Group size influences behavioral plasticity in responses to thermoregulation-foraging trade-offs by a socially cohesive bird

Behavioral plasticity, such as changes in habitat use and activity, can be a critical modulator of thermal pressures on endotherms. However, shifts in behaviors can meet diversified costs such as missed feeding opportunities. Individual decision-making should therefore capture the trade-off between the costs and the benefits of thermoregulation. In the case of social species, the decision process could also be facilitated or slowed down by the modulation of costs arising through the social environment. In this study, we tested how vulturine guineafowl (Acryllium vulturinum) change their use of open areas (where they predominately forage) according to heat using GPS data from 105 birds collected every 5 minutes for 6 months. Because animals vary in their sensitivity to risk according to group size—e.g. due to the dilution effect—we compared the responses of individuals depending on the size of the group they belong to. We also analyzed if behavioral responses translate into less precise thermoregulation by recording the body temperature of a subset of birds in two groups. We foundd that birds avoid heat by selectively using open areas and moving to cover, as well as reducing activity when temperature increases, birds use open areas less and move less. Individuals from intermediate-sized groups seemed to be able to use open areas during warmer conditions compared to individuals from small and large groups (10% higher probability of use). However, active birds in the open did not present hyperthermia, suggesting that behavioral changes are efficient or that individuals have other efficient strategies such as physiological cooling. Together, these results indicated that responses to high temperatures are complex, as they included not only a range of environmental constraints but that responses can also vary according to social context.

behavioral thermoregulation – group size – heat – movement ecology – foraging 3 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint

Climate change is predicted to not only drive an increase in average temperature but also to increase the frequency and the intensity of the occurrence of extremely high temperatures (IPCC 2021). High temperatures can strongly constrain the physiology of organisms, ultimately leading to lethal hyperthermia and dehydration if temperatures exceed the organism’s performance range (J. B. Williams and Tieleman 2005; Angilletta 2009; McKechnie and Wolf 2010). Animals can, however, use short-term behavioral adjustments which buffer the effect of high temperatures on physiology (Marais and Chown 2008; Huey et al. 2012; Long et al. 2014; Sunday et al. 2014). This behavioral plasticity, often called “behavioral buffering”, can range from postural changes limiting the exposure of the body to hot surfaces or solar radiation, to changes in habitat use and activity patterns (Cunningham, Martin, and Hockey 2015). Typically, the heterogeneity of the landscape—featuring a mosaic of sunny hot micro-habitats and shaded shelter—can allow animals to avoid high temperatures and facilitate thermoregulation by selecting microhabitats close to their physiological thermal preferences (Long et al. 2014; Sears et al. 2016). This flexibility is likely to make behavior particularly critical for ectotherm thermoregulation (Bauwens, Hertz, and Castilla 1996; Angilletta 2009). It is also a modulator of body temperature in endotherms by reducing water losses through evaporative cooling (Wolf and Walsberg 1996; Cain III et al. 2006; Hetem et al. 2012; Smit et al. 2016; Oswald et al. 2019). Thus, in principle animals can maintain their body temperature constant in an optimal range through behavior when the environment allows them to. 4 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Behavioral thermoregulation is, however, energetically costly (Angilletta 2009). In addition to energetic expenditure in movement, shifts in activity patterns and micro-habitat use in response to heat can lead to “missed-opportunity” costs, such as a suboptimal use of food or water resources, an increased exposure to predators, or a reduction in the expression of reproductive or social behaviors—especially when behavioral decisions are mutually exclusive (du Plessis et al. 2012; Cunningham, Gardner, and Martin 2021; Stofberg et al. 2022). When faced with such thermoregulation trades-offs, we expect behavioral decisions to maximize the overall benefits and minimize overall costs of the constraints of the environment. As a consequence, a behavior might appear to be suboptimal when studied under the frame of thermoregulation alone. Examples of this are large savanna grazers that avoid foraging in cooler periods of the day when faced with higher predation pressure (Veldhuis et al. 2020) and lizards that shift their micro- habitat preferences to cooler and moister places under water restriction at the cost of taking advantage of better thermal conditions for thermoregulation (Rozen Rechels et al. 2019; 2020)‐ . But, in other cases, heat avoidance must be prioritized. For example, southern yellow-billed hornbills (Tockus leucomelas) and Alpine ibex (Capra ibex) engage in thermorgulation at the expense of foraging and body condition (Mason et al. 2017; van de Ven, McKechnie, and Cunningham 2019). Therefore, behavioral adjustments for thermoregulatory needs under heat stress are not trivial to predict, especially when those implies a trade-off with meeting foraging needs. If heat avoidance is prioritized in the animal’s decision-making, we expect body temperature to remain constant and close to the physiological optimum. However, the costs associated with 5 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint water losses under high temperatures might enhance different physiological responses. Water conservation behaviors for example would reduce dehydration at the expense of suboptimal body temperatures (Anderson and Andrade 2017; Rozen Rechels et al. 2019; 2020)‐ . This is not only true in ectotherms but also in endotherms, if the latter can relax the need to maintain a precise body temperature under thermal conditions that challenge their water and energetic balances (Wooden and Walsberg 2002; Hetem et al. 2016; McKechnie and Wolf 2019). Some arid habitat birds could even tolerate a facultative hyperthermia, thus limiting dehydration (Gerson et al. 2019). The same way as ectotherms’ behavioral strategies could be studied indirectly through the analyses of individuals body temperatures (Blouin-Demers and Nadeau 2005; Rozen Rechels et ‐ al. 2021), the efficiency of endotherms’ behavioral decision facing heat could also be understood by studying the variations of their body temperatures (Thompson, Cunningham, and McKechnie 2018). Until recently, behavioral responses to heat were predominately studied in solitary individuals. While this is relevant in many species, it neglects that social interactions are critical in movement and habitat use decision-making processes (Couzin et al. 2005; Strandburg-Peshkin et al. 2015; Hansen et al. 2024). Conspecifics can have a substantial influence on foraging behaviors. Individuals in groups benefit from information sharing, which can not only improve access to resources such as food but also thermal refuges in heterogeneous environments (Dall et al. 2005; Cantor, Aplin, and Farine 2020). Group living also dilutes predation pressures and allows the emergence of collective antipredator strategies such as sentinel behaviors and mobbing (Kenward 1978; Graw and Manser 2007; Lehtonen and Jaatinen 2016). We might then expect 6 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint food, temperature and predation constraints to synergistically favor larger groups if these can better offset trade-offs with predation. However, larger groups also tend to slow down the decision-making process and reduce the group movement speed due to coordination challenges (Herbert-Read et al. 2013; Strandburg-Peshkin et al. 2015; 2017; Papageorgiou and Farine 2020b; Klarevas-Irby, Nyaguthii, and Farine 2025), potentially limiting their ability to take advantage of heterogeneous landscape features. The trade-off between group advantages in information sharing and protection, and the constraints in movement, might then favor an intermediate—or ‘optimal’—group size. However, group size effects on micro-habitat selection and behavioral buffering of heat stress have not been explored. In this study, we investigated the interplay between foraging and heat constraints on the habitat use of a large social bird, the vulturine guineafowl Acryllium vulturinum (Hardwicke 1834). Given their relatively large body size (1.5–1.8 kg), we expected heat dissipation to be challenging for vulturine guineafowl and, thus, that they should rely on behavior for thermoregulation (Smit et al. 2016; Pattinson et al. 2020) as has been shown in the closely related helmeted guineafowl Numida meleagris (Rakowski et al. 2019). However, thermoregulation is likely to be traded off against foraging, the latter being predominately done in open grassy areas exposed to the sun and rich nutrients from being used overnight by large herbivores (Young, Patridge, and Macrae 1995). We used large-scale and long-term data from GPS trackers fitted to approximately 10% of a population of wild vulturine guineafowl to test how birds prioritize between different measures of heat constraints when deciding to use these foraging areas. We also tested whether individuals shift their habitat use differently according to 7 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint the size of their group, balancing the complexity of decision-making with the anti-predation and collective intelligence benefits that larger groups experience. As a result of the trade-off with the ability to effect rapid decision-making, we expected that individuals living in large groups may be more risk-averse in terms of temperature. We also expected that groups of intermediate size will spend more time in the open under heat relative to individuals in small or large groups. Finally, given that vulturine guineafowl are endemic to the arid savannahs of east Africa, we expected to observe birds managing trade-offs under higher heat conditions by exhibiting hyperthermia, which we measure using internal body temperature loggers on a subset of GPS-tracked birds. We lay out all of our predictions in Table 1.

