Your privacy, your choice

We use essential cookies to make sure the site can function. We also use optional cookies for advertising, personalisation of content, usage analysis, and social media.

By accepting optional cookies, you consent to the processing of your personal data - including transfers to third parties. Some third parties are outside of the European Economic Area, with varying standards of data protection.

See our privacy policy for more information on the use of your personal data.

for further information and to change your choices.

You are viewing the site in preview mode

Skip to main content
Download PDF
Download ePub

Seasonal movement behavior of domestic goats in response to environmental variability and time of day using Hidden Markov Models

Movement Ecology volume 13, Article number: 28 (2025) Cite this article

Abstract

Background

Current research on livestock movement ecology focuses on quantifying the factors that trigger alterations in movement behavior and understanding hidden mechanisms. Modern tracking technologies and robust statistical analysis models deliver new opportunities for investigating how individual animals cope with the joint effect of biotic and abiotic factors at different time scales.

Methods

We applied multivariate Hidden Markov Models (HMMs) to characterize the fine-scale movement behavior (30-second intervals) of GPS-tracked domestic Zhongwei goats (Capra aegagrus hircus) for 124 days and analyzed the combined influence of biotic and abiotic factors and specific time of day on their seasonal movement behavioral transition in a predator-free, semi-arid mountain grassland in China.

Results

We classified the behaviors of goats into two states: foraging (low step length, varied turning angle) and travelling (long step lengths, small turning angles). The terrain slopes had the most impact on their movement behavioral transition in the full year, spring, autumn, and winter. However, in the summer with hotter temperatures, the specific time of day explains their movement behavior most. Forage resources indicated by the Normalized Difference Vegetation Index (NDVI), and terrain ruggedness measured by the Vector Ruggedness Measure (VRM), had less impact on their behavior transitions compared to terrain slope and specific time of day. Elevation and solar radiation could not explain their movement behavior in different seasons, nor could NDVI in winter or VRM in spring and autumn. Across different seasons, the probability of foraging behavior increased with the later times of day, steeper terrain slopes, and higher NDVI, while it decreased with increasing VRM. The impact of NDVI on the probability of foraging behavior was largest during the early onset of vegetation growth in spring, and lowest in winter coinciding with a lower availability of food resources. The movement speed was lower, and the daily foraging percentage was higher in spring and winter due to lower food resources and shorter daylight hours. In contrast, movement speed was higher, and the daily foraging percentage was lower in summer and autumn with more food resources and longer daylight hours. The percentage of time allocated to foraging increases hourly from 9:00 am to 8:00 pm across various seasons.

Conclusions

HMMs were found to be useful for disentangling the movement behavior of goats. Our approach provides new insights into the seasonal and daily behavioral strategies of goats. Results demonstrate that in the mountain region, terrain slopes and specific times of the day more effectively trigger domestic goat behavioral transition from one state to the next compared with biotic factors, represented herein by NDVI, across different seasons. The early onset of vegetation growth and a shorter period of available high-quality forage in spring, significantly influenced goat behavioral transitions. Overall, these results are important for designing appropriate grazing management strategies that satisfy the ecological and socioeconomic demands of semi-arid grassland ecosystems.

Background

Livestock directly impacts 25% of the terrestrial land area through their grazing and browsing behaviors [1], and products derived from domesticated livestock support 1.3 billion people worldwide [2, 3] and make up 17% of the total global energy intake [4]. The movement state is a key determinant of organism-environment interaction [5], with distinct movement patterns such as “foraging” (e.g. short step lengths and high turning angles) and “travelling” (e.g. long step lengths and small turning angles) [6, 7]. Behavioral transition in livestock involves switching their movement states on the basis of a complex interplay of behavioral decisions, motion capacity, navigation capacity, and landscape characteristics, aiming to optimize their growth and enhance their prospects of survival and reproduction [7,8,9,10]. In rangelands, domestic livestock play a critical role in shaping the soil health, vegetation structure, and biodiversity [11, 12]. With 61% of global natural grasslands overgrazed [13] and land degradation being a major global issue [14], understanding livestock behavioral responses and decision-making, particularly in conjunction with changing environments, is important for effective rangeland grazing management and the formulation of conservation policies.

