Food competition is a fundamental ecological interaction. Individuals compete indirectly by consuming a limited resource and depleting its availability for others (Petren and Case, 1996) or they can directly prevent competitors from accessing a resource through interference competition (Amarasekare, 2002). Direct agonistic encounters between individuals can have costs that increase with the frequency of interactions. For example, frequent territorial encounters between lizards can lead to suppressed immune function (Svensson et al., 2001). Therefore, it may be predicted that the frequency of competitive behaviors changes depending on the density of competitors. However, little research has been conducted to test this hypothesis.

Bats (order Chiroptera) provide an interesting test case for understanding mechanisms of food competition. Bat flight abilities allow them to exploit a wide range of food resources (Voigt et al., 2017). Echolocation provides bats the ability to detect food items in complete darkness, but it operates over a limited range of several meters for finding most food items (Schnitzler et al., 2003). Bats emitting echolocation calls also alert nearby competitors to their presence. Insectivorous bats increase their call repetition rate to produce a feeding buzz in the final moments before attacking insect prey. Other bats eavesdrop on these feeding buzzes to find ephemeral food patches, thus setting the stage for potential interference competition between bats hunting in the same food patch (Gillam, 2007; Dechmann et al., 2009).

In addition to echolocation calls, bats emit social calls to communicate to other bats (Bohn and Gillam, 2018). Social call repertoires are varied and serve many functions, including as signals used in food competition. For example, two closely related species of Pipistrellus bats in Britain each have their own distinct social calls (Barlow and Jones, 1997b). These calls are produced more often at lower insect densities and playbacks of social calls cause a reduction of bat activity for conspecifics on foraging grounds (Barlow and Jones, 1997a). Male big brown bats (Eptesicus fuscus) also produce social calls during food competitions (Wright et al., 2014). These calls have individual-specific acoustic features, and they cause competitors to fly away from available prey in food competitions staged in a laboratory. Big brown bats that made more social calls were more successful capturing prey, indicating that they were being used by the bats for defending or claiming food items (Wright et al., 2014).

Here, I study food competition at natural foraging sites in Arizona and New Mexico, United States by Mexican free-tailed bats (Tadarida brasiliensis), which live in colonies that can exceed one million individuals (Betke et al., 2008). Bats of this species disperse from colonies across the landscape to search for ephemeral patches of insects while traveling distances of over 100 km (Best and Geluso, 2003). Mexican free-tailed bats eavesdrop on feeding buzzes of conspecifics to find food (Gillam, 2007). They also produce sinusoidal Frequency Modulated (sinFM) signals to jam the echolocation of conspecifics (Corcoran and Conner, 2014). These sinFM calls are produced at the same time and frequency as conspecific feeding buzzes and they cause other bats to miss prey items. In addition to sinFM calls, Mexican free-tailed bats have a large repertoire of social calls (Bohn et al., 2008), including song, which are used by male bats in roosts to attract mates and fend off male competitors (Bohn et al., 2009).

The objective of this study was to determine what food defense mechanisms Mexican free-tailed bats use under different conditions at foraging patches in the field. Initial observations indicated that these bats use both song and sinFM calls at food patches. My main hypothesis was that Mexican free-tailed bats use song as a food claiming defense at low bat densities but switch to jamming conspecifics when more competitors are present because it would not be possible to defend food patch from many competitors. I tested this hypothesis by analyzing continuous acoustic recordings made at two foraging sites for a total of nine nights. I used number of echolocation calls as an index of total bat activity, feeding buzzes as indicators of foraging activity and numbers of songs and sinFM calls as indicators of different competitive behaviors. I also recorded interactions at foraging sites using four calibrated infrared cameras to reconstruct interactions in 3-D to examine behavioral interactions with finer detail.