Study system Vulturine guineafowl (hereafter called VGFs) are predominately terrestrial birds that live in highly cohesive social groups ranging from a dozen to over 60 individuals (Papageorgiou et al. 2019). Group membership is stable, but groups can respond to ecological conditions by temporarily splitting up (to breed; Nyaguthii et al. 2025) or by temporarily associating with other groups (during dry periods, Ogino, Strauss, and Farine 2023). Herein our references to ‘group’ correspond to the stable social groups that individuals spend most of their lives in (Farine in review), which reflects a stable social unit at the intermediate level of their multilevel society. In this paper, group size refers to the number of conspecifics that an individual is currently with (which can include individuals from other stable groups). Groups follow shared decision-making 8 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint processes to decide where to move, meaning that rather than following a dominant leader, all individuals can contribute to the group’s next actions (Papageorgiou and Farine 2020a; Papageorgiou, Nyaguthii, and Farine 2024). Their diet is composed of grass root buds, seeds and small arthropods that the birds mainly find on glades, i.e. open areas rich in nutrients (Young, Patridge, and Macrae 1995), showing a higher propensity to forage and greater foraging effort on these open areas (Appendix 1). Our study site is a c. 12km2 dry woody-savanna in the south of Mpala Research Center (MRC), in Laikipia, Kenya. This tropical ecosystem is typically described by two main wet seasons (October to November and from April to June) separated by dry seasons characterized by low rainfall and little to no green vegetation. VGFs home range size as well as the distance moved every day drastically increase during the dry seasons when resources are more clumped into open areas (Papageorgiou et al. 2021). The current study focuses on an extraordinarily long dry period that lasted from December 2020 to July 2021 (Appendix 2: Figure S2). VGFs are a prey species that faces pressure from both aerial (raptors such as martial eagles Polemaetus bellicosus, Heine 1890; tawny eagles Aquila rapax, Temminck 1828; and African hawk-eagles Aquila spilogaster, Bonaparte 1850) and terrestrial (e.g. jackals, hyenas) predators. While encounters with opportunistic terrestrial predators occur supposedly at random (Cooper, Holekamp, and Smale 1999; Moehlman 2014), i.e. indiscriminately in open and covered patches, VGFs are particularly visible (and sensitive) to soaring predators (raptors) in open areas. 9 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint GPS tracking data From early February 2021 to the end of July 2021, high-resolution solar-powered GPS tags (15g Bird Solar, e-obs Digital Telemetry, Grünwald, Germany) were fitted to 118 different individuals (from both sexes and all adults) across 20 identified social groups. In order to prevent feathers from covering the tags’ solar panel, loggers were elevated using neoprene pads. The tag, Teflon harness, and the platform used to raise the tag above the feathers together weighted c. 20.5g, less than 2% of the weight of the bird (below the 3% recommendation for animal welfare, see Bodey et al. 2018; Portugal and White 2018 for further discussion). The birds tracked in this study were tagged between October 2016 and June 2021. Data were downloaded remotely every 2 days using a BaseStation II (e-obs Digital Telemetry, Grünwald, Germany) and uploaded on the Movebank repository (Kays et al. 2022). The majority of the birds in the study population (including all GPS-tagged birds) were fitted with a unique combination of colour leg bands, allowing identification in the field. We excluded data collected on the day of deployment for birds that joined the study along the way, as well as data collected during periods when a bird’s group was being trapped. We also excluded birds living in the group habituated to human living in the research center as they use a fenced, highly vegetated area with different predation pressure and resource availability, as well as subadult birds which were tagged as part of a study on dispersal (Klarevas Irby, Wikelski, and ‐ Farine 2021). We collected individual locations from 07:00 to 18:00 (EAT; when the birds are out of the roosts), with a burst of 10 locations (1Hz) every 5 minutes. We identified and excluded outliers’ location when two consecutive locations were farther than 600m (this threshold have 10 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint been defined by visualizing the tracks and observing the distribution of consecutive locations distance). In this project, we only kept the 10th location of each burst which is supposed to be the most accurate one (He et al. 2023), thus down-sampling our tracking data to one measure every 5 minutes. Finally, we removed birds with fewer than 500 total locations (3 birds). Our final dataset consisted of movement trajectories from 105 different birds. Community composition and group size Birds were censused daily by driving along a network of vehicle tracks crossing the study area. When encountering a cohesive set of individuals, we recorded the number of individuals and identified all the marked birds present (or until birds moved into vegetation). Membership can be dynamic during dry periods—such as in our study period—as the long-term stable groups can temporarily merge as part of the multilevel society. Thus, we estimated the group size experienced by each tagged individual each month using a network community detection approach as described in Ogino, Strauss, and Farine (2023). Specifically, we extracted in census data all observations in which tagged birds were identified each month. For each observation, we compared the birds identified to the network community membership (based on all the data, per Ogino et al. 2023) and counted (1) the number of observed birds that belong to the community that month, (2) the number of observed birds that do not belong to the community that month and (3) the number of known birds in the community based on community detection. We calculated the ratio (1) / [(2) + (3)] and defined each individual’s group size based on the number of individuals identified in their observation with the highest ratio. If two group observations showed the best ratio, we used the observation with the maximum number of birds. If an 11 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint individual had not been censused during a given month, we predicted its group size using linear interpolation from the month before to the month after. If it had not been observed the month before or after, we considered group size to be the same as the one counted when the data was available. We then defined “small groups” for a given month to be those with group sizes in the lowest tercile (average: 34 ± 9 SD individuals, from 16 to 47 individuals), “big groups” to be those with group sizes in the upper tercile (average: 74 ± 13 SD individuals, from 58 to 109 individuals), and “median groups” those with intermediate group sizes (average: 52 ± 3 SD individuals, from 48 to 57 individuals). Body temperature Between February 25th and March 1st 2021, 23 VGFs from two groups were implanted with an ECG-logger (ECG-tag 1AA2, e-obs Digital Telemetry, Grünwald, Germany; 25g; 38 x 23 x 19mm) in addition to GPS tags (the sum of the weights of the loggers ranged from 2.6 to 3.4% of the weight of the birds). ECG-loggers record heart-rate data (not used in this study, see Brandl et al. 2025) and internal body temperature with an inaccuracy of maximum 0.5°C. The ECG logger was implanted in the thoraco-abdominal cavity of anesthetized birds. The logger was fixed to the abdominal wall with an absorbable suture (Monosyn® 4/0, B. Braun AG, Melsungen, Germany), with the longer electrode placed close to the heart. All implantation details are described in Brandl et al. (2025) Data were downloaded remotely every fourth night. We excluded 5 birds that were predated or died in the month following implantation. Five loggers that showed unexpected body temperature 12 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint records below 34°C or higher than 47°C were also excluded from the analysis. Body temperature was sometimes recorded once every 5 minutes or once every 20 seconds. We homogenized the dataset by averaging body temperature every 5 minutes in the second case, hereafter called Tbody (103679 occurrences for 16 birds, Appendix 3: Figure S1). We then matched Tbody with recorded locations of 14 birds (66399 locations as some birds’ locations were only recorded once every 4 days and the locations of one bird were not recorded during the study due to a faulty GPS tag). For each individual we then calculated the resting body temperature which we defined as the average body temperature of the individual when it is resting in the shade across the whole study period. A bird was defined as inactive when its speed (here distance moved between two locations divided by the location) was below 0.01 m/s (i.e. less than 3m moved in 5 minutes, Appendix 3: Figure S2) and it was located under cover. We did not use instant speed as body temperature would not change as fast as a bird would change its behavior. This measure of activity is thus an integration of what the bird was doing in the last 5 minutes. For active birds, we then calculated the deviation between Tbody and the average resting body temperature per individual, hereafter called the body temperature anomaly ΔTbody. Remote sensing and landscape description We classified habitat within the home range of our study population as open (where VGFs can find food but where they are also subject to both aerial predation pressure and higher thermal constraints) and closed (with lower food availability but also lower predation and thermal constraints). To do so, we used the ESA Sentinel-2 multi-spectral images to calculate vegetation presence metrics such as NDVI (Papageorgiou et al. 2021). We downloaded every 10m- 13 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint resolution satellite image of MRC (tile 37N BA) since they were recorded and made available (October 2017) until the end of this project (early August 2021) and cropped it to restrict the analysis to the zone used by the individuals of our study. Some