A role for lakes in revealing the nature of animal movement using high dimensional telemetry systems - IFishMan
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Lennox et al. Movement Ecology (2021) 9:40
https://doi.org/10.1186/s40462-021-00244-y
REVIEW Open Access
A role for lakes in revealing the nature of
animal movement using high dimensional
telemetry systems
Robert J. Lennox1* , Samuel Westrelin2, Allan T. Souza3, Marek Šmejkal3, Milan Říha3, Marie Prchalová3,
Ran Nathan4, Barbara Koeck5, Shaun Killen5, Ivan Jarić3,6, Karl Gjelland7, Jack Hollins5,8, Gustav Hellstrom9,
Henry Hansen10, Steven J. Cooke11, David Boukal6,12, Jill L. Brooks11, Tomas Brodin9, Henrik Baktoft13,
Timo Adam14 and Robert Arlinghaus10,15,16
Abstract
Movement ecology is increasingly relying on experimental approaches and hypothesis testing to reveal how, when,
where, why, and which animals move. Movement of megafauna is inherently interesting but many of the
fundamental questions of movement ecology can be efficiently tested in study systems with high degrees of
control. Lakes can be seen as microcosms for studying ecological processes and the use of high-resolution
positioning systems to triangulate exact coordinates of fish, along with sensors that relay information about depth,
temperature, acceleration, predation, and more, can be used to answer some of movement ecology’s most pressing
questions. We describe how key questions in animal movement have been approached and how experiments can
be designed to gather information about movement processes to answer questions about the physiological,
genetic, and environmental drivers of movement using lakes. We submit that whole lake telemetry studies have a
key role to play not only in movement ecology but more broadly in biology as key scientific arenas for knowledge
advancement. New hardware for tracking aquatic animals and statistical tools for understanding the processes
underlying detection data will continue to advance the potential for revealing the paradigms that govern
movement and biological phenomena not just within lakes but in other realms spanning lands and oceans.
Keywords: Telemetry, Sensor, Biologging, Movement ecology, Fish ecology
Introduction when, why, and how? These are foundational eco-
Animals are born, they move and reproduce, and then logical questions and the answers have significant im-
they die. This simple model of life supports all ecological plications for our understanding of the natural world and
processes and movement has therefore emerged as a the management of resources that we depend upon [172,
frontier for animal research [131, 145, 200]. Movement 200, 210].
ecology is a multiscale branch of ecology operating from Significant and rapid advances have been made in our
cells to whole animals, populations, and communities understanding of movement ecology coincident with the
across short or long distances for brief intervals or introduction and proliferation of electronic tags to remotely
even spanning generations. Where do animals move, measure animal behaviour and physiology [131, 145]. The
capacity to simultaneously monitor movement and the
* Correspondence: robertlennox9@gmail.com environment yields great opportunity but also significant
1
Laboratory for Freshwater Ecology and Inland Fisheries (LFI) at NORCE
Norwegian Research Centre, Nygårdsporten 112, 5008 Bergen, Norway
responsibility to identify focal systems with which to make
Full list of author information is available at the end of the article inferences [117]. To this end, Hays et al. [117] presented a
© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data.
Lennox et al. Movement Ecology (2021) 9:40 Page 2 of 28
list of research priorities related to megafaunal movement, acoustically instrumented lake environments to discuss
specific to a system where research is inherently challenging the vast opportunities these systems have to address 15
and limited by the vast scale of latitudinal and longitudinal movement ecology questions identified by Hays et al.
connectivity coupled with profound depths: the marine [117] that we agree will drive the movement ecology
environment. This daring focus renders many studies, field forward in coming years. Each section is divided
particularly those that concentrate on community into three paragraphs in which we first describe key ex-
scales and consider interactions among species, logistic- amples and potential connections, followed by questions
ally challenging. using lakes as focal systems that could advance under-
Lakes are ideal study systems for testing ecological standing, and finally the approaches that could accom-
paradigms, including for movement ecology. For over a plish this. We conclude this essay with a synthesis where
century, lakes have been acknowledged for providing we discuss the tools and approaches that we envision re-
ample opportunities to investigate ecological, behavioral searchers applying to better understand the complexities
and evolutionary questions at manageable scales [86]. of aquatic life for better habitat management, ecosystem
Lakes are highly important venues for studying ecology conservation, and fundamental science.
because freshwater habitats are among Earth’s most
valuable, rare, and threatened ecosystems [79, 240]. As How can movement data be used to support
relatively closed ecosystems with less influence from dis- conservation and management?
tant processes [192], animal movement can be linked Aquatic biodiversity is in steep decline due to a range of
more directly to local phenomena, including weather anthropogenic factors, including habitat alterations
patterns and the immediate ecological community. Lakes [240]. There are also increasing examples of overfishing
offer a great diversity of structural and physical pro- of freshwater stocks [232] and of other exploitation-
cesses with similarity at local scales but substantial vari- induced issues [10, 167]. Movement data are key, yet
ation in fish assemblages and aquatic communities underutilized to design effective conservation and man-
across latitudes and longitudes. Small lakes can effect- agement strategies, e.g., in the context of fisheries and
ively be covered by an array of acoustic receivers in a conservation of freshwater fish and freshwater habitats
comparable design to a bay or coastal area in the ocean [14, 71]. Lake tracking data can be used to identify
or a great lake but with higher resolution of the processes seasonal and daily movements, dispersal, connectivity of
operating within. Replication of studies in multiple lakes habitats [115, 198], e.g., after stocking [193], behavioural
offers the potential for robust inferences from ecological diversification and its relation to individual fitness [150],
and manipulative experiments [50, 255], including how capture probability [193], spawning site fidelity [149],
environmental stressors and ecological interactions mod- stock boundaries among connected ecosystems and
ify movement behaviour. For these reasons, lakes have within ecosystems [67, 116], reactions to human influ-
long provided essential venues for ecological inquiry and ences, such as boat movement [135] or catch-and-
many paradigms have emerged from the flexibility, release [15], and degree of fishing-induced mortality
observability, and replicability of research in lakes, includ- [120]. An obvious further application example from a
ing ecological regime shifts [253], predation risk effects conservation context is applying telemetry to examine
[304], predator-prey-habitat complexity relationships [95], the ability of freshwater protected areas to help heavily
trophic positioning from stable isotopes [293], habitat deg- exploited fishes recover from heavy fishing pressure [236].
radation [256], and ecological speciation [258, 262]. In this context, telemetry is useful to identify sites where
encounters with fishing gears are rare.
