Individual differences in activity and habitat selection of juvenile queen conch evaluated using acceleration biologgers

Fine-scale differences in behaviour and habitat use have important ecological implications, but have rarely been examined in marine gastropods. We used tri-axial accelerometer loggers to estimate activity levels and movement patterns of the juvenile queen conch Lobatus gigas (n = 11) in 2 habitat types in Eleuthera, The Bahamas. In 2 manipulations in nearshore areas, queen conchs were equipped with accelerometers and released in adjacent coral rubble or seagrass habitats. Queen conchs were located approximately every 6 h during daylight by snorkeling, to measure individual differences in linear distance moved, and after 24 h they were relocated to an alternate habitat (24 h in each habitat). We found significant inter-individual variability in activity levels, but more consistent levels of activity between the 2 habitat types within individual queen conchs. Four (36%) of the individuals placed in seagrass moved back to the adjacent coral rubble habitat, suggesting selectivity for coral rubble. Individuals showed variable behavioural responses when relocated to the less preferable seagrass habitat, which may be related to differing stress-coping styles. Our results suggest that behavioural variability between individuals may be an important factor driving movement and habitat use in queen conch and, potentially, their susceptibility to human stressors. This study provides evidence of diverse behavioural (activity) patterns and habitat selectivity in a marine gastropod and highlights the utility of accelero meter biologgers for continuously monitoring animal behaviour in the wild.


INTRODUCTION
Animal movement is a central mechanism driving habitat selection, population, and community dynamics (Nathan et al. 2008).It is the result of internal and external stimuli and is commonly aimed at improving foraging opportunities, environmental conditions, avoiding predation, or finding a mate.Movement is limited by the navigational and locomotory capacities of the animal, but even relatively simple invertebrate species are capable of advanced perception of their environment, resulting in complex patterns of movement and a diversity of behaviours (Matthews 1969, Mather & Logue 2013).Animal behaviour can be highly flexible, resulting in high variability between individuals even at the population level, but is often consistent within individuals across time and contexts (i.e.personality; Sih et al. 2004, Réale et al. 2007, Bergmüller 2010).The degree and nature of this variability has important implications for understanding population dynamics, spatial ecology, and how populations will respond to anthropogenic stressors and environmental changes (Wolf & Weissing 2012).Despite the fact that personality provides important insight into ecological and evolutionary dynamics, it has rarely been examined in invertebrate species, many of which are both ecologically and economically important, as well as being good models for behavioural research (Kralj-Fišer & Schuett 2014).
The queen conch Lobatus gigas is a large gastropod found throughout the tropical western Atlantic and Caribbean Seas.Highly popular for human consumption and ornamental purposes, it is a socially and economically valuable fisheries resource (Appeldoorn 1994).However, it is currently listed in Appendix II of the Convention on International Trade in Endangered Species (CITES), and concerns are growing that it will be threatened with extinction due to continued exploitation rates throughout its range (Stoner 2003), though it has not yet been formally evaluated for Red List status by the International Union for the Conservation of Nature.Juvenile queen conchs typically occupy shallow (<10 m) nearshore habitats, including coral rubble, seagrass beds, and sand flats (Stoner & Ray 1993, Glazer & Berg 1994, Stoner 2003) and often form distinct aggregations that move through nursery habitats foraging on macroalgae, diatoms, and detritus (Stoner & Waite 1991, Stoner & Ray 1993).As queen conchs reach sexual maturity (~3.5 to 4 yr; Appeldoorn 1988) they migrate to deeper water habitat (up to 40 m), occupying a diversity of substrates.Queen conchs frequently move between substrate types to improve foraging or reproductive opportunities (Stoner & Sandt 1992).Rates of movement can be surprisingly high (1 to 2+ km yr −1 ), with juveniles generally having smaller home ranges (<1000 m 2 ) than adults (2500 to 5000 m 2 ; Hesse 1979).Like most gastropods, queen conch movement is facilitated by a strong foot used to 'crawl' or 'leap' across the substrate (Berg 1975).
While previous studies have provided coarse estimates of queen conch movement over large temporal scales (e.g.Hesse 1979, Stoner & Sandt 1992), fine-scale activity and movement have never been quantified in the wild.Further, while many studies have examined forms of locomotor activity in gastropods (e.g.Berg 1974, 1975, Perron 1978), individual-level differences in behavioural traits have yet to be quantified in the wild or with respect to environmental factors such as photoperiod or habitat use.Queen conchs occupy a diversity of habi-tats and often exhibit variable preferential habitat use between hard bottoms and seagrass beds across space and time (Randall 1964, Hesse 1979, Stoner & Sandt 1992, Stoner et al. 1996).Given the importance of movement and habitat use for foraging, survival, and reproduction of queen conch, estimates of behaviour in relation to environmental factors are useful for their conservation (Caro 1999).Recent advances in technology have made such measurements possible (Cooke et al. 2004).Accelerometer biologgers collect sub-second measurements of animal movement and posture that enable quantification of fine-scale behaviour in the wild in a range of species (Shepard et al. 2008, Brown et al. 2013, Brownscombe et al. 2014), including molluscs (Robson et al. 2012).Here, our objective was to examine individual juvenile queen conch activity and movement in the wild across 2 commonly occupied habitat types using tri-axial accelerometer biologgers.We predicted that queen conchs would exhibit diverse behavioural patterns between individuals and across habitat types, with consistent behaviours within individuals across space and time (personality).