additional parts of the landscape, which contain black cotton soils, were also removed because of darker soil color made habitat classification less accurate (Appendix 2: Figure S1); the parts removed contained 15100 GPS locations of 20 different individuals (i.e. 1.4% of the dataset). Clouds were cropped from the image and images with more than 20% cloud cover were removed from the time series. We calculated NDVI raster maps (Normalized Difference Vegetation Index, eqn. 1), NDWI maps (Normalized Difference Water Index, water content in vegetation, eqn. 2, Gao 1996) and brightness maps (eqn. 3, Valero et al. 2016) maps for each satellite image. with NIR meaning Near InfraRed, SWIR meaning Short-Wave InfraRed. We then summarized these indices from October 2017 to the beginning of August 2021 into average, standard deviation, minimum, maximum and median raster maps. We then performed a simple Random Forest procedure to classify each pixel into either open or closed habitat (Breiman 2001). We observed that the landscape of our study gives access to open and cover zones all over its area at the VGF movement scale, i.e. all groups can equally access both habitats wherever they are (Appendix 2: Figure S4 and S5). We also calculated the proportion of open pixels in a buffer of 14 NDVI = NIR − Red NIR+Red Eqn. 1 NDWI = NIR − SWIR NIR+SWIR Eqn. 2 Brig h tness=√Green 2 +Red 2 +NIR 2 +SWIR 2 Eqn. 3 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint 120m-radius (99%-quantile of the distances walked by birds in 5 minutes) centered on each focal pixel, hereafter called “open patch availability”. This value allowed us to account for open vs. cover availability around each bird location in statistical models. The whole procedure is detailed in Appendix 2 and code is provided in Zenodo (the url will be provided upon publication). Operative temperature variations To approach the thermal constraints experienced by VGFs in cover and in the open, we measured proxies of operative temperature (Dzialowski 2005). “Operative temperatures” are measures of the temperature of an object with the same thermal properties as the target organism, integrating heat exchanges such as conduction, convection, solar radiation, and wind cooling, thus capturing the body temperature of the organism without thermoregulation. Black copper spheres allow the measurement of relatively good proxies of operative temperatures (Walsberg and Weathers 1986) despite several issues making them unsuitable for precise thermal physiology studies (Bakken and Angilletta 2014). As our measurement errors were homogeneously distributed (i.e. because we made contrasts between microhabitats) these should not have impacted our conclusions. Technical limitations with materials available at the field site led us to build hollow black iron cylinders which we painted with white dots to mimic VGFs (ca 20x20cm, hereafter called “operative models”, Figure 1A). Operative models were made of two parts: the lower one was welded to a ca 35cm iron bar which will be inserted 15cm deep in the soil in order to maintain the model ca 20cm above it such as a VGF body. We stretched and fixed a gauze across the center of each cylinder on which we mounted a DS1922L iButton (Maxim Integrated) temperature logger programmed to record temperature every 5 minutes. We deployed 20 15 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint operative models across the landscape used by the birds, during the whole duration of the study, set up in pairs space a few meters apart such that one model was in the sun and the other in the shade (Figure 1A). iButtons were replaced and downloaded every two weeks. Damaged operative models (e.g. by elephants) were repaired or replaced throughout the study and data collected by these models were excluded from the last time when it had been observed intact to the time of replacement. Data were visually monitored in order to identify operative models which could have been in the wrong micro-climatic conditions at some hours of the day (in the sun instead of shade for example). In this case, data were excluded throughout the period when the operative model was not at the right place (this could change during the course of the study due to changes in the sun course). We then calculated time series of average operative temperature in the sun (Tsun, Figure 1B and Appendix 4: Figure S1) and in the shade as well as time series of the difference of average operative temperature in the sun and in the shade (Tdiff) every 5 minutes. Each bird location was then associated to an instant and concomitant Tsun and Tdiff averaged for all operative temperature models. We used instant temperature conditions (instead of hourly or other type of averaged temperature conditions) to predict shifts between open and cover habitats in this study to detect possible responses to short-term drop in temperature as the consequence transient cloud cover. We also calculated the average Tsun and Tdiff in the last 15 minutes and in the last hour before each location record to test the robustness of our results to the temporal scale of thermal conditions variables. Statistical analyses 16 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint All statistical analyses were conducted in R (4.2.1 “Funny-Looking Kid”). Generalized linear mixed models (LMM or GLMM) were fitted with the “glmmTMB” library. Models in model comparison procedures were fitted using a Maximum Likelihood (ML) estimation. Final models were analyzed after being fitted with a Restricted Maximum Likelihood (REML) estimation. Residuals distribution and homoscedasticity as well as the leverage effects of outliers or overdispersion of data were assessed in the selected (final) models with the “DHARMa” library. If we detected a significant outliers leverage effect, we tested the robustness of our results after outlier removal. If we detected a significant overdispersion of data in binomial error models, we added an observation-level random effect and tested the robustness of our results after correction. In both outliers and overdispersion case, if the problem was not solved after correction but results remained qualitatively the same, we considered our model to be robust. We visually analyzed QQ- plots considering that the very large size of our data would always lead to significant difference between the residuals distribution and a Gaussian distribution. Variations of habitat use We analyzed which prediction (see Table 1) best explained the probability of being in the open (i.e. under sun; 1,018,675 locations, 9909 individual.day trajectories), the probability of shifting from a closed habitat to the open (562639 locations), and the probability of shifting from the open to a closed habitat (365684 locations). In all cases, we first tested if open patch availability around individuals significantly explained the variations of the response variable with a likelihood ratio test. If so, it was added in the null model of our model comparison. We then compared what were the best predictors of the variations of the response variable by comparing 17 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint the null model, the models including the linear variation of average operative temperature in the sun Tsun, and the difference of average operative temperature the sun and the shade Tdiff, the models including the quadratic variations of Tsun, and Tdiff, and the same models including the interaction of the environmental variable with estimated group size (Table 1). When estimating the probability of shifting from a closed habitat to the open or the probability of shifting from the open to a closed habitat, we fitted the variations of Tsun, Tdiff and open patch availability of the previous location (5 minutes earlier) considering that these were more informative of the decision of the bird to shift habitat or not. The best model was identified based on AICc with the “aictab” function from the “AICcmodavg” library (Mazerolle 2025). We accounted for inter- individual and inter-group variations in all response variables by fitting an individual random effect and a group random effect (additively as an individual could change group during the study). Note that we tried to account for temporal autocorrelation between the successive locations of an individual during a day with an AR1 covariance structure when fitting the probability of being in the open. We removed it from models at the end as models did not manage to converge when both temporal autocorrelation and open patch availability were included in the model; our results were robust to the use of one or the other. Probability of being in the open and shifting from one habitat to another were fitted with GLMMs with a binomial error. The same procedures have been run with Tsun and Tdiff averaged in the 15 minutes and in the last hour before each location record. Variations of distance moved 18 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint We calculated the distance moved in one hour by summing the distance between all consecutive locations in that hour to reduce the impact of any noise (e.g. running away from a predator) on the response variable (82324 measurements). The distance moved was square-rooted to approach a Gaussian distribution and fitted with an LMM with a Gaussian error. Tsun and Tdiff were averaged per hour, then scaled and centered to facilitate model convergence. Only hour-bouts including 6 or more locations were included and the number of locations in the hour-bout was included in the model, as the fewer the locations the shortest one single displacement may look like (McCann et al. 2021). Model comparison protocol and random effects included in the model were similar as used in the models fitting shifts from one habitat to another. Variations of body temperature anomaly In order to investigate if being active in the sun might impose a cost on thermoregulation, we analyzed the variations of ΔTbody as a function of the quadratic variations of Tdiff (as an index of the different temperatures available in the environment) in interaction with the habitat. We accounted for inter-individual variability with a random effect of the individual on the intercept but also for potential effects of the meteorology with a date random effect on the intercept. Tdiff was scaled and centered in order to facilitate model convergence.

