total trials across social conditioning
2 treatments (x = 2.19, P > 0.05; q(0.05),∞,3 = 3.31).
Agonism
Agonistic encounters in close proximity to shelter were significantly more likely to occur near the resident shelter than near the intruder shelter (MANOVA; F (1,126) = 14.72, P < 0.001).
The proportion of fights that were shelter-related was equal across all treatments (MANOVA
(distance x conditioned status); F (4,63) = 0.86, P = 0.49). The total number of agonistic encounters observed were also similar across conditioned status and distance treatments (MANOVA; F (4,63) 62
= 1.06, P = 0.38). Crayfish engaged in more fights around resident shelter and neither inter- shelter distance, nor resident social conditioning affected the overall occurrence of agonism.
The occurrence of shelter evictions was strongly related to resident/intruder dominance relationships. Residents that performed shelter evictions were all dominant to intruders except for two individuals (2 out of 32). These two were conditioned subordinate. No effect of inter- shelter distance was observed (MANOVA; F (2,29) = 0.23, P = 0.79). Additionally, both residents and intruders evicted both shelters evenly (MANOVA; resident: F (2,46) = 2.71, P = 0.08; intruder:
F(1,52) = 0.93, P = 0.34).
Effect of social status and inter-shelter distance on shelter use
Residents
Inter-shelter distance significantly influenced shelter approaches by resident crayfish
(MANOVA; F (2,126) = 3.29, P = 0.04). Residents approached resident shelters significantly more often than intruder shelters in 60 cm distance treatments (Tukey post hoc: P < 0.01); resident and intruder shelters were approached an equal proportion of times at 20 cm and 120 cm (Tukey post hoc; 20 cm: P = 0.21; 120 cm: P = 0.99). Status-conditioned residents approached both shelter equally (MANOVA; conditioned status: F (2,126) = 2.08, P = 0.13; trials status: F (1,132) = 2.67, P =
0.11). The proportion of shelter entries made out of total approaches (including entries) were also similar between resident and intruder shelters across status conditions and distance treatments
(MANOVA; F (4,126) = 1.02, P = 0.40); a similar result was observed for resident trial status
(MANOVA; F (1,132) = 0.76, P = 0.39).
Both inter-shelter distance and trial status significantly predicted shelter occupation.
Residents overall spent more time occupying shelter in the 120 cm treatment (MANOVA; F (2,66)
= 3.15, P < 0.05) and when residents were dominant to intruders (MANOVA; F (1,66) = 14.45, P < 63
0.001). Shelters were occupied equally with regard to status condition (MANOVA; F (2,63) = 2.45,
P = 0.09). Additionally, no significant interaction between distance and trial status was observed
(MANOVA; F (2,66) = 0.93, P = 0.39). Resident crayfish overall spent a larger proportion of time in the resident shelter (ANOVA; F (1,126) = 11.13, P < 0.001).
Intruders
Approaches to shelter by intruders in experimental trials also differed depending on inter- shelter distance (MANOVA; F (2,126) = 8.50, P << 0.001). At 20 cm, intruders approached the resident shelter significantly more often than intruder shelters (Tukey post hoc; P < 0.001). As inter-shelter distance increased resident and intruder shelters were approached evenly (120 cm: P
= 0.46). Dominant and subordinate intruders differed in the percent of entries made between resident and intruder shelters (MANOVA; F (1,132) = 9.73, P < 0.005). Dominant intruders entered shelters the same amount (Tukey post hoc, P = 0.95), but subordinate intruders entered the intruder shelter significantly more than resident shelter (Tukey post hoc, P < 0.001). Dominant and subordinate intruders also differed in the percent of shelter approaches that resulted in entries to shelter overall (MANOVA; F (1,132) = 13.24, P < 0.001). Dominant intruders entered shelter a higher percent of time when approached than subordinate intruders (Tukey post hoc; P <<
0.001). Intruders overall spent an equal proportion of time in resident versus intruder shelters
(MANOVA; F (1,132) =0.05, P = 0.82).
Discussion
Results summary
In this study, we tested how social history, resource ownership and distribution interact to alter the ability of crayfish to control resources and succeed in aggressive contests. We found that the distance between shelters exhibits the strongest influence over fight outcomes (i.e. RHP) 64 and thus, the formation of subsequent dominance relationships. The social status of resident crayfish was reversed significantly more often when shelters were close together (FIGURE 14).
As inter-shelter distance increased, social conditioning and prior residence effects became more predictive of resident/intruder dominance relationships (FIGURE 13). The occurrence of agonism was equal across all treatments but a significantly higher proportion of fights was associated with the resident, as opposed to the intruder shelter. Residents also exhibited a preference for resident shelters and intruders approached resident shelters more when shelters were closely spaced. Some subtle differences were observed in crayfish shelter use depending on social conditioning; but ultimately, dominant crayfish were better able to control shelter resources as subordinate individuals showed decreased occupancy and entries. Thus, status- specific resource use may be more a product of proximal dominance relationships (i.e. winner/loser effects) than additive effects from previous social history. Ultimately, social history, prior residence and resource distribution all interact (along with other intrinsic and extrinsic factors) to compose the terrain of a dominance landscape (sensu fitness landscape: Wright 1932) in which an individual traverses through peaks and valleys of potential social outcomes.
Dominance relationships
The likelihood that socially conditioned residents were dominant to naïve intruders was highly dependent upon shelter spacing. Control residents were dominant to intruders significantly more frequently as inter-shelter distance increased. Thus, prior residence impacts dominance establishment but the effect appears to be highly dependent on contextual factors
(e.g. resource distribution). Additionally, social conditioning appeared to negate any residency effects as subordinate conditioned residents were subordinate, not dominant, in the 120 cm treatment. Prior residence effects in Orconectes rusticus have been shown to be non-existent in 65 light of body size asymmetries that more reliably predict social status (Martin & Moore 2007). In
Pacifasticus leniusculus , similar effects are observed as size asymmetries predict shelter turnover but prior residence does not (Ranta & Lindtrom 1992). Even in taxa where prior residence is strongly predictive of fight outcome, social experience can override residency (spider crabs:
Hoefler 2002). Prior residence for O. rusticus males may only significantly impact fight outcome in the absence of other asymmetries and thus, may not be ecologically relevant.
Inter-shelter distance also mediated the extent to which social conditioning predicted dominance relationships among residents and intruders. Both dominant and subordinate- conditioned residents were less likely to maintain their social status against intruders when shelters were closer together. Crayfish with a history of losing may thus fair better with close neighbors as the odds of being dominant to an opponent increase; whereas those with a history of winning are disadvantaged. Studies examining social spacing typically find that individuals with lower RHP or lower dominance rank are in the periphery with respect to a valued resource
(Ranta & Lindstrom 1992), which may be due to avoidance behavior in response to dominant individuals (Hemerlijk 2000). In a previous study, we had found that social spacing in
Orconectes propinquuis is correlated with dominance such that high ranking individuals had more distant neighbors (Fero & Moore 2008). Such spacing may reflect dominant preferences for relatively isolated shelters as their ability to maintain dominance is improved when shelters are less dense. Possible explanations for this effect may include decreased defensibility of shelters that are in close proximity or an increase in crayfish motivation to compete for shelter. Social conditioning may also either confound or even negate prior residence effects as subordinate conditioned crayfish maintained their low status in 120 cm distance trials. While prior residence 66 may assist an individual in reaching a peak in the dominance landscape, social history exerts a much greater influence on where in the landscape an individual is to be found.