Study site
This study was conducted in a shallow embayment of the western Atlantic Ocean in Eleuthera, The Bahamas (24.83981°N,76.33651°W).This embayment consisted primarily of seagrass, with adjacent coral rubble substrate 5 to 10 m in width along the shoreline.The seagrass habitat was predominately (65% ground cover) turtle grass Thalassia testudinum in sand substrate.The coral rubble habitat consisted primarily of coral debris, but also contained Batophora algae (30% ground cover).Queen conchs were released 10 m apart from each other in a linear line adjacent to the shoreline at a mean (± SD) depth of 90 cm (± 3.6) in the seagrass and 16 cm (±1.6) in the coral rubble.Within 8 m of release locations, habitat consisted of 988 m 2 (71% of study area) of seagrass and 405 m 2 (29%) of coral rubble.This study was conducted in the wild; hence, external factors such as the density of untagged queen conchs and the presence of their predators could not be controlled.Initial surveys of the study site via snorkeling found queen conchs in the coral rubble habitat (~0.1 to 0.2 m −2 ), but none in the adjacent seagrass habitat.

Behavioural measures: activity and movement patterns
Juvenile queen conchs (n = 12; mean ± SD: 16.7 ± 2.0 cm length, no flared shell lip indicative of maturation; Appeldoorn 1988) were collected from coral rubble habitat and fitted with tri-axial accelerometer loggers (Model X8M-3, 500 mAh battery, 15 g in air, 25 Hz recording frequency; Gulf Coast Data Concepts), which were fastened with electrical tape to the dorsal surface of the body whorl (Fig. 1).Colouration of the tape was unique to allow individual identification during recapture and displacement trials.Capture and release of tagged queen conchs occurred in 2 trials: on 30 January 2013 at 17:00 h and on 4 February 2013 at 15:00 h (n = 6 per release; different individuals for each release).Tagged queen conchs were released into either the coral rubble or adjacent seagrass habitats (n = 3 per habitat per release).Releases oc curred a minimum of 10 m apart from each other, and the location of release was marked with a flag.Researchers returned to the site approximately every 6 h during daylight hours to locate (but not disturb) tagged queen conchs and measure the linear distance moved between observations.Upon location, the marker flag was moved to the new location, and subsequent movement distances were measured from that location.After 24 h, queen conchs were relocated to the alternate habitat.After 48 h, accelerometers were removed, and queen conchs were released back into the coral rubble habitat from which they had initially been collected, but outside of the study area so as to avoid reusing the same individuals in the second trial.One individual released in the coral rubble habitat could not be found at the end of the second trial, resulting in the collection of acceleration data from 11 queen conchs.