Model comparisons results are listed in Appendix 5. The temporal scale of our temperature measurements does not impact our results Appendix 6 and 7. 19 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Probability of using open habitats The probability of selecting open patches significantly and strongly increased with open patch availability (χ² = 221286, df = 1, p < 0.001; effect: 7.4 ± 0.02). This predicts that if the availability of open patches around the location is very low, the probability of selecting open patches is close to 0, if average it is around 60%, if very high it is almost 100%. We found that the interaction between group size and the instant difference of operative temperatures between sun and shade Tdiff best explained the probability to use open habitats (Appendix 5: Table S1). Specifically, the interactions between group size and both the linear term and the quadratic term of Tdiff significantly explained the variations in the probability of being in the open (group size×Tdiff: χ² = 827.8, df = 2, p < 0.001; group size×Tdiff²: χ² = 14.3, df = 2, p < 0.001). The probability of being in the open was around 60- 65% when temperatures in the sun and in the shade were similar, and decreased with Tdiff (Figure 2A). At high Tdiff, the probability of being in the open was slightly higher for median groups ( c. 25%) than for large and small groups (c. 15%; Figure 2A). Probability of moving from cover to open patch The probability of shifting from a cover to an open patch significantly increased with open patch availability (χ² = 40903, df = 2, p < 0.001; effect: 5.1 ± 0.03). The interaction between group size and Tdiff best explained the probability to shift to the open when under cover (Appendix 5: Table S2), with the interactions between group size and both the linear term and the quadratic term of Tdiff being significant (Tdiff: χ² = 106.8, df = 2, p < 0.001 ; Tdiff²: χ² = 32.4, df = 2, p < 0.001). When temperatures in the sun and in the shade were similar, the probability to transition to the open when in the shade was 20-25%, independently of group size (Figure 2B). This probability dropped 20 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint significantly when Tdiff increased. At high Tdiff, the probability of shifting to the open was around 6% for small and median groups but was around 3% for large groups. At intermediate Tdiff, median sized groups had the highest probability of shifting to open habitats. Probability of moving from open to cover patch The probability of shifting from an open to a cover patch significantly decreased with open patch availability (χ² = 34549, df = 2, p < 0.001; effect: -4.5 ± 0.02). The interaction between group size and the quadratic variations of Tdiff best explained the probability to shift to a cover habitat from open habitats (Appendix 5: Table S3), with the interactions between group size and both the linear term and the quadratic term of Tdiff being significant (Tdiff: χ² = 122.5, df = 2, p < 0.001 ; Tdiff²: χ² = 28.3, df = 2, p < 0.001). The probability to transition under cover when in the open increased with Tdiff until a threshold of c. Tdiff=10°C (Figure 2C). Based on confidence intervals, this probability was significantly higher in median sized groups at intermediate Tdiff. Big groups had a lower probability to transition from the open to cover at low Tdiff compared to median and small groups (Figure 2C). However, small groups and big groups had similar probability to transition from the open to cover at high Tdiff (Figure 2C). Distance moved The interaction between group size and the quadratic variations of Tsun best explained the variations in the hourly distance moved by individuals (Appendix 5: Table S4), with the interactions between group size and both the linear term and the quadratic term of Tsun being significant (Tsun: χ² = 87.2, df = 2, p < 0.001; Tsun²: χ² = 7.6, df = 2, p = 0.02). The distance moved by an individual in one hour 21 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint was maximal between 20 and 25°C across all group sizes (Figure 2D). Individuals in median group sizes moved larger distances than other group sizes between 10 and 35°C, whereas distances moved were similar for all group sizes at higher temperatures (Figure 2D). As expected, the number of locations in an hour-bout was positively and significantly correlated with the distance walked in an hour (χ² = 124.3, df = 2, p < 0.001; effect: 0.16 ± 0.01). Body temperature anomaly Changes in body temperature (ΔTbody) were significantly predicted by the quadratic value of Tdiff in interaction with the habitat ( Tdiff: χ² = 32.8, df = 1, p < 0.001 ; Tdiff²: χ² = 13.8, df = 1, p 0; Figure 3). Δ Tbody increased significantly with Tdiff and increased more steeply for birds active in the open, resulting in a ΔTbody of c. 0.1°C higher for birds that were active in the open than under cover (Figure 3). At high Tdiff, ΔTbody peaked around 0.4°C whatever the habitat where the bird is active (Figure 3).

We showed that VGFs reduce their use of open habitats—which they rely on for foraging—as temperatures in the sun become higher than temperatures in shade. This effect was captured by a lower probability of presence in the open, a lower probability of transitioning from cover to the open, and higher probability of transitioning from the open to cover under elevated temperatures in the sun. The direction of the effects of thermal constraints on the different habitat use and movement patterns align with a heat avoidance strategy. The use of thermal shelters such as 22 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint shade to limit overheating is now well-known and expected in larger birds that cannot rely on physiological responses as easily as small passerine birds (Smit et al. 2016; Oswald et al. 2019; Pattinson et al. 2020). Such behaviors mostly involve limiting the exposure of the organism to solar radiations (Wolf and Walsberg 1996, see the operative temperature variations in Figure 1). The main predictor of changes in thermal conditions is the time of the day, with the highest temperatures recorded in the afternoon. Our results mostly suggest that VGFs take advantage of early hours of the day as well as the late ones, but also during cloud cover (pers. obs.; see June 5th compared to previous days and the following one in Figure 1; also see Appendix 4: Figure S1 for days with more variable weather and more contrasted behavioral responses). In addition, the reduction of the distance moved with increasing temperature indicates that birds spend less time commuting and probably more time resting when foraging patches have high heat exposure, which is consistent with a strategy aiming at avoiding both hot micro-habitats but also when lowering metabolic rate and thus heat production (Clark 1987; Pattinson et al. 2020). We also observed VGFs panting (or doing other heat dissipation behavior such as wing dropping or fluttering) while resting under shade in the hottest hours of the day, as expressed by many other arid habitats birds (Pattinson et al. 2020). This suggests that VGFs compensate foraging activity under heat by more resting and heat dissipation behaviors, as a result of the trade-off between foraging and heat dissipation. However, the lower distance moved observed at lower temperatures (when birds were mostly in the open) also suggests that birds move less when actively foraging in low heat conditions (i.e. because they remain on the relatively small glades). These observations together suggest that physiological responses (especially evaporative water losses, EWL) alone are insufficient for the regulation of body temperature under the high and 23 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint frequent heat conditions of the dry season or that high limitations in water availability during the dry season impose a higher cost on evaporative water losses and shift bird thermoregulation processes from physiology to behavior (Pattinson et al. 2020). Our results also show that active birds have an increase of their body temperature of c. 0.3°C compared to period of inactivity, as a result of the trade-off between heat dissipation and activity. This predicted hyperthermia is, however, small compared to the variations observed in body temperature across the day or amongst individuals, which can range to several degrees (Figure S5). The 0.1°C difference of body temperatures between bouts of activity in the open or under cover also confirm the impact of radiation in the heat constraints experienced by individuals (Wolf and Walsberg 1996). Overall, the range of body temperatures observed in our study do not exceed expected body temperature of such a large bird (McKechnie and Wolf 2019) as well as it is coherent with the range of temperatures observed in a closely related species (Withers and Crowe 1980) and far from the range of temperatures observed under both pathological or facultative hyperthermia (Gerson et al. 2019). The behavioral modulation of heat constraints is thus likely to be efficient at mitigating the costs of heat on the organism. Heat management in VGFs is also likely to be supported by other strategies, such as panting, and morphological adaptations, such as bare heads allowing heat dissipation despite their dark plumage and reduces the costs of foraging under heat (Galván, Palacios, and Negro 2017). However, the consequences of these behaviors in water losses and thus dehydration—and in turn individual fitness—remains a blind spot of our study (Albright et al. 2017; Rozen Rechels et al. 2019; Czenze et al. 2020)‐ . 24 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint We observed that the changes in birds’ behaviors were also modulated by their group size. We found consistent group size effects, with intermediate (i.e. optimal; Bertram 1978) group sizes often expressing more flexible responses (e.g. higher transition probabilities to and from cover; Sibly 1983; Papageorgiou and Farine 2020b). Large groups seemed more inclined to avoid open habitats at high heat constraints while intermediate-size groups (median group sizes) seemed to have more opportunities to move and shift habitats when heat increased, and more particularly at intermediate temperatures. This is consistent with recent observations that intermediate-size groups of the same population are the most mobile (Papageorgiou and Farine 2020b). This is likely due to the time costs and coordination challenged associated with making decisions by larger groups (Davis, Crofoot, and Farine 2022). In other words, when faced with heat constraints in the sun, large groups are likely to be less able to respond quickly by shifting between habitat conditions. By contrast, small groups show similar habitat use patterns as intermediate-size groups under high heat constraints, suggesting that they can benefit from making faster decisions in response to dynamic changes in their (thermal) environment. Overall, we therefore expect that an increase in the length and frequency of hot dry seasons in such savanna habitats could favor smaller groups despite the benefits against predation in large groups. Alternatively, we could also interpret the tendency by larger groups to avoid open patches at higher temperatures as an ability to take a more optimal decision in terms of thermoregulation and avoid warmer habitats more efficiently.