Shelter related agonism
Territories or resources that are closer together may be more vulnerable to theft or susceptible to intrusion by neighboring conspecifics (Hamilton 2004). Shelter-related agonism was affected by both resident/intruder dominance relationships and by inter-shelter distance, even though overall agonism was not. The significantly increased eviction behavior exhibited by conditioned subordinate residents (dominant to intruders) in 20 cm trials suggests that eviction is facilitated at this distance. Additionally, more fights were associated with resident shelter than with intruder shelter. Resident shelter appears to be more contested, even though there is no difference in physical quality between the two shelters and shelter resources were not limiting for the crayfish. As there were no differences observed in the occurrence of agonism, closely spaced shelters may alter the dynamics of agonistic interactions in a way that impacts outcome (e.g. escalation, intensity, or willingness to retreat). Conflicts between resident and intruders may escalate faster and begin at higher levels of intensity (Pratt & McLain 2006). Unfortunately, fight intensity and initiation could not be determined in this study due to time lapse video recording of behavior.
Intruders performed more evictions of resident shelter only when residents underwent subordinate conditioning. Interestingly, residents did not evict intruders from one shelter more than the other. Since social status treatments did not differ overall in the proportion of residents that were dominant to intruders, this observed increase in eviction of resident shelters cannot be attributed to increased eviction success. Shelter eviction may function more for dominance reinforcement than as a means of shelter defense or acquisition (Karnofsky et al. 1989). For 67 example, lobsters, in a natural context, have been shown to evict other males from shelters and then subsequently return to their home shelter; evicted individuals also return after some time to their shelter (Karnofsky et al. 1989). Resident shelters were associated with increased agonism but the equal occurrence of evictions and shelter entries between resident and intruder shelter suggests that inter-shelter distance does not impact defensibility. In lizards, Anolis sagrei , resident territorial signaling was found to be higher than newly established neighbors when they were close, but less when neighbors were farther away (McMann 2000). Eviction and other shelter-related agonism may act as mechanisms by which crayfish maintain their location in the dominance landscape.
Shelter Ownership
Crayfish may fight more around resident shelter as a result of higher relative use of resident shelter. Resident crayfish spent more time occupying the shelter with which they had the most prior experience. Intruders, which had no prior experience with shelters, spent the same amount of time in both. These differences in resident and intruder occupancy indicate that residents maintain a preference or sense of ownership over their shelter. Residents did approach the resident shelter more often, but only in 60cm trials. Intruders approached the resident shelter more when shelters were closest together. Inter-shelter distance correlated with both intruder approaches to resident shelter and with overall time residents spent in shelter. Crayfish motivation to either explore novel shelter or initiate encounters may differ depending on neighbor proximity or resource distribution. Thus, the likelihood of positive fight outcomes may be influenced by willingness to engage in agonistic interactions and to a willingness to defend valued resources. 68
Since no aspect of shelter use that was examined correlated with conditioned social status, status-specific resource use may be more a product of proximal dominance relationships
(i.e. winner/loser effects) than additive effects from previous social history. Dominant crayfish appear to be better able at controlling preferred resources. Dominant residents may exhibit more control over their shelters in the sense that dominants occupied shelter for longer and performed more evictions than subordinate residents. Dominant and subordinate residents entered both shelters the same amount and with similar frequency, as did dominant intruders. However, subordinate intruders made more entries to intruder shelter. These results indicate that subordinate intruders are less successful in accessing shelter that either dominants or residents.
Crayfish can differentiate chemical cues from individuals with differing social histories well after the occurrence of an agonistic encounter (at least 3 hrs) (Zulandt-Schneider et al.
1999). Thus social conditioning of residents was likely perceived by intruders. In lobsters, females have been shown to use chemical cues to differentiate between male shelter residents with differing social status and subsequently show entry preferences for shelters belonging to males with higher dominance (Bushmann & Atema 2000). Additionally, crayfish exhibit increased locomotion while being exposed to conspecific odor, regardless of social history
(Zulandt-Schneider et al. 1999). Differences in behavior with regard to inter-shelter distance may have arisen due to different exposure to chemical cues of conspecifics. Crayfish separated by shorter distances may have been responding to an increased presence of conspecific odor in close proximity to their own shelter. The presence of lingering chemical cues may also explain why residents preferred their own shelters, or how they differentiated between the two, and also may explain why intruders approached resident shelter more often. 69
Taken together, our results suggest that in O. rusticus , prior residence may induce shelter preference but does not afford residents a competitive advantage. Instead, we surprisingly found that the formation of dominance relationships is considerably influenced by the distance between shelter resources, which also overrides effects of previous social history. Shelters do not necessarily become less defensible at shorter distances as shelter related agonism does not differ across contexts; but apparent interest in a resident’s shelter is high for both residents and intruders. Subsequent relationships formed between crayfish do more to predict resource control than the other factors examined. This study demonstrates how the dynamic interaction between social and environmental factors significantly contributes to the social reality of crayfish in setting the odds of winning aggressive contests along a landscape of possibilities. 70
CHAPTER VI
SUMMARY AND GENERAL CONCLUSIONS
The use of crayfish as a model organism for aggression necessitates an understanding of the adaptive value of agonistic interactions in this system. The consequences of dominance establishment are assumed to be associated with increased access to resources (increased relative fitness). The equating of dominance with resource advantages is thus incorporated into explanations of behavior without much regard for whether this assumption has been formally tested. Thus, the goal of this dissertation was to determine how social dominance relates to resource access and use. The work presented here has shown that: social status has persisting behavioral consequences with regard to shelter use, which are modulated by social context
(Chapter II); dominance relationships influence the spatial distribution of crayfish in natural environments such that dominant individuals possess access to more space (Chapter III); resource use strategies differ depending on social history and these strategies may influence larger scale segregation across habitats (Chapter IV); and finally, shelter distribution modulates the extent to which social history and shelter ownership influence the formation of subsequent dominance relationships (Chapter V). Taken together, these results demonstrate that dominance has significant consequences for crayfish resource acquisition and holding. However, a complex picture has been revealed as to the nature of dominance establishment and the potential resource benefits associated with dominance. The likelihood of becoming dominant and the resource use consequences of dominance are largely dependent on social and environmental context.
In Chapter II, the occurrence of status-specific resource use was examined in different social contexts. Three experiments were conducted in order to determine how social history and social context influence shelter use. Crayfish were observed in all-male populations, mixed-sex 71 populations, and were observed alone. Crayfish established dominance status while in populations or through status conditioning and had access to variable resources (food, mates, and/or shelters) in each experiment. Subsequent resource use was quantified and compared to dominance rank. Our results did not match conventional predictions that dominance would confer increased access to resources. Top ranked dominant crayfish occupied shelter significantly less than lower ranks. This differential shelter use may be due to dominant motivation to reinforce status as dominants also participated in the most agonistic interactions.
When dominant crayfish had access to resources in the absence of conspecifics, dominant crayfish occupied shelter significantly more than subordinate and naïve crayfish. This result illustrates that present social context has a significant impact on behavioral decisions in crayfish.
Social history and social context interact to determine shelter occupancy in this case. Feeding and mating was unaffected by social status in our populations. This chapter shows that status- specific shelter use arises from intrinsic changes that occur upon status formation. This status- specific use subsequently varies due to changing behavioral motivations in different social contexts.
