Abstract
Pathogens can impose substantial ecological costs on infected individuals, including reduced cognition, foraging ability, and predator avoidance. In a prior experiment, gerbils infected with Mycoplasma haemomuris-like bacteria had higher giving up densities, despite spending more time foraging, and were more likely to be killed by predators. One hypothesis to explain this is that infected gerbils suffer from cognitive impairment that reduces their ability to forage efficiently, causing them to spend more time foraging and placing themselves at higher risk of predation. We tested this hypothesis by comparing the ability of gerbils uninfected, acutely infected, or chronically infected with Mycoplasma to equalize giving up densities (GUDs) in seed patches with different initial abundances and detect micropatches within seed patches in a semi-natural vivarium enclosure. We predicted that uninfected gerbils would equalize GUDs and detect micropatches better than infected gerbils. Contrary to our predictions, infected gerbils performed equally well as uninfected gerbils on both tasks. These experiments did not support the hypothesis that cognitive impairment explains past results regarding Mycoplasma and gerbil foraging.
Introduction
Animals afflicted with disease and parasites often seem to forage less efficiently than non-diseased animals. For example, songbirds infected with Mycoplasma are less efficient at obtaining seeds from birdfeeders and spend more time at the feeder to meet their energetic needs (Hotchkiss et al., 2005). Tadpoles infected with Batrachochytrium dendrobatidis consume less food in foraging trails than uninfected tadpoles (Hanlon et al., 2015; Venesky et al., 2009). And gerbils infected with Mycoplasma leave higher giving up densities in trays containing seeds mixed into a sand matrix than uninfected gerbils (Makin et al., 2020). These examples suggest that disease can reduce animals’ ability to locate food within patches or to optimally choose when to leave patches.
Optimal foraging theory predicts that animals exhibit foraging behaviour that maximizes their rate of energy intake while minimizing the costs of foraging (Brown, 1988; Charnov, 1976). To do this, animals should forage a patch until the benefit of the current harvest rate is less than the costs of foraging. Suppose animals can accurately estimate the amount of food in each patch prior to foraging and forage every patch optimally. In that case, they will equalize GUDs across patches regardless of the initial richness of the patches (Valone and Brown, 1989). However, animals often do not have perfect information about either the amount of food in a specific patch or the boundaries of the patch. Many strategies exist for choosing when to leave a patch when patch richness is not perfectly known, including spending fixed time in each patch (McNair, 1983; Valone and Brown, 1989), using giving-up time where the animal leaves after a set time not encountering food (Krebs et al., 1974; McNair, 1983), using a giving-up yield where the animal leaves after acquiring a set number of food items (McNair, 1983), tracking instantaneous forage rates (McNair, 1983; Valone and Brown, 1989), or employing Bayesian methods of estimating patch richness, where animals choose to leave a patch based on a combination of their knowledge of overall average food availability in patches and their current harvesting rate in the patch they are in (Green, 1984). These strategies present trade-offs between the amount of information an animal must process about a patch and how well the animal equalizes GUDs between patches. The simplest strategy, spending fixed time in each patch, requires no information about the patch and is the least cognitively demanding. However, this strategy results in the worst performance of equalizing GUDs, as the initial ratio of food in the rich and poor patch is preserved in the GUDs. The Bayesian strategy outperforms fixed time, giving up time and tracking instantaneous forage rates alone. However, it also requires animals to process information about their current forage rate and information about the average richness of patches in the environment (Green, 1984). Diseased animals may be expected to struggle with more cognitively demanding strategies of equalizing GUDs and to shift towards strategies that require processing less information (Townsend et al., 2022).