Lakes as venues for movement ecology research Despite the opportunities, there are limited examples
We submit that lakes provide perfect venues in which to of fine-scale, whole-lake tracking studies that have real-
investigate many of the most fundamental questions of ized the potential of informed management and conser-
movement ecology with results that are scalable to larger vation. The few systems that were or are in place have
systems. To that end, we turn to the key questions of generated a number of highly relevant results. Baktoft
marine megafaunal movement ecology presented by et al. [15] used whole lake telemetry to assess the reac-
Hays et al. [117] and suggest that many of these ques- tions of northern pike (Esox lucius) to handling, includ-
tions can also be applied to whole lake studies. We ing catch-and-release. Jacobsen et al. [135] studied the
interpret these questions as relevant across mobile taxa response of different freshwater fish to boating, revealing
and not limited to the marine environment or to mega- limited impacts on the behaviour of freshwater fish.
fauna specifically. We posit that answering these ques- O’Connor et al. [214] showed that a one-time intensive
tions will yield significant advances in our understanding stressor can have carry-over effects many months later
of movement ecology independent of the system. Our during hypoxia in largemouth bass (Micropterus sal-
approach is to draw on our experiences working in moides). Work in a small lake in Germany has revealed
Lennox et al. Movement Ecology (2021) 9:40 Page 3 of 28 how angling can directly select on behavioural traits, Before-after-control-impact studies are a gold-standard such as habitat choice in perch (Perca fluviatilis) [194]. in the applied environmental sciences, particularly in Similar research has been conducted in “lake-like” coastal freshwater ecology, and are particularly useful to identify systems where small-bodied coastal fish with limited how common conservation and management actions home range were exposed to angling, revealing how an- operate at ecologically realistic scales. Lakes offer excellent gling could be a selective force on home range, activity, experimental arenas for such types of studies. Experiments and chronotypes [7]. A ground baiting experiment at a could, for example, tackle questions of habitat enhance- whole lake scale showed how omnivorous fish respond to ment or degradation, stocking and introductions, selective angler-induced bait and how this novel energy is embed- harvesting and effectiveness of protected areas. Smaller ded in certain trophic levels elevating secondary produc- pond ecosystems could also be experimentally warmed tion [187]. Fine-scale acoustic telemetry has also been to study impacts of climate change. Replicated lakes used to study restoration success in Toronto, Canada could be used to study impacts of invasive species, the [297] and how exposure to pollutants affects the behav- release of chemicals, light pollution, and exploitation iour of Eurasian perch in the wild [148]. pressures. Stock assessment methods could be calibrated Compared to the oceans, spatially finite ecosystems and gear biases and estimation of catchability could be such as ponds or lakes can offer replication and allow quantified in situ using telemetry. Indeed, whole lake tel- whole-ecosystem type experiments to be conducted with emetry constitutes an excellent opportunity to estimate appropriate replicates (either in space or time) and with the otherwise “unmeasurable” (Fig. 2), such as size- controls (e.g. manipulated vs. unmanipulated; Fig. 1). dependent mortality, predator-prey interactions (e.g. Fig. 1 Lakes come in many shapes and sizes, all of which have the potential to be monitored using environmental sensors and telemetry to reveal the nature of animal movement. In this grid we show lake size scales between small (left half) and large (right half) and experiments can be conducted in isolation (a single lake, lower half) or in a replicated design (upper half). Finding matching lakes to replicate experiments allows a degree of control that is difficult or impossible to achieve in other systems. Moreover, scaling lakes from small to large allows a degree of environmental realism desired for the experiments, with animals in small lakes using all habitats but in large lakes habitat segregation and different competitive mechanisms emerging
Lennox et al. Movement Ecology (2021) 9:40 Page 4 of 28
Raw positions Home range
Receivers - green dots • Space use
• Species
Tench - blue dots spatiotemporal
(5 843 locations) overlap
Wels - red dots • Predator-prey
(5 971 locations) interactions
• Competition
Period: • Lanscape ecology
7- 15 July 2015 (fear, energy)
• Protected areas
design
Trajectories Activity
(one day only) (one day only)
• Navigation • Circadial rhytms
abilities • Energy budget
• Movement • inter/intra-
pattern specific activity
• Movement rules interactions
• Memory
• Direct individual
interactions
Temperature Habitat
(one day only) • Habitat
• Temprature preferences
preferences • Habitat partioning
• Metabolism • Space use
regulation • Predator-prey
• Energy budget interactions
• Glogal climate • Competition
changes • Ecosystem role
Fig. 2 Acoustic telemetry yield data on the instrumented animal’s positions, path, space use, activity levels, temperature use, and habitat
selection in up to four dimensions. Here, we illustrate how detections on a grid of acoustic receivers can be used to investigate patterns in the
behaviour and physiology of free living fish to describe where, when, how, and why animals are moving. Together, lake telemetry studies are
powerful tools for inquiry about processes and patterns in ecology
after stocking of piscivores), or ecosystem reactions to (e.g. aeration), or alternatively, how the movements of fish
changes in fish populations (e.g. invasions). In this context, affect turbidity and water quality. Telemetry may also in-
the success of common restoration measures, such as bio- form eradication of pest species, should this be desired [14].
