A review of conspecific attraction for habitat selection across taxa

Abstract Many species across taxa select habitat based on conspecific presence, known as conspecific attraction. Studies that document conspecific attraction typically provide social information (i.e., cues that indicate the presence of a given species) and then determine if a given species is more likely to settle at locations where the social information is provided compared to those locations that do not. Although the number of studies examining conspecific attraction has grown in recent years, a comprehensive review has not yet been undertaken. Here, we conducted a review of the literature and found 151 studies investigating conspecific attraction across eight taxa. We found that conspecific attraction is widespread with between 80% and 100% of studies, depending on taxa, documenting positive associations between habitat selection and the presence of conspecific cues. Conspecific attraction has been documented more frequently in bird and fish species with less attention given to invertebrate and mammal species. We use the patterns we found to (a) provide an overview of the current state of research on conspecific attraction and (b) discuss how important factors, such as cue characteristics and life history traits, may play a role in shaping conspecific attraction patterns within and across taxa.


| INTRODUC TI ON
Habitat selection by animals is not random, with evidence suggesting that animals use information to select locations with resources necessary for survival and reproduction (Schmidt et al., 2010). This information may be gathered either from physical cues, termed "non-social information," or through interactions with or observations of others in the environment, termed "social information" (Seppänen et al., 2007;Wagner & Danchin, 2010). Social information, either unintentionally or intentionally conveyed to others, can be gathered from many cues, including the physical presence of an individual, chemical cues, and/or acoustic vocalizations (Danchin et al., 2004;Wagner & Danchin, 2010). Indeed, many species across multiple taxa locate conspecifics with social cues and preferentially settle in these locations, often resulting in the phenomenon of conspecific attraction (reviews in Reed & Dobson, 1993;Stamps, 2001).
However, the prevalence of conspecific attraction in relation to habitat selection and the differences in proximate mechanisms used to detect conspecifics across taxa have not yet been explored in a single review.
Using the presence of conspecifics to select habitat is considered widespread across taxa (e.g., Danchin et al., 2004) and is often conspecific attraction actually is for habitat selection and what selective pressures are shaping the behavior within and across taxa.
The topic of conspecific attraction has been frequently reviewed within a taxa (e.g., birds: Ahlering & Faaborg, 2006;Ahlering et al., 2010;Szymkowiak, 2013;amphibians: Buxton & Sperry, 2017) or within special interests such as across species of conservation concern (Putman & Blumstein, 2019). On a similar note, heterospecific attraction, or using the presence of other species to select habitat, has been reviewed extensively across taxa (Mönkkönen et al., 1999;Putman & Blumstein, 2019;Seppänen et al., 2007). To our knowledge, however, there is no single literature review on conspecific attraction that synthesizes studies from multiple taxa to explore patterns of the habitat selection strategy across taxa. Such a literature review would greatly improve our understanding of why conspecific attraction occurs, and also identify knowledge gaps that future research should address.
Here, we conducted a literature review to explore and discuss patterns of conspecific attraction for habitat selection across several taxa. We use the resulting patterns generated to qualitatively investigate the following three questions: 1. Which broad taxonomic groups does conspecific attraction for habitat selection occur most frequently in? 2. How do the proximate mechanisms (e.g., chemical, acoustic, or visual cue) of conspecific attraction vary within and among taxa?
3. What are the fitness benefits gained by individuals using conspecific attraction (e.g., increased number of offspring), both within and across taxa?
We then discuss our results in light of how life history traits (such as mobility, dispersal, and sex-or age-specific characteristics) could shape the proximate mechanisms of conspecific attraction. We also discuss how selection for or against the behavior could occur due to variation in life history traits both within and across taxa. Finally, within these discussion topics, we highlight research needs that would help fill in the gaps discovered in this review. This approach will therefore aid researchers in where to focus study efforts to better our understanding of conspecific attraction as a habitat selection strategy.

