Rodents Inhabiting the Southeastern Mu Us Desert May Not Have Experienced Prolonged Heat Stress in Summer 2022

Simple Summary Rodents in the Mu Us Desert may respond to increasing summer heat through behavioral or physiological adjustments. We studied the ambient temperatures that Meriones meridianus and Meriones unguiculatus might experience in their habitats. The results showed that both the heat-sensitive Meriones meridianus and the heat-tolerant Meriones unguiculatus were threatened by extreme summer heat. In addition, although vegetation and rodent burrows increased the variety of ambient temperature in the habitat, only rodent burrows were effective in providing protection from the stress of extreme summer heat. Therefore, behavioral adjustments may be the major way in which small rodents in this region avoid continuous heat stress in summer. Abstract Climate change combined with human activities has altered the spatial and temporal patterns of summer extreme heat in the Mu Us Desert. To determine how those rodents living in the desert respond to increased extreme heat in summer, in July 2022, during the hottest month, we examined the rodent species, vegetation coverage, and small-scale heterogeneity in ambient temperature in the southeastern Mu Us Desert. The results showed that Meriones meridianus, Meriones unguiculatus, and Cricetulus longicaudatus were found in the study area, where the vegetation coverage is 33.5–40.8%. Moreover, the maximum temperature of the desert surface was 61.8 °C. The maximum air temperature at 5 cm above the desert surface was 41.3 °C. The maximum temperature in the burrow at a depth of 15 cm was 31 °C. M. unguiculatus might experience 4–9.3 h of heat stress in a day when exposed outside the burrow, whereas M. meridianus would experience 8.5–10.8 h of heat stress. Yet, inside the burrow, both species were barely exposed to heat stress. In conclusion, adjustments in behavioral patterns can be the main way that rodents in the Mu Us Desert adapt to the extreme heat in the summer.


Introduction
Hot summer is becoming a deadly season for many mammals on Earth [1,2]. Many regions of the world experienced record-breaking heat events in 2022 [3][4][5]. Moreover, the frequency and severity of extreme heat events are predicted to increase in the future due to global climate change [6]. It is necessary to understand how mammals in the wild will respond to increasing extreme summer heat.
Generally, facing extreme heat, small mammals will first reduce the stress on their thermoregulatory system through behavioral adjustments. For example, desert rodents tend to search shaded cooler places by moving around flexibly during the hottest part of the day [7]. Furthermore, some thermoregulating behaviors, such as specific postures,

Vegetation Analysis
Vegetation surveys were conducted at three study sites. The size of the study site was 90 m × 60 m. A square of 5 m × 5 m was randomly sampled in the site to survey shrub communities. In addition, 3 squares of 1 m × 1 m across the square for herbs were sampled to investigate herbaceous plants [35]. The height and coverage of the plants were measured using a metric tap (to 1 mm). Plant species were identified by their morphologies.

Rodent Surveys
Rodents were captured using animal live traps [32]. In total, 50 wire-mesh traps (28 cm × 13 cm × 10 cm) were placed at each study site with 2-3 m spacing between traps. They were located close to rodent burrows or where there were animal tracks, such as rodent footprints or feces. For traps not under vegetation, a shade area was created with Salix cheilophila branches and leaves to protect the animals from excessive exposure to sunlight. All traps were set at 6 pm with peanuts as bait. The trap was triggered when the rodent ate the bait. All traps were checked on 2 consecutive days at 6 am, 12 am, and 6 pm. All captured rodents were placed separately in a clear plastic box (43 cm × 34 cm × 30 cm) to record their body mass and morphology. The species of each rodent was identified based on characteristic morphological features. Finally, all rodents were released at the place where they were caught.