manipulation [188], depends on risk-sensitive foraging [3],
which in the past was indirectly inferred from the capture Are there simple rules underlying seemingly complex
of fish in gill nets and other gears or was simply inferred movement patterns and, hence, common drivers for
from prey responses to introduced predators. Telemetry movement across species?
could be used to directly measure how zooplanktivorous Common rules underlying seemingly complex movement
fish respond to stocking of predators, to the removal of fish, patterns have been identified in a number of aquatic
to fish-eating birds or otters, or to technological measures animals, including seabirds, sharks, turtles [117, 269], andLennox et al. Movement Ecology (2021) 9:40 Page 5 of 28 freshwater fish [177]. It has been extensively studied how, models (HMMs; [159, 181, 222]), general state-space where, and when individuals move, from which the fol- models (SSMs; [11, 141, 221]), and diffusion processes lowing common drivers for movement have been sug- (e.g. Ornstein-Uhlenbeck position models or stochastic gested: optimal foraging, site fidelity and revisitation, and differential equations; cf. [222] for an overview of the temporal patterning. For the first, optimal foraging, Lévy available methods). Fueled by increasingly large and walks [269, 299], Brownian motions [129], or similarly complex telemetry data sets, several methodological ex- simple random walk-type models have been proposed as a tensions towards a more unified picture of movement simple evolutionary trait that has been adopted by many (cf. [200]) have recently been proposed. For example, species when searching for sparsely distributed prey. In hierarchical HMMs provide a versatile framework for recent years, however, this randomness paradigm [200] jointly inferring movement patterns at multiple time was the subject of controversial discussions (cf. [28, 234]). scales (e.g. fine-scale variation in activity vs. coarse-scale In fact, conclusive evidence for the Lévy walk and related migration patterns; [1, 166]), energy budgets and recharge hypotheses is still lacking, and it is now regarded as overly dynamics have been explicitly incorporated into individual- simplistic. This perspective has catalyzed a shift towards level movement models [125], and group dynamics have explaining specific movement paths rather than move- been modeled by relating individuals’ movement decisions ment behavior in general [222]. For site fidelity and revisi- to herd-level movement patterns [160, 205]. tation patterns, home range or homing affinities have Testing comprehensive models of animal movement in been identified in various freshwater fishes, for which lar- which movement is assumed to be generated by many ger individuals were found to generally have larger home different factors interacting with each other, against ranges [177, 315]. Yet, simple random walk-type models simple null models such as (truncated) Lévy walks, such as (truncated) Lévy walks or Brownian motions are Brownian motions, or related random walk-type models, generally inadequate to resolve the patterns [126]. In may provide a promising avenue for confirming (or addition to an individual’s size, the shape of the water rejecting) simple rules that have been suggested in the body was suggested to affect movement [315], emphasiz- past. This approach can also test the validity of patterns ing how environmental conditions can be regarded as a and rules discovered with state-of-art laboratory tracking common driver for movement. For temporal patterns, diel techniques of aquatic invertebrates (e.g. [59]) for fishes variation as well as daily and seasonal movement patterns, in the wild. In addition, the unprecedented opportunities particularly regarding the times of feeding, breeding, offered by high-resolution, three-dimensional lake fish aggregating, and resting behavior, have been found in telemetry - most notably the possibility to observe an numerous aquatic species [118]. The majority of fresh- individual’s movement throughout an entire ecosystem water fish tend to be predominantly diurnal [17, 52, 58], at fine temporal resolution while being able to control although marine top predators tend to be more nocturnal for multiple variables (Fig. 2) that can affect its behavior [121]. Time is linked to both temperature and photo- in replicated designs, may help to identify new common period, which influence the individual’s physiology and drivers for movement across species. motivation for movement. Temperature, for example, has been shown to control activity timing in juvenile salmon How do learning and memory versus innate behaviours [87]. Time and photoperiod can be regarded as a common influence movement patterns, including ontogenetic driver for movement, either affecting movement directly changes? or indirectly by affecting the prey’s behavior, which is then Animals moving in their natural environments are typic- adopted by its predator. ally exposed to a variety of factors and conditions that Movement is often assumed to be the result of a single span from highly beneficial (e.g. food or mates) to highly paradigm that neglects its complex nature. An alterna- detrimental (e.g. toxic items or predators). The ability of tive, more comprehensive perspective on movement ad- animals to optimize fitness gain by adjusting their move- dresses the animal’s internal state (“why does an animal ment in response to complexities depends on both in- move?”), its motion (“how does it move?”), and naviga- nate and learned skills that enable animals to perceive, tion (“when and where does it move?”) capacities, and respond, learn, and remember the structure and dynam- external factors all interact to generate movement [200]. ics of such factors in their environment. Studies of Lakes provide a nearly ideal environment to collect de- animal cognition have yielded numerous insights into tailed data that inform complex statistical models and the mechanisms affecting spatial learning and memory more comprehensive pictures of an animal’s behavior. in various taxa [227, 270] and fish in particular [30, To fully exploit the complex detection data, powerful 40, 78, 142, 146, 158, 215, 298, 303]. These insights statistical methods are needed. Popular models for infer- divulged the role of ontogenetic and cognitive pro- ring behavioral patterns from high-resolution bio- cesses in shaping movement patterns and their fitness logging data include discrete-time hidden Markov consequences, stressing the critical role of learning from