Data analysis
Tri-axial accelerometer loggers continuously recorded acceleration (g = 9.8 m s −2 ) at 25 Hz frequency along 3 axes.Total acceleration comprises static (gravity) and dynamic (queen conch movement) acceleration, with maximum values of ± 8 g.Dynamic and static acceleration were separated using a smoothing filter with a 2 s (50 samples) moving window.Dynamic acceleration along the x-axis (representing the strongest dynamic acceleration signal along the longitudinal axis of the animal; Fig. 1) was analyzed using continuous wavelet transformation with a band-pass filter including frequencies of 0.1 to 2 Hz.A k-means clustering algorithm was used to categorize spectra into 6 categories, which were then grouped into either active or inactive behavioural states (Fig. 2).All accelerometer data analyses were conducted with Igor Pro 6.32 software (WaveMetrics) and Ethographer (see Sakamoto et al. 2009).
Queen conch activity levels (% time active) were analyzed with generalized least squares regression.Queen conch identification (ID), habitat type (coral rubble, seagrass), photoperiod (day, night), and the 2-way interactions ID × photoperiod, ID × habitat, and photoperiod × habitat were included as predictors.Backward model selection with single-term de -  et al. 1974, León & Bjorndal 2002).Habitat availability was calculated as the area of each habitat within 8 m of release locations (mean movement distance by queen conchs during this study).All statistical analyses were conducted using RStudio (Version 0.98.501;R Core Team).

Activity
Queen conchs were active an average of 5% (3 min h −1 ) of the time, and were most active in the late afternoon until midnight, with similar levels of activity in coral rubble and seagrass habitats (Fig. 3).Activity levels were highly variable among individuals, with some more active during daylight hours and others at night (Fig. 4).Individual queen conch and the interaction between individual and time of day were significant predictors of queen conch activity (Table 1).Some individuals were  consistent in their activity levels across diel periods and habitats, ranging from highly active (e.g.Conchs 4 and 5) to less active (e.g.Conchs 2, 7, 11).Some queen conchs became highly active in seagrass habitat compared to coral rubble (e.g.Conchs 5 and 9), while others were the opposite (e.g.Conchs 3 and 6).While there was a high diversity of behavioural patterns, queen conchs did not exhibit consistent levels of activity across habitats (Fig. 5A; W = 0.70, χ 2 10 = 13.9, p = 0.18), or across day and night (Fig. 5B) in either seagrass (W = 0.50, χ 2 10 = 9.9, p = 0.45) or coral rubble (W = 0.52, χ 2 10 = 10.6, p = 0.39).

Activity and movement
Queen conchs moved on average 8.7 m (3.9 m h −1 active) in seagrass habitat and 6.6 m (3.1 m h −1 active) in coral rubble over the 24 h periods (Fig. 6).Linear distance moved generally increased with the amount of time queen conchs were active, and there was a nearly significant effect of activity level on linear distance moved (LME; t = 2.1, p = 0.056, df = 13), no effect of habitat type (t = 1.3, p = 0.2, df = 13), but a significant interaction between time active and habitat type (t = 3.1, p = 0.009, df = 13), reflecting a stronger positive relationship between time active and linear distance moved in seagrass habitat than in coral rubble (Fig. 6).Of the queen conchs placed in the seagrass habitat, 4 (36%) moved back to the coral rubble habitat, while no individuals placed in coral rubble moved to seagrass.After the 24 h study periods, there was a significant difference in the number of queen conchs in each habitat type compared to the number expected based on habitat availability (χ 2 1 = 5.3, p = 0.02).

DISCUSSION
Queen conchs exhibited a high diversity of activity and movement levels among individuals, and across habitat types and diel periods in this study.Interindividual variability within a population was common due to genetic, physiological, and experiential factors; however, individual consistency along key behavioural axes (e.g.activity, boldness, aggression) are being documented more and more often both across time and ecological contexts (Sih et al. 2004, Réale et al. 2007, Bell et al. 2009).Indeed, a diversity of behavioural types has been observed not only in vertebrates, but in a few cases also in invertebrates (e.g.Mowles et al. 2012, Mather & Logue 2013).
Locomotory activity is an important behavioural metric that is often correlated with other ecologically relevant traits, including foraging, reproduction, and predator avoidance behaviours (Sih et al. 2004, Dingemanse & Réale 2005).In this study, we did not find statistically significant consistency in activity levels within individual queen conchs across habitat types or diel periods.However, Kendall's concordance values were relatively high, especially when comparing activity levels between habitats, and a lack of statistical significance may be a symptom of small sample sizes.In order to further interpret observed queen conch behaviours, it is important to consider that they exhibited selectivity for coral rubble habitat over seagrass in this study.This is not surprising given that queen conchs were collected from coral rubble, while initial surveys of the study site found no queen conch in the seagrass habitat.In many areas of The Caribbean, including The Bahamas, queen conchs have been primarily associated with seagrass habitat (Hesse 1979, Stoner et al. 1996, Stoner 2003); however, some studies have found selectivity for coral rubble as well, areas which often represent important foraging grounds (Randall 1964, Stoner & Sandt 1992).Here, queen conchs also moved greater distances when placed in seagrass than in coral rubble, and there was a greater positive relationship between activity levels and distance moved.This would suggest that either (1) queen conchs were moving more linearly in seagrass, presumably in search of more preferred coral rubble habitat, or (2) movement was more efficient in seagrass than in coral rubble.Coral rubble is more rugose than seagrass, but dense seagrass would also likely pose challenges for movement, so it is unclear which of the above explanations, or a combination of both, accounts for this pattern.Future research on the energetics of movement through diverse habitat types could provide insight into the space use of this species, especially in the context of the 'energy landscapes' framework (Wilson et al. 2012).
A lack of strong evidence for within-individual consistency in this study does not preclude the existence of behavioural types in this species, especially given the short time period of this study and the chal-lenges posed to these animals.For example, displacing queen conchs to the less preferable seagrass habitat triggered a range of behavioural responsessome individuals became more active, many of which were able to return to the coral rubble (~10 m away) within 24 h of displacement, while others responded by becoming relatively inactive (Fig. 4).This could reflect proactive and reactive coping styles (Koolhaas et al. 1999).Proactive individuals tend to be bolder, recover from stress more quickly (i.e. from handling), and explore new environments more readily than reactive individuals (Sih et al. 2004).Similarly, coping styles could influence the response of conch populations to environmental changes or anthropogenic stressors.Highly mobile individuals are more likely to be dispersers and thus play key roles in the colonization of new habitats.These individuals might also be more resilient to environmental change, such as habitat loss and degradation.However, prolonged inconsistent environmental conditions might favour more reactive individuals in the long term (Sih et al. 2004).
Understanding the potential role of behavioural types in queen conch population dynamics and habitat use is an important avenue for future research.For example, queen conchs form distinct foraging aggregations that move through a diversity of substrates, yet Stoner & Ray (1993) found that they only occupy a small proportion of seemingly suitable habitats.Individual differences in activity might reflect an underlying behavioural mechanism tied to differences in spatial distributions, which are highly dynamic in this species.More active individuals may disperse further and exploit the higher resource availability outside the population core, but, in doing so, they may also experience higher predation risk (Stoner & Ray 1993).With tradeoffs between higher food availability and higher predation risk at aggregation edges, behavioural types could be tightly linked with the diversity of the life-history characteristics observed in this species (Appeldoorn 1988).
With increasing harvest pressure, reductions in habitat quality and quantity, and declining population numbers, the interest in queen conch population, habitat, and harvest management is increasing (Stoner 2003, Delgado et al. 2004).Here we have demonstrated that queen conchs exhibit a diversity of behavioural patterns, which may have important implications for conservation efforts.For example, Delgado et al. (2004) explored the potential for relocating queen conchs to more suitable habitats when faced with habitat degradation.Behavioural responses could influence the success of such initiatives in the long term.For example, more active, bolder individuals could respond better to these types of displacement and might fare better when colonizing new areas.
Queen conchs exhibited significant inter-individual variability in activity levels across habitat types and diel periods.Further, queen conchs showed selectivity for coral rubble habitat (from which they were collected) during this study; relocating them to less preferable seagrass habitat revealed a diversity of behavioural responses which might reflect individual coping styles.This diversity may have important implications for population dynamics and spatial ecology, particularly in the face of increasing anthropogenic and environmental stressors.Given the high variability in behavioural patterns observed in queen conchs in this study, future research should aim at examining the potential of behavioural types in this species, which may have important implications for ecological and evolutionary processes, as well as in situ and ex situ conservation efforts.

Fig. 3 .
Fig. 3. Time active (% of total observation time) for queen conchs Lobatus gigas fitted with a Loess smoother by hour of the day in seagrass and coral rubble habitats

Fig. 5 .
Fig. 5. (A) Relationship between queen conch (Lobatus gigas) activity in coral rubble and seagrass habitats as a % of total time observed.(B) Relationship between queen conch activity during the day and night in coral rubble (n) and seagrass (d) habitats as a % of total time observed