In addition to determining the amount of food in a patch, animals must determine the boundaries of a food patch. If they underestimate patch size, they will miss food in parts of the patch. If they overestimate patch size, they will waste time searching areas that do not contain food. When patches have discrete boundaries separating the patch from the surrounding environment, assessing patch size is easy. However, in a study of seed distribution in the Sonoran desert, Reichman (1984) found that the seeds foraged on by granivorous rodents can vary up to 78-fold across space despite few distinct boundaries marking where the seeds are located. As such, detecting patch boundaries accurately is difficult, but the reward for doing so is high. ‘Micro-patch’ experiments, where food is concentrated in a portion of a seed tray rather than evenly mixed in the sand throughout, have been used to investigate how rodents assess patch boundaries (Fierer and Kotler, 2000; Kotler et al., 2002; Schmidt and Brown, 1996). Rodents are able to estimate micro-patches within a seed tray. Still, their ability to do so declines as the number of vague boundaries of patches increases (Fierer and Kotler, 2000; Schmidt and Brown, 1996) and as apprehension from predation risk increases (Kotler et al., 2002). Infected rodents often show impaired spatial learning (Gibertini et al., 1995; Kavaliers and Colwell, 1995; Schleich et al., 2015, but see Braithwaite et al., 1998), and might be expected to struggle with identifying patch boundaries.
Wild gerbils are commonly infected with Mycoplasma haemomuris-like bacteria, which spreads through direct contact with infected individuals (Cohen et al., 2018; Messika et al., 2017). In a prior experiment, Makin et al. (2020) found that infection with Mycoplasma led to higher giving up densities and increased risk of predation in Allenby’s gerbils (Gerbillus andersoni allenbyi). Paradoxically, they discovered that infected gerbils had higher GUDs despite foraging for longer in seed trays and foraging under greater risk than uninfected gerbils. These results suggest that infected gerbils foraged less effectively than uninfected gerbils and may have experienced cognitive impairment that limited their ability to assess patch richness or locate seeds within a patch successfully. As a result, they are forced to forage under conditions of greater risk than uninfected gerbils and suffer greater predation costs.
We tested this hypothesis by performing two experiments assessing the ability of uninfected and infected gerbils to equalize giving up densities and detect micro-patches. In the first, we measured the ratio of GUDs between rich trays and poor trays to determine which strategy gerbils employed for resource trays of varying richness. Previous results have found that Allenby’s gerbils use Bayesian methods to estimate patch richness and equalize GUDs (Garb et al., 2000). We expected infected gerbils to be worse at tracking information needed to successfully use a Bayesian strategy, including their current harvesting rate and average patch richness. Thus, we expected infected gerbils to shift away from a Bayesian strategy towards a fixed-time strategy and to equalize GUDs less effectively. Alternatively, infected gerbils may experience an increase in foraging costs but not a deficit in assessing patch richness. In this case, we would expect to see higher GUDs but no difference in equalization. In the second experiment, we measured GUDs in trays with seeds mixed evenly throughout or with seeds mixed into half the tray, creating distinct boundaries set by the edges of the tray and vague boundaries set by changes in seed abundances in the sand. With a perfect assessment of the micropatches, the GUDs in the divided trays should equal one-half the full tray. Prior results show that gerbils imperfectly assess micropatches (Fierer and Kotler, 2000), and we expected uninfected gerbils to forage the one vague boundary tray to the lowest GUD, followed by the two-vague-boundary tray and the fully mixed tray. We expected infected gerbils to be less able to detect micropatches than uninfected gerbils and that infected gerbils would forage the three tray types to similar GUDs. Alternatively, if infected gerbils only experience an increase in foraging costs but do not lose their ability to detect micropatches, we would expect higher GUDs from infected gerbils but no change in the ratio of GUDs from the micropatch tray and full tray.
Methods
Animal care, inoculation, and testing
A total of 29 Allenby’s gerbils were used for this experiment, 15 of which were infected with Mycoplasma (see paragraph on infection methods below) and the remaining 14 of which were injected with a saline control. Prior to the experiment, and in between the phases of the experiment, gerbils were housed individually or with same-sex siblings in an animal room at about 25 °C with a window to provide natural light. They were fed 3 g of millet seed per day and provided with alfalfa and a mealworm once per week.
For the vivarium experiment, the gerbils were subcutaneously inoculated with 150-300 ul of blood from gerbils known to be infected with Mycoplasma. Control gerbils were injected with saline. To confirm the Mycoplasma-infected status of the inoculated gerbils and the uninfected status of the control gerbils, we conducted two rounds of PCR testing. One was performed prior to their entry into the vivarium, and the other was conducted between the two phases of the vivarium experiment. Additional details are available in the Supplementary materials. Protocols for this experiment were approved by the IACUC at Ben-Gurion University of the Negev (Approval number: IL-09-02-19(D)).