Lennox et al. Movement Ecology (2021) 9:40 Page 6 of 28 experience during early life. Across species, details were advanced by implementing high-throughput field tel- revealed mostly from controlled laboratory experiments emetry approaches. on captive animals [30, 215, 270], whereas field studies Understanding how early-life processes shape animal have been much less frequent, and studies based on move- movement and behavior through learning and memory ment data collected from free-ranging animals in the wild is also important for managing populations, for example have been scarce and focused on terrestrial systems (e.g. of fishes in lakes and rivers. Better understanding of [108, 218, 219, 267, 285]). these processes can guide the development of infrastruc- Studies of fish in their natural environment have ture to facilitate fish migration and survival in light of yielded important insights in the ontogeny of spatial anthropogenic disturbances such as river dams [100], or learning and memory. Whereas much of the literature by enriching the relevant early-life environment of has come from marine species, there is great opportunity captive-reared fish [142]. Studies examining fish re- to use lakes as a study system to test and advance move- sponse to capture by hooks can also largely benefit from ment ecology paradigms. Such studies have shown, for high-resolution fish tracking. For example, movements example, that the remarkable homing ability of adult of both fish and fisher might be tracked rather exhaust- salmon depends on long-term olfactory memory of their ively in a closed lake system, to accurately estimate the natal streams learned during early stages of life [113, probability of captures and encounters and to elucidate 259]. Although the basic formulation of this salmon the factors affecting these probabilities [7, 164, 193]. olfactory imprinting hypothesis received further support More generally, high-throughput wildlife tracking sys- from later studies and has been broadly accepted, some tems such as acoustic telemetry in lakes can unravel important details remain controversial [235]. For some of the most basic relationships between animal example, does olfactory imprinting occur exclusively in a cognition/memory and movement (Fig. 2). This has re- limited time (the smolt stage) or at specific sites [259], cently been shown through the use of ATLAS, a new or as a learned sequence of odors acquired during differ- reverse-GPS tracking system that is principally very ent early-life stages at different times and sites [114]? similar to acoustic lake telemetry, to reveal the first field Furthermore, fish might learn other cues and in a more evidence for a cognitive map and spatial memory of complex manner. For example, juvenile reef fish multiple specific targets by free-ranging animals within responded to cues sensed through different mechanisms their large (100 km2) natural foraging area (Toledo et al. (olfaction, hearing and vision) at different sites experi- [285]). Furthermore, such tracking projects can be enced during their early-life movements [128]. Tracking coupled with methods providing complementary infor- fish movements throughout their life cycle, and espe- mation on behavioral, physiological and environmental cially during early stages of life, offers a unique oppor- changes, as well as experimental manipulations of learn- tunity to tackle such complexities. Earlier studies of fish ing and memory by altering landmarks, fishing habits movement mechanisms have used boats to follow indi- (e.g. bait type), sensory cues, and the presence of in- vidual fish marked by a tethered float [111], ultrasonic formed vs. naïve fish. [112], or radio [14] tags, resulting in relatively limited datasets of few individuals tracked at low frequency and To what degree do social interactions influence for short durations. Although these studies made some movements? important propositions – that wild white bass (Morone The study of animal social behaviour is fundamental to chrysops) can swim directly homeward in open water our understanding of behavioural, physiological, and presumably by using a sun compass [111] and other cues evolutionary ecology. Group living is key for predator- [112], and that wild carps can quickly learn and remem- avoidance, foraging, and reproduction in most animal ber the location of new food resources [14] – more taxa. This directly affects organismal fitness but also conclusive insights and more in-depth investigation of modulates the outcome of numerous life-history and the mechanisms underlying the observed tracks were evolutionary trade-offs. There is also increasing evidence still rather limited. This powerful research system has that sociality plays a key role in the maintenance or just started to be applied to study topics related to erosion of within-species phenotypic variation in behav- ontogeny of spatial learning and memory. Topics ioural and physiological traits [140]. Fish display numer- strongly related to ontogeny, movement, and spatial ous forms of complex social behaviours including social learning and memory, such as personality traits [2], networks, dominance hierarchies, social learning, and cognitive flexibility, and inter-individual variation in coordinated group movements with leader-follower space use [174], time-place associations [241], land- dynamics (Fig. 3) As such, fish are often used as models mark use [303], and various other orientation and to study animal social behaviour and form the basis of a navigation mechanisms [31], have been predominantly large proportion of our knowledge about emergent studied in the laboratory, and now can be critically group behaviours. Notably, however, most of this