Determination and Analysis of Ambient Temperature
Near-surface ambient temperatures at the study sites were determined daily and at various heights over a period of two weeks. Ambient temperatures at 50 cm, 5 cm, and 0 cm above the ground (T50cm, T5cm, and T0cm) were measured using temperature loggers (RC-4, Elitech, Xuzhou, China). The ambient temperatures were recorded every 15 min. The data obtained were used for further analysis when the weather conditions were sunny and clear for 36 consecutive hours. The ambient temperature extremes were calculated by averaging 3 consecutive maximum or minimum ambient temperatures over a 24 h period.

Rodent Surveys
Rodents were captured using animal live traps [32]. In total, 50 wire-mesh traps (28 cm × 13 cm × 10 cm) were placed at each study site with 2-3 m spacing between traps. They were located close to rodent burrows or where there were animal tracks, such as rodent footprints or feces. For traps not under vegetation, a shade area was created with Salix cheilophila branches and leaves to protect the animals from excessive exposure to sunlight. All traps were set at 6 p.m. with peanuts as bait. The trap was triggered when the rodent ate the bait. All traps were checked on 2 consecutive days at 6 a.m., 12 a.m., and 6 p.m. All captured rodents were placed separately in a clear plastic box (43 cm × 34 cm × 30 cm) to record their body mass and morphology. The species of each rodent was identified based on characteristic morphological features. Finally, all rodents were released at the place where they were caught.

Determination and Analysis of Ambient Temperature
Near-surface ambient temperatures at the study sites were determined daily and at various heights over a period of two weeks. Ambient temperatures at 50 cm, 5 cm, and 0 cm above the ground (T 50cm , T 5cm , and T 0cm ) were measured using temperature loggers (RC-4, Elitech, Xuzhou, China). The ambient temperatures were recorded every 15 min. The data obtained were used for further analysis when the weather conditions were sunny and clear for 36 consecutive hours. The ambient temperature extremes were calculated by averaging 3 consecutive maximum or minimum ambient temperatures over a 24 h period.
At the same time, the desert surface temperature was measured with an infrared thermal imager (C200 Pro, InfiRay, Yantai, China) and temperature logger. The ratio of the desert surface temperature under vegetation (T UV ) extremes to the desert surface temperature with no vegetation (T NV ) extremes was used to measure the effect of vegetation cover on desert surface temperature. Moreover, correlations between T 5cm , T NV , and T UV were evaluated by simple linear regression analysis.
Ambient temperatures inside the burrow (T burrow ) were measured using temperature loggers. The temperature probe was placed at a depth of 15 cm from the desert surface,  20 cm away from the opening of the rodent burrow. The depth of the temperature probe was determined by averaging the 10 rodent burrows surveyed in the field. In addition, to understand the ambient temperature characteristics of the rodent burrow, desert soil temperatures were measured with a digital thermometer (TES-1310, TES Electrical Electronic Corp., Taipei, China).
Linear regression analysis was performed to assess the linearity between air temperature (T air ) and T 50cm , T 5cm , and T 0cm , and T UV . T air data were obtained from the official website (www.nmc.cn, accessed on 29 July 2022). The mean values of Shenmu and Ejin Horo Banner T air were used in this study.

Determination of the Extreme Heat
Extreme heat was determined by the thermoneutral zone (TNZ) of the rodents [31]. The M. unguiculatus has a TNZ of 26 to 38 • C [24]. Therefore, the species was experiencing extreme heat when the ambient temperature was above 38 • C. In addition, the TNZ of the M. meridianus ranges from 25.5 to 30.1 • C [27]. This species was experiencing extreme heat when the ambient temperature was greater than 30.1 • C.

Statistical Analyses
All statistical analyses were performed using IBM SPSS.v.20.0 software. Plant heights, the difference between ambient temperature extremes and T air extremes, the ratio of T UV extremes/T NV extremes, and the ratio of T burrow extremes/T 0cm extremes were provided as mean ± SD. The general linear model was used to regress measures of ambient temperature against T air . Differences in other indicators between 2 groups were analyzed using the independent sample t-test. p values of less than 0.5 were considered statistically significant.

Plant Species, Height, and Coverage
The plant species recorded at each site are shown in Table 1. In site 1, the vegetation coverage was 40.82%. Salix cheilophila was the major species, with a height of 1.904 ± 0.256 m and coverage of 40.44%. A small amount of Artemisia ordosica was found, with a height of 0.7 m and coverage of 0.38%. In site 2, the vegetation coverage was 33.53%. Caragana intermedia was the major species, with a height of 1.625 ± 0.178 m and coverage of 22.17%. Artemisia ordosica was the minor species, with a height of 0.535 ± 0.176 m and coverage of 11.36%. A small amount of Cynanchum thesioides was distributed in the interspace of shrubs. In site 3, the vegetation coverage was 36.67%. Artemisia ordosica was the minor species, with a height of 0.610 ± 0.188 and coverage of 26.40%. Caragana intermedia was the minor species, with a height of 1.165 ± 0.191 and coverage of 7.85%. Amygdalus pedunculata was another minor species, with a height of 0.700 ± 0.100 and coverage of 2.42%. There were also small amounts of Cynanchum thesioides and Psammochloa villosa mixed among the bushes.

Small Mammals Living in the Study Area
Mongolian gerbil, Long-tailed hamster, and Midday gerbil were found in the study area ( Figure 2). No animals were captured at site 1. Three Midday gerbils and two Longtailed hamsters were captured at site 2. Two Mongolian gerbils, two Midday gerbils, and one Long-tailed hamster were captured at site 3. All rodents were captured before 6 a.m.

Small Mammals Living in the Study Area
Mongolian gerbil, Long-tailed hamster, and Midday gerbil were found in the study area ( Figure 2). No animals were captured at site 1. Three Midday gerbils and two Longtailed hamsters were captured at site 2. Two Mongolian gerbils, two Midday gerbils, and one Long-tailed hamster were captured at site 3. All rodents were captured before 6 am.

Ambient Temperature Variation in Near-Surface Space
The daily changes in near-surface ambient temperatures are shown in Figure 3a-c. The near-surface ambient temperatures decreased during the night. They started to increase after sunrise and reach a maximum between 1 and 3 pm. Compared to the Tair from the official website, the near-surface ambient temperatures were lower at night (Figure 3d-f). The difference between the lowest T50cm, T5cm, and T0cm and the lowest Tair were −2.5 °C, −3.3 °C, and −3.5 °C, respectively ( Figure 3d). In contrast, during the day, the near-surface temperatures were higher and showed a greater variation (Figure 3d

Ambient Temperature Variation in Near-Surface Space
The daily changes in near-surface ambient temperatures are shown in Figure 3a-c. The near-surface ambient temperatures decreased during the night. They started to increase after sunrise and reach a maximum between 1 and 3 pm. Compared to the T air from the official website, the near-surface ambient temperatures were lower at night (Figure 3d-f). The difference between the lowest T 50cm , T 5cm , and T 0cm and the lowest T air were −2.5 • C, −3.3 • C, and −3.5 • C, respectively (Figure 3d). In contrast, during the day, the near-surface temperatures were higher and showed a greater variation (Figure 3d-f). The difference between the highest T 50cm , T 5cm , and T 0cm and the highest T air were 6.  (d-f) The difference (mean ± SD) between ambient temperature extremes (n = 3) and Tair extremes. Txcm (x = 50, 5, 0), ambient temperature measured at x cm above the ground; Tair, air temperature from the official website.

Effect of Vegetation on Desert Surface Temperature
Vegetation cover had a regulatory function of desert surface temperature ( Figure  4a  (d-f) The difference (mean ± SD) between ambient temperature extremes (n = 3) and T air extremes. T xcm (x = 50, 5, 0), ambient temperature measured at x cm above the ground; T air , air temperature from the official website.

Effect of Vegetation on Desert Surface Temperature
Vegetation cover had a regulatory function of desert surface temperature (Figure 4a

Ambient Temperature in Rodent Burrow
The daily changes in ambient temperature on the desert surface and in the rodent burrow are shown in Figure 6a. At night, the ratio of the lowest Tburrow to the lowest T0cm was 1.35-1.66 (Figure 6b). During the day, the ratio of the highest Tburrow to the highest T0cm was 0.47-0.49 (Figure 6b). Moreover, daily variations in desert soil temperature declined with increasing soil depth (Figure 6c). At 9 am and 7 pm, the effect of vegetation on the soil temperature was not significant at depths greater than 10 cm (Figure 6d,f). At 2 pm, the difference in soil temperature due to vegetation cover was less than 2.3 °C at depths greater than 10 cm (Figure 6e).

Figure 5.
Linear regression between T air and T 50cm (a), T 5cm (b), T 0cm (c), and T UV (d). T air , air temperature from the official website; T xcm (x = 50, 5, 0), ambient temperature measured at x cm above the ground; T UV , desert surface temperature under vegetation.

Ambient Temperature in Rodent Burrow
The daily changes in ambient temperature on the desert surface and in the rodent burrow are shown in Figure 6a. At night, the ratio of the lowest T burrow to the lowest T 0cm was 1.35-1.66 (Figure 6b). During the day, the ratio of the highest T burrow to the highest T 0cm was 0.47-0.49 (Figure 6b). Moreover, daily variations in desert soil temperature declined with increasing soil depth (Figure 6c). At 9 a.m. and 7 p.m., the effect of vegetation on the soil temperature was not significant at depths greater than 10 cm (Figure 6d,f). At 2 p.m., the difference in soil temperature due to vegetation cover was less than 2.3 • C at depths greater than 10 cm (Figure 6e). was 1.35-1.66 (Figure 6b). During the day, the ratio of the highest Tburrow to the highest T0cm was 0.47-0.49 (Figure 6b). Moreover, daily variations in desert soil temperature declined with increasing soil depth (Figure 6c). At 9 am and 7 pm, the effect of vegetation on the soil temperature was not significant at depths greater than 10 cm (Figure 6d,f). At 2 pm, the difference in soil temperature due to vegetation cover was less than 2.3 °C at depths greater than 10 cm (Figure 6e). (d-f) Effect of vegetation cover on desert soil temperature (mean ± SD, n = 3) during the day. S1-3, site 1-3. T9am, T2pm, and T7pm desert soil temperature measured at 9 am, 2 pm, and 7 pm. Tburrow, ambient temperature inside the burrow; T0cm, ambient temperature measured at 0 cm above the ground. Statistical significance is indicated by asterisks.

Distribution of Extreme Heat
The distribution of extreme heat in the study area over the day is shown in Figure 7. For M. unguiculatus, extreme heat persisted from 11:45 am to 3:45 pm at 5 cm from the ground. On the desert surface, the extreme heat lasted from 9:15 am to 6:30 pm. On the vegetation-covered desert surface, the extreme heat continued from 12:00 am to 4:00 pm. In the burrow, no extreme heat was measured. In addition, for M. meridianus, extreme heat persisted from 9:15 am to 6:30 pm at 5 cm from the ground. On the desert surface, the extreme heat lasted from 8:15 am to 7:00 pm. On the vegetation-covered desert surface, the extreme heat continued from 10:00 am to 6:30 pm. In the burrow, no extreme heat was measured. (d-f) Effect of vegetation cover on desert soil temperature (mean ± SD, n = 3) during the day. S 1-3 , site 1-3. T 9am , T 2pm , and T 7pm desert soil temperature measured at 9 a.m., 2 p.m., and 7 p.m. T burrow , ambient temperature inside the burrow; T 0cm , ambient temperature measured at 0 cm above the ground. Statistical significance is indicated by asterisks.

Distribution of Extreme Heat
The distribution of extreme heat in the study area over the day is shown in Figure 7.