I recorded 356,159 echolocation calls, 4,491 feeding buzzes, 336 songs and 144 sinFM calls during the nine nights of this study. During one-minute intervals including song, rates of echolocation recordings were slightly, but significantly higher than periods containing only echolocation (Figure 2A). In comparison, intervals with sinFM calls had substantially higher numbers of echolocation calls recorded, indicating the presence of many more bats (Figure 2A). Only 9% of recordings (17 of 186) made within one-minute of sinFM calls were classified manually as one bat being present during most of the recording, compared to 38% of recordings (171 of 453) made within one minute of song, a highly significant difference (Fisher’s exact test, P = 2.5e-14). This result and inspection of echolocation recordings (Supplementary Figure 1) confirms that the observed differences in echolocation rates reflected a difference in the number of individual bats present at the field site.

Feeding buzz rates were substantially higher during 1-minute intervals containing sinFM calls, but not intervals containing song, compared to baseline feeding buzz activity (Figure 2B). This demonstrates that bats make song with relatively few bats present that are feeding at low rates. In contrast, bats produce sinFM calls most often when many bats that are present and feeding at high rates.

Insect abundance had no detectable effect on echolocation rates, feeding buzzes, song or sinFM call production (Table 1). However, the addition of a light at the SWRS field location caused a dramatic increase in the number of insects present (Personal observation). It should be noted that the method used was not able to document the increase of insects within the foraging patch caused by supplemental lighting. Supplemental lighting at SWRS also led to a significant increase in echolocation recordings, feeding buzzes and sinFM calls, but not song (Figure 2C-F and Table 1). This further demonstrates that sinFM calls, but not song, are produced when many bats are feeding at high rates in a food patch. Bats produced more song later in the night, but echolocation, feeding buzzes and sinFM production did not vary with time of night (Table 1).

I documented nine cases of bats chasing conspecifics on video (Supplementary Video 1). Seven of these nine events included a bat producing song, but none involved a bat producing sinFM calls, a statistically significant difference (binomial probability test; P = 0.0078). Three-dimensional reconstructions of these events show a bat flying closely behind another bat and following the leading bat’s repeated turning maneuvers (Figure 3). The trailing bat’s flight path shows wider turns and a longer path length, which is indicative of the trailing bat flying at a higher speed than the leading bat. These are all indicators of chasing behavior rather than following behavior (Shelton et al., 2014).

Mexican free-tailed bats exhibited two distinct behaviors for competing with conspecifics at foraging sites. When the density of conspecifics and feeding events were low (as indicated by echolocation recordings and feeding buzzes), bats produced song and chased conspecifics (Figures 2A,B, 3 and Supplementary Video 1). In contrast, bats produced sinFM calls to jam competitors (Corcoran and Conner, 2014) when numerous bats were present and feeding at high rates.

Singing and jamming within food patches have different contexts and apparent aims. Singing and chasing conspecifics appears to be an effort to deter conspecifics and force them to leave the food patch. The most direct evidence for this was the observations of bats producing song while chasing conspecifics in the food patch (Figure 3 and Supplementary Video 1). This is consistent with previous research conducted in a laboratory showing that big brown bats use food claiming signals to deter competitors from foraging (Wright et al., 2014). Mexican free-tailed bats also produce song when defending mates in the roost (Bohn et al., 2009). Therefore, it appears that song is used for both defending food and mates in this species. Song was produced more often later in the night (Table 1), which could result from bats having more time to establish foraging areas to defend from competitors.

In contrast to singing, jamming does not appear to be aimed at deterring conspecifics as much as directly preventing them from sensing and capturing individual prey items. SinFM calls are produced only during the last moments of prey capture, when a conspecific is trying to intercept the prey (Corcoran and Conner, 2014). Acoustic 3-D reconstructions from earlier research (Corcoran and Conner, 2014) showed that after jamming a conspecific, the jamming bat would often attempt to capture the prey item instead of chasing the other bat. The bat that was jammed would often turn the tables and try to jam its competitor. This sometimes went back and forth several times until one bat gave up and the other had a chance to capture the prey without interference (Corcoran and Conner, 2014). SinFM calls are very similar to the “herding” calls used by dominant males forcefully pushing females into a cluster in his territory (Bohn et al., 2008). However, aggressive behavior was not associated with sinFM calls being used in the field, therefore this similar acoustic signal appears to have very different function when used in food competition versus in a roost with potential mates. In summary, singing appears to be an effort to chase conspecifics out of a food patch and deter them through aggression, whereas jamming directly prevents conspecifics from acquiring a specific food item by interfering with its sensory system.