Lennox et al. Movement Ecology (2021) 9:40 Page 7 of 28 Fig. 3 Data from acoustic telemetry will greatly enhance our analysis of social and collective behaviour in fish, as well as allow new forms of analysis that have previously been impossible in the wild. The analysis of leader-follower dynamics, social networks, and group cohesion can now be performed at much greater temporal and spatial scales using telemetry data. This will allow study of how these social factors affect ecological phenomena including group foraging, migrations, and predator avoidance, and how changing environments further modulate these effects. Telemetry data will revolutionize the study of the interactions between habitat use (e.g. in response to physical structure or factors such as temperature of oxygen availability) and passive and active assortment of phenotypes among and within groups. In addition, an opportunity now exists to examine among-group variation in space use, territoriality, and changes in social group membership, with possible effects on individual fitness research with fish has been done in the laboratory, behaviour in the wild [147]. Our knowledge of how mainly because of the extreme difficulty associated fish social groups function in the wild, and how they with long-term measurements of individual fish are affected by environmental conditions, has
Lennox et al. Movement Ecology (2021) 9:40 Page 8 of 28 therefore been hindered by this basic constraint in fish are being influenced by their social environment, we our research capabilities [103]. must have data for all or at least the vast majority of fish Tools are now available to begin addressing detailed within a natural system. This is extremely difficult questions of social interactions and animal movement. because in most cases it will be impossible to know if all Lab-based observations of fish social behaviour can be fish within a system have been captured and tagged. A realized with sophisticated software for automatically possible solution may be the removal of most fish, tracking the trajectories of multiple individuals from followed by stocking with a known number of tagged recorded video [226, 245]. These data are a series of x-y individuals, or the use of dedicated fishless lakes or coordinates for each individual within a group that can artificial ponds. An additional challenge will be the de- in turn be used to quantify: 1) group-level metrics, velopment of a statistical and analytical framework for including group cohesion and polarity; 2) the behaviour studying the desired social behaviours and emergent of individuals within groups such as individual speed, phenomena. To be most useful for social analyses, tel- alignment, spatial positioning, distance from group emetry data must have a high spatial and temporal reso- mates, and social network position; and 3) the propaga- lution and low error. Lab-based work can provide tion of changes in movement metrics throughout social precise positions of individual fish dozens of times per groups. The spatial and temporal resolution of telemetry second [226, 245]. This is not possible with even the systems in the field can now advance basic forms of most advanced forms of acoustic telemetry, and so we these analyses on freely roaming fish in their natural will need to work back to uncover the minimal adequate habitats, with the coordinates of individual detections spatial and temporal resolutions needed for basic ana- being analogous to the x-y coordinates captured by lyses of individual interactions, spatial positioning within automated software in laboratory behavioural arenas. groups, and group fission-fusion processes. Enhanced Analysis of movement propagation in the lab is used to resolution also greatly increases the required computing inform leader-follower dynamics in fish social groups power and analyses time, and so it might initially only ([144]; Fig. 3), and in the wild could provide information be possible to perform the most sophisticated analyses on migrations and other phenomena related to collective on subsets of data. movement [25, 306]. Telemetry data are currently being used to infer differences in individual space use and habitat preferences within species [83, 199], but it is How does the distribution of prey impact movement? highly likely that these are also affected by social dynam- Prey distribution and availability can highly alter the be- ics in ways that we are yet to understand but that will haviour and movement of predators [19]. Initially, preda- now be possible. Increased knowledge of fish social sys- tion concepts focused on the optimality of foraging tems will also provide knowledge on how group move- behaviour, i.e. maximization of the rate of energy intake, ment and behaviour affect individual vulnerability to in relation to prey density and distribution [175]. Later, different fishing methods [122, 287]. Perhaps most im- predation risk [305], competition among conspecifics portantly, increased knowledge of fish social systems in [88], effects of environmental abiotic factors [4], level of the wild will help us understand their responses to nat- individual [290], or individual state [185] were intro- ural and human-associated changes in environmental duced into the models explaining the effects of prey on factors such as temperature, oxygen availability, turbid- predator distribution. These concepts mostly targeted ul- ity, and food availability. A promising opportunity also timate causes of predator-prey distribution interactions exists to combine telemetry movement data with other and their effect on life history traits and fitness of both forms of logged or transmitted data from individual fish predators and prey [96]. Concepts such as optimal for- (e.g. heart rate data, temperature) to carefully dissect the aging, game theory and ideal free distribution further interplay among animal movements, their social envir- considered that individuals tend to optimize their onment, their physiological state, and the external envir- foraging strategies based on all relevant environmental onment [65, 66]. It will also be possible to combine all factors (such as the amount of prey, predation risk, and of this information with established theoretical move- number of conspecifics) and internal physiological state, ment models from lab-based work to more fully under- and chose the behaviour that maximizes individual fit- stand fish social dynamics, emergent group behaviours, ness and future reproduction [96]. However, it is now and then predict their responses in the wild and empir- widely recognized that wild animals are limited by in- ically test these predictions. complete information and imperfect ability to analyze Despite these exciting opportunities, there remain information and foresee consequences of alternative be- many challenges that must be addressed before we can havioural options [9]. Consequently, recent research has fully take advantage of acoustic telemetry in the study of shifted more towards individual level and proximate fish social behaviour. In order to fully understand how causes of predator-prey distribution interactions.
Lennox et al. Movement Ecology (2021) 9:40 Page 9 of 28 Much attention has been given to the role of different statistical techniques to identify and analyze patterns in decision-making processes for involving individual deci- multidimensional big data will help understand predator- sions and their regulation into foraging behavior [9, prey interactions in great detail. Current technology of 200]. Current thinking frames individuals as units de- high-resolution tracking dramatically expands our abilities termined by various properties (individual genotype, to uncover predator-prey spatiotemporal overlay and use physiological state, age, or size) and moving in space it to infer their direct and indirect interactions (Fig. 2). defined by multilevel landscapes of, e.g., fear [162] or Such sampling can be accompanied with measurements of energy [246]. Individual space use then depends on suitable individual traits before or after tracking and use the overlay of these landscape ‘bricks’ (e.g., infrequent these traits in possible proximate or ultimate explanations use of locations with rich food and high predation of their behavioural strategies and predator-prey inter- risk) and actual cognitive and physiological state of action strengths. The main limitation for such studies cur- an individual [89, 101]. Both individual state and rently seems to be the need for a carefully planned landscape topography are affected by environmental protocol with a large number of tracked fish to obtain ro- factors (e.g. temperature and light in aquatic environ- bust patterns. ments) and change dynamically in time [89]. Yet, many important questions are poorly understood in What sensory information do animals use to sense prey, these fields and high-resolution movement data can breeding partners, and environmental conditions? be a key component in their understanding: e.g. The sensory perception of the abiotic and biotic envir- proper matching of the landscapes of fear and energy onment is the basic input for fish behaviour. Fish may with resulting movement trajectory [89]; effects of use a wide array of senses (gustation, olfaction, vision, prey availability on predator behaviour under different lateral line, hearing, magnetoreception, and electrorecep- environmental contexts [47]; predator-prey personality tion) for orientation in the environment and one or mul- interactions in forage/escape behaviour; mismatch in tiple senses may be used as a basis for their behavioural the timing of predator-prey activity peaks [8]; tem- decisions. To disentangle which sensory system is used poral individual variation in the forage/hide behav- for assessing particular situations, experimental designs iour, the role of individual traits in ontogenetic shifts using sensory blocking, nerve suppression, nerve in space and resource use [197, 242]; and the causes transection or ablation experiments are frequently used and triggers of diel vertical and horizontal migrations [201, 213, 229]. Sensory ecology of aquatic organisms is [186, 242, 250]. predominantly studied under controlled laboratory con- Even relatively large water bodies up to several hun- ditions [196, 279]. Due to the ability to precisely track dreds of hectares can be fully covered by positioning sys- animal movement beyond laboratory environments, tems [20, 307] to provide fine-scale positioning of both novel research designs using 3D telemetry technologies predator and prey over long periods of time that can be have the potential to shed light on many research topics used to answer a variety of questions related to dealing with sensory perception important to predator- predator-prey interactions. For example, Jacobsen et al. prey interactions, communication among conspecifics, [134] identified alternative foraging strategies in acous- and animal orientation within the visually limited space tically tagged Eurasian perch in mesotrophic and hyper- of aquatic environments. Such a design was imple- eutrophic conditions. In a long-term movement study, mented to discriminate among visual, magnetic and ol- Nakayama et al. [198] found distinct diel horizontal mi- factory navigation to natal stream in sockeye and masu gration of Eurasian perch likely related to foraging op- salmons Oncorhynchus nerka and O. masou [291]. portunities. Baktoft et al. [16] used tagged Eurasian The ability to precisely track individual fish opens new perch to quantify the links between metabolic rate and opportunities to test hypotheses and validate laboratory activity patterns. Kobler et al. [150] studied behavioural findings linking sensory information to individual behav- types of pike using radio-telemetry in a lake and found iour (e.g., [59]) in ponds and lakes. Two approaches can distinct differences in habitat use and activity levels, be used for experimental study designs in lakes: which they related to an ideal free distribution pattern. manipulation with the environment and manipulation Madenjian et al. [176] demonstrated a positive effect of with fish physiology and sensory ability. Using multiple food availability on consumption rate in walleye Sander small lakes or ponds (or dividing them with curtains) and vitreus. In the same species, Raby et al. [237] concluded manipulating variables (e.g. turbidity, anthropogenic noise, that drivers such as temperature and food availability in- pH or light pollution) may help disentangle the effect of fluence migratory behaviour. All these studies show the tested variables on the fish behaviour and fitness [263]. high potential of telemetry in studying predator and prey Study design may alternatively involve manipulation with space use and their spatial interactions. We believe that fish physiology by using slow-release implants and the development of high-resolution telemetry and comparing it to non-manipulated individuals [180, 182].
Lennox et al. Movement Ecology (2021) 9:40 Page 10 of 28
Finally, experimental designs using sensory blocking, nerve out in lakes to reveal relationships among lake morphology,
suppression, nerve transection or ablation experiments may productivity, and fish biomass (e.g. [51, 265]) and with tel-
help determine which sense provides critical input for the emetry tools we have the capacity to expand this knowledge
observed behaviour. Novel approaches using depth, with finer-scale details of the functional roles that fish have
temperature, acceleration, predation, or metabolism-level in these systems and the feedbacks between consumers and
sensors may be integrated in the study design, thereby en- producers in the ecosystem. Throughout the field of ecol-
abling a wider interpretation of the data [5, 106, 238]. ogy, there is broad interest in understanding how roles are
Study designs using 3D telemetry to differentiate among partitioned among species in an ecosystem, and how the
senses used for observed behaviour would require careful system responds under stress such as when challenged by
study design using one of the above-mentioned options. invasive species, climate change, or pollution. Understand-
As an example of such an experiment in semi-wild condi- ing roles and identifying pathways through which ecosys-
tions, disabling a selected sensory input in selected prey tem services are generated is therefore a key question to
individuals and comparing them to controls may help dis- ecology, albeit one that has been afforded less consideration
entangle the role of sensory information in predator in the context of movement ecology [117]. In lakes, prod-
avoidance and quantify the role of each sensory input. uctivity scales with the perimeter/area ratio, suggesting that
Manipulation of the sensory ability of predators can be small lakes, rather than great lakes or seas, would be ideal
used to discriminate which senses are important in which venues for investigating habitat coupling and ecological
part of the predator-prey cycle [201, 229]. Uncertainty in roles with replicated whole lake experiments including ma-
the data interpretation may be further minimized by mon- nipulations of the fish assemblage and experimental alter-
itoring all potential prey and predator individuals. Given ations of lake productivity [257, 307].