Movement state of animals in their natural habitat can be influenced by both biotic and abiotic factors. Several studies have presented findings concerning the abiotic-driven resource selection of goats, indicating that they generally tend to save energy by selecting gentle slopes [15, 16] and comfortable thermal conditions [17, 18], and escape predators by selecting rugged areas [19]. What is less well understood is how abiotic factors influence the behavioral transition of goats that may switch their behavior from one movement state (e.g. foraging) to another (e.g. travelling). Other investigations have emphasized that biotic factors such as forage quality (e.g. dry matter digestibility) and forage biomass, primarily trigger animal behavior to switch from one state to the next and create multiple movement states [8, 20]. To complicate the generalization of goat movement behavior, the impact of biotic and abiotic factors is strongly context-dependent, introducing considerable uncertainty and seasonal variability. For instance, the availability of increased forage resources can either decrease [21], increase [21,22,23], or have no effect on the movement velocity of goats [24]. To date, relatively few studies have explored the impact of the joint effect of biotic and abiotic factors on the transition between the foraging and travelling behavior of domestic goats at different time scales.

Goats are highly selective, adaptive, and plastic browsers [25,26,27]. They exhibit high thermal and drought resilience, the capability to survive on sparsely vegetated pastures, and strong disease resistance, all of which make them ideal climate-resilient livestock in drylands [28, 29]. Domestic Zhongwei goats (Capra aegagrus hircus), the breed explored in the present study, are mainly distributed in the provinces of Ningxia, Gansu, and Inner Mongolia in China and are famous for their fur and climbing abilities [29]. They are often allowed to range freely over large areas, with minimal supervision. Except for the young kids and pregnant and nursing ewes, who receive supplementary feed, the goats graze freely in the mountains throughout the year. There is a common phenomenon for Zhongwei goats known as “summer full, autumn fat, winter and spring fatigue”, which reflects their physiological responses to changing environmental conditions and management challenges in semi-arid regions.

Recent technological advances in the form of movement data collection at high sampling rates, increased availability of remotely-sensed data and imagery, along with the utility of statistical models, have provided a capacity to enhance our understanding of how an organism perceives and operates within its environment [9, 30]. Improvements in tracking technologies enable the recording of high-precision animal movements for extended periods, regardless of weather and lighting conditions, even in remote and inaccessible regions [31]. Remotely sensed vegetation indices, including the Normalized Difference Vegetation Index (NDVI), can provide spatially extensive vegetation phenology information to establish connections with animal movements [30, 32]. In addition to these data-based insights, a variety of modelling approaches, such as behavioral mode models, Hidden Markov Model (HMM), random walk models, Markov processes, and movement metrics and state–space models, have all evolved to analyze and provide insights into animal movement processes [6, 7, 33]. The HMM is a particularly popular model due to its flexibility, straightforward interpretability, and computational feasibility [10, 34, 35]. A HMM is a time series model that consists of two processes: an observed movement process and a hidden state process [6, 36]. By fitting the observed series of step lengths and turn angles into the HMM, a set of hidden behaviors (e.g., foraging and travelling) can be derived while incorporating a series of covariates. Collectively, these improvements in data collection and statistical analysis help to improve our capacity to evaluate interactions between organisms and their environment.

In predator-free environments of many livestock systems, free-ranging livestock decisions are driven by environmental factors such as forage availability, thermal conditions, and topography. This study provides a unique opportunity to investigate animal movement in response to abiotic and biotic factors without the confounding effects of predation risk, allowing for a more focused exploration of how animals optimize resource use in managed landscapes. Here, we integrate GPS-tracking of free-ranging goats, Landsat 8 imagery, and multivariate HMMs, complemented with fitted environmental covariates (e.g. slope, elevation, Vector Ruggedness Measure (VRM), solar radiation, and NDVI) and a time covariate (e.g. specific time of day) to assess the seasonal movement behavior of goats in the semi-arid mountain grassland of the Loess Plateau, China. The specific objectives were to: (1) identify and characterize the behaviors of goats in the absence of predators in a semi-arid natural mountain grassland area; (2) quantify and compare the importance of environmental and time covariates on goat behavior across different seasons; and (3) understand how the seasonal movement behavioral transition of goats is affected by these covariates. These findings contribute to unraveling the behavioral decision-making of free-ranging livestock, providing fundamental information and implications for grazing management and ethology.