In Chapter III, the impact of dominance on crayfish social spacing and resource control was examined. Spatial distributions of individual crayfish, Orconectes propinquus , were recorded from five sample sites in Douglas Lake, MI. Crayfish populations from each site were collected and then immediately transferred to artificial ponds in order to reproduce potential dominance hierarchies. Spatial data from field sites were compared with dominance relationships and shelter use observed from reconstructed hierarchies. Dominant crayfish were found to have greater nearest neighbor distances than lower ranking crayfish. In addition, as the difference in dominance score between nearest neighbors increased, the distance between them also increased. 72
Although claw size was an accurate predictor of dominance, size did not correlate with nearest neighbor distance. Factors such as social dynamics may thus play a larger role in natural crayfish populations than previously thought. Dominant crayfish also performed more shelter evictions during hierarchy formation, which were correlated with nearest neighbor distance, suggesting that eviction by dominant crayfish may enforce spacing. Social status appears to significantly impact crayfish spatial distribution and shelter acquisition such that more dominant crayfish exhibit increased control over space and shelter. In addition, these findings suggest the possibility that stable crayfish dominance hierarchies exist in nature.
In Chapter IV, the influence of resource abundance and social status on habitat selection was examined. A semi-natural study was conducted in order to quantify agonism in a controlled setting while providing a large enough scale that habitat use was observable. The presence of shelter was associated with a significant increase in crayfish habitat use. Overall agonism also increased in habitats where shelter was present. Status specific differences in shelter use and habitat use were observed and were dependent upon environmental context (e.g. day versus night). Lower ranking crayfish were found to occupy and defend shelter less frequently than dominant crayfish and low ranking crayfish were also found more often in shelter-poor habitat.
These status specific differences in behavior were only observed during the day when shelter- related behaviors were increased overall. Distinct shelter use strategies were also apparent between top ranked (alpha) and bottom ranked (omega) crayfish. Taken together, these results suggest that shelter use and habitat use are, in part, mediated by social status. Additionally, these results indicate that habitat selection may arise as a consequence of status-specific differences in resource use. 73
Finally, Chapter V examined how social history, resource ownership and resource distribution interact to alter the ability of crayfish to control resources and succeed in aggressive contests. Agonism and shelter use was quantified in resident/intruder pairings in contexts with varying resident social experience and varying distances between resident and intruder shelters.
Inter-shelter distance was found to modulate the extent to which social history and shelter ownership determine RHP. The social status of resident crayfish was reversed significantly more often when shelters were close together. As inter-shelter distance increased, social conditioning and prior residence effects became more predictive of resident/intruder dominance relationships.
Both residents and intruders exhibited more interest in resident shelter than intruder shelter and increased relative agonism was also associated with resident shelter. Status specific resource use corresponded with resident/intruder dominance relationships and was not influenced by previous social history. Shelter ownership and distribution likely influence crayfish motivation to compete for a shelter, and in this way, shelter distribution may factor into RHP. Interactions between social history, resource ownership and resource distribution produced different scenarios for dominance establishment depending on the how these factors varied. Crayfish dominance establishment is thus a dynamic process involving many factors that push and pull against each other.
What is evident from this work is that crayfish employ different resource use strategies as a consequence of dominance relationships and social context. Crayfish exhibit status-specific shelter use when no other crayfish are present (Chapter 2), indicating that these differences do not arise solely as a product of interference competition for shelter. Largely, the observed trend is that crayfish with winning experience (dominants) exhibit shelter fidelity, whereas crayfish with losing experience (subordinates) exhibit a pattern of transient and shorter lived shelter 74 occupation. This occupancy pattern was directly observed in Chapter 4 for the top (alpha) and bottom (omega) dominance ranked individuals and additionally, dominant crayfish were consistently observed occupying shelter for relatively longer periods of time, overall and per entry (Chapter 2, 3, 4, and 5). Different contexts may involve different payoffs and thus, it appears status-specific shelter use is modulated by factors such as social context (Chapter 2), as well as resource distribution (Chapter 3 and 4) and prior residence (Chapter 5).
The crayfish literature is replete with studies describing intrinsic and extrinsic factors affecting fight outcome and dominance establishment (Bovbjerg 1956; Daws et al. 2002; Garvey
& Stein 1993; Moore & Bergman 2005). With the multitude of factors that influence fight outcome, it becomes difficult to determine which of these are the most relevant for crayfish in natural environments. Given that crayfish are aggressive animals that may be limited by space or shelter resources (e.g. in small streams where crayfish densities can be high), available space and shelter may play a critical role in crayfish social interaction. This dissertation shows that space and shelter are closely associated with crayfish dominance. Subordinate (low ranking) crayfish may actually benefit from close spacing to neighbors (Chapter 3) due to increased odds of reversing their status. This may not only be due to having neighbors that are of similar competitive ability (Chapter 5), but close shelter spacing, in and of itself, improves the likelihood that subordinate individuals will establish dominance in subsequent interactions (Chapter 5).
Retention of both social status and of access to more space by dominant individuals may be contingent upon being farther away from neighbors (Chapter 3 and 5). Potential consequences for crayfish distributions on a larger scale may result in habitat segregation due to preferences for habitats with differing resource distributions. Lower ranking crayfish were found more frequently in shelter-poor habitat (Chapter 4), but this result is more likely due to the 75 consequences of local scale shelter use. Subordinate individuals may not be entirely excluded from local areas as the presence of status-specific resource use strategies may offset the potential of increased agonism over limited resources. Taken together, spatial structure in crayfish populations may arise as byproducts of status-specific resource use and defense and by self- reinforcing factors that dictate the likelihood of positive fight outcomes.
Does dominance give crayfish a resource advantage? In general, dominant individuals exhibit more shelter-related behavior (e.g. occupation, eviction, and defense) than subordinate crayfish. Since shelter is not associated with reproductive behavior in crayfish (Gherardi 2002), the value and potential benefits are likely associated with risk-avoidance (e.g. predation).
Subordinates appear to employ an alternate shelter use strategy which may allow them to access shelter, even in the midst of dominant control over shelter resources, but subordinates face increased risk as they are consequently more exposed and vulnerable to the surrounding environment. Dominant control over shelter resources and over space likely occurs through a combination of active enforcement through agonistic interaction and shelter/space use preferences that are related to the likelihood of retaining dominance status. Dominants may consequently obtain uncontested access to other resources such as food. Ultimately, dominance confers a significant advantage in terms of resource holding.
It’s good to be dominant, but it may not be so bad to be subordinate. The results presented in this dissertation essentially show that dominance in the crayfish system is dynamic in the sense that changes in one factor, such as social context or prior residence, can significantly alter the probability that an individual will win a fight. Fighting is costly and dominance is often thought of as being advantageous in the sense that animals may use it as a cue of RHP and consequently reduces the occurrence of aggression in populations (Hsu et al. 2006; Maynard 76
Smith & Parker 1976). While this effect occurs in static laboratory populations (Goessman et al.
2000; Fero et al. 2007), crayfish in nature are subject to constantly fluctuating social and environmental contexts. Crayfish engage frequently and repeatedly in costly fighting and the reason may be that the resource consequences of dominance are substantial and depend more on proximate relationships than previous social history. Given that dominance establishment is highly contingent on context, it may benefit an individual that is subordinate to continue to engage in fights as the odds of winning are variable. Such selection pressure to fight may be balanced by other factors such as RHP assessment (e.g. ritualized fighting and visual and chemical cues of RHP), dominance status recognition through signaling. A complex picture arises as this work demonstrates the intricate and flexible role of dominance in modulating social and resource-related behavior. This picture fits nicely with what we know concerning the flexibility of the nervous system and its dynamic changes in response to social interaction in crayfish (Panksepp et al. 2003; Edwards et al. 2003). This work provides a much needed ecological context for the nature of crayfish dominance and aggression, thus lending significant weight to the use of this organism as a model for understanding the evolution of aggression and dominance systems in general.