the cost limitations, preference should be given to simple Whole-lake studies have contributed in substantial
systems with limited predator-prey species interactions to ways to our understanding of energy landscapes and
enable thorough interpretation of the results and to ecosystem services. Predation and competition are the
minimize the risk of study failure [170]. While we argue key biotic processes that structure lake fish communities
above that purely behavioural studies would benefit from and manipulative experiments in lakes have illuminated
as many tagged fish as possible, we partly take the oppos- how these processes operate [133]. Replicated whole-
ite stance here because experiments targeting sensory in- lake experiments have been conducted by modifying the
formation are potentially of an invasive nature. Such fish community and observing changes in abundance
experiments should be planned carefully to minimize the and growth to reveal mechanisms that structure assem-
number of individuals used for the study and maximize blages (e.g. [46, 51]). However, existing studies have
their welfare [38, 247]. Therefore, the questions should be lacked the resolution to observe competition and preda-
addressed primarily using non-invasive methods such as tion in situ. Manipulative experiments in whole lakes
environmental manipulation or temporary sensory sup- provide ideal templates for research on ecosystem roles
pression by chemical treatment [201]. Joint efforts of when coupled with tools that allow direct inference of
physiologists and behavioral ecologists respecting these material and energy cycling, such as stable isotopes
limitations can still provide novel insights in the use of [294] or chlorophyll measurements in situ [51]. Stable
sensory information in fish behaviour in lakes. isotopes have revealed transmission of carbon and nitro-
gen within lakes and the terrestrial-aquatic interface
Can movement data provide information on the [220] as well as shifts in the trophic network as a conse-
ecosystem role of megafauna? quence of species invasions [293]. Measurement of
Ecosystems are built upon matter and energy, the move- stable isotopes linked with movement data can illustrate
ment of which generates ecosystem services [69]. In how matter is transferred within the lake and what func-
lakes, matter and energy cycle among riparian, benthic, tional movement classes exist within species and
littoral, and pelagic zones; gravity and flow create con- whether movement syndromes (i.e. consistent individual
nections but organismal movement is critical to creating differences) exist. Movement syndromes may be key to
linkages and generating ecosystem services. Rates at determining how intraspecific differences in behaviour
which these processes occur vary as a function of a drive ecosystem roles. Acoustic telemetry in replicated
variety of factors operating at broad spatial scales such whole-lake experiments will reveal how individuals, pop-
as those driven by temperature as well as shorter scales ulations, and communities shift their patterns of space
such as depth and nutrient loads [264]. Organisms carry use across days, seasons, and years to incorporate and
out ecosystem services by cycling matter and energy deposit matter and energy within their confined land-
through their bodies, as such, they develop functional scape. Layering this information with abiotic data will
roles in the ecosystem as producers, consumers, decom- reveal drivers of migration and dispersal within habitats
posers, etc. [22, 123]. Valuable research has been carried across time scales [22, 37]. We can then link where andLennox et al. Movement Ecology (2021) 9:40 Page 11 of 28
when animals move with the consequences of that and scales [269]. One encompassing pattern deals with
movement for the ecosystem, established from site- how much an environment influences movement
specific sampling of lake productivity and contrasts patterns, and whether collected trajectories are represen-
among species under investigation. Multispecies studies tative of an animal’s full potential for movement [12, 21,
in whole lakes can also reveal dynamic niche partitioning 35]. Movement data for such comparative problems are
and species interactions including predation, competi- typically collected from a wide range of environments
tion, and parasitism when multiple species are tagged that are often assumed comparable rather than explicitly
(Fig. 2). Critical to this is considering scale by contrast- tested. These limitations are an artefact of early move-
ing results from lakes of different size: we will likely find ment tracking technologies and their relatively small
increased sympatry and decreased connectivity with in- sample sizes, whereas contemporary technology allows
creasing habitat size, a factor that can easily be investi- for greater scalability and replication. Many of the lar-
gated in these closed systems [133]. gest lakes on the planet have hosted extensive tracking
We envision replicated whole-lake experiments that spe- networks, suggesting that the gap between technology
cifically investigate multi-species dynamics in habitat use and scale-appropriate studies continues to narrow. But
and the nature of connectivity within lake ecosystems. there is ample room to investigate ecological phenomena
Instrumented individuals moving within an array of acous- at smaller scales that encompass a greater diversity of
tic receivers will reveal patterns and drivers of movement lake types and ages and thus physical environments
across spatial, temporal, and ontogenic scales. Spatial over- [133]. Such a broad variety of smaller and usually self-
lap of individuals and species can be calculated using kernel contained ecosystems gives researchers the ability to
density or convex hulls from two- or three-dimensional po- perform either observational or experimental studies.
sitions within arrays (e.g. [104]; Fig. 2). Detection data from The field of limnology consistently takes full advantage
acoustic receivers can be investigated using network ana- of small lake attributes to investigate fundamental pat-
lysis (Fig. 3) to determine which species are central to con- terns of abiotic interactions (e.g., biological, chemical,
necting the ecosystem across space and time [136] and and physical). The morphometry of smaller lakes can
functional movement classes can be identified within and range from simple gradual depressions with circular
across systems from cluster analysis [35]. Contextual data boundaries to complex depth profiles with asymmetrical
can be derived from biologging sensors including acceler- boundaries. Where a lake is located will affect how its
ometers that measure fine-scale behaviours that can be morphometry limits utilization of light and thus thermal
interpreted as foraging or reproduction to reveal the fre- input and stratification. There are many other physical
quency and spatiotemporal distribution of these exchanges environment modifiers (e.g., wind, geothermal, under-
of matter and energy (e.g. [43, 289]). Novel tag sensors and water springs) that can also be influenced by location
analytical models can also be used to remotely reveal preda- and have the potential to affect fish movement. Unco-
tion in lakes with smaller risk of a predator evading detec- vering how the physical environment influences organis-
tion than in marine systems but the tag size still limits the mal movement across and within gradients of change
size of fish that can be studied [93, 106]. Telemetry data can (e.g., aging, disturbances) is another avenue to consider
predominantly be derived from fish but interactions with that is also understudied. In summary, lakes can provide
other species such as ducks [209], crayfishes [308], semi- the necessary scalability to investigate the relations be-
aquatic mustelids, turtles, frogs, snakes, or crocodilians are tween physical environment and movement, through
also certain to be important and some of these species could both observational and experimental means in stable or
be tagged as part of a broad community study. Investigating dynamic contexts.