| ME THODS
We used Google Scholar and Web of Science to identify studies on conspecific attraction from 1960 to 2017 by specifying the keywords "conspecific attraction," "conspecific information," "conspecific cue," and "social information." From these publications, we also identified other relevant studies from their references. We limited our review to studies that identified conspecific attraction for purposes of breeding and/or nonbreeding habitat selection or the act of finding an area that is suitable to meet all of an organism's resource needs (i.e., the organism's "habitat"; Piper, 2011). We excluded studies that examined conspecific attraction to pursue specific intraspecific interactions, such as mate selection (e.g., Michelena et al., 2005;Pearl et al., 2000) and intrasexual territory defense (e.g., Campos et al., 2017), or to improve performance during a specific task, such as foraging decisions (e.g., van Bergen et al., 2004). We only reviewed experimental studies that manipulated conspecific presence with cue treatments to a habitat; these studies provide more direct evidence of conspecific attraction than observational studies (e.g., observing clustered distributions in a species; reviewed in Campomizzi et al. 2008).
We did not conduct a formal meta-analysis, but instead used tallies from each publication to generate percentages of studies that found evidence of conspecific attraction. To address Question 1, we investigated eight taxonomic groups, with species grouped together by class: birds (Aves), ray-finned fishes (Actinopterygii), reptiles (Reptilia), amphibians (Amphibia), mammals (Mammalia), insects (Insecta), arachnids (Arachnida), and crabs and lobsters (Malacostraca, hereafter referred to as "crustaceans"). We chose these groups because they had at least five published studies on conspecific attraction meeting the criteria described above. To address Question 2, within each study we categorized the cue(s) used to attract conspecifics that were experimentally manipulated. These cues included: acoustic cue (e.g., song or call playback), chemical cue, visual cue (e.g., decoys or models), conspecific presence (e.g., tethered or caged individuals), and indirect cues of conspecific presence (e.g., conspecific web and burrow). The conspecific presence cue included those studies in which the physical presence of conspecifics was experimentally used as a stimulus, and thus, one specific cue type was impossible to isolate. Some studies tested multiple cue types independently. For these studies, we considered each cue presented individually as a single "test." We also categorized studies that used a combination of cues (e.g., acoustic playback and visual decoys combined).
To address Question 3, for those papers that found conspecific attraction, we categorized the proposed ultimate mechanisms (i.e., fitness benefits) of using conspecific cues for habitat selection based on discussion text in each study. We found that ultimate mechanisms could be categorized into the following six benefits to habitat selection via conspecific attraction: (1) location/identification of suitable habitat, (2) protection benefits (e.g., group defense from predators), (3) mating benefits (e.g., increased access to mates), (4) foraging benefits (e.g., enhanced foraging efficiency and sharing information about foraging locations), (5) thermoregulatory benefits, and (6) kin selection. To assess fitness benefits of conspecific attraction further and more directly, we identified all studies that quantified measures of fitness (e.g., survival, number of offspring produced, and clutch success/failure).
We acknowledge that we have not identified all studies fitting our criteria in this review, as the literature on conspecific attraction is vast.
Moreover, a potential shortcoming of this review is that there is likely a positive-results bias in publication, such that negative results from experimental tests are less likely to be published. We return to how this publication bias may affect our findings in the Discussion. Nonetheless, our search likely identified the majority of studies in the taxa we considered and provides an unbiased overview of the current state of published conspecific attraction research for the taxa of interest.
In reptiles, crustaceans, and fish, the dominant cue type tested was chemical cues (reptiles: 81% of tests; crustaceans: 53%; fish: 48%). In insects, the dominant cue type tested was visual cues (45% of tests), while in arachnids, the dominant cue type was indirect cues (e.g., presence of webs or silks; 50%). Across all cue types, chemical cues and presence cues were tested in 100% of taxa, acoustic and visual cues in 50% of taxa, and indirect cues in 38% of taxa. Only 12 studies used a combination of cue types in their experimental paradigm, and only in birds (20% of tests), amphibians (8% of tests), and insects (5% of tests).
When considering acoustic cues only (i.e., not presented in combination with other cues), most studies on birds, mammals, and amphibians exhibited conspecific attraction in habitat treated with these cues (birds: 89% of tests; mammals: 88%; amphibians: 82%;

| Question 3: What are the fitness benefits of conspecific attraction for habitat selection?
For seven of eight taxa, we found that location or identification of suitable habitat was the most commonly cited ultimate mechanism

F I G U R E 2
Percentage of species showing attraction, repulsion, no response (neutral), or conflicting response to conspecific cues grouped by taxa. Conflicting response indicates that a species showed a differential response to conspecific cues in two or more studies (e.g., one study showed attraction, while another showed repulsion). Species sample size for each category is listed in bold on the respective bar segment  et al., 2012;Fletcher, 2009;Harrison et al., 2009). Two bird studies found fledging success was greater at sites treated with conspecific cues compared to control sites (Anich & Ward, 2017;Ward & Schlossberg, 2004), whereas one found lower fledging success at conspecific cue sites compared to controls (Grendelmeier et al., 2017). In crustacean studies, porcelain crab (Petrolisthes cintipes) fitness was maximized at intermediate conspecific densities based on predation and growth rate (Donahue, 2006), whereas no relationship was found between Caribbean spiny lobster (Panulirus argus) survival and conspecific density (Childress & Herrnkind, 2001). Similarly, no relationship was found between survival and conspecific presence for common lizards (Lactera vivipara), although growth rates for juveniles were higher when conspecifics were absent (le Galliard et al., 2005).