77
LITERATURE CITED
Adams, E.S. 2001. Approaches to the study of territory size and shape. Annu. Rev. Ecol. Syst. 32: 277-303.
Alberstadt, P.J., Steele, C.W. and Skinner, C. 1995. Cover-seeking behavior in juvenile and adult crayfish, Orconectes rusticus : effects of darkness and thigmotactic cues. J. Crust. Biol. 15: 537-541.
Alexander, R.D. 1974. The evolution of social behaviour. Annu. Rev. Ecol. Syst. 5: 325-383.
Amano, T., Ushiyama, K., Fugita, G., Higuchi, H. 2006. Costs and benefits of flocking in foraging white-fronted geese ( Anser albifrons ): effects of resource depletion. J. Zool. 269: 111-115.
Bartos, L. 1986. Dominance and aggression in various sized groups of red deer stags. Aggressive Behavior 12: 175-182.
Bay, L., Jones, G. and McCormick, M. 2001. Habitat selection and aggression as determinants of spatial segregation among damselfish on a coral reef. Coral Reefs 20: 289-298.
Beauchamp, G. 1998. The effect of group size on mean food intake rate in birds. Biol. Rev. 73: 449-472.
Bergman, D. A., Kozlowski, C. P., McIntyre, J. C., Huber, R., Daws, A. G., & Moore, P. A. 2003. Temporal dynamics and communication of winner-effects in the crayfish, Orconectes rusticus . Behaviour , 140: 805-825.
Bergman, D.A. and Moore, P.A. 2003. Field observations of intraspecific agonistic behavior of two crayfish species, Orconectes rusticus and Orconectes virilis , in different habitats. Biol. Bull. 205: 26-35.
Berrill, M. and Arsenault, M. 1982. Spring breeding of a northern temperate crayfish, Orconectes rusticus . Can. J. of Zool. 60: 2641-2645.
Bovbjerg, R.V. 1953. Dominance order in the crayfish Orconectes virilis (Hagen). Physiol. Zool. 29: 127-136.
Bovbjerg, R.V. 1956. Some factors affecting aggressive behavior in crayfish. Physiol. Zool. 10: 485.
Bovbjerg, R.V. 1970. Ecological isolation and competitive exclusion in two crayfish ( Orconectes virilis and Orconectes immunis ). Ecology 51: 225-236. 78
Bruinzeel, L.W., de Pol, M.V. and Trierweiler, C.2006. Competitive abilities of oystercatchers (Haematopus ostralegus ) occupying territories of different quality. J. Ornith. 147: 457-463.
Bruski, C.A. and D.W. Dunham. 1987. The importance of vision in agonistic communication of the crayfish Orconectes rusticus . Behaviour 103: 83-107.
Bushmann, P.J. and Atema, J. 2000. Chemically mediated mate location and evaluation in the lobster, Homarus americanus . J. Chem. Ecol . 26: 883-899.
Case, T. J. & Gilpin, M. E. 1974. Interference competition and niche theory. Proc. Natl. Acad. Sci. USA 71: 3073-3077.
Chase, I.D., Tovey, C., Spangler-Martin, D. and Manfredonia, M. 2002. Individual differences versus social dynamics in the formation of animal dominance hierarchies. Proc. Natl. Acad. Sci. USA 99: 5744-5749.
Ciuti, S., Davini, S., Luccarini, S. and Apollinio, M. 2004. Could the predation risk hypothesis explain large-scale spatial sexual segregation in fallow deer ( Dama dama )? Behav. Ecol. Sociobiol . 56: 552-564.
Clutton-Brock, T. H. & Harvey, P. H. 1976. Evolutionary rules and primate societies. In: Growing points in ethology, P. P. G. Bateson & R. A. Hinde, eds. The University Press, Cambridge.
Clutton-Brock, T.H., Albon, S.D., Gibson, R.M., and Guinness, F.E. 1979. The logical stag: adaptive aspects of fighting in red deer ( Cervus elaphus L.). Anim. Behav. 27: 211-225.
Conradt, L., Clutton-Brock, T.H. and Thomson, T.H. 1999. Habitat segregation in ungulates: are males forced into suboptimal foraging habitats through indirect competition by females? Oecologia 119: 336-377.
Daws, A.G., Grills, J.L., Konzen, K. and Moore, P.A. 2002. Previous experiences alter the outcome of aggressive interactions between males in the crayfish, Procambarus clarkii . Mar. Freshw. Behav. Physiol. 35(3): 139-148.
79
David BO, Closs GP, Crow SK, & Hansen EA. 2007. Is diel activity determined by social rank in a drift-feeding stream fish dominance hierarchy? Anim. Behav. 74:2, 259-263.
Davies, N. B. 1978. Ecological questions about territorial behaviour. In: Behavioural ecology: an evolutionary approach (J.R. Krebs and N.B. Davies, eds.) Blackwell Scientific Publications, Sunderland. Pp. 317-350.
Debuse, V.J., Addison, J.T. and Reynolds, J.D. 2003. Effects of breeding site density in competition and sexual selection in the European lobster. Behav. Ecol . 14:396-402.
Dennenmoser S & Thiel M. 2007. Competition for food and mates by dominant and subordinate male rock shrimp, Rhyncocinetes typus . Behaviour . 144, 33-59.
Dingle, H. 1983. Strategies of agonistic behavior in Crustacea. In: Studies in adaptation: the behavior of higher crustacean (S. Rebach and D.W. Dunham, eds) John Wiley and Sons, New York. Pp. 85-111.
DiStefano, R. J., Gale, C. M., Wagner, B. A., & Zweifel, R. D. 2003. A sampling method to assess lotic crayfish communities. J. Crust. Biol . 23(3): 678-690.
Drews, C. 1993. The concept and definition of dominance in animal behaviour. Behaviour 125: 283-313.
Dugatkin, L.A. 1997. Winner and loser effects and the structure of dominance hierarchies. Behav. Ecol. 8: 583-587.
Dugatkin, L.A. 2001. Bystander effects and the structure of dominance hierarchies. Behav. Ecol. 12: 348-352.
Edwards, D.H. and Kravitz, E.A. 1997. Serotonin, social status and aggression. Curr. Opin. Neurobiol. 7: 812-819.
Edwards, D.H., Issa, F.A., Herberholz, J. 2003. The neural basis of dominance hierarchy formation in crayfish. Microsc. Res. Tech. 60: 369-376.
Eggleston, D.B. and Lipcius, R.N. 1992. Shelter selection by spiny lobster under variable predation risk, social conditions, and shelter size. Ecology 73: 992-1011.
80
Elwood, R.W., Wood, K.E., Gallagher, M.B. and Dick J.T.A. 1998. Probing motivational state during agonistic encounters in animals. Nature 393: 66-68.
Emlen, S.T. and Oring, L.W. 1977. Ecology, sexual selection, and the evolution of mating systems. Science 197: 215-223.
Englund, G., & Krupa, J. J. 2000. Habitat use by crayfish in stream pools: influence of predators, depth and body size. Freshw. Biol. 43: 75-83.