Discussion
Rodents inhabiting the Mu Us Desert were influenced by extreme heat events during the summer. Ambient temperatures in the near-surface space showed a clear variance within small scales due to the vegetation cover and burrow system. Moreover, high temperatures beyond thermal tolerance limits were detected in the near-surface space frequented by rodents.

Vegetation Coverage and Rodent Species in the Mu Us Desert
Desert vegetation provides rodents with a habitat and the resources they need to survive. Our study showed that the vegetation coverage in the study area is 33.5-40.8%. The vegetation was composed of shrubs dominated by Salix cheilophila, Caragana intermedia, and Artemisia ordosica and a small number of herbaceous plants. Moreover, the vegetation cover in most areas of the Mu Us Desert has increased rapidly in the last 30 years, driven by both climate change and human activities [12,17,35]. Climate change may alter the seasonal temperature and precipitation patterns. Studies have shown an increasing trend in precipitation and temperature in northwest China [36][37][38], which will likely favor desert vegetation restoration. In addition, human activities, such as fencing and vegetation construction, are important forces of desert vegetation rehabilitation [39,40]. In conclusion, future trends in the vegetation coverage of the Mu Us Desert may be favorable for rodent survival.
The Mu Us Desert is a habitat for many species of rodents. In the present study, M. meridianus, M. unguiculatus, and C. longicaudatus were found. However, our data do not provide a full picture of the rodent species in this desert. In fact, Phodopus roborovskii, Cricetulus barabensis, Allactaga sibirica, and Spermophilus dauricus also live there [20,41]. In addition, no rodents were caught in simple Salix cheilophila-dominated habitats, while 10 rodents were caught in Caragana intermedia-and Artemisia ordosica-dominated habitats. This may indicate that, together with vegetation coverage, the increased plant species in the habitat also favor the survival of rodents [42]. Moreover, the number of rodents captured in this study was quite low. It is hard to provide effective information on population density. Yet, rodents of all 3 species were captured before 6 am, suggesting that relationships among rodent species may be affected in the face of the combined pressures of climate change and habitat modification [43]. Taken together, although the trends in the vegetation cover of the Mu Us Desert are positive for rodents, changes in the rodent population density and interspecific relationships require further study.

Discussion
Rodents inhabiting the Mu Us Desert were influenced by extreme heat events during the summer. Ambient temperatures in the near-surface space showed a clear variance within small scales due to the vegetation cover and burrow system. Moreover, high temperatures beyond thermal tolerance limits were detected in the near-surface space frequented by rodents.

Vegetation Coverage and Rodent Species in the Mu Us Desert
Desert vegetation provides rodents with a habitat and the resources they need to survive. Our study showed that the vegetation coverage in the study area is 33.5-40.8%. The vegetation was composed of shrubs dominated by Salix cheilophila, Caragana intermedia, and Artemisia ordosica and a small number of herbaceous plants. Moreover, the vegetation cover in most areas of the Mu Us Desert has increased rapidly in the last 30 years, driven by both climate change and human activities [12,17,35]. Climate change may alter the seasonal temperature and precipitation patterns. Studies have shown an increasing trend in precipitation and temperature in northwest China [36][37][38], which will likely favor desert vegetation restoration. In addition, human activities, such as fencing and vegetation construction, are important forces of desert vegetation rehabilitation [39,40]. In conclusion, future trends in the vegetation coverage of the Mu Us Desert may be favorable for rodent survival.
The Mu Us Desert is a habitat for many species of rodents. In the present study, M. meridianus, M. unguiculatus, and C. longicaudatus were found. However, our data do not provide a full picture of the rodent species in this desert. In fact, Phodopus roborovskii, Cricetulus barabensis, Allactaga sibirica, and Spermophilus dauricus also live there [20,41]. In addition, no rodents were caught in simple Salix cheilophila-dominated habitats, while 10 rodents were caught in Caragana intermediaand Artemisia ordosica-dominated habitats. This may indicate that, together with vegetation coverage, the increased plant species in the habitat also favor the survival of rodents [42]. Moreover, the number of rodents captured in this study was quite low. It is hard to provide effective information on population density. Yet, rodents of all 3 species were captured before 6 am, suggesting that relationships among rodent species may be affected in the face of the combined pressures of climate change and habitat modification [43]. Taken together, although the trends in the vegetation cover of the Mu Us Desert are positive for rodents, changes in the rodent population density and interspecific relationships require further study.