| Proximate mechanisms of conspecific attraction
Our review found that a broad range of taxa exhibits conspecific attraction when selecting habitat. Depending on the taxa, ~80%-100% of studies documented conspecific attraction, suggesting the habitat selection strategy may be widespread. Notably, taxonomy diversity is considerably low among studies testing for conspecific attraction during habitat selection: To date, most experimental studies are on birds or fish, whereas crustaceans and arachnids are largely underrepresented (Figure 2). Although sample sizes within F I G U R E 3 Percentage of tests in which taxa have shown conspecific attraction, repulsion, or no response (neutral) to (a) acoustic, (b) chemical, (c) visual, and (d) presence/abundance cues. If a study tested species response by multiple age classes or sexes to a given cue type or species response to multiple variations of a cue type and found attraction in any of these instances, then it was counted as a single test showing attraction. Sample size for each category is listed in bold on the respective bar segment. The absence of a bar for a taxon represents that there were no tests of that cue type Conspecific attraction was most often documented in response to cue types that matched the taxa-specific communication systems (Bradbury & Vehrencamp, 2011). For example, birds primarily communicate with acoustic cues (Kroodsma & Miller, 1996), which was the conspecific cue type that most commonly elicited conspecific attraction in this taxon (Figure 3a). Acoustic cues also elicited re- larders (e.g., Hromada et al., 2008). The exception is for arachnids, but species samples sizes were small (n = 5; Figure 2) and only three studies used this cue type. More research is needed to explore how widespread conspecific attraction is in response to these indirect cues not used for communication.
We found that conspecific attraction was evident in taxa with several migratory species (such as birds and fish), or with species that exhibit breeding dispersal (such as birds, fish, and anurans).
Several migratory species must quickly locate habitat upon arrival and reproduce as soon as possible, which conspecific cues could greatly assist with (Buxton et al., 2015;Danchin et al., 2004). For example, in anurans (Gatz, 1981;Wells, 2007) and birds (Amrhein et al., 2007), earlier-arriving males are more likely to encounter and/ or attract females than later-arriving males. Similarly, spawning fish may reduce the search costs associated with long-distance migration to spawning habitat by orienting toward conspecific cues (Bett & Hinch, 2015). Considering nonbreeding dispersal, our review found many examples of fishes (e.g., Galbraith et al., 2017;Lecchini et al., 2005;Schmucker et al., 2016;Sweatman, 1985), snakes (e.g., Burger et al., 1991;Hileman et al., 2015), and bats (Ruczyński et al., 2009;Schöner et al., 2010)   Fish are also highly mobile, with studies indicating that larvae can travel many kilometers when prospecting for breeding habitat (see Lecchini et al., 2007). Paired with a sensory bias toward potentially far-reaching acoustic and chemical cues (discussed above), the cost to sampling conspecific cues is likely low for birds and fish. Similarly, conspecific attraction studies on mammals are heavily biased toward bats (n = 7 of 11 studies), which are also highly mobile and able to disperse long distances (Krauel & McCracken, 2013). A useful future direction would be to determine the perceptual ranges of species known to exhibit conspecific attraction for habitat selection, as well as how far the conspecific cue used propagates from its source in the environment. Doing so would help elucidate how accessibility of social information influences the prevalence of conspecific attraction among species that utilize it for habitat selection.
It is notable that demographic traits such as sex and age structure could influence conspecific attraction prevalence within and across taxa. However, our review revealed that few studies explicitly examined conspecific attraction by multiple age classes (n = 20) or by both sexes (n = 23) within the same study. Given that dispersal and habitat selection pressures can differ among sex and age (reviewed in Dobson, 2013), it would stand to reason that using conspecific