Enquist, M. and Leimar, O. 1990. The evolution of fatal fighting. Anim. Behav. 39: 1-9.
Fayed, S.A., Jennion, M.D. and Backwell, P.R.Y. 2008. What factors contribute to an ownership advantage? Biol. Lett . 4: 143-145.
Fero, K., Simon, J.L., Jourdie, V., Moore, P.A. 2007. Consequences of social dominance on crayfish resource use. Behaviour 144: 61-82.
Fero, K. and Moore, P.A. 2008. Social spacing of crayfish in natural habitats: what role does dominance play? Behav. Ecol. Sociobiol . 62: 1119-1125.
Figler, M.H., Cheverton, H.M., and Blank, G.S. 1999. Shelter competition in juvenile red swamp crayfish ( Procambarus clarkii): the influences of sex differences, relative size, and prior residence. Aquaculture 178: 63-75.
Figler, M.H., Blank, G.S. and Peeke, H.V.S. 2005. Shelter competition between resident male red swamp crayfish Procambarus clarkii (Girard) and conspecific intruders varying by sex and reproductive status. Mar. Freshw. Behav. Physiol. 38: 237-248.
Flaxman SM & deRoos CA. 2007. Different modes of resource variation provide a critical test of ideal free distribution models. Behav Ecol Sociobiol . 61, 877-886.
Franke H-D, Gutow L & Janke M. 2007. Flexible habitat selection and interactive habitat segregation in the marine congeners Idotea baltica and Idotea emarginata (Crustacea, Isopoda). Mar Biol . 150, 929-939.
Fretwell, S.D and Lucas, H.L. 1970. On territorial behavior and other factors influencing habitat distribution in birds. Acta Biotheor 19: 17-36.
Garvey, J. E. and Stein, R.A. 1993. Evaluating how chelae size influences the invasion potential of an introduced crayfish. Am. Midl. Nat. 129: 172-181.
81
Gherardi, F. 2002. Behavior. In: Biology of freshwater crayfish (D.M. Holdich, ed.) Blackwell Science Ltd., Oxford. Pp. 258-290.
Gherardi, F. and Cioni, A. 2004. Agonism and interference competition in freshwater decapods. Behaviour 141: 1297-1324.
Gherardi, F. and Daniels, W.H. 2004. Agonism and shelter competition between invasive and indigenous crayfish species. Can. J. Zool. 82: 1923-1932.
Goessmann, C., Hemelrijk, C. and Huber, R. 2000. The formation and maintenance of crayfish hierarchies: behavioral and self-structuring properties. Behav. Ecol. Sociobiol. 48: 418-428.
Grabowski, T.B. and Isely, J.J. 2007. Spatial and temporal segregation of spawning habitat by catostomids in the Savannah River, Georgia and South Carolina, USA. J. Fish Biol. 70: 782-798.
Grant, J.W.A. 1993. Whether or not to defend? The influence of resource distribution. Mar. Behav. Physiol. 23: 137-153.
Grant JWA, Gaboury CL, & Levitt HL 2000. Compeitior-to-resource ratio, a general formulation of operational sex ratio, as a predictor of competitive aggression in Japanese medaka (Pisces: Oryziidea). Behav. Ecol. 11: 670-675.
Guiasu, R.C. and Dunham, D.W. 1997. Initiation and outcome of agonistic contests in male form I Cambarus robustus (Girard, 1852) crayfish. Crustaceana 70: 480-496.
Haenel, G.J., Smith, L.C., and John-Alder, H.B. 2003. Home-range analysis in Sceloporus undulates (Eastern Fence Lizard). I. Spacing patterns and the context of territorial behavior. Copeia 1: 99-112.
Hall, C.L. and Fedigan, L.M. 1997. Spatial benefits afforded by high rank in white-faced capuchins. Anim. Behav. 53: 1069-1082.
Hamilton, I.M. 2004. Distance to neighbours influences the trade-off between hiding after disturbance and defending food patches in convict cichlids ( Archocentrus nigrofasciatus ). Behav. Ecol. Sociobiol. 56: 530-538.
Hamilton, W.D. 1971. Geometry for the selfish herd. J. Theor. Biol. 31: 295-311.
Hazlett, B.A., Rittschof, D., and Rubenstein, D. 1974. Behavioral biology of the crayfish Orconectes virilis . I. Home range. Amer. Midl. Nat. 92: 301-19.
82
Hazlett, B.A., Anderson, F.E., Esman, L.A., Stafford, C., and Munro, E. 1992. Interspecific behavioral ecology of the crayfish Orconectes rusticus . J. Freshw. Ecol. 7: 69-76.
Hediger, H. 1955. Studies of the psychology and behaviour of captive animals in zoos and circuses. Dover Publications Inc, New York.
Hemelrijk, C.K. 2000. Towards the integration of social dominance and spatial structure. Anim. Behav. 59: 1035-1048.
Hemelrijk, C.K. and Kunz, H. 2005. Density distribution and size sorting in fish schools: an individual-based model. Behav. Ecol. 16: 178-187.
Herberholz, J., Sen, M.M. and Edwards, D.H. 2003. Parallel changes in agonistic and non- agonistic behaviors during dominance hierarchy formation in crayfish. J. Comp. Physiol. A 189: 321-325.
Herrera, E.A. and Macdonald, D.W. 1993. Aggression, dominance, and mating success among capybara males ( Hydrochaeris hydrochaeris ). Behav. Ecol. 4: 114-119.
Hill, A.M. and Lodge, D.M. 1994. Diel changes in resource demand: competition and predation in species replacement among crayfishes. Ecology 75: 2118-2126.
Hock, K. and Huber, H. 2006. Modeling the acquisition of social rank in crayfish: winner and loser effects and self-structuring properties. Behaviour 143: 325-346.
Hoefler, C.D. 2002. Is contest experience a trump card? The interaction of residency status, experience, and body size on fighting success in Misumenoides formosipes (Araneae: Thomisidae). J. Ins. Behav . 15: 779-790.
Hsu, Y.Y., Earley, R.L., and Wolf, L.L. 2006. Modulation of aggressive behaviour by fighting experience: mechanisms and contest outcomes. Biol. Rev. 81: 33-74.
Issa, F.A., Adamson, D.J. and Edwards, D.H. 1999. Dominance hierarchy formation in juvenile crayfish Procambarus clarkii . J. Exp. Biol. 202: 3497-3506.
Jennions, M.D. and Backwell, P.R.Y. 1996. Residency and size affect fight duration and outcome in the fiddler crab Uca annulipes. Biol. J. Linn. Soc . 57: 293-306.
Jensen, S.P., Gray, S.J. and Hurst J.L. 2003. How does habitat structure affect activity and use of space among house mice? Anim. Behav . 66: 239-250.
Jensen, S.P., Gray, S.J. and Hurst, J.L. 2005. Excluding neighbors from territories: effects of habitat structure and resource distribution. Anim. Behav. 69: 785-795. 83
Johnson, D.D.P., Kays, R., Blackwell, P.G. and Macdonald, D.W. 2002. Does the resource dispersion hypothesis explain group living? TREE 17: 563-570.
Karnofsky, E.B., Atema, J. and Elgin, R.H. 1989. Field observations of social behavior, shelter use, and foraging in the lobster, Homarus americanus. Biol. Bull. 176: 239-246.
Kemp, D.J. and Wiklund, C. 2004. Residency effects in animal contests. Proc. Roy. Soc. Lon. Series B 271: 1707-1711.