movement responses of fish to experimental manipulations There are relatively few lake studies that specifically
such as nutrient subsidy (e.g. [220]), introduction of novel examine the physical environment using telemetry and
species [46], change in water quality (e.g. temperature, even fewer that study multiple lakes simultaneously.
clarity, pH) can then be used to establish mechanisms Often, studies will characterize an entire lake’s physical
explaining movements observed in telemetry data. Repli- environment (e.g., temperature, light) with relatively
cated experimental designs will be critical to establish caus- coarse sampling resolution, either spatially or tempor-
ality and determine whether movement phenotypes drive ally. Yet, lakes are perfect arenas for detailed fine-scale
ecosystem services or whether characteristics of the ecosys- sampling of processes that cannot easily be detected in
tem shape the movements of animals that reside within. the vast marine environment. Gerking [91] described the
variability of individual fish movement behaviour as an
How much does the physical environment influence association between an individual and its surroundings
movement? that is informed by sensory stimuli and driven by
Ecologists are continually searching for fundamental pat- recognition of familiar areas. A more modern perspec-
terns of movement that are predictable across organisms tive also suggests that physical environments oftenLennox et al. Movement Ecology (2021) 9:40 Page 12 of 28 contain recognizable landmarks so fish can learn and include but are not limited to different forms of pollu- generate spatial maps [31]. What is not clear is what tion (e.g light, sound), boating traffic and shipping, habi- drives shifts in fish home ranges, which stimuli inform tat modification (e.g., aeration, weed removal, shoreline movements more than others, and how to respond to development, thermal effluent). Comparing the differ- changes – all as a function of their physical environ- ences between altered and unaltered environments is ment. At a coarse scale, studies have shown that fish can particularly suitable for urban areas. Alternative distur- consistently find the same food patches, discriminate bances could be drought and severe water level decrease, among habitats using multiple cues, and optimize for- prolonged ice coverage and increased ice thickness, hyp- aging strategies in heterogeneous physical environments oxic events driven by algal blooms, and introduction of [30, 128, 211]. Interestingly, when multiple connected an invasive species that specifically modifies the physical lakes are considered, fish dispersal seems to be more af- environment. Overall, all the recommendations here fected by spatial distribution of lakes, number of connec- only scratch the surface of possibilities but provide a tions, and suitability of corridors as opposed to local template for an unexplored research area that can be en- environmental factors [27, 216]. At a finer scale, studies hanced with other experimental design techniques such have shown that lake morphology (simple basin vs. com- as transplanting fish and manipulating physical plex) can influence habitat use, spatial distribution, and environments. activity [239]. Furthermore, lakes with stratification can influence vertical movement patterns [102, 208]. As un- How will climate change impact animal movements? derstanding of individual lakes and their physical charac- Climate change is a ubiquitous process affecting all teristics continues to grow, so too will the opportunities ecosystems and one of the major drivers of species to link such phenomena with fish movement ecology. extinctions [132, 292]. In response to climatic change, Lakes are ideal for revealing relationships between the geographic range and distribution shifts have been physical environment and animal movement, particularly observed in a number of species [161, 277]. Ectotherms when considering using multiple lakes simultaneously. are particularly sensitive to environmental temperature There are unlimited ways to design movement studies extremes [231], explaining the conformity found be- using lakes but to disentangle the physical environment tween their latitudinal ranges and thermal tolerance from organismal movement, we have four recommenda- [277]. In freshwater teleosts, moving away and dispersing tions. Our recommendations consist of different types of to find a more suitable environment, matching with their studies 1) before and after, 2) gradients (longitudinal or own biological constraints, is indeed commonly observed latitudinal), 3) replicated, and 4) stable vs. dynamic com- in response to climatic change [62], with a general ten- parison. Before and after type studies can take a lake or dency of range contractions at warm range edges and multiple lakes monitored before and after some eco- shifts to higher altitudes or latitudes [277]. However, an- logical phenomenon, alteration in lake morphology, or imals are constrained by system boundaries with limited physical change occurs but the sample unit is the lake opportunities to disperse and relying upon alternative (e.g., some lakes are controlled while others represent strategies to cope with climate change [61]. This is espe- treatments). The second is the same but the sample unit cially true for lake teleosts, for which climate-induced is the lake in a nested design (e.g., the lake is subdivided changes of lake properties and phenology, such as catch- with an impermeable barrier). Often, these studies ment hydrology, lake ice phenology, thermal characteris- emphasize using lakes with similar physical characteris- tics, nutrient supply and cycling, primary production, tics and are in close proximity of one another. Longitu- and bacterial blooms [92] can create challenging condi- dinal and latitudinal gradients are simply studies where tions for development and survival. Additionally, cli- lake choice is spread along a coordinate axis (e.g., north- matic effects often coincide with other anthropogenic south, east-west) so variations of light and thermal re- stressors affecting lake ecosystems such as eutrophica- gimes can be incorporated. These studies are character- tion, pollution, biological invasions, habitat degradation, ized by long distances between study areas where each and direct exploitation of organisms [51, 109]. lake is arranged at the furthest and opposite edges of the Beyond distribution shifts, teleosts strongly rely on study organism’s distribution. For example, one lake in their phenotypic plasticity, i.e. ability to adjust their this study may be affected by ice coverage in winter behavior and physiology, to cope with new climatic re- while another lake in the study has year-round open gimes and associated ecosystem changes (for a general water. Additionally, gradients along elevations are also review see [24]), in particular, under a rapid climate possible. The third recommendation is focused on lakes change that is limiting the capacity for evolutionary where anthropogenic activity manipulates the physical adaptation [295]. Changes in the abiotic environment environment intermittently or frequently to introduce al- can directly affect the metabolic processes of fish, more tered physical environments. Examples of alterations specifically, warming water temperatures accelerates
You can also read