Vivarium set up and seed tray treatments
The vivarium experiment was conducted in two phases. The acute infection phase, which started 10 days after the animals were inoculated with Mycoplasma, was run for four moon phases from December 2022 – January 2023. Acute infection with Mycoplasma occurs about 15-35 days after exposure to the microbe (Cohen et al., 2018; Makin et al., 2020). The chronic infection phase, which started 75 days after inoculation, ran from February 2023 – March 2023. Chronic infection occurs after 65 days post-exposure to the microbe and lasts indefinitely (Cohen et al., 2018; Makin et al., 2020).
Both phases of the experiment were carried out in an outdoor vivarium in Sde Boker divided into four quadrants separated by rodent-proof fences. During the acute phase, six acutely infected animals (3 male, 3 female) were placed in two quadrants, and six uninfected gerbils (3 male, 3 female) were placed in the other two quadrants. During the chronic phase, the same setup was used, with the infected and uninfected quadrants switched. However, due to a high number of deaths among infected individuals during the acute phase, only a single quadrant contained chronically infected individuals during the chronic phase. In each quadrant, we set up 12 trellises, evenly spaced in a 3 × 4 arrangement across the quadrant, to provide cover for the gerbils. Under three of these trellises, we put a set of three seed trays, plastic trays (28 × 17 × 10 cm) containing 3 litres of sand. The trays were each half covered lengthwise by the trellis. The result was that each quadrant contained three tray sets, each containing three trays, for a total of 9 trays.
During each phase, animals in the vivarium were exposed to two foraging tasks. The first was an ‘equalizing’ task to determine if Mycoplasma impairs the ability of gerbils to optimize their patch use based on harvest rate. For this task, one tray of each set was filled with 6 g of millet seed, the second with 3 g, and the third was left empty and covered to prevent gerbils from foraging in it. The experiment was carried out in each quadrant for three nights of each moon phase, and the treatments of each tray rotated so that by the end of the three nights, each tray was exposed to every treatment. The rotation of trays was done to prevent idiosyncrasies associated with a particular tray from affecting the results and to ensure that the gerbils did not learn that one tray is always better than the other.
The second task was a ‘micropatch’ task (Fierer and Kotler, 2000; Schmidt and Brown, 1996), used to determine if Mycoplasma infection impaired the ability of gerbils to detect ‘micropatches’ or areas within a patch of high food density. One tray of each tray set contained 3 g of millet seed mixed evenly in 3 litres of sand. The second contained 3 g of millet seed mixed into 1.5 litres of sand in one half of the tray, with 1.5 litres of sand with no seeds in the other half. This tray contained a micropatch with a single vague boundary. The final tray contained 3 g of millet seed mixed into 1.5 litres of sand in the center of the tray, with 0.75 litres of sand with no millet seed on either side. This tray contained a micropatch with two vague boundaries. This experiment was also carried out in each quadrant for three nights of each moon phase, and we rotated the tray treatments across the three nights.
Measurements of GUDs
To measure giving up densities (GUDs) after a night of foraging, we first examined seed trays for gerbil tracks and evidence of foraging. We then sifted the remaining seeds out of each foraged tray, brought them back to the lab, cleaned off excess sand and debris, and weighed them to the nearest 0.001 g on an electronic balance (Ohaus Adventurer, Model AR3130).
Data analysis
To determine how well gerbils equalized GUDs between rich and poor patches, we calculated the ratio GUD6g: GUD 3g. If the GUD of the 6 g tray was >3 g, it would not be possible for the gerbils to equalize GUDs, but if the gerbils took all of the seed from the 6 g tray (i.e. they foraged less than 3 g of seed from the tray set, but foraged 100% from the rich tray), they would still be foraging optimally. To deal with this problem, we set the numerator to 3 if the GUD of the 6 g tray was greater than 3. This ensured the ratio for an optimal forager remained equal to 1. To determine how well the gerbils detected the micropatches in the micropatch experiment, we calculated the COMP scores following (Schmidt and Brown, 1996). These scores were equal to the GUD of each micropatch tray divided by the GUD of the full tray for each set. A gerbil with perfect information about the location of the micropatch that equalizes GUD between the full tray and micropatch is expected to have a COMP score of 0.5. A gerbil with no information that treats the trays the same is expected to have a COMP of 1.0.