Fig. 1
figure 1

Elevation (a) and mountain landscape (terracette micro-topography formed by goat trampling) (b) in the semi-arid hilly rangeland of the western Loess Plateau, China

Full size image

Methods

Study site

The study was conducted near the Wanchan River in the semi-arid hilly rangeland of the western Loess Plateau, China (Fig. 1), a region characterized by severe soil erosion [37]. The region has been grazed for thousands of years [38], resulting in widespread terracette landscapes formed by goat trampling [39, 40]. The mean annual precipitation is 378 mm, with 80% of rainfall occurring from May to September. The mean annual temperature is 7.2℃, with the warmest and coldest months being July (19.4℃) and January (-7.5℃), respectively. The soil is loess, and the hilly terrain has very steep slopes up to 45°. The temperate grassland is dominated by dwarf shrubs, and perennial and annual herbaceous plants.

Collection of goat movement data

Between April 22, 2016 and May 7, 2017, a total of 124 sets of GPS tracking data of Zhongwei goats were collected at 1-second intervals (Supplementary Table S1). For each of the 124 days with GPS recordings, goat herds grazed along different pathways in the mountain region. The shepherd fitted one GPS unit (HOLUX GPSport 245) to the stomach of randomly selected adult males every morning at the den and unloaded the unit when the goats returned to the den in the evening. The GPS units weighed 72 g and represented less than 0.16% of the goats’ body weight. The goats entered the mountain at about 8 am and returned at about 8 pm. During the period when the goat flock was equipped with GPS receivers, there were no incidents of attacks by predators.

Pre-processing of goat movement data

GPS locations were acquired at a fixed interval more than  99.99% of the time. For the data analysis, we removed the GPS locations between the den and the inlet of the mountain to focus on the movement of goats in the mountain. We also removed data that exceeded a movement speed of 5 m/s to eliminate unrealistic movements. Finally, we screened the data based on autocorrelation functions of step length [41], which represents the distance between two sequential GPS locations [42]. Autocorrelation is often considered a problem in GPS tracking because consecutive spatial and temporal locations are not independent, leading to inaccuracies in the estimation of behaviors inferred from their movements [41, 43]. The autocorrelation coefficient of the step length of all movement locations gradually declined from 0.52 (1-second interval) to 0.2 (30-second interval), indicating weak autocorrelation (Supplementary Fig. 1). Therefore, we subsampled the GPS tracking data to include every 30-second entry. Hence, the number of GPS locations was reduced from 3,666,166 at a 1-second interval to 112,785 at a 30-second interval (Supplementary Table S1).

Environmental covariates

Topography can influence the movement characteristics of goats [15, 17, 18]. A 5-m Digital Elevation Model (DEM) was acquired from the Geographical Information Monitoring Cloud Platform (http://www.dsac.cn/). Elevation (m), slope (°), Vector Ruggedness Measure (VRM), and solar radiation (Wh/m2) were extracted based on the DEM data using ArcMap 10.6 (Redlands, 2013). VRM, ranging between 0 (flat) and 1 (most rugged), provides a way to measure terrain ruggedness as the variation in the three-dimensional orientation of grid cells within a neighborhood [44, 45]. VRM measures the heterogeneity of terrain independently of slope [44, 45]. Corresponding to the GPS sampling dates, the cumulative solar radiation of the study area was calculated from 8:00 am to 8:00 pm for each pixel using the ArcPy Python package. Here, NDVI is used as a surrogate for vegetation biomass, and its time series can provide information on phenological changes and plant growth throughout the year [46]. Twelve Landsat 8 OLI/TIRS C1 Level-1 (16-bit images at 30 m resolution for multispectral bands; see https://earthexplorer.usgs.gov/) cloud-free datasets covering the study area were downloaded and paired with coincident GPS data (Supplementary Table S1). After atmospheric correction of the Landsat 8 images in the Environment for Visualizing Images Software [47], the NDVI values were calculated from the red and near-infrared bands in ArcMap 10.6 [48].

Seasonality

Seasonal periods were defined according to the current climatic seasons [49], reflecting changes in forage resource availability. Spring and summer begin when the average temperatures over 5 consecutive days have a moving average that is equal to or greater than 10 °C and 22 °C, while the start of autumn and winter requires five consecutive days with a moving average lower than 22 °C and 10 °C, respectively. This approach resulted in four distinct temperature-based seasons (Supplementary Table S1).

Hidden Markov models (HMMs)

HