Where do floaters settle? An experimental approach

Animal Behaviour 86 (2013) 1069e1075

Contents lists available at ScienceDirect

Animal Behaviour journal homepage: www.elsevier.com/locate/anbehav

Where do floaters settle? An experimental approach in odonates a Maria J. Go1a˛b a, *, Szymon Sniegula , Szymon M. Drobniak b, Tadeusz Zaja˛c a, c M. A. Serrano-Meneses a

Institute of Nature Conservation, Polish Academy of Sciences, Kraków, Poland ski, Kraków, Poland Instytut Nauk o Srodowisku, Uniwersytet Jagiellon c Laboratorio de Biología Evolutiva, Centro Tlaxcala de Biología de la Conducta, Universidad Autónoma de Tlaxcala, Tlaxcala, Mexico b

a r t i c l e i n f o Article history: Received 14 May 2013 Initial acceptance 18 July 2013 Final acceptance 20 August 2013 Available online 2 October 2013 MS. number: 13-00409 Keywords: Calopteryx splendens floater habitat change habitat quality nonterritorial reproductive behaviour settlement territoriality

According to classic ecological models, nonterritorial males should settle in low-quality habitats as a result of losing competition over reproductive sites (‘defeated male’ hypothesis). Alternatively, according to evolutionary game theory models, nonterritorial males should settle in the vicinity of high-quality sites and ‘choose’ to delay breeding until these habitats are vacant for them (‘male player’ hypothesis). However, nonterritorial male spatial distribution has not been experimentally tested. If the defeated male hypothesis is true (1) deterioration of high-quality sites should increase the number of nonterritorial males in a population and (2) vacated low-quality territories should be taken over by new territorial males. If the male player hypothesis is true, a similar manipulation should (1) decrease the number of nonterritorial males and (2) vacated low-quality territories should not be taken over. We performed two types of field experiment to test these hypotheses: male removal and patch quality manipulation. Our study species was the territorial damselfly Calopteryx splendens; males of this species exhibit both territorial and nonterritorial behaviour. Our results suggest that deterioration of highquality habitats significantly reduced the number of nonterritorial males. The proportion of take-overs of the high-quality territories was significantly higher than that of low-quality territories. Our study supports the assumptions of the male player hypothesis and indicates that nonterritorial damselflies are more sensitive to habitat quality changes than territorial ones. Because nonterritorial individuals exist in most populations of territorial taxa, a better understanding of their settlement rules may be relevant for population dynamics and modelling. Ó 2013 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

Darwin (1876) argued that sexual selection depends on a struggle between individuals within one sex over mating success. The unsuccessful competitor leaves few or no offspring, which is equivalent to his ‘evolutionary death’. Despite this, many studies on territorial species have shown that some individuals do not reproduce at a given time and do not possess territories, even though they are capable of doing so (e.g. Newton 1992). Therefore two questions arise: (1) why do nonterritorial males (‘floaters’) ‘decide’ not to reproduce; and (2) where do they settle? In our study a ‘floater’ is any nonterritorial male adopting a tactic alternative to territoriality (sneaking, wandering, queuing). Different studies of floaters’ lifestyles have produced diverse results; however, there are two general explanations for why individuals adopt nonterritoriality. According to classic ecological models, nonterritorial males represent a reservoir of future breeders and should

* Correspondence: M. J. Go1a˛ b, Polish Academy of Sciences, Institute of Nature Conservation, al. Mickiewicza 33, 31-120 Kraków, Poland. E-mail address: [email protected] (M. J. Go1a˛ b).

settle in any vacated habitats regardless of their quality (Brown 1969; Fretwell & Lucas 1969). Alternatively, floaters, using evolved decision-making rules, ‘decide’ to delay reproduction, even though there are vacant territories available (Kokko & Sutherland 1998) and settle in the vicinity of high-quality habitats until these become available for them. According to the latter approach, which is based on evolutionary game theory, only territories with a high expected reproductive success are occupied while territories below a given threshold quality are not (Zack & Stutchbury 1992; Kokko & Sutherland 1998). It has been shown that bird floaters may settle in the vicinity of a given territory in order to gain knowledge about that place and conspecifics present (Smith 1978). Nonterritorial individuals could also attempt to reproduce by performing extrapair copulations (EPCs; Platek & Shackelford 2006). Intuitively, for nonterritorial individuals to succeed in obtaining EPCs, they have to appear at the breeding site for a period of time that guarantees them a mate, but it still remains unclear whether floaters choose to settle at the site or not. In contrast, other authors report that nonterritorial individuals disperse and lead a transient life at some distance from

0003-3472/$38.00 Ó 2013 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.anbehav.2013.09.013

1070

M. J. Goła˛ b et al. / Animal Behaviour 86 (2013) 1069e1075

the breeding territories (Campioni et al. 2010), ‘serving’ in a population as potential replacement individuals (e.g. Brown 1969). Finally, it has been shown that some individuals prefer to defend poor-quality territories rather than taking the risk of floating (e.g. Bayne & Hobson 2001). In contrast to the wealth of studies focused on the consequences of habitat loss for territorial individuals, our knowledge of the reaction of nonterritorial individuals to habitat deterioration is limited (Kokko & Sutherland 1998; Penteriani et al. 2011). In their theoretical model, Kokko & Sutherland (1998) demonstrated that the loss of high-quality habitats causes not only a reduction in the reproductive population but also a greater reduction in the number of nonterritorial males. To our knowledge, experimental evidence of the settlement rules of nonterritorial males is lacking. In addition, no experiments have tested the influence of habitat quality changes on the behaviour of nonterritorial individuals. Although widespread, territorial behaviour in insects has received less emphasis than that in vertebrates (Matthews & Matthews 2010). In territorial odonate species, nongenetically based nonterritoriality is usually classified as an inferior tactic (reviewed by Suhonen et al. 2008). However, the settlement of nonterritorial odonates has not been studied in detail. Our study organism was the riverine damselfly Calopteryx splendens. Sexually mature males spend their whole life at a riverside, and territorial males are attached strictly to their territories, which consist of relatively small floating vegetation clumps. Both males and females exhibit alternative reproductive tactics (Rüppell & Hilfert 1997). With regard to the approaches presented above, we set out to test empirically two alternative hypotheses: (1) the defeated male hypothesis, which assumes that nonterritorial males should settle in low-quality patches of a habitat as a result of losing territorial contests (Brown 1969; Fretwell & Lucas 1969); and (2) the male player hypothesis, which predicts that nonterritorial males should settle in high-quality patches of a given habitat because they decide to delay breeding until these sites are available to them (Zack & Stutchbury 1992; Kokko & Sutherland 1998). The aim of this study was to test experimentally which of the two models best describes the mechanism of habitat selection in nonterritorial individuals, as well as to assess the importance of the deterioration of either high- (HQ) or low-quality (LQ) patches on the number of nonterritorial males. If the defeated male hypothesis is true, we predicted that (1) destruction (habitat patch manipulation) of the best sites should increase the number of nonterritorial males in a given area, and (2) any vacated LQ territory should be taken over. If the male player hypothesis is true (1) a similar habitat patch manipulation should decrease the number of nonterritorial males and (2) vacated LQ territory should not be taken over. We performed two types of experiment: first, male removal, to test for differences between two types of habitat (LQ and HQ) in the probability of the take-over of a vacated territory by a new territorial male and the probability of reacquisition of a territory by an original territorial male, and second, manipulation of patch quality to test for changes in the number of territorial and nonterritorial males in response to deterioration of HQ and LQ patches.

the water surface and defends a territory (floating vegetation clumps) within 2 m of its perching site (Tynkkynen et al. 2006). Nonterritorial males usually perch on higher parts of the riparian vegetation and can exhibit both wandering and satellite tactics. Wandering males patrol large sections of a river, whereas satellite males are active along a border of one or more territories (Marden & Cobb 2004; Koskimäki et al. 2009). In our study, satellite nonterritorial males were assigned to a particular patch when they were perching up to 0.5 m from its border. Males fight with each other to gain access to a territory and also court females before copulating, and guard them during oviposition (detailed descriptions of reproductive behaviours can be found in Marden & Waage 1990; Córdoba-Aguilar & Cordero-Rivera 2005; Rüppell et al. 2005; Go1a˛ b & Sniegula 2012). Experimental Set-up The study was conducted between 1 July and 5 August 2011 and 2012, on the river Bia1a Nida (Fig. 1), in southern Poland. It is a narrow (about 8 m wide) lowland river with a sandy bottom. The river section chosen for experiments is regulated and located in a homogeneous landscape (agricultural meadows). Hydrological conditions (water depth, velocity and temperature) of the section are homogeneous because of a weir situated 400 m downstream. The studied section was 50 m long and separated from the other parts of the river by 25 m sections that had been cleared of aquatic vegetation. Riparian vegetation was mowed regularly to prevent changes in its height (Ward & Mill 2005). All studied floating vegetation clumps (which are used by the females as oviposition patches) were cut with a pair of scissors so that the only physical factor that differentiated them was size. Every patch consisted of live Potamogeton natans and was situated directly adjacent to the riverbank. The patches were separated from one another by about 1.5 m. Since calopterygid females are known to prefer ovipositing on larger vegetation clumps (Waage 1987; Meek & Herman 1991), the patches were divided into two groups: HQ (large patch) and LQ (small patch). Also, based on the conspecific attraction hypothesis (e.g. Stamps 1987) and public information hypothesis (e.g. Danchin & Doligez 2001), we assumed that the presence of conspecifics reflects patch quality. Therefore, prior to manipulations, we conducted 40 min of continuous observations (see below) of every patch to assess general damselfly activity and site attractiveness to damselflies for further evaluation of patch quality (Switzer 2002a; Córdoba-Aguilar & Cordero-Rivera 2005; Guillermo-Ferreira & DelClaro 2011). All patches were marked with a marker post (patch ID) and their areas were measured. A single patch consisted of one to four territories, depending on its size (Fig. 1). Experiments were conducted between 1000 and 1600 hours and only under favourable and comparable weather conditions since the weather influences damselfly behaviour (Rüppell et al. 2005; Go1a˛ b & Sniegula 2012). To avoid multiple counting, all

METHODS Study Species Calopteryx splendens is a common European damselfly, which inhabits lowland rivers (Askew 1988). The highest population density occurs in July in Central Europe (Rüppell et al. 2005; Go1a˛ b 2012). The average life span of a mature male is & Sniegula approximately 1e2 weeks (Svensson et al. 2006; Tynkkynen et al. 2009). A male is classified as territorial when it remains close to

Figure 1. Bia1a Nida river. Six of the patches used in the study are visible on the righthand side of the river.


specimens were marked individually with a unique combination of three-digit numbers (silver marking pen Artline 999XRF) at the beginning of every working day (marking sessions lasted about 1 h). Two types of experiments, (1) male removal and (2) patch manipulation (patch depletion and patch release; see below), were performed on both HQ and LQ patches. (1) For the male removal experiment we caught an original territorial male with an insect net and kept him for 10 min in a coolbox to avoid handling trauma (e.g. Stettmer 1996). In the meantime, his territory was observed until a new territorial male took it over. The time from removal of the territory owner until a new male took over was noted. After the original territory owner was released, the time until reacquisition of the territory was noted. (2) The patch manipulation experiment was preceded by control observations. Each patch was observed for 20 min. Every other minute, we recorded the number of individuals displaying specific behaviours (territorial and nonterritorial tactic, fighting). For further analyses we counted the average number of both territorial and nonterritorial males present at each patch and the total number of fights observed on each patch per set of observations. After the control observation, we depleted the patch by sinking half of it with ballast, which consisted of pair of steel pipes connected with fishing net. We then let the damselflies adjust to the new situation by waiting 5 min (our preliminary studies showed that 5 min is enough for damselflies to adjust to a given new situation). After this, another 20 min observation was conducted. The ballast was removed from the previously depleted patch and its original size was restored. The third set of 20 min observations was repeated after another 5 min break. These two types of experiment for a given patch were repeated after a break of at least 4 days. Statistical Analyses Statistical analyses were performed using R version 2.15.1 (R Development Core Team 2012). In total we studied two HQ and two LQ patches in 2011, and five HQ and seven LQ patches in 2012. Overall, we analysed 95 territorial and 492 nonterritorial males in 2011, and 112 territorial and 595 nonterritorial males in 2012. To assess a patch’s quality (HQ or LQ) we took into account its size and general damselfly activity. Principal component analysis (PCA) was applied to four variables: number of territorial males, number of nonterritorial males, number of fights (in total we observed 196 fights in 2011 and 234 in 2012) and patch size. PC1 was used as a proxy of patch quality (see Results). The k-means clustering algorithm was applied (Hartigan & Wong 1979) to classify patches into two groups (kmeans command in R, stats package, assuming division of patches into k distinct groups). This algorithm divides observations into k ¼ 2 groups minimizing the sum of squared distances of each observation from the centre of the group to which it is assigned. In the male removal experiment (three to four replicates for every patch) we constructed two generalized linear mixed models (GLMM) with binomial error distribution and logit link function (glmer command; lme4 package in R; Bolker et al. 2009). We first tested for differences between HQ and LQ patches in the probability of a new territorial male taking over a territory. Patch quality and year were used as fixed effects and patch ID was used as a random effect. Second, we tested for differences between HQ and LQ patches in the probability of reacquisition of a territory. We further constructed two linear mixed-effect models (lme command; nlme package) to test for differences between HQ and LQ patches in the time needed for a new territorial male to arrive and in the time the original territory owner needed for reacquisition of his former place. Patch quality and year were used as

1071

fixed effects and patch ID was used as a random effect. At the end of the daily observations (1600 hours) additional quick observations were performed to check whether original territory owners that did not return to their territories during the observation (10 min) did so afterwards. We found nine such situations. These latearriving males were included in the analysis and assigned as those that arrived back in 20 min. The experiment with patch deterioration and patch release (the experiment was repeated three to four times on every patch) was analysed using a GLMM with Poisson error distribution (ASReml command in R; Gilmour et al. 2002). We tested for differences in the number of territorial and nonterritorial males between the two types of patches as a response to our manipulations. Patch quality, manipulation and year were used as fixed effects and patch ID was used as a random effect. We also attempted to express quantitatively (i.e. by the change in relevant numbers) the impact of patch manipulation on the numbers of nonterritorial and territorial males in HQ patches. This impact was expressed as the difference between the relevant number of individuals before and after manipulation. To assess differences between territorial and nonterritorial males we calculated the standard errors of these differences using the delta method (e.g. Lynch & Walsh 1998) after extracting relevant means and their covariance matrices from the output of the lme procedure. RESULTS Patch Quality The PCA of four patch characteristics (number of territorial males, number of nonterritorial males, number of fights and patch size) showed that PC1 explained 64.80% of total variance, while PC2 explained 15.65%. Therefore PC1 was chosen as a measure of patch quality for further analyses; loadings for the respective variables were as follows: number of territorial males: 0.85; number of nonterritorial males: 0.77; number of fights: 0.75; patch size: 0.84. The k-means analysis successfully identified two groups of patches: first, one of high quality (HQ) and, second, one of low quality (LQ). The average PC1 for cluster HQ was PC1 þ SE ¼ 1.15 þ 0.16 and for LQ PC1 þ SE ¼ 1.48 þ 0.06. Male Removal Experiment The probability of a new male taking over a territory was significantly higher in HQ (100% of take-overs) than in LQ patches (57.14% of take-overs). There was also a significant difference between the 2 years (Table 1). By contrast, we found no difference in the probability of reacquisition of a territory by an original territorial male (68.00% of HQ and 32.14% of LQ territories were regained) between HQ and LQ patches. Also, there were no differences between the 2 years. Both models were simplified since we found no interactions between fixed effects (Table 1). Table 1 Generalized linear model with binomial error used to analyse differences in the proportion of territory take-over and reacquisition events between two patch qualities (high or low) in 2 years Variables Response: territory take-over Fixed effects: patch quality Year Response: territory reacquisition Fixed effects: patch quality Year

SE 0.64 0.65 3543.77 0.78

Z

P

2.54 2.30

0.011 0.021

0.005 0.22

0.996 0.823


Patch Manipulation Experiment The number of territorial males did not change in response to patch manipulation (Wald X2 ¼ 2.88, P ¼ 0.237; Fig. 3a) but it was higher in HQ than in LQ patches (Wald X2 ¼ 23.71, P < 0.001; Fig. 3a). There were no differences between the 2 years (Wald X2 ¼ 0.24, P ¼ 0.622). The interaction term patch quality*manipulation was not significant (P ¼ 0.679; Fig. 3), which suggests that patch manipulation had the same effect on the number of territorial males in both HQ and LQ patches. The number of nonterritorial males was always higher in HQ than in LQ patches (Wald X2 ¼ 51.38, P < 0.001; Fig. 3b) and decreased significantly after our manipulations (Wald X2 ¼ 14.07, P < 0.001; Fig. 3b). We found no differences between the 2 years (Wald X2 ¼ 0.11, P ¼ 0.733). The interaction term patch quality*manipulation was not statistically significant (P ¼ 0.224; Fig. 3b), which suggests that the number of nonterritorial males changed similarly after experimental manipulations (patch depletion and release) in both HQ and LQ patches. The estimated difference in the abundance of males in control of manipulated HQ patches was mean SD ¼ 0.96 0.29 (on the scale of the link function) for territorial males and 1.71 0.27 for nonterritorial males. DISCUSSION We found that the number of nonterritorial males declined by almost twice as much as the number of territorial males after the best habitats were depleted. We further demonstrated that nonterritorial males settled in the vicinity of the HQ sites to improve their chances of territory ownership. Our results suggest that floaters are able to defend territories and that nonterritorial males probably settle in the vicinity of highquality territories since the time needed to establish the new residency in the male removal experiment was shorter in HQ patches. According to the defeated male hypothesis, any vacated territories located in both LQ and HQ patches should be taken over. In our experiments, males were more interested in HQ patches. After the original territorial males were removed from their territories, over 42% of the LQ territories were left empty, while all HQ patches were taken over. This supports the male player hypothesis. Rapid replacement of displaced territorial males was first described by Darwin (1871), and by many more since (e.g. Alcock 1995; Kempenaers et al. 2001; Switzer 2002b; Baird et al. 2012; Ota et al. 2012). These studies, together with our results, confirm that males of territorial species aspire to a tactic that guarantees the highest mating success, that is, territoriality (Mills et al. 2010). Males present near experimentally vacated territories could also carry out EPCs and, in our study, these EPCs might have influenced a male’s decision to take over a territory whose owner had been removed. Territorial species of odonates adopt multiple mating tactics (Conrad & Pritchard 1992) with two types of oviposition: (1) preceded by copulation with a territory owner (such males can sire up to 95% of the offspring) and (2) not preceded by copulation with a territory

(a)

1000 Time for take-over (s)

New territorial males needed significantly less time to take over territories located in HQ patches than in LQ patches (F1,13 ¼ 25.69, P < 0.001). Also, the time needed for the reacquisition of territories by original owners was shorter in HQ patches than in LQ patches (F1,13 ¼ 6.46, P ¼ 0.024). These patterns were similar between years in both models (F1,13 ¼ 4.38, P ¼ 0.056 and F1,13 ¼ 0.69, P ¼ 0.422, respectively). The interaction term patch quality*year was not statistically significant in any of the analyses (P ¼ 0.151 for takeover and P ¼ 0.932 for reacquisition); therefore both models were simplified by removing the interaction term (Fig. 2).