25 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Beyond the result that vulturine guineafowl habitat-use behavior is driven by heat avoidance, which might therefore constrain foraging behaviors, we have shown that the behavioral responses of individuals to heat are complex. One particularly important dimension of our findings is that thermoregulatory behaviours—or strategies—appear to be related to group size. At the relatively short time scale of our study, we found that VGFs cope with heat efficiently, however we lack information on the fitness consequences of these behavioral modulations on individual survival and recruitment (but note that the subsequent drought that continued until 2023 resulted in total suppression of reproductive behaviors due to a lack of dense cover for nesting or the lack of food resources). Thus, our results underline the importance of sociality on individual-level responses to unfavorable weather conditions, a factor that is often overlooked in studies of the responses of individual organisms to heat. A better understanding of the consequences of climate changes on animals, and more generally on biodiversity, will also benefit from greater consideration of the social landscape (Webber et al. 2023). Acknowledgments All procedures were approved by the Ethics Council of the Max Planck Society (2016_13/1), the National Commission for Science, Technology and Innovation (annual permits for all contributors of this study who worked in Kenya, licence NACOSTI/P/21/7442 for David Rozen-Rechels in 2021), Kenya Wildlife Service (annual permits for research and capture, KWS/904), the National Environment Management Authority (NEMA/AGR/68/2017), the Wildlife Research and Training Institute (WRTI-0026-02-21), and the National Museums of Kenya (NMK/ZLG/TRN/6.5). The procedure of i mplantations of ECG loggers was reviewed by the Animal Welfare Officer at the 26 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint University of Zurich. The procedure was realized by Dr. Daniel Zuñiga and performed under the supervision of Dr. Maureen Kamau from the Kenyan Wildlife Service. The procedure was reviewed and Dr. Daniel Zuñiga was approved to perform the surgeries by the Kenya Veterinary Board (KVB/FVS/V oll/6). We thank Alex Baiywa, Wismer Cherono, Janet Wangare Kariuki, Mary Waithira Ngugi, Edel Awour Odhiambo, Monicah Wambui, John Wanjala, and all the field assistants and all the rangers, administrative people in Mpala Research Center allowing this research to be conducted since 2016. This research was funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 850859 awarded to DRF), the Swedish Research Council (Grant Number 2019-06407 awarded to CHW), and an Eccellenza Professorship Grant of the Swiss National Science Foundation (Grant Number PCEFP3_187058 awarded to DRF). DRR received additional funding from a Humboldt Postdoctoral Fellowship granted to David Rozen-Rechels. Authors contributions DRR elaborated the hypotheses, set-up the thermal study, implemented the remote sensing mapping of the landscape and the temperatures measurements, collected behavioral data, analyzed the data and wrote the manuscript. BN advised the implementation of the protocols and managed the collection of GPS, census and operative temperature measurements. DP gave expertise on the hypotheses and the implementation of the study as well as analyses. MO managed the determination of communities. NB analyzed all videos. CHW implemented the ECG loggers study. JKI helped calculating group size. PN supervised the long-term monitoring of guineafowl. DRF 27 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint established the guineafowl project, led the project, and supervised the study. All authors contributed to the writing of the manuscript.

Albright, Thomas P., Denis Mutiibwa, Alexander. R. Gerson, Eric Krabbe Smith, William A. Talbot, Jacqueline J. O’Neill, Andrew E. McKechnie, and Blair O. Wolf. 2017. “Mapping Evaporative Water Loss in Desert Passerines Reveals an Expanding Threat of Lethal Dehydration.” Proceedings of the National Academy of Sciences 114 (9): 2283–88. https://doi.org/10.1073/pnas.1613625114. Anderson, Rodolfo C. O., and Denis V . Andrade. 2017. “Trading Heat and Hops for Water: Dehydration Effects on Locomotor Performance, Thermal Limits, and Thermoregulatory Behavior of a Terrestrial Toad.” Ecology and Evolution 7 (21): 9066–75. https://doi.org/10.1002/ece3.3219. Angilletta, Michael J. 2009. Thermal Adaptation: A Theoretical and Empirical Synthesis. Oxford University Press. Bakken, George S., and Michael J. Angilletta. 2014. “How to Avoid Errors When Quantifying Thermal Environments.” Edited by Marek Konarzewski. Functional Ecology 28 (1): 96–107. https://doi.org/10.1111/1365-2435.12149. Bauwens, Dirk, Paul E. Hertz, and Aurora M. Castilla. 1996. “Thermoregulation in a Lacertid Lizard: The Relative Contributions of Distinct Behavioral Mechanisms.” Ecology 77 (6): 1818– 30. https://doi.org/10.2307/2265786. 28 574 575 576 577 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Bertram, B. C. R. 1978. “Living in Groups: Predators and Prey.” In Behavioural Ecology: An Evolutionary Approach, edited by J. R. Krebs and N. B. Davies, Oxford: Blackwell Science., 64– 96. Blackwell Science. Blouin-Demers, Gabriel, and Patrick Nadeau. 2005. “The Cost–Benefit Model of Thermoregulation Does Not Predict Lizard Thermoregulatory Behavior.” Ecology 86 (3): 560– 66. https://doi.org/10.1890/04-1403. Bodey, Thomas W., Ian R. Cleasby, Fraser Bell, Nicole Parr, Anthony Schultz, Stephen C. V otier, and Stuart Bearhop. 2018. “A Phylogenetically Controlled Meta analysis of Biologging Device ‐ Effects on Birds: Deleterious Effects and a Call for More Standardized Reporting of Study Data.” Methods in Ecology and Evolution 9 (4): 946–55. https://doi.org/10.1111/2041- 210X.12934. Brandl, Hanja B., James A. Klarevas-Irby, Daniel Zuñiga, Christina Hansen Wheat, Charlotte Christensen, Fred Omengo, Cosmas Nzomo, Wismer Cherono, Brendah Nyaguthii, and Damien R. Farine. 2025. “The Physiological Costs of Leadership in Collective Movements.” bioRxiv. https://doi.org/10.1101/2023.11.22.567987. Breiman, Leo. 2001. “Random Forests.” Machine Learning 45 (1): 5–32. https://doi.org/10.1023/A:1010933404324. Cain III, James W., Paul R. Krausman, Steven S. Rosenstock, and Jack C. Turner. 2006. “Mechanisms of Thermoregulation and Water Balance in Desert Ungulates.” Wildlife Society Bulletin 34 (3): 570–81. https://doi.org/10.2193/0091-7648(2006)34[570:MOTAWB]2.0.CO;2. 29 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Cantor, Mauricio, Lucy M. Aplin, and Damien R. Farine. 2020. “A Primer on the Relationship between Group Size and Group Performance.” Animal Behaviour 166:139–46. https://doi.org/10.1016/j.anbehav.2020.06.017. Clark, L. 1987. “Thermal Constraints on Foraging in Adult European Starlings.” Oecologia 71 (2): 233–38. https://doi.org/10.1007/BF00377289. Cooper, Susan M., Kay E. Holekamp, and Laura Smale. 1999. “A Seasonal Feast: Long-Term Analysis of Feeding Behaviour in the Spotted Hyaena (Crocuta Crocuta).” African Journal of Ecology 37 (2): 149–60. https://doi.org/10.1046/j.1365-2028.1999.00161.x. Couzin, Iain D., Jens Krause, Nigel R. Franks, and Simon A. Levin. 2005. “Effective Leadership and Decision-Making in Animal Groups on the Move.” Nature 433 (7025): 513–16. https://doi.org/10.1038/nature03236. Cunningham, Susan J, Janet L Gardner, and Rowan O Martin. 2021. “Opportunity Costs and the Response of Birds and Mammals to Climate Warming.” Frontiers in Ecology and the Environment 19 (5): 300–307. https://doi.org/10.1002/fee.2324. Cunningham, Susan J, Rowan O Martin, and Philip AR Hockey. 2015. “Can Behaviour Buffer the Impacts of Climate Change on an Arid-Zone Bird?” Ostrich 86 (1–2): 119–26. https://doi.org/10.2989/00306525.2015.1016469. Czenze, Zenon J., Ryno Kemp, Barry Van Jaarsveld, Marc T. Freeman, Ben Smit, Blair O. Wolf, and Andrew E. McKechnie. 2020. “Regularly Drinking Desert Birds Have Greater Evaporative Cooling Capacity and Higher Heat Tolerance Limits than Non drinking Species.” ‐ Functional Ecology 34 (8): 1589–1600. https://doi.org/10.1111/1365-2435.13573. 30 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Dall, S, L Giraldeau, O Olsson, J Mcnamara, and D Stephens. 2005. “Information and Its Use by Animals in Evolutionary Ecology.” Trends in Ecology & Evolution 20 (4): 187–93. https://doi.org/10.1016/j.tree.2005.01.010. Davis, Grace H., Margaret C. Crofoot, and Damien R. Farine. 2022. “Using Optimal Foraging Theory to Infer How Groups Make Collective Decisions.” Trends in Ecology & Evolution 37 (11): 942–52. https://doi.org/10.1016/j.tree.2022.06.010. Dzialowski, Edward M. 2005. “Use of Operative Temperature and Standard Operative Temperature Models in Thermal Biology.” Journal of Thermal Biology 30 (4): 317–34. https://doi.org/10.1016/j.jtherbio.2005.01.005. Galván, Ismael, Daniel Palacios, and Juan José Negro. 2017. “The Bare Head of the Northern Bald Ibis (Geronticus Eremita) Fulfills a Thermoregulatory Function.” Frontiers in Zoology 14 (1). https://doi.org/10.1186/s12983-017-0201-5. Gao, Bo-cai. 1996. “NDWI—A Normalized Difference Water Index for Remote Sensing of Vegetation Liquid Water from Space.” Remote Sensing of Environment 58 (3): 257–66. https://doi.org/10.1016/S0034-4257(96)00067-3. Gerson, Alexander R., Andrew E. McKechnie, Ben Smit, Maxine C. Whitfield, Eric K. Smith, William A. Talbot, Todd J. McWhorter, and Blair O. Wolf. 2019. “The Functional Significance of Facultative Hyperthermia Varies with Body Size and Phylogeny in Birds.” Edited by Tony Williams. Functional Ecology 33 (4): 597–607. https://doi.org/10.1111/1365-2435.13274. Graw, Beke, and Marta B. Manser. 2007. “The Function of Mobbing in Cooperative Meerkats.” Animal Behaviour 74 (3): 507–17. https://doi.org/10.1016/j.anbehav.2006.11.021. 31 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Hansen, K. Whitney, Nathan Ranc, John Morgan, Neil R. Jordan, J. Weldon McNutt, Alan Wilson, and Christopher C. Wilmers. 2024. “How Territoriality and Sociality Influence the Habitat Selection and Movements of a Large Carnivore.” Ecology and Evolution 14 (4): e11217. https://doi.org/10.1002/ece3.11217. He, Peng, James A. Klarevas Irby, Danai Papageorgiou, Charlotte Christensen, Eli D. Strauss, ‐ and Damien R. Farine. 2023. “A Guide to Sampling Design for GPS based Studies of Animal ‐ Societies.” Methods in Ecology and Evolution 14 (8): 1887–1905. https://doi.org/10.1111/2041- 210X.13999. Herbert-Read, J. E., S. Krause, L. J. Morrell, T. M. Schaerf, J. Krause, and A. J. W. Ward. 2013. “The Role of Individuality in Collective Group Movement.” Proceedings of the Royal Society B: Biological Sciences 280 (1752): 20122564. https://doi.org/10.1098/rspb.2012.2564. Hetem, Robyn S., Shane K. Maloney, Andrea Fuller, and Duncan Mitchell. 2016. “Heterothermy in Large Mammals: Inevitable or Implemented?” Biological Reviews 91 (1): 187–205. https://doi.org/10.1111/brv.12166. Hetem, Robyn S., W. Maartin Strauss, Linda G. Fick, Shane K. Maloney, Leith C. R. Meyer, Mohammed Shobrak, Andrea Fuller, and Duncan Mitchell. 2012. “Activity Re-Assignment and Microclimate Selection of Free-Living Arabian Oryx: Responses That Could Minimise the Effects of Climate Change on Homeostasis?” Zoology 115 (6): 411–16. https://doi.org/10.1016/j.zool.2012.04.005. Huey, Raymond B., Michael R. Kearney, Andrew Krockenberger, Joseph AM Holtum, Mellissa Jess, and Stephen E. Williams. 2012. “Predicting Organismal Vulnerability to Climate Warming: Roles of Behaviour, Physiology and Adaptation.” Philosophical Transactions of the Royal 32 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Society of London B: Biological Sciences 367 (1596): 1665–79. https://doi.org/10.1098/rstb.2012.0005. IPCC. 2021. “Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change.” In , edited by Valérie Masson-Delmotte, Panmao Zhai, Anna Pirani, Sarah L. Connors, Clotilde Péan, Sophie Berger, Nada Caud, Y . Chen, L. Goldfarb, and M. I. Gomis, 2391 pp. Cambridge University Press, Cambridge, United Kingdom and New York, NY , USA. https://doi.org/10.1017/9781009157896. Kays, Roland, Sarah C. Davidson, Matthias Berger, Gil Bohrer, Wolfgang Fiedler, Andrea Flack, Julian Hirt, et al. 2022. “The Movebank System for Studying Global Animal Movement and Demography.” Methods in Ecology and Evolution 13 (2): 419–31. https://doi.org/10.1111/2041- 210X.13767. Kenward, R. E. 1978. “Hawks and Doves: Factors Affecting Success and Selection in Goshawk Attacks on Woodpigeons.” The Journal of Animal Ecology 47 (2): 449. https://doi.org/10.2307/3793. Klarevas-Irby, James A., Brendah Nyaguthii, and Damien R. Farine. 2025. “Moving as a Group Imposes Constraints on the Energetic Efficiency of Movement.” Proceedings of the Royal Society B: Biological Sciences 292 (2041): 20242760. https://doi.org/10.1098/rspb.2024.2760. Klarevas Irby, James A., Martin Wikelski, and Damien R. Farine. 2021. “Efficient Movement ‐ Strategies Mitigate the Energetic Cost of Dispersal.” Ecology Letters 24 (7): 1432–42. https://doi.org/10.1111/ele.13763. 33 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Lehtonen, Jussi, and Kim Jaatinen. 2016. “Safety in Numbers: The Dilution Effect and Other Drivers of Group Life in the Face of Danger.” Behavioral Ecology and Sociobiology 70 (4): 449– 58. https://doi.org/10.1007/s00265-016-2075-5. Long, Ryan A., R. Terry Bowyer, Warren P. Porter, Paul Mathewson, Kevin L. Monteith, and John G. Kie. 2014. “Behavior and Nutritional Condition Buffer a Large-Bodied Endotherm against Direct and Indirect Effects of Climate.” Ecological Monographs 84 (3): 513–32. https://doi.org/10.1890/13-1273.1. Marais, Elrike, and Steven L. Chown. 2008. “Beneficial Acclimation and the Bogert Effect.” Ecology Letters 11 (10): 1027–36. https://doi.org/10.1111/j.1461-0248.2008.01213.x. Mason, Tom H.E., Francesca Brivio, Philip A. Stephens, Marco Apollonio, and Stefano Grignolio. 2017. “The Behavioral Trade-off between Thermoregulation and Foraging in a Heat- Sensitive Species.” Behavioral Ecology 28 (3): 908–18. https://doi.org/10.1093/beheco/arx057. Mazerolle, Marc J. 2025. “AICcmodavg: Model Selection and Multimodel Inference Based on (Q)AIC(c).” https://cran.r-project.org/web/packages/AICcmodavg/index.html. McCann, R., A. M. Bracken, C. Christensen, I. Fürtbauer, and A. J. King. 2021. “The Relationship between GPS Sampling Interval and Estimated Daily Travel Distances in Chacma Baboons (Papio Ursinus).” International Journal of Primatology 42 (4): 589–99. https://doi.org/10.1007/s10764-021-00220-8. McKechnie, Andrew E., and Blair O. Wolf. 2010. “Climate Change Increases the Likelihood of Catastrophic Avian Mortality Events during Extreme Heat Waves.” Biology Letters 6 (2): 253– 56. https://doi.org/10.1098/rsbl.2009.0702. 34 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint ———. 2019. “The Physiology of Heat Tolerance in Small Endotherms.” Physiology 34 (5): 302–13. https://doi.org/10.1152/physiol.00011.2019. Moehlman, Patricia D. 2014. “Ecology of Cooperation in Canids.” In Ecological Aspects of Social Evolution: Birds and Mammals, edited by Daniel I. Rubenstein and Richard W. Wrangham, 64–86. Princeton University Press. https://doi.org/10.1515/9781400858149.64. Nyaguthii, Brendah, Tobit Dehnen, James A. Klarevas Irby, Danai Papageorgiou, Joseph ‐ Kosgey, and Damien R. Farine. 2025. “Cooperative and Plural Breeding by the Precocial Vulturine Guineafowl.” Ibis, January, in press. https://doi.org/10.1111/ibi.13393. Ogino, Mina, Eli D. Strauss, and Damien R. Farine. 2023. “Challenges of Mismatching Timescales in Longitudinal Studies of Collective Behaviour.” Philosophical Transactions of the Royal Society B: Biological Sciences 378 (1874): 20220064. https://doi.org/10.1098/rstb.2022.0064. Oswald, Krista N., Ben Smit, Alan T.K. Lee, and Susan J. Cunningham. 2019. “Behaviour of an Alpine Range-Restricted Species Is Described by Interactions between Microsite Use and Temperature.” Animal Behaviour 157:177–87. https://doi.org/10.1016/j.anbehav.2019.09.006. Papageorgiou, Danai, Charlotte Christensen, Gabriella E. C. Gall, James A. Klarevas-Irby, Brendah Nyaguthii, Iain D. Couzin, and Damien R. Farine. 2019. “The Multilevel Society of a Small-Brained Bird.” Current Biology 29 (21): R1120–21. https://doi.org/10.1016/j.cub.2019.09.072. Papageorgiou, Danai, and Damien R. Farine. 2020a. “Shared Decision-Making Allows Subordinates to Lead When Dominants Monopolize Resources.” Science Advances 6 (48): eaba5881. https://doi.org/10.1126/sciadv.aba5881. 35 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Papageorgiou, Danai, and Damien Roger Farine. 2020b. “Group Size and Composition Influence Collective Movement in a Highly Social Terrestrial Bird.” eLife 9:e59902. https://doi.org/10.7554/eLife.59902. Papageorgiou, Danai, Brendah Nyaguthii, and Damien R. Farine. 2024. “Compromise or Choose: Shared Movement Decisions in Wild Vulturine Guineafowl.” Communications Biology 7 (1): 95. https://doi.org/10.1038/s42003-024-05782-w. Papageorgiou, Danai, David Rozen-Rechels, Brendah Nyaguthii, and Damien R. Farine. 2021. “Seasonality Impacts Collective Movements in a Wild Group-Living Bird.” Movement Ecology 9 (1): 38. https://doi.org/10.1186/s40462-021-00271-9. Pattinson, Nicholas B., Michelle L. Thompson, Michael Griego, Grace Russell, Nicola J. Mitchell, Rowan O. Martin, Blair O. Wolf, et al. 2020. “Heat Dissipation Behaviour of Birds in Seasonally Hot Arid zones: Are There Global Patterns?” ‐ Journal of Avian Biology 51 (2): jav.02350. https://doi.org/10.1111/jav.02350. Plessis, Katherine L. du, Rowan O. Martin, Philip A. R. Hockey, Susan J. Cunningham, and Amanda R. Ridley. 2012. “The Costs of Keeping Cool in a Warming World: Implications of High Temperatures for Foraging, Thermoregulation and Body Condition of an Arid-Zone Bird.” Global Change Biology 18 (10): 3063–70. https://doi.org/10.1111/j.1365-2486.2012.02778.x. Portugal, Steven J., and Craig R. White. 2018. “Miniaturization of Biologgers Is Not Alleviating the 5% Rule.” Edited by Luca Börger. Methods in Ecology and Evolution 9 (7): 1662–66. https://doi.org/10.1111/2041-210X.13013. Rakowski, Allison E., R. Dwayne Elmore, Craig A. Davis, Samuel D. Fuhlendorf, and J. Matthew Carroll. 2019. “Thermal Refuge Affects Space Use and Movement of a Large-Bodied 36 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Galliform.” Journal of Thermal Biology 80 (February):37–44. https://doi.org/10.1016/j.jtherbio.2018.12.024. Rozen Rechels, David, Arnaud Badiane, Simon Agostini, Sandrine Meylan, and Jean François ‐ ‐ Le Galliard. 2020. “Water Restriction Induces Behavioral Fight but Impairs Thermoregulation in a Dry skinned Ectotherm.” ‐ Oikos 129 (4): 572–84. https://doi.org/10.1111/oik.06910. Rozen Rechels, David, Andréaz Dupoué, Olivier Lourdais, Simon Chamaillé Jammes, Sandrine ‐ ‐ Meylan, Jean Clobert, and Jean François Le Galliard. 2019. “When Water Interacts with ‐ Temperature: Ecological and Evolutionary Implications of Thermo hydroregulation in Terrestrial‐ Ectotherms.” Ecology and Evolution 9 (17): 10029–43. https://doi.org/10.1002/ece3.5440. Rozen Rechels, David, Alexis Rutschmann, Andréaz Dupoué, Pauline Blaimont, Victor ‐ Chauveau, Donald B. Miles, Michael Guillon, et al. 2021. “Interaction of Hydric and Thermal Conditions Drive Geographic Variation in Thermoregulation in a Widespread Lizard.” Ecological Monographs 91 (2): e01440. https://doi.org/10.1002/ecm.1440. Sears, Michael W., Michael J. Angilletta, Matthew S. Schuler, Jason Borchert, Katherine F. Dilliplane, Monica Stegman, Travis W. Rusch, and William A. Mitchell. 2016. “Configuration of the Thermal Landscape Determines Thermoregulatory Performance of Ectotherms.” Proceedings of the National Academy of Sciences 113 (38): 10595–600. https://doi.org/10.1073/pnas.1604824113. Sibly, Richard M. 1983. “Optimal Group Size Is Unstable.” Animal Behaviour 31 (3): 947–48. Smit, B., G. Zietsman, R. O. Martin, S. J. Cunningham, A. E. McKechnie, and P. A. R. Hockey. 2016. “Behavioural Responses to Heat in Desert Birds: Implications for Predicting Vulnerability 37 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint to Climate Warming.” Climate Change Responses 3 (1): 9. https://doi.org/10.1186/s40665-016- 0023-2. Stofberg, Miqkayla, Arjun Amar, Petra Sumasgutner, and Susan J. Cunningham. 2022. “Staying Cool and Eating Junk: Influence of Heat Dissipation and Anthropogenic Food on Foraging and Body Condition in an Urban Passerine.” Landscape and Urban Planning 226 (October):104465. https://doi.org/10.1016/j.landurbplan.2022.104465. Strandburg-Peshkin, Ariana, Damien R. Farine, Iain D. Couzin, and Margaret C. Crofoot. 2015. “Shared Decision-Making Drives Collective Movement in Wild Baboons.” Science 348 (6241): 1358–61. https://doi.org/10.1126/science.aaa5099. Strandburg-Peshkin, Ariana, Damien R Farine, Margaret C Crofoot, and Iain D Couzin. 2017. “Habitat and Social Factors Shape Individual Decisions and Emergent Group Structure during Baboon Collective Movement.” eLife 6 (January). https://doi.org/10.7554/eLife.19505. Sunday, Jennifer M., Amanda E. Bates, Michael R. Kearney, Robert K. Colwell, Nicholas K. Dulvy, John T. Longino, and Raymond B. Huey. 2014. “Thermal-Safety Margins and the Necessity of Thermoregulatory Behavior across Latitude and Elevation.” Proceedings of the National Academy of Sciences 111 (15): 5610–15. https://doi.org/10.1073/pnas.1316145111. Thompson, Michelle L., Susan J. Cunningham, and Andrew E. McKechnie. 2018. “Interspecific Variation in Avian Thermoregulatory Patterns and Heat Dissipation Behaviours in a Subtropical Desert.” Physiology & Behavior 188 (May):311–23. https://doi.org/10.1016/j.physbeh.2018.02.029. Valero, Silvia, David Morin, Jordi Inglada, Guadalupe Sepulcre, Marcela Arias, Olivier Hagolle, Gérard Dedieu, Sophie Bontemps, Pierre Defourny, and Benjamin Koetz. 