King, J.A. 1973. The ecology of aggressive behavior. Annu. Rev. Ecol. Syst. 4: 117-138.
Klocker, C.A. and Strayer, D.L. 2004. Interactions among an invasive crayfish ( Orconectes rusticus ), a native crayfish ( Orconectes limosus ), and native bivalves (Sphaeriidae and Unionidae). Northeastern Nat. 11: 167-178.
Koczaja, C., McCall, L., Fitch, E., Gloriosos, B., Hanna, C., Kyzar, J., Niemiller, M., Spiess, J., Tolley, A., Wyckoff, R. and Mullen, D. 2005. Size-specific habitat segregation and intraspecific interactions in banded sculpin ( Cottus carolinae ). Southeast. Nat. 4:207-218.
Kokko, H., Lopez-Sepulcre, A. and Morrell, L.J. 2006. From hawks and doves to self-consistent games of territorial behavior. Amer. Nat . 167: 901-912.
Langkilde, T., Lance, V.A. and Shine, R. 2005. Ecological consequences of agonistic interactions in lizards. Ecology 86: 1650-1659.
Lawton, P. 1987. Diel activity and foraging behavior of juvenile american lobsters, Homarus americanus. Can. J. Fish. Aquat. Sci. 44: 1195-1205.
Levenbach, S. and Hazlett, B.A. 1996. Habitat displacement and the mechanical and display functions of chelae in crayfish. J. Freshw. Ecol. 11: 485-492.
Light, T. 2003. Success and failure in a lotic crayfish invasion: the roles of hydrologic variability and habitat alteration. Freshw. Biol. 48: 1886-1897.
Lindstrom, K. and Pampoulie, C. 2005. Effects of resource holding potential and resource value on tenure at nest sites in sand gobies. Behav. Ecol . 16: 70-74.
84
MacPherson, E. 1998. Ontogenetic shifts in habitat use and aggregation in juvenile sparid fishes. J. Exp. Mar. Biol. Ecol. 220: 127-150.
Maher, C.R. and Lott, D.F. 1995. Definitions of territoriality used in the study of variation in vertebrate spacing systems. Anim. Behav. 49: 1581-1597.
Mares, M.A., Lacher Jr., T.E., Willig, M.R., Bitar, N.A., Adams, R., Klinger, A. and Tazik, D. 1982. An experimental analysis of social spacing in Tamias striatus . Ecology 63: 267-273.
Marra PP & Holmes RT. 2001. Consequences of dominance-mediated habitat segregation in American redstarts during the non-breeding season. The Auk. 188: 92-104.
Martin, T.E. 2001. Abiotic vs. biotic influences on habitat selection of coexisting species: climate change impacts? Ecology 82: 175-188.
Martin, A.L. and Moore, P.A. 2007. Field observations of agonism in the crayfish, Orconectes rusticus : shelter use in a natural environment. Ethology 113: 1192-1201.
Mason, J.C. 1970. Copulatory behavior of the crayfish, Pacifasticus trowbridgii (Stimpson). Can. J. Zool. 48: 969-976.
Maynard Smith, J. and Parker, G.A. 1976. The logic of asymmetric contests. Anim. Behav. 24: 159-175.
McMann, S. 2000. Effects of residence time on displays during territory establishment in a lizard. Anim. Behav . 59: 513-522.
Moore, P.A. and Bergman, D.A. 2005. The smell of success and failure: the role of intrinsic and extrinsic chemical signals on the social behavior of crayfish. Integ. Comp. Biol. 45: 650- 657.
Morrell, L.J. and Kokko, H. 2003. Adaptive strategies of territory formation. Behav. Ecol. Sociobiol. 54: 385-395.
Nakano S. 1995. Individual differences in resource use, growth and emigration under the influence of a dominance hierarchy in fluvial red-spotted masu salmon in a natural habitat . J. Anim. Ecol. 64: 75-84.
Noel, M.V., Grant, J.W.A. and Carrigan, J.G. 2005. Effects of competitor-to-resource ratio on aggression and size variation within group of convict cichlids. Anim. Behav. 69: 1157- 1163.
Pagel, M. and Dawkins, M.S. 1997. Peck orders and group size in laying hens: ‘future contacts’ for non-aggression. Behav. Proc. 40: 13-25.
85
Panksepp, J.B., Yue, Z.X., Drerup, C. and Huber, R. 2003. Amine neurochemistry and aggression in crayfish. Microsc. Res. Tech. 60: 360-368.
Parker, G. A. 1974. Assessment strategy and the evolution of fighting behaviour. J . Theoret. Biol. 47: 223-243.
Pavey, C.R. and Fielder, D.R. 1996. The influence of size differential on agonistic behaviour in the freshwater crayfish, Cherax cuspidatus (Decapoda: Parastacidae). J. Zool. London 238: 445-457.
Pecor, K.W. and Hazlett, B.A. 2003. Frequency of encounter with risk and the tradeoff between pursuit and antipredator behaviors in crayfish: a test of the risk allocation hypothesis. Ethology 109: 97-106.
Peeke, H.V.S., Sippel, J. and Figler, M. H. 1995. Prior residence effects in shelter defense in adult signal crayfish ( Pacifastacus leniusculus (Dana)): results in same and mixed-sex dyads. Crustaceana 68: 873-881.
Poisbleau, M., Jenouvrier, S. and Fritz, H. 2006. Assessing the reliability of dominance scores for assigning individual ranks in a hierarchy. Anim. Behav. 72: 835-842.
Pratt, A.E. and McLain, D.K. 2006. How dear is my enemy: intruder-resident encounters in male sand fiddler crabs ( Uca pugilator ). Behaviour 143: 597-617.
Rahman, N., Dunham, D.W. and Govind, C.K. 2004. Mate choice in the big-clawed snapping shrimp, Alpheus heterochaelissay , 1818. Crustaceana 77: 95-111.
Rands, S.A., Pettifor, R.A., Rowcliffe, J.M. and Cowlishaw, G. 2006. Social foraging and dominance relationships: the effects of socially mediated interference. Behav. Ecol. Sociobiol. 60: 572-581.
Ranta, E. and Lindstrom, K. 1992. Power to hold sheltering burrows by juveniles of signal crayfish, Pasifasticus leniusculus . Ethology 92: 217-226.
Robb, S. E. & Grant, J. W. A. 1998. Interactions between the spatial and temporal clumping of food affect the intensity of aggression in Japanese medaka. Anim. Behav. 56: 29-34.
Robinson, C.A., Thom, T.J. and Lucas, M.C. 2000. Ranging behaviour of a large freshwater invertebrate, the white-clawed crayfish Austropotamobius pallipes . Freshw. Biol. 44: 509- 521. 86
Ruckstuhl, K.E. and Neuhaus, P. 2002. Sexual segregation in ungulates: a comparative test of three hypotheses. Biol. Rev. 77: 77-96.
Rutherford, P.L., Dunham, D.W. and Allison, V. 1995. Winning agonistic encounters by male crayfish Orconectes rusticus (Girard) (Decapoda, Cambaridae)- chela size matters but chela symmetry does not. Crustaceana 68: 526-529.
Rutte, C., Taborsky, M. and Brinkhof, M.W.G. 2006. What sets the odds of winning and losing? TREE 21: 16-21.
Schapker, H., Breithaupt, T., Shuranova, Z., Burmistrov, Y. and Cooper, R.L. 2002. Heart and ventilatory measures in crayfish during environmental disturbances and social interactions. Comp. Biochem. Physiol. A-Molec. Integr. Physiol. 131: 397-407.