600

200

HQ

1000

Time for reacquisition (s)

1072

LQ

(b)

800

600

400

HQ

LQ Patch quality

Figure 2. Male removal experiment. Differences between high-quality (HQ) and lowquality (LQ) patches in (a) the time new territorial males needed to take over a vacant territory and (b) the time original territorial males needed to reacquire a territory upon release. Solid lines/circles (2011) and dashed lines/triangles (2012) are added to facilitate interpretation. Error bars denote 95% CI.

owner. The latter tactic is equivalent to EPC; the female is fertilized by an unknown, probably nonterritorial male (Corbet 1999; CórdobaAguilar 2008). In our study, males could gather in the vicinity of the best sites for at least two reasons: EPC and territory take-over. The difference in the probability of take-overs between the 2 years could be the result of low sample sizes or seasonal fluctuations in some environmental and/or population conditions. Since we were unable to find significant differences between years in all other analyses, however, it would be speculative to draw conclusions from these data.


(a) 2.5

2

1.5

Mean number of males

1

Before

During

After (b)

3

2.5

2

1.5

1

Before

During Manipulation

After

Figure 3. Patch manipulation experiment. Difference between HQ (solid lines/circles) and LQ (dashed lines/triangles) patches in the number of (a) territorial and (b) nonterritorial males of Calopteryx splendens before, during and after a temporary reduction in patch quality. Error bars refer to 95% CI.

We found no indication that reacquisition of territories by the original owners depended on patch quality and suggest that predation risk could be responsible for this result. Many species assess habitats in terms of predation risk (Lima & Dill 1990; Hughes et al. 1994; Rettie & Messier 2000; Kiflawi et al. 2003; Whittingham & Evans 2004; Chalfoun & Martin 2009). For instance, it has been shown that animals, including damselflies, can learn about dangerous situations and that their former predation experiences can influence their future decisions regarding territory ownership (Chivers et al. 1996; Hilfert-Rüppell 1999; Barcellos et al. 2010; Breed & Moore 2011). Hilfert-Rüppell (1999) observed that some calopterygid males left their territories immediately after a

1073

predator’s attack. In our study, we collected damselflies using insect nets. Although all males were released within 10 min of their capture, our treatment was probably perceived as an unsuccessful predatory attack. All patches were equally affected by this treatment, regardless of their quality. It is possible that the defence costs associated with maintaining some of these territories probably exceeded their potential benefits and therefore some territorial males decided to abandon their sites. Prior to the patch manipulation experiment, the number of territorial males was always higher at HQ than at LQ patches. This result was expected, since patches of higher quality usually attract more territorial and nonterritorial damselflies (Switzer 2002a; Córdoba-Aguilar & Cordero-Rivera 2005; Guillermo-Ferreira & Del-Claro 2011). The number of territorial males did not change after patch manipulation, and patch manipulation influenced territorials from HQ and LQ patches similarly (the interaction term patch quality*manipulation was not significant; see Results). Several conclusions can be drawn regarding these two results. First, we can assume that territorial males invested their time and energy to obtain and hold their territories for a relatively long time (e.g. a few hours). It is also likely that there were no vacant territories in the neighbourhood and that territorial males may have perceived this as a visual cue that the chance of obtaining a new territory after abandoning the original one was very low. In such a situation it was not cost effective for them to abandon the patch after its quality decreased, even though the patch quality was not initially very high. Second, the experimental patch deterioration may not have influenced the most valuable parts of patches. Third, our manipulation increased male density per patch (in the case of HQ patches) and territorial males could have reconstructed territory partitioning to avoid unnecessary intrusions by neighbours (Smith 2011) instead of leaving the patch. Calopteryx splendens males can adjust their territory sizes depending on population density and the availability of macrophytes (Rüppell et al. 2005; Go1a˛ b & Sniegula 2012). However, this should not be the case for LQ patches, where at most one territorial male was present. It could also be claimed that the time we set for manipulations and observations could have been too short and damselflies did not have enough time to react. However, this argument may not be correct since C. splendens individuals are known to make rapid behavioural decisions (Corbet 1999; Hilfert-Rüppell 1999; Rüppell et al. 2005). Obviously, patch release (reconstruction of the original size and quality) did not influence the number of territorial males in either HQ or LQ patches. This is a natural consequence of the fact that most territorial males stayed at their patches after our partial removal of the patch. Based on the fact that the year effect on the number of territorial males was not significant, we infer that the patterns found in our study are independent of season-to-season environmental fluctuations. In the case of nonterritorial males, the nonsignificant interaction term patch quality*manipulation suggests that floaters are sensitive to patch manipulation, regardless of its quality. On the one hand, this outcome supports the prediction that destruction of the HQ sites should decrease the number of nonterritorial males (male player hypothesis). On the other hand, a similar pattern was found for LQ patches. This implies that nonterritorial males were also present in the vicinity of LQ patches. It should be stressed that our LQ patches usually held one territory, which means these were not of the worst possible quality. It is therefore probable that some (small) number of nonterritorial males could have visited or even settled near them. We are not aware of any studies that distinguish between the number of territorial and nonterritorial males of C. splendens in relation to the quality of vegetation clumps. It is also hard to make any sensible comparisons with other taxa in this case because of differences in territory characteristics. However, since