Why do Mexican free-tailed bats have multiple strategies for competing with conspecifics? At low densities of competitors, the benefits of potentially removing a competitor from a food patch may outweigh the costs. With fewer competitors present, the remaining bat can hunt insects without a competitor present that can distract attention (Cvikel et al., 2015) and capture food items. Song conveys information on the identity of the individual (Bohn et al., 2008), so it may be possible for individuals to establish dominance over others that they encounter frequently at foraging sites. When many competitors are present, food claiming via song might not be effective or energetically beneficial because the bats would spend too much time and energy chasing competitors instead of pursuing prey. Increased costs of territorial behavior with higher competitor density is known to occur, for example, in the form of reduced immune responsiveness in some lizards (Svensson et al., 2001). When more competitors are present, Mexican free-tailed bats appear to compete directly for individual prey items, with sonar jamming being a key strategy for preventing competitors from depleting the immediate food resource.

It should be noted that this study was not able to track individuals over time. Therefore, observed differences in competitive behavior may result from individuals switching competitive behaviors under different contexts. Alternatively, different individuals could specialize on different behaviors that are used more often under different competition densities. For example, dominant males could preferentially use song to establish food defense territories where competition is kept at a low density, whereas less dominant males and females may preferentially use sinFM calls at food patches with higher densities of competitors. Additional research is needed to differentiate between these possibilities.

Contrary to my prediction, measured insect densities did not correlate with differences of any acoustic behaviors (Table 1). The method used—placing insect traps 200 m away from foraging sites—was designed to assay insect abundance in the vicinity of the foraging areas without removing potential prey from foraging patches. However, this method was unable to measure the increases of insect activity caused by the addition of supplemental lighting. Observations at the field site indicated that supplemental lighting had a dramatic effect on insect abundance. Supplemental lighting also increased bat echolocation recordings, feeding buzzes and sinFM calls (Figures 2C–F). Therefore, insect abundance within the food patch appears to be an important factor for bat competitive behavior and moreso than insect activity in the surrounding area. Additional research is needed to confirm this finding and determine how the combination of overall insect abundance and insect abundance within a food patch influences bat behavior. This could be achieved through a combination of insect traps and imaging of the night sky (Ruczyński et al., 2019).

This study provides novel detail into the natural competitive behaviors of bats. Mexican free-tailed bats may be under particularly high selective pressure from food competition. Flight is energetically costly and Mexican free-tailed bats fly long distances in pursuit of prey (Best and Geluso, 2003) while competing with as many as one million individuals that share a colony (Betke et al., 2008). It may be more energetically favorable for individuals of this species to stay in a food patch and compete for prey via song or jamming rather than to depart in search of unoccupied food patches.

In addition to being competitors, other bats can improve efficiency of prey search. Bats can hear one another’s calls at much longer distances than they can detect prey with echolocation (Dechmann et al., 2009). Therefore, groups of bats can function as a distributed array of sensors in the environment (Egert-Berg et al., 2018). This can lead to a situation where individuals simultaneously experience costs and benefits to social foraging. This may mean that there is an optimal intermediate foraging density for insectivorous bats (Cvikel et al., 2015). The behaviors of singing and jamming documented here are therefore only part of a broader set of social interactions that likely occur in this species and more broadly within bats.

Food competition is one of the most common ecological interactions that organisms encounter. This study demonstrates that the density of competitors is a key factor in determining what competitive strategies will be effective. This research also highlights how animal sensory and locomotor abilities are important for determining the effectiveness of competition strategies.

The original contributions presented in the study are publicly available. This data can be found here: https://osf.io/nzqtv/?view_only=1befa9e597634c12a9bfa526878a86c2.

The animal study was reviewed and approved by Wake Forest University IACUC.

AC designed and conducted the experiments, conducted the data analysis and wrote the manuscript.

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.