We averaged ratios and COMP scores between tray sets in the same quadrant for each night, resulting in a single ratio or COMP score for each quadrant on each night. In the acute phase, we had a total of 23 nights of data collection, with two quadrants infected and two quadrants uninfected, for a total of 23 nightly mean ratios and COMP scores for the infected gerbils and 23 for the acutely infected gerbils. In the chronic phase, we had a total of 22 nights of data collection with one quadrant of chronically infected gerbils and two quadrants of uninfected gerbils. This resulted in a total of 11 nightly mean ratios and COMPs for the chronically infected gerbils and 22 for the uninfected gerbils. The grand total of nightly mean COMPs and ratios for the uninfected gerbils was 45 (23 + 22).
To determine if infection status affected the equalization of GUDs, we fitted a linear model with the ratio of GUD6g: GUD3g as a dependent variable, infection status as an independent variable, and moon phase and minimum nightly temperature as covariates to control for their potential effects (Kotler et al., 2010). We then ran a type II ANOVA on the model to assess the significance of the effect of infection on ratio. We evaluated the effect of infection status on COMPs by fitting a linear model with ln(COMP) as a dependent variable, infection status, number of vague boundaries, and their interaction as independent variables, and moon phase and minimum nightly temperature as covariates. We ran a type III ANOVA on the model to determine the significance of number of vague boundaries, infection status, and their interaction on COMPs. Analyses were conducted in R (version 4.3.0) using the car package. Means and 95% confidence intervals (Cis) were calculated using Emmeans, and plots were generated with ggplot2.
To assess the effect of infection on GUDs overall, we summed the GUDs of all the trays in a set (this was done for both the micropatch and equalizing experiment) and then fit a linear model to see how infection affected the GUD, with moon phase and night-time minimum temperatures as covariates. We ran a type II ANOVA on the model to determine the significance of the effect of infection on giving up density. We had significant death in one of the acute infection quadrants during the acute phase that resulted in only two survivors out of the six animals initially released. Possibly because of this, the quadrant had higher GUDs than the other quadrants, and we eliminated this quadrant from the analysis of overall GUDs.
Results
Effect of infection status on equalization of GUDs
We found no significant effects of infection status on GUD equalization (Table 1, Figure 1, Ratio Uninfected = 1.36, 95%CI = 1.25–1.47; Ratio acute infection = 1.36, 95%CI = 1.21–1.51, Ratio Chronic infection = 1.32, 95%CI = 1.10–1.53). We also found no significant effects of infection status on the sum GUD in the equalizing experiment, although there was a non-significant trend towards chronically infected gerbils having lower GUDs than acutely infected gerbils (Table 2, Figure 2; GUD Uninfected= 0.829 g, 95%CI = 0.698–0.960 g, Acute infection = 1.082 g, 95%CI = 0.825–1.338 g, Chronic infection = 0.664 g, 95%CI = 0.401–0.928 g).
GUD ratio across the three infection statuses. Each point represents the mean ratio of a quadrant with the given treatment on one night. The dashed line at 1.0 is the expected ratio if the gerbils equalized GUDs perfectly. The dotted line at 2.0 is the expected ratio if the gerbils spent equal time in the trays regardless of richness.
Citation: Israel Journal of Ecology and Evolution 70, 4 (2024) ; 10.1163/22244662-bja10082
ANOVA table for the effect of infection status on GUD ratio between rich and poor trays.
Citation: Israel Journal of Ecology and Evolution 70, 4 (2024) ; 10.1163/22244662-bja10082
ANOVA table for the effect of infection status on the sum GUD per tray set in the equalization experiment and micropatch experiment.
Citation: Israel Journal of Ecology and Evolution 70, 4 (2024) ; 10.1163/22244662-bja10082
Effect of infection status on micropatch detection
We found no significant effect of infection status on COMP scores (Table 3, Figure 3; COMP uninfected = 0.622, 95%CI = 0.580–0.667, COMP Acute infection = 0.688, 95%CI = 0.621–0.761, COMP chronic infection = 0.662, 95%CI = 0.568–0.770, back-transformed from log). COMP scores were higher with two vague boundaries than one vague boundary (Table 3; COMP 1 boundary trays = 0.604, 95%CI 0.580–0.657, back-transformed from log; COMP 2 boundary trays = 0.679, 95%CI 0.632–0.730, back-transformed from log, contrast COMP1/COMP2 = 0.857, 95%CI = 0.750–0.963). However, while this increase primarily seems to have occurred in the chronically infected gerbils (Figure 3), the infection by boundary number interaction was not significant (Table 3). There was not a significant effect of infection status, boundary number, or their interaction on the GUDs at tray sets (Table 2, Figure 4).
ANOVA for the effect of infection status and a number of vague boundaries on COMP scores in micro-patch detection experiment.1
Citation: Israel Journal of Ecology and Evolution 70, 4 (2024) ; 10.1163/22244662-bja10082
The sum of GUDs in rich and poor patches by infection status in the equalization experiment. Each point is the mean GUD for the three tray sets in each quadrant per night.
Citation: Israel Journal of Ecology and Evolution 70, 4 (2024) ; 10.1163/22244662-bja10082
COMP scores for 1 boundary (white boxes) and 2 boundary (grey boxes) trays across infection statuses. Each point is the mean COMP score for the three tray sets in a quadrant per night. COMP = GUD micropatch tray/GUD full tray. 1.0 (dotted line) is the expected value if the gerbils do not detect the micropatch and treat the micropatch trays the same as the full trays. 0.5 (dashed line) is the expected COMP if the animals detect the micropatches and equalize GUDs between the micropatch and full tray.
Citation: Israel Journal of Ecology and Evolution 70, 4 (2024) ; 10.1163/22244662-bja10082
The sum of GUDs in the three trays of the micropatch experiment by infection status. Each point is the mean GUD for the three tray sets in each quadrant per night.
Citation: Israel Journal of Ecology and Evolution 70, 4 (2024) ; 10.1163/22244662-bja10082
Discussion
Overall, our results did not support the prediction that Mycoplasma-infected gerbils, either in the acute or chronic phase of infection, are worse at equalizing GUDs or detecting micropatches (Table 1 and 3, Figure 1 and 3). Thus, our hypothesis that gerbils experience cognitive impairment as a result of infection was not supported. We also failed to observe an increase in GUDs in infected animals (Table 2, Figure 2 and 4). Thus, we were also unable to support the hypothesis that infected gerbils have higher foraging costs than uninfected gerbils.
The performance of the gerbils in both experiments was consistent with prior experiments. On the GUD equalization task, the gerbils underutilized rich patches but equalized GUDs better than expected from a fixed-time strategy. This result is similar to results from pocket mice and ground squirrels. It confirms that Allenby’s gerbils use Bayesian foraging strategy to determine how much time they spend in patches of varying richness (Valone and Brown, 1989). On the micropatch task, the COMPs of the one boundary trays (0.604) and two boundary trays (0.679) similarly confirmed previous results that rodents are capable of detecting vague patch boundaries, but that their ability to do so decreases as the number of vague boundaries increases. However, the COMPs we observed in this experiment were lower than what was observed by Fierer and Kotler (2000) for G. allenbyi. Gerbils search more systematically for food when they are not apprehensive about predation (Dall et al., 2001), and the improved performance we observed suggests that the gerbils in this experiment were less concerned about predation. The lack of concern about predation was possibly a combination of the fact we used lab-raised gerbils with no experience with predation and the time of year we ran the experiment. Gerbils increase foraging and decrease responsiveness to predators during the winter and spring, corresponding to their breeding season (Turovets, 2023).