1074


the habitat quality of C. splendens varies continuously within patches, we may suppose that LQ territories may be attractive to some males under some circumstances (e.g. at high population density). Population overcrowding is known to influence the behavioural decisions of C. splendens (Cordero 1999; Ward & Mill 2007; Go1a˛ b & Sniegula 2012) and our study was conducted during the peak of the flying season, when the population density is at its highest. A trait that was not measured in our study was the quality of males. Possibly some of the nonterritorial males observed in this study were teneral or low-quality adults, which might have been more prone to fly away after our manipulations. Quantifying the condition of calopterygid males involves killing and dissecting them (Córdoba-Aguilar & Cordero-Rivera 2005; Córdoba-Aguilar 2008). In our study killing every individual would have influenced both the population density and damselfly behaviour (see impact of predation discussed above). Finally, our last analysis revealed that nonterritorial males declined in abundance almost twice as much as territorial males, as a response to patch deterioration. This result gave additional support to the male player hypothesis and showed that deterioration of the best habitats affects mostly the nonterritorial fraction of a population. Nonterritorial individuals are regarded as a very important component of a population and their behavioural decisions have a key role in shaping the dynamics of the population and are expected to have considerable implications for conservation (Gordon 1997; Penteriani et al. 2011). Here we have provided the first empirical test of their settlement rules. We confirmed (1) that not only are floaters unsuccessful, subordinate individuals but also that male players can actively improve their reproductive chances, (2) floaters settle mostly in the vicinity of HQ sites and (3) habitat disturbance affects mostly the nonterritorial part of a population. Acknowledgments We are very grateful to Paulina Go1a˛ b for her field assistance and two anonymous referees for their valuable comments on the manuscript. M.J.G. was funded by a Polish State Committee for Scientific Research/National Science Centre grant (Project No. NN3042945537) and partly by the Institute of Nature Conservation PAS. References Alcock, J. 1995. Body-size and its effect on maleemale competition in Hylaeus alcyoneus (Hymenoptera, Colletidae). Journal of Insect Behavior, 8, 149e159. Askew, R. R. 1988. The Dragonflies of Europe. Colchester: Harley Books. Baird, T. A., Baird, T. D. & Shine, R. 2012. Aggressive transition between alternative male social tactics in a long-lived Australian dragon (Physignathus lesueurii) living at high density. PLoS One, 7, e41819. Barcellos, L. J. G., Ritter, F., Kreutz, L. C. & Cericato, L. 2010. Can zebrafish Danio rerio learn about predation risk? The effect of a previous experience on the cortisol response in subsequent encounters with a predator. Journal of Fish Biology, 76, 1032e1038. Bayne, E. M. & Hobson, K. A. 2001. Effects of habitat fragmentation on pairing success of ovenbirds: importance of male age and floater behavior. The Auk, 118, 380e388. Bolker, B. M., Brooks, M. E., Clark, C. J., Geange, S. W., Poulsen, J. R., Stevens, M. H. H. & White, J.-S. S. 2009. Generalized linear mixed models: a practical guide for ecology and evolution. Trends in Ecology & Evolution, 24, 127e135. Breed, M. & Moore, J. 2011. Animal Behaviour. Burlington, Massachusetts: Academic Press. Brown, J. 1969. Territorial behavior and population regulation in birds. Wilson Bulletin, 81, 293e329. Conrad, K. F. & Pritchard, G. 1992. An ecological classification of odonate mating systems: the relative influence of natural, inter- and intra-sexual selection on males. Biological Journal of the Linnean Society, 45, 255e269. Campioni, L., Delgado, M. D. M. & Penteriani, V. 2010. Social status influences microhabitat selection: breeder and floater eagle owls Bubo bubo use different post sites. Ibis, 152, 569e579. Chalfoun, A. D. & Martin, T. E. 2009. Habitat structure mediates predation risk for sedentary prey: experimental tests of alternative hypotheses. Journal of Animal Ecology, 78, 497e503.