Temperature Variation of Near-Surface Space in Summer
Extreme heat in the summer is becoming more frequent, longer lasting, and hotter. Evidence can be seen from both climate model predictions and weather station detections [6]. Although the temperature data from weather stations differ from the real ambient temperatures experienced by rodents, data from nearby weather stations can be used to further predict the ambient temperatures in rodent habitats. In this study, we found a significant difference between the near-surface ambient temperature and data from the official website (T air ). The daily maximum T air was 10.2 • C lower than T 5cm and 29.6 • C lower than the desert surface temperature (T 0cm ) in July 2022. Moreover, a linear relation between T air and the near-surface ambient temperature was established. These results indicated that the extreme heat experienced by desert rodents in summer was likely underestimated.
In the Mu Us Desert, the vegetation cover and burrow system are effective in regulating the temperature surrounding rodents. In the present study, the vegetation cover reduced the maximum daily desert surface temperature to 65 percent, whereas it had little effect on the air temperature near the desert surface. In addition, the daily maximum temperature in the rodent burrow was 47% of the desert surface temperature. It had a good insulation function at night. However, the study presented measured temperature changes at a depth of 15 cm in the burrow alone. In fact, the structure of rodent burrows is complex. The temperature environment in the burrows is also very diverse. Further research is needed on the role of burrows in rodent summer heat adaptation.

Strategies for Rodent Response to Summer Heat
TNZ is defined as a range of ambient temperature, at which the metabolic rate is at basal or resting levels [44]. This concept provides us with a lot of information on the ability of species to survive in different ambient temperatures. Rodents will be faced with severe physical and physiological limitations when the ambient temperature is above the TNZ. Here, the concept of the TNZ was used to analyze whether rodents experience heat stress during the summer. M. unguiculatus exhibits a wide TNZ, and its upper critical temperature is 38 ± 1 • C [21,22,24]. Moreover, this species is able to survive for the long term at 37 • C [23,45]. In contrast, the range of the TNZ in M. meridianus is only 38.3% of that in M. unguiculatus, and the upper critical temperature is 30.1 • C. In brief, the small desert rodents vary widely in their sensitivity and regulatory limits to variations in ambient temperature [31].
Desert surface temperature, air temperature near the desert surface, and T burrow may reflect the true temperature experienced by small desert rodents. In the study area, the maximum desert surface temperature was 61.8 • C, with a diurnal temperature difference of 46.1 • C. The maximum air temperature near the desert surface was 41.3 • C, with a diurnal temperature difference of 26.2 • C. In contrast, the maximum T burrow was 31 • C, with a diurnal temperature difference of 7.5 • C. Furthermore, outside the burrows, both M. meridianus and M. unguiculatus were at risk from extreme summer heat. Therefore, summer in the Mu Us Desert is a season of stress and survival for rodents, but burrows may provide them with a place to escape the extreme heat.

Conclusions
Climate change is changing the pattern of distribution of extreme summer heat in the Mu Us Desert. The evidence suggested that both M. meridianus and M. unguiculatus living in the southeastern Mu Us Desert were affected by extreme summer heat. This also implies that the behavior, physiology, and interspecies relationships of rodents in this region are expected to change significantly in the near future. Furthermore, burrows can help rodents escape the stress of summer heat, suggesting that, at least in the present study, desert rodents may respond to increased summer heat extremes primarily through behavioral adjustments. Informed Consent Statement: Not applicable.

Data Availability Statement:
The data used in this study are available from the first author upon reasonable request.