F I G U R E 4
Percentage of studies citing a given ultimate explanation for conspecific attraction. Some studies included multiple explanations, while others gave no explanation (i.e., none given). Only studies that found evidence of conspecific attraction were included, as studies that did not find evidence of conspecific attraction often did not give an explanation for the behaviors cues for habitat selection would similarly differ. We might expect, for example, that juveniles may be more receptive to conspecific cue use because they lack experience to draw personal information from about a given habitat's suitability (e.g., Nocera et al., 2006;Ward & Schlossberg, 2004 Chemical cues may also contain information on conspecific body size, as Scott et al. (2013) demonstrated with small-eyed snakes (Cryptophis nigrescens). In some songbirds, males sing two distinct song categories during the breeding season, with first category song sung by unpaired males and second category song sung by paired males (Spector, 1992). Kelly and Ward (2017) found that yellow warblers (Setophaga petechia) were more abundant at sites where second category song was experimentally broadcast, possibly because successfully paired males provide better information on habitat quality.
We expect that future conspecific attraction research will advance toward understanding how information contained within a cue, and how the number of cues given influences the magnitude of attrac- Future research should also address how multiple cue types exposed at once influence conspecific attraction, rather than one conspecific cue type at a time. Many organisms use multimodal signaling for communication to make fitness-related decisions (reviewed in Hebets et al., 2016;Higham & Hebets, 2013), notably in arachnids (Herberstein et al., 2014;Uetz & Roberts, 2002), amphibians (Starnberger et al., 2014), and some birds, reptiles, and fishes are on birds where both visual cues (i.e., decoys) and acoustic cues (i.e., conspecific song) are presented. This approach would be more appropriate for species that require multimodal signaling to recognize conspecifics, for example, in some insects (South et al., 2008), fishes (Hankison & Morris, 2003), and bird species (Uy et al., 2009).
For species such as these, conspecific attraction may only occur if various biologically salient cues are presented in the habitat at once.

| Fitness benefits of conspecific attraction
We found the ultimate mechanisms driving conspecific attraction, as suggested by authors for using conspecific cues, were largely the same across taxa. In practice, however, studies rarely measure the suggested fitness benefits of conspecific attraction. Indeed, of the 12 studies that attempted to measure some fitness benefit, only four found enhanced fitness in relation to conspecific presence or density. Even fewer studies discussed the costs of settling with or near conspecifics, which can include increased competition for resources (Grand & Dill, 1999), increased parasite transmission (Brown & Brown, 2004), and higher likelihood of being detected by predators (McGuire et al., 2002). This paucity of studies could be due to positive-results bias in publication, where studies unable to find fitness benefits to conspecific attraction are less likely to get published. Lack of publications on negative results limits our ability to understand or predict directional selection for the use of conspecifics when selecting a habitat. It is thus at present difficult to tell whether (a) there are indeed fitness benefits associated with conspecific attraction that make the behavior an adaptive habitat selection strategy, or (b) conspecific attraction is not adaptive at all and is simply an exaptive byproduct of other behaviors requiring interactions with conspecifics (e.g., mating and territory defense), in line with Gould and Lewontin's famous critique of the adaptationist approach in behavioral ecology (Gould & Lewontin, 1979).
Fitness trade-offs framed around density dependence have been discussed extensively as what has "shaped" conspecific attraction (Doligez et al., 2003;Fletcher, 2006;Szymkowiak, 2013). Given that a likely cost of settling near conspecifics is the price of having to compete with conspecific neighbors, one might expect that prospecting cue users would also assess the number of conspecifics present as part of the decision-making process. If cue users do assess density using conspecific cues, do they prefer to settle where cues indicate high-or low-density locations? Moreover, the cost of competition could preclude some individuals from using conspecific cues in high-density locations, such as less competitive juveniles (sensu Szymkowiak et al., 2016). Surprisingly, few studies take conspecific cues revealing density into account (Kelly et al., 2018, and references therein), as most studies provide a cue that simulates one individual present at the treatment site. Fewer still compare the costs/benefits accrued by conspecific cue users to nonusers (Grendelmeier et al., 2017), making it difficult to evaluate whether conspecific attraction is an adaptive habitat selection strategy actually shaped by density dependence.
Whatever the fitness payoffs to conspecific attraction are, they are likely the sum of several information sources and not just conspecific presence. Indeed, individuals collect information from many different sources (Danchin et al., 2004;Seppänen et al., 2007;Wagner & Danchin, 2010), and in some cases, decision-making based on information sources other than conspecifics leads to fitness higher payoffs (e.g., Doligez et al., 2003).
Additionally, interactions with heterospecifics can also constrain habitat selection using conspecific cues (Fletcher, 2008;Parejo et al., 2018), which is largely ignored in the conspecific attrac-

| CON CLUS IONS
The phenomenon of conspecific attraction has been a topic of interest in the scientific literature for decades, if not longer (e.g., Denton, 1889;Liley, 1982;Solomon, 1977;Stamps, 1988

CO N FLI C T O F I NTE R E S T S
The authors have no competing interests regarding this review.

DATA AVA I L A B I L I T Y S TAT E M E N T
A list of all publications included in this review is provided in Appendix S1. Data that were collated from these publications and included in the review are available at: https://doi.org/10.13012/ B2IDB -86374 11_V1 .