Statzner, B., Fievet, E., Champagne, J. Y., Morel, R. and Herouin, E. 2000. Crayfish as geomorphic agents and ecosystem engineers: Biological behavior affects sand and gravel erosion in experimental streams. Limnol. Oceanogr. 45: 1030-1040.
Stein, R.A. 1976. Sexual dimorphism in crayfish chelae: functional significance linked to reproductive activities. Can. J. of Zool. 54: 220-227.
Stocker, A.M. and Huber, R. 2001. Fighting strategies in crayfish Orconectes rusticus (Decapoda, Cambaridae) differ with hunger state and the presence of food cues. Ethology 107: 727-736.
Takahashi, M., Suzuki, N. and Koga, Tsunenori, K. 2001. Burrow defense behaviors in a sand- bubbler crab, Scopimera globosa , in relation to body size and prior residence. J. Ethol. 19: 93-96.
Tierney, A.J., Godleski, M.S., & Massanari, J.R. (2000) Comparative analysis of agonistic behavior in four crayfish species. J. Crust. Biol. 20: 54-66.
Tobias, J. 1997. Asymmetric territorial contests in the European robin: the role of settlement costs. Anim. Behav. 54: 9-21.
Treisman, M. 1975. Predation and the evolution of gregariousness. I. Models for concealment and evasion. Anim. Behav. 23: 779-800.
87
Usio, N., Konishi, M. and Nakano, S. 2001. Species displacement between an introduced and a 'vulnerable' crayfish: the role of aggressive interactions and shelter competition. Biol. Inv. 3: 179-185. van Oort, H., Otter, K.A. and Fort, K.T. 2007. Habitat, dominance, and the phenotypic quality of male Black-capped Chikadees. Condor 109: 88-96.
Verhulst, S. and Salomons, H. M. 2004. Why fight? Socially dominant jackdaws, Corvus monedula , have low fitness. Anim. Behav. 68: 777-783.
Walker, D., Porter, B.A. and Avise, J.C. 2002. Genetic parentage assessment in the crayfish Orconectes placidus , a high-fecundity invertebrate with extended maternal brood care. Molec. Ecol. 11: 2115-2122.
Wasserberg G, Kotler BP & Abramsky Z. 2006. The role of site, habitat, seasonality and competition in determining the nightly activity patterns of psammophilic gerbils in a centrifugally organized community. Oikos 112: 573-579.
Webster, M.M. and Hart, P.J. 2006. Subhabitat by foraging threespine stickleback ( Gasterosteus aculeatus ): previous experience and social conformity. Behav. Ecol. Sociobiol. 60: 77-86.
Weir, L.K. and Grant, J.W.A. 2004. The causes of resource monopolization: interaction between resource dispersion and mode of competition. Ethology 110: 63-74
Wilson, E. O. 1975. Sociobiology: The New Synthesis. Harvard University Press, Cambridge, MA.
Wright, S. 1932. The roles of mutation, inbreeding, crossbreeding, and selection in evolution. Proc. 6th Internatl. Congr. Genet. 1: 356-366.
Yeh, S.R., Musolf, B.E. and Edwards, D.H. 1997. Neuronal adaptations to changes in the social dominance status of crayfish. J. Neurosci . 17: 697-708.
Zar, J. H. 1999. Biostatistical analysis, 4 th edn. Prentic-Hall Inc. Simon and Schuster, Upper Saddle River, NJ.
Zulandt-Schneider, R.A., Schneider, R.W.S. and Moore, P.A. 1999. Recognition of dominance status by chemoreception in the red swamp crayfish, Procambarus clarkii. J. Chem. Ecol. 25: 781-794.
88
Zulandt-Schneider, Huber, R., & Moore, P A. 2001. Individual and status recognition in the crayfish, Orconectes rusticus : the effects of urine release on fight dynamics. Behaviour, 138: 137-153.
89
APPENDIX A: FIGURES AND TABLES
3
2
1 Intensity of Interactions of Intensity 0
1 2 3 4 Day
FIGURE 1. Mean (± SE) intensity of interactions for all male groups ( N = 10) over 96 hrs. Intensity of interactions is measured on an hourly basis. Nighttime periods are represented by the black boxes. 90
1.0 0.8 0.6 0.4 0.2 0.0
1 10 100 1.0 0.8 0.6 0.4 0.2 0.0
1 10 100
Dominance Activity Index Activity Dominance 1.0 0.8 0.6 0.4 0.2 0.0 1 10 100 Time (hr)
FIGURE 2. Dominance activity indices for individual male crayfish calculated at different time intervals over a 96 hr period. Each graph represents one group of five male crayfish. DAI values indicate hierarchy rank with the most dominant crayfish possessing the highest value. Symbols correspond to initial ranks after the first hour. 91
0.45 *
0.40
0.35
0.30
0.25
0.20
0.15
0.10 Proportion of encounters of Proportion
0.05
0.00 1 2 3 4 5 Final Rank
FIGURE 3. Proportion of agonistic interactions that males within a dominance hierarchy ( N = 10) participated in. Significant differences between bars are indicated by an asterisk. 92
20 c 18 b b 16 b
14 a 12
10
8
6
4 Time spent in shelters (hrs) shelters in spent Time 2
0 1 2 3 4 5 Final hierarchical position
FIGURE 4. Mean (± SE) number of hours that crayfish of differing social ranks spent in shelter over a 96 hr period. Represented ranks were calculated at the end of 96 hrs. Rank is in descending order with '1' indicating most dominant. Bars with the same letter are not significantly different ( N = 10, ANOVA; p < 0.05). 93
80 c Control c,d Dominant Subordinate 70 d
60 b 50
40 a 30 a
Mean percent time percent Mean 20
10 e e e 0 In Shelter Out Feeding Position within tank
FIGURE 5. Mean (± SE) percentage of time crayfish spent either in or out of the shelter, or feeding. White bars represent naïve control, hatched bars represent conditioned dominant, and black bars represent conditioned subordinate treatments. Arcsine transformation of percentages were performed for ANOVA and LSD post hoc test ( N = 20). Bars with the same letter are not significantly different (p < 0.05). 94
0.035 a
0.030 a,b 0.025
0.020 b
DS / hr 5 0.015
0.010
0.005
0.000 1-5 5-10 10-15 Hours In Population
FIGURE 6. Mean shifts (± SE) in dominance score over 5 hour intervals while in population ( n = 48) . The shift in dominance is calculated as (Dominance Score at t – Dominance Score at t-1) / time. Hierarchies were considered stable when mean shifts in dominance score over time decreased to ± 0.015 ∆ DS / 5 hr, which occurred by the 15th hr. Differing letters above bars indicate statistical significance (ANOVA; P = 0.004). 95
1.0 b
0.8
0.6
a,b
0.4
0.2 TotalDS shifts over 15 hrs a
0.0 Alpha Intermediate Omega Ordinal Hierarchy Rank
FIGURE 7. Mean of total shifts (± SE) in dominance score over 15 hrs for three different hierarchy ranks in each population ( n = 5). ‘Alpha’ refers to the most dominant crayfish in a hierarchy. ‘Intermediate’ refers to the crayfish with the median dominance score. ‘Omega’ refers to the least dominant (i.e. most subordinate) crayfish. Differing letters above the bars indicate statistical significance (ANOVA; P = 0.01). 96
HIGH
4 m ABSENT
LOW