Chivers, D. P., Wisenden, B. D. & Smith, R. J. F. 1996. Damselfly larvae learn to recognize predators from chemical cues in the predator’s diet. Animal Behaviour, 52, 315e320. Corbet, P. 1999. Dragonflies: Behavior and Ecology of Odonata. Ithaca, N.Y.: Comstock Cordero, A. 1999. Forced copulations and female contact guarding at a high male density in a Calopterygid damselfly. Journal of Insect Behavior, 12, 27e37. Córdoba-Aguilar, A. 2008. Dragonflies and Damselflies: Model Organisms for Ecological and Evolutionary Research. New York: Oxford University Press. Córdoba-Aguilar, A. & Cordero-Rivera, A. 2005. Evolution and ecology of Calopterygidae (Zygoptera: Odonata): status of knowledge and research perspectives. Neotropical Entomology, 34, 861e879. Danchin, E. & Doligez, B. 2001. Public information and breeding habitat selection. In: Dispersal (Ed. by J. Clobert, E. Danchin, A. Dhondt & J. Nichols), pp. 243e258. Oxford: Oxford University Press. Darwin, C. R. 1871. The Descent of Man and Selection in Relation to Sex. New York: Appleton. Darwin, C. R. 1876. The Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. 6th edn. London: J. Murray. Fretwell, S. D. & Lucas, H. L. 1969. On territorial behavior and other factors influencing habitat distribution in birds. I Theoretical development. Acta Biotheoretica, 19, 45e52. Gilmour, A., Gogel, B., Cullis, B., Welham, S. & Thompson, R. 2002. ASReml User Guide Release 1.0. Hemel Hempstead: VSN International Ltd. Goła˛ b, M. J. & Sniegula, S. 2012. Changes in reproductive behavior in adult damselfly Calopteryx splendens (Odonata: Calopterygidae) in response to flood. Entomological Science, 15, 280e287. Gordon, D. M. 1997. The population consequences of territorial behaviour. Trends in Ecology & Evolution, 12, 63e66. Guillermo-Ferreira, R. & Del-Claro, K. 2011. Territoriality and male-biased sexual size dimorphism in Argia reclusa (Odonata: Zygoptera). Acta Ethologica, 15, 101e105. Hartigan, J. A. & Wong, M. A. 1979. A K-means clustering algorithm. Applied Statistics, 28, 100e108. Hilfert-Rüppell, D. 1999. To stay or not to stay: decision-making during territorial behaviour of Calopteryx haemorrhoidalis and Calopteryx splendens splendens (Zygoptera: Calopterygidae). International Journal of Odonatology, 2, 167e175. Hughes, J. J., Ward, D. & Perrin, M. R. 1994. Predation risk and competition affect habitat selection and activity of Namib desert gerbils. Ecology, 75, 1397. Kempenaers, B., Everding, S., Bishop, C., Boag, P. & Robertson, R. J. 2001. Extrapair paternity and the reproductive role of male floaters in the tree swallow (Tachycineta bicolor). Behavioral Ecology and Sociobiology, 49, 251e259. Kiflawi, M., Blaustein, L. & Mangel, M. 2003. Oviposition habitat selection by the mosquito Culiseta longiareolata in response to risk of predation and conspecific larval density. Ecological Entomology, 28, 168e173. Kokko, H. & Sutherland, W. 1998. Optimal floating and queuing strategies: consequences for density dependence and habitat loss. American Naturalist,152, 354e366. Koskimäki, J., Rantala, M. & Suhonen, J. 2009. Wandering males are smaller than territorial males in the damselfly Calopteryx virgo (L.) (Zygoptera: Calopterigidae). Odonatologica, 38, 159e165. Lima, S. L. & Dill, L. M. 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology, 68, 619e640. Lynch, M. & Walsh, B. 1998. Genetics and Analysis of Quantitative Traits. Sunderland, Massachusetts: Sinauer. Marden, J. H. & Cobb, J. R. 2004. Territorial and mating success of dragonflies that vary in muscle power output and presence of gregarine gut parasites. Animal Behaviour, 68, 857e865. Marden, J. H. & Waage, J. K. 1990. Escalated damselfly territorial contests are energetic wars of attrition. Animal Behaviour, 39, 954e959. Matthews, R. W. & Matthews, J. R. 2010. Insect Behavior. New York: Springer. Meek, S. B. & Herman, T. B. 1991. The influence of oviposition resources on the dispersion and behaviour of calopterygid damselflies. Canadian Journal of Zoology, 69, 835e839. Mills, D. S., Marchant-Forde, J. N., McGreevy, P. D., Morton, D. B., Nicol, C. J., Phillips, C. J. C., Sandoe, P. & Swaisgood, R. R. 2010. The Encyclopedia of Applied Animal Behavior and Welfare. Wallingford: CAB International. Newton, I. 1992. Experiments on the limitation of bird numbers by territorial behaviour. Biological Reviews, 67, 129e173. Ota, K., Hori, M. & Kohda, M. 2012. Changes in reproductive life-history strategies in response to nest density in a shell-brooding cichlid, Telmatochromis vittatus. Naturwissenschaften, 99, 23e31. Penteriani, V., Ferrer, M. & Delgado, M. M. 2011. Floater strategies and dynamics in birds, and their importance in conservation biology: towards an understanding of nonbreeders in avian populations. Animal Conservation, 14, 233e241. Platek, S. M. & Shackelford, T. K. 2006. Female Infidelity and Paternal Uncertainty: Evolutionary Perspectives on Male Anti-cuckoldry Tactics. Cambridge: Cambridge University Press. R Development Core Team 2012. R: A Language and Environment for Statistical Computing. Vienna: The R Foundation for Statistical Computing. Rettie, W. J. & Messier, F. 2000. Hierarchical habitat selection by woodland caribou: its relationship to limiting factors. Ecography, 23, 466e478. Rüppell, G. & Hilfert, D. 1997. Alternative mating tactics in Calopteryx splendens (Odonata: Calopterygidae). In: Advances in Ethology. Vol. 32: Contributions to the XXV International Ethological Conference, Vienna, Austria (Ed. by M. Taborsky & B. Taborsky), p. 47. Berlin: Blackwell Wissenschafts-Verlag GmbH. Rüppell, G., Rehfeldt, G., Schütte, C. & Hilfert-Rüppell, D. 2005. Die Prachtlibellen Europas: Gattung Calopteryx. Hohenwarsleben: Westarp Wissenschaften.