Neither this study nor Makin et al. (2021) observed an increase in GUDs in infected gerbils, despite a substantial increase in Makin et al. (2020). The key difference between these studies is the presence of a predator in Makin et al. (2020), and this suggests ecological costs of Mycoplasma infection may only occur in the presence of predators. One way this can occur is for the pathogen to reduce anti-predator behaviour. Toxoplasma gondii, for instance, reduces anxiety and behavioural responses to predation cues (Berdoy et al., 2000; Gonzalez et al., 2007) but does not reduce performance on cognitive tasks (Gulinello et al., 2010). Similarly, parasitized sticklebacks are less cautious around predators (Giles, 1987; Milinski, 1985), and finches infected with Mycoplasma show fewer anti-predator behaviours (Adelman et al., 2017). Infected gerbils show less of a change in their foraging behaviour when predators are present than uninfected gerbils, which suggests they may be worse at detecting if predators are present (Makin et al., 2020). We also did not collect time in tray data in this study. Thus, even though the infected animals ultimately had similar success to the control animals in equalizing GUDs and detecting micropatches, we cannot rule out that it took them longer to assess patch richness or patch boundaries. This would not have much of an effect in the predator-free environment of this experiment but could leave infected animals more vulnerable to predation when predators are present.
Another possibility is that we failed to detect an effect because the infection was not severe enough to affect foraging. While we confirmed that animals were infected, we did not measure the intensity of infection. Some evidence exists that the ecological consequences of infection may depend on the severity of infection. For example, red grouse that are severely infected with nematodes have higher mortality from predation than those with lower parasite loads (Hudson et al., 1992). There is less research on how the severity of Mycoplasma infection affects gerbil behaviour. Still, Mycoplasma infection loads vary across seasons (Cohen et al., 2018), suggesting this may be an avenue for future study.
Combined research on the interactions between Mycoplasma infection, parasites, and predators suggests complex interactions may exist in the wild that are difficult to replicate in controlled settings. For example, Messika et al. (2017) found that fleas produce a larger number of smaller offspring when feeding on Mycoplasma-infected gerbils. Gerbils infested with fleas are less able to use vigilance to detect predators and respond by decreasing the amount of time they spend foraging when predators are present (Raveh et al., 2011). Thus, infection with Mycoplasma could affect how likely gerbils are to encounter fleas, which in turn may alter their foraging behaviour and their behaviour around predators. These complex interactions will require further study to unravel.
Overall, we did not find evidence that Mycoplasma infection decreases the cognitive or foraging abilities of Allenby’s gerbils. These results suggest that either the ecological costs of Mycoplasma infection are constrained to situations when a predator is present or that the costs depend on the severity of the infection.
Supplementary material
Supplementary material can be found online at https://doi.org/10.6084/m9.figshare.25688730
Supplementary methods. Inoculation procedures, PCR, PCR post vivarium.
Table S1. List of donor and recipient gerbils from a single source (Ga-255).
Table S2. List of donor gerbils that inoculated gerbils for vivarium experiment and gerbils that received saline for the vivarium experiment.
Figure S1. Ga-624 and 625 show bands eight days post inoculation using HM-16s (489 bp) primers.
Figure S2. Ga- 624 and 625 show bands eight days post inoculation using HM-16s (180 bp).
Figure S3. Nested PCR on DNA samples from saline inoculated gerbils using HM-16s 180 bp primers.
Figure S4. Nested PCR on Mycoplasma positive gerbils.
Acknowledgements
We thank Gideon Grafi for providing lab space and materials for PCR and gel electrophoresis. We also thank Michal Turovets for helping conduct the vivarium experiments and Stu Summerfield for providing technical assistance. This work was funded by a grant from the Israel Science Foundation, Effects of the microbe Mycoplasma haemomuris-like bacterium on the foraging behaviour, risk management, and species interactions of desert gerbils awarded to BPK.
References
Adelman, J. S., Mayer, C. and Hawley, D. M. (2017). Infection reduces anti-predator behaviors in house finches. Journal of Avian Biology 48(4): 519–528. https://doi.org/10.1111/jav.01058.
Berdoy, M., Webster, J. P. and McDonald, D. W. (2000). Fatal attraction in rats infected with Toxoplasma gondii. Proceedings of the Royal Society B: Biological Sciences 267(1452): 1591–1594. https://doi.org/10.1098/rspb.2000.1182.