FIGURE 8. Schematic of the experimental arena consisting of a 4 × 4 m pool and three habitat types. Grey rectangles represent individual crayfish shelters. The ‘absent’ habitat was positioned in the center of the pool for all trials. The positions of ‘high’ and ‘low’ habitats were switched at random relative to the pool inflow and outflow at the beginning of each trial. 97
High Shelter a Low Shelter 70 Shelter Absent
60
50
40 b 30
b b b 20 b b b 10 c
Percent of total time spent in each habitat c c c 0 Day Night Day Night In shelter Out of shelter
FIGURE 9. Percent of time that crayfish spent in shelter in ‘high,’ ‘low’ and ‘shelter absent’ habitats. Grey and black bars represent shelter occupation during the day (6:00:00 - 21:00:00) and night (21:00:00 – 6:00:00) respectively. Differing letters above the bars indicate statistical significance (ANOVA; P = 0.0013). 98
120
110
100
90
80
70
60
50
40 ** 30
20
Totalnumber of agonistic encounters 10
0 High Low Absent Shelter Habitat
FIGURE 10. Number of agonistic encounters that individuals engaged in across habitats over the total trial time. Differing letters above the bars indicate statistical significance (ANOVA; P < 0.0001). 99
80 **
60
40 Alpha
20
0
-20
Omega -40 % time spent inthe same shelter
Difference betweenalpha andomega -60
Shelters ranked by alpha shelter occupation
FIGURE 11. Alpha dominant (highest ranking) and omega subordinate (lowest ranking) differences in shelter use. The difference in the proportion of time that the alpha and omega crayfish of each trial spent in specific shelters was calculated (e.g. alpha % time in shelter A – omega % time in shelter A). Shelters of each trial were then ranked (1-7) by the magnitude of difference with the highest difference in shelter occupation ranked as 1. The adjusted percent time spent in shelter for ranked shelters equals alpha minus subordinate shelter use. The asterisk above the bar indicates statistical significance (ANOVA; P < 0.00001). 100
FIGURE 12. Experimental arena consisting of a 1.8 m diameter pool with resident shelter (R) and an intruder shelter at one of three distances: 20 cm, 60 cm, or 120 cm. White boxes represent shelters made of halved 7.5 diameter PVC pipe, 11.0 cm long. 101
20 cm 0.9 b b 60 cm b 120 cm 0.8 b
0.7
0.6 b,c b,c 0.5
0.4 a,c a,c
0.3 c
isdominant to intruder 0.2
0.1 Proportionof trials where resident
0.0 DOM SUB CONTROL Conditioned Status
FIGURE 13. Proportion of trials in each treatment (conditioned status × inter-shelter distance) where focal residents were dominant to naïve intruders. Hatched bars represent 20 cm inter- shelter distance treatments; gray bars represent 60 cm treatments; and black bars represent 120 cm treatments. The likelihood of a resident being dominant to an intruder was dependent upon 2 the distance between shelters (x = 16.39, P < 0.05; q(0.05),∞,9 = 4.39). Letters above the bars differ when bars are significantly different from each other. 102
20 cm 0.8 c 60 cm 120 cm 0.7 a,c 0.6 a,c 0.5
0.4
0.3 a,b
0.2 b b 0.1 Proportion oftrials with statusreversal 0.0 DOM SUB Conditioned Status
FIGURE 14. Proportion of dominant and subordinate status condition trials where dominance reversals occurred. Reversals are defined as resident/intruder dominance relationships that deviate from those predicted by status conditioning. Hatched bars represent 20 cm inter-shelter distance treatments; gray bars represent 60 cm treatments; black bars represent 120 cm treatments. The occurrence of reversals depended on inter-shelter distance (x 2 = 11.94, P < 0.05; q(0.05) ,∞,6 = 4.03). Letters above the bars differ when bars are significantly different from each other.
103
TABLE 1. Comparisons between observed behaviors, crayfish dominance activity index (DAI), and size (males: N = 30; females: N = 80).
Male Male chelae Female – carapace Male DAI length male size length across trials difference across trials
# of mating p = 0.2738 p = 0.2091 p = 0.4594 p < 0.01 events r2 = 0.0426 r2 = 0.0557 r2 = 0.0197 r2 = 0.0885
Total t spent p = 0.4417 p = 0.1970 p = 0.1640 p < 0.01 mating r2 = 0.0213 r2 = 0.0587 r2 = 0.0680 r2 = 0.0837
# of p = 0.6470 p = 0.9557 p = 0.5757 ------mates/hr r2 = 0.0076 r2 = 0.0001 r 2 = 0.0113
% t spent p = 0.7555 p < 0.05 p < 0.05 ------mating r2 = 0.0035 r2 = 0.1704 r2 = 0.1645
Mean t per p = 0.5346 ------mating event r2 = 0.0139
p = 0.7547 p < 0.005 p < 0.005 t in shelter ------r2 = 0.0035 r2 = 0.2924 r2 = 0.2790
104
TABLE 2. Results of linear regressions between dominance score (DS), difference in DS between an individual and its nearest neighbor ( │DS Individual – DS Neighbor │= absolute difference in DS), number of shelter evictions performed, chelae length (cm), and nearest neighbor distance (cm) ( n = 48).
2 Factors F1,46 R P
Nearest Neighbor DS 11.2518 0.1965 0.0016 Distance (cm)
│DS Individual – 4.2517 0.0846 0.0448 DS Neighbor │
Evictions 10.7730 0.1898 0.0019 Performed
Chelae 0.2948 0.0064 0.5897 Length (cm)
Evictions Dominance Score 40.6654 0.4692 < 0.0001 Performed
Chelae 45.8722 0.4993 < 0.0001 Length (cm)
105
TABLE 3. Results of linear regressions between dominance score (DS) and diurnal actitivity ( n = 40).
Shelter Activity R2 P Habitat
% time in HIGH 0.0693 0.1008 shelter
% time out of 0.0509 0.1615 shelter
LOW % time in 0.0156 0.4422 shelter
% time out of 0.0041 0.6951 shelter
% time out of ABSENT 0.1133 0.0337 shelter
Agonistic ALL 0.1645 0.0094 encounters
Number of shelter 0.0229 0.3509 entries
Mean time per shelter 0.1150 0.0323 entry
106
TABLE 4. Results of linear regressions between dominance score (DS) and nocturnal actitivity (n = 40).
Shelter Activity R2 P Habitat
% time out of HIGH 0.0043 0.6886 shelter
% time out of LOW 0.0034 0.7126 shelter
% time out of ABSENT 0.0005 0.8937 shelter
Agonistic ALL 0.0844 0.0689 encounters
Number of shelter 0.1199 0.0286 entries
107
TABLE 5. Term descriptions for crayfish agonistic and shelter use behaviors.
Behavior Description
Agonistic encounter Begins with approach within one body length; ends with retreat (e.g. tail flip) and separation by 2 body lengths.
Shelter-related encounter Agonistic encounters occurring within 2 body lengths of a shelter (including evictions).
Shelter entry Crayfish enters a shelter and is completely obstructed from view.
Shelter occupancy Begins with shelter entry and ends when the crayfish moves ½ body length away from the shelter.
Shelter eviction Crayfish leaves a shelter in direct response to an agonistic encounter.
Shelter approach Crayfish comes within one body length of a shelter without entering.
Percent shelter entries Number of entries into shelter X / (Number of approaches to shelter X + Number of entries into X)
Status reversal Social status of the focal resident relative to the naïve intruder is opposite of that predicted by status conditioning.