2016. “Production of a 38 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Dynamic Cropland Mask by Processing Remote Sensing Image Series at High Temporal and Spatial Resolutions.” Remote Sensing 8 (1): 55. https://doi.org/10.3390/rs8010055. Veldhuis, Michiel P., Tim R. Hofmeester, Guy Balme, Dave J. Druce, Ross T. Pitman, and Joris P.G.M. Cromsigt. 2020. “Predation Risk Constrains Herbivores’ Adaptive Capacity to Warming.” Nature Ecology & Evolution 4 (8): 1069–74. https://doi.org/10.1038/s41559-020-1218-2. Ven, T. M. F. N. van de, A. E. McKechnie, and S. J. Cunningham. 2019. “The Costs of Keeping Cool: Behavioural Trade-Offs between Foraging and Thermoregulation Are Associated with Significant Mass Losses in an Arid-Zone Bird.” Oecologia 191 (1): 205–15. https://doi.org/10.1007/s00442-019-04486-x. Walsberg, Glenn E., and Wesley W. Weathers. 1986. “A Simple Technique for Estimating Operative Environmental Temperature.” Journal of Thermal Biology 11 (1): 67–72. https://doi.org/10.1016/0306-4565(86)90020-3. Webber, Quinn M. R., Gregory F. Albery, Damien R. Farine, Noa Pinter Wollman, Nitika ‐ Sharma, Orr Spiegel, Eric Vander Wal, and Kezia Manlove. 2023. “Behavioural Ecology at the Spatial–Social Interface.” Biological Reviews 98 (3): 868–86. https://doi.org/10.1111/brv.12934. Williams, Joseph B., and B. Irene Tieleman. 2005. “Physiological Adaptation in Desert Birds.” BioScience 55 (5): 416–25. https://doi.org/10.1641/0006-3568(2005)055[0416:PAIDB]2.0.CO;2. Withers, Philip C., and Timothy M. Crowe. 1980. “Brain Temperature Fluctuations in Helmeted Guineafowl under Semi-Natural Conditions.” The Condor 82 (1): 99–100. https://doi.org/10.2307/1366793. 39 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Wolf, Blair O, and Glenn E Walsberg. 1996. “Respiratory and Cutaneous Evaporative Water Loss at High Environmental Temperatures in a Small Bird.” Journal of Experimental Biology 199:451–57. https://doi.org/10.1242/jeb.199.2.451. Wooden, K. Mark, and Glenn E. Walsberg. 2002. “Effect of Environmental Temperature on Body Temperature and Metabolic Heat Production in a Heterothermic Rodent, Spermophilus Tereticaudus.” Journal of Experimental Biology 205 (14): 2099–2105. https://doi.org/10.1242/jeb.205.14.2099. Young, Truman P., Nathaniel Patridge, and Alison Macrae. 1995. “Long term Glades in Acacia ‐ Bushland and Their Edge Effects in Laikipia, Kenya.” Ecological Applications 5 (1): 97–108. https://doi.org/10.2307/1942055. 40 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Table 1 Hypotheses and predictions on the priorities of the vulturine guineafowl between heat and food avoidance and foraging which are compared in the study. VGFs: vulturine guineafowl. Tsun: operative temperature in the sun. Tdiff: difference of operative temperature the sun and the shade. 41 578 579 580 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint # Hypothesis Predictions Model formula 1 Random foraging behavior VGFs select randomly the habitat to use Habitat use index ~ 1 2 Temperature-driven foraging behavior VGFs use open habitats less with increasing sun temperature in order to avoid hyperthermia Habitat use index ~ Tsun 3 VGFs use open habitats the most at intermediate sun temperature in order to avoid hyperthermia but also to take advantage of warmer temperatures at the start of the day Habitat use index ~ Tsun + Tsun² 4 VGFs use open habitats the most when the temperatures in open and cover habitats are similar, i.e. when heat constraints are the lowest or when there is no increased thermal cost to go in the sun compared to staying in the shade Habitat use index ~ Tdiff 5 VGFs use open habitats the most when the temperature in the open is higher than under cover when heat constraints are low (early or late in the day, cloudy days), but avoid open habitats when the temperature gets too high while staying in the shade allow cooling at reduced costs Habitat use index ~ Tdiff + Tdiff² 6 Group foraging behavior The use of open habitats is maximal at medium group size as the benefits of predation risk dilution in large groups and low competition in small groups trade-off Habitat use index ~ Group size 7 Temperature-driven group foraging behavior Prediction #2 + Medium-sized groups use open habitats at higher sun temperatures than other groups Habitat use index ~ Tsun × Group size 8 Prediction #3 + Medium-sized groups use open habitats at higher sun temperatures than other groups Habitat use index ~ [Tsun + Tsun²] × Group size 9 Prediction #4 + Medium-sized groups use open habitats at higher sun temperatures than other groups Habitat use index ~ Tdiff × Group size 10 Prediction #5 + Medium-sized groups use open habitats at higher sun temperatures than other groups Habitat use index ~ [Tdiff + Tdiff²] × Group size 1 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Figure 1 Operative temperature models. The orange arrow shows the model in open habitat and the gray arrow showing the model in cover. The car is provided for scale. The plot shows the variations of the operative temperature in the open averaged for all models in the open (i.e. Tsun) and operative temperature in cover averaged for all models in cover, for a subset of dates (for clarity). Ribbons show the standard deviation. Points are showing the proportion of occurrence in the open of VGFs for 3 separated moment of the day indicated by shape: early in the morning (circles, 07:00 to 08:59 EAT), midday (squares, 12:00 to 13:59 EAT), and late afternoon (triangles, 16:00 to 17:59 EAT). Figure 2 Predictions of the best models identified by model selection for each studied behavior: (A) probability to be in the open vs. cover as a function of Tdiff; (B) probability to shift to the open while under cover as a function of Tdiff; (C) probability to shift to cover while in the open as a function of Tdiff and (D) the distance moved in one hour as a function of Tsun. In all graphs, points are raw proportion of the behavior for each group size for bins of 1°C of Tdiff (A,B,C) or Tsun (D). Error bars show the calculated 95% confidence intervals around the calculated proportions. Lines are the predicted probability to do the behavior according to group size. Ribbons are the predicted 95% confidence intervals of this probability. Predictions in A, B and C have been made for an average open habitat availability calculated in the dataset used in the statistical model. The strong predicted effect of open habitat availability explains the deviance of the predictions to the points. In D, predictions have been made for hour-bouts including 12 different locations. Figure 3 Predicted changes in the deviation between body temperature and resting body temperature (body temperature anomaly Δ tbody) according to Tdiff and the habitat (open or cover) 43 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint where the bird is active. Points are the average ΔTbody for bins of 1°C of Tdiff for each habitat. Error bars show the calculated 95% confidence interval around the calculated mean. Lines are the predicted probability to do the behavior according to group size. Ribbons are the predicted 95% confidence intervals of this probability. 44 603 604 605 606 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Figure 1 45 607 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Figure 2 46 609 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint Figure 3 611 612 2 .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 24, 2025. ; https://doi.org/10.1101/2025.04.24.650232doi: bioRxiv preprint