M. J. Goła˛ b et al. / Animal Behaviour 86 (2013) 1069e1075 Smith, C. 2011. Good fences make good neighbours: the role of landmarks in territory partitioning in the rose bitterling (Rhodeus ocellatus). Behaviour, 148, 233e246. Smith, S. 1978. The ‘underworld’ in a territorial sparrow: adaptive strategy for floaters. American Naturalist, 112, 571e582. Stamps, J. 1987. Conspecifics as cues to territory quality: a preference of juvenile lizards (Anolis aeneus) for previously used territories. American Naturalist, 129, 629e642. Stettmer, C. 1996. Colonisation and dispersal patterns of banded (Calopteryx splendens) and beautiful demoiselles (C. virgo) (Odonata: Calopterygidae) in south-east German streams. European Journal of Entomology, 93, 579e593. Suhonen, J., Rantala, M. J. & Honkavaara, J. 2008. Territoriality in odonates. In: Dragonflies and Damselflies. Model Organisms for Ecological and Evolutionary Research (Ed. by A. Córdoba-Aguilar), pp. 203e218. New York: Oxford University Press. Svensson, E., Eroukhmanoff, F. & Friberg, M. 2006. Effects of natural and sexual selection on adaptive population divergence and premating isolation in a damselfly. Evolution, 60, 1242e1253. Switzer, P. 2002a. Territory quality, habitat selection, and competition in the amberwing dragonfly, Perithemis tenera (Say) (Odonata: Libellulidae): population patterns as a consequence of individual behavior. Journal of the Kansas Entomological Society, 75, 145e157.

1075

Switzer, P. 2002b. Individual variation in the duration of territory occupation by males of the dragonfly Perithemis tenera (Odonata: Libellulidae). Annals of the Entomological Society of America, 95, 628e636. Tynkkynen, K., Kotiaho, J., Luojumaki, M. & Suhonen, J. 2006. Interspecific territoriality in Calopteryx damselflies: the role of secondary sexual characters. Animal Behaviour, 71, 299e306. Tynkkynen, K., Raatikainen, K. J., Hakkila, M., Haukilehto, E. & Kotiaho, J. S. 2009. Alternative reproductive tactics and the propensity of hybridization. Journal of Evolutionary Biology, 22, 2512e2518. Waage, J. K. 1987. Choice and utilization of oviposition sites by female Calopteryx maculata (Odonata: Calopterygidae). I. Influence of site size and the presence of other females. Behavioral Ecology and Sociobiology, 20, 439e446. Ward, L. & Mill, P. J. 2005. Habitat factors influencing the presence of adult Calopteryx splendens (Odonata: Zygoptera). European Journal of Entomology, 102, 47e51. Ward, L. & Mill, P. 2007. Long range movements by individuals as a vehicle for range expansion in Calopteryx splendens (Odonata: Zygoptera). European Journal of Entomology, 104, 195e198. Whittingham, M. J. & Evans, K. L. 2004. The effects of habitat structure on predation risk of birds in agricultural landscapes. Ibis, 146, 210e220. Zack, S. & Stutchbury, B. J. 1992. Delayed breeding in avian social systems: the role of territory quality and ‘floater’ tactics. Behaviour, 123, 194e219.