Braithwaite V. A., Salkeld D. J., Mcadam H. M., Hockings C. G., Ludlow A. M. and Read A. F. (1998). Spatial and discrimination learning in rodents infected with the nematode Strongyloides ratti. Parasitology 117(2): 145–154. https://doi.org/10.1017/S003118209800290X.
Brown, J. S. (1988). Patch use as an indicator of habitat preference, predation risk and competition. Behavioral Ecology and Sociobiology 22(1): 37–47. https://doi.org/10.1007/BF00395696.
Charnov, E. L. (1976). Optimal foraging, the marginal value theorem. Theoretical Population Biology 9(2): 129–136. https://doi.org/10.1016/0040-5809(76)90040-X.
Cohen, C., Shemesh, M., Garrido, M., Messika, I., Einav, M., Khokhlova, I., Tasker, S. and Hawlena, H. (2018). Haemoplasmas in wild rodents: Routes of transmission and infection dynamics. Molecular Ecology 27(18): 3714–3726. https://doi.org/10.1111/mec.14826.
Dall, S. R. X., Kotler, B. P. and Bouskila, A. (2001). Attention, ‘apprehension’ and gerbils searching in patches. Annales Zoologici Fennici 38(1): 15–23.
Fierer, N. and Kotler, B. P. (2000). Evidence for micropatch partitioning and effects of boundaries on patch use in two species of gerbils. Functional Ecology 14(2): 176–182. https://doi.org/10.1046/j.1365-2435.2000.00405.x.
Garb, J., Kotler, B. P. and Brown, J. S. (2000). Foraging and community consequences of seed size for coexisting Negev Desert granivores. Oikos 88(2): 291–300. https://doi.org/10.1034/j.1600-0706.2000.880207.x.
Gibertini, M., Newton, C., Friedman, H. and Klein, T. W. (1995). Spatial learning impairment in mice infected with Legionella pneumophila or administered exogenous interleukin-1-β. Brain, Behavior and Immunity 9(2): 113–128. https://doi.org/10.1006/brbi.1995.1012.
Giles, N. (1987 ). Predation risk and reduced foraging activity in fish: experiments with parasitized and non-parasitized three-spined sticklebacks, Gasterosteus aculeatus L. Journal of Fish Biology 31(1): 37–44. https://doi.org/10.1111/j.1095-8649.1987.tb05212.x.
Gonzalez, L. E., Rojnik, B., Urrea, F., Urdaneta, H., Petrosino, P., Colasante, C., Pino, S. and Hernandez, L. (2007). Toxoplasma gondii infection lower anxiety as measured in the plus-maze and social interaction tests in rats. A behavioral analysis. Behavioural Brain Research 177(1): 70–79. https://doi.org/10.1016/j.bbr.2006.11.012.
Green, R. F. (1984). Stopping rules for optimal foragers. American Naturalist 123(1): 30–43. https://doi.org/10.1086/284184.
Gulinello, M., Acquarone, M., Kim, J. H., Spray, D. C., Barbosa, H. S., Sellers, R., Tanowitz, H. B. and Weiss, L. M. (2010). Acquired infection with Toxoplasma gondii in adult mice results in sensorimotor deficits but normal cognitive behavior despite widespread brain pathology. Microbes and Infection 12(7): 528–537. https://doi.org/10.1016/j.micinf.2010.03.009.
Hanlon, S. M., Lynch, K. J., Kerby, J. and Parris, M. J. (2015). Batrachochytrium dendrobatidis exposure effects on foraging efficiencies and body size in anuran tadpoles. Diseases of Aquatic Organism. 112(3): 237–242. https://doi.org/10.3354/dao02810.
Hotchkiss, E. R., Davis, A. K., Cherry, J. J. and Altizer, S. (2005). Mycoplasmal conjunctivitis and the behavior of wild house finches (Carpodacus mexicanus) at bird feeders. Bird Behavior 17(1): 1–8.
Hudson, P. J., Dobson, A. P. and Newborn, D. (1992). Do parasites make prey vulnerable to predation? Red Grouse and parasites. Journal of Animal Ecology 61(3): 681–692. https://doi.org/10.2307/5623.
Kavaliers, M. and Colwell, D. D. (1995). Reduced spatial learning in mice infected with the nematode, Heligmosomoides polygyrus. Parasitology 110(5): 591–597. https://doi.org/DOI:10.1017/S0031182000065318.
Kotler, B. P., Brown, J. S., Dall, S. R. X., Gresser, S., Ganey, D. and Bouskila, A. (2002). Foraging games between gerbils and their predators: Temporal dynamics of resource depletion and apprehension in gerbils. Evolutionary Ecology Research 4(4): 495–518.
Kotler, B. P., Brown, J., Mukherjee, S., Berger-Tal, O. and Bouskila, A. (2010). Moonlight avoidance in gerbils reveals a sophisticated interplay among time allocation, vigilance and state-dependent foraging. Proceedings of the Royal Society B: Biological Sciences 277(1687): 1469–1474. https://doi.org/10.1098/rspb.2009.2036.
Krebs, J. R., Ryan, J. C. and Charnov, E. L. (1974). Hunting by expectation or optimal foraging? A study of patch use by chickadees. Animal Behaviour 22(4): 953–964. https://doi.org/10.1016/0003-3472(74)90018-9.
Makin, D. F., Kotler, B. P., Brown, J. S., Garrido, M. and Menezes, J. F. S. (2020). The enemy within: How does a bacterium inhibit the foraging aptitude and risk management behavior of allenby’s gerbils? American Naturalist 196(6): 717–729. https://doi.org/10.1086/711397.
Makin, D. F., Kotler, B. P. and Sargunaraj, F. (2021). Does a bacterial infection lead to a nutritional deficiency in foraging Allenby’s gerbils? Israel Journal of Ecology and Evolution 67(3–4): 121–126. https://doi.org/10.1163/22244662-bja10024.
McNair, J. N. (1983). A class of patch use strategies. American Zoologist 23: 303–313.
Messika, I., Garrido, M., Kedem, H., China, V., Gavish, Y., Dong, Q., Fuqua, C., Clay, K. and Hawlena, H. (2017). From endosymbionts to host communities: factors determining the reproductive success of arthropod vectors. Oecologia 184(4): 859–871. https://doi.org/10.1007/s00442-017-3906-4.
Milinski, M. (1985). Risk of predation of parasitized sticklebacks (Gasterosteus aculeatus L.) under competition for food. Behaviour 93(1): 203–216.
Raveh, A., Kotler, B. P., Abramsky, Z. and Krasnov, B. R. (2011). Driven to distraction: Detecting the hidden costs of flea parasitism through foraging behaviour in gerbils. Ecology Letters 14(1): 47–51. https://doi.org/10.1111/j.1461-0248.2010.01549.x.
Schleich, C. E., Zenuto, R. R. and Cutrera, A. P. (2015). Immune challenge but not dietary restriction affects spatial learning in the wild subterranean rodent Ctenomys talarum. Physiology & Behavior 139: 150–156. https://doi.org/https://doi.org/10.1016/j.physbeh.2014.11.023.
Schmidt, K. A. and Brown, J. S. (1996). Patch assessment in fox squirrels: The role of resource density, patch size and patch boundaries. American Naturalist 147(3): 360–380. https://doi.org/10.1086/285856.
Townsend, A. K., Sewall, K. B., Leonard, A. S. and Hawley, D. M. (2022). Infectious disease and cognition in wild populations. Trends in Ecology and Evolution 37(10): 899–910. https://doi.org/10.1016/j.tree.2022.06.005.
Turovets, M. (2023). Effects of season and sex on foraging aptitude and risk management in Allenby’s gerbil (Gerbillus andersoni allenbyi). M.S. Thesis, Mitrani Department of Desert Ecology, Ben-Gurion University of the Negev, Israel.
Valone, T. J. and Brown, J. S. (1989). Measuring patch assessment abilities of desert granivores. Ecology 70(6): 1800–1810. https://doi.org/https://doi.org/10.2307/1938113.
Venesky, M. D., Parris, M. J. and Storfer, A. (2009). Impacts of Batrachochytrium dendrobatidis infection on tadpole foraging performance. EcoHealth 6(4): 565–575. https://doi.org/10.1007/s10393-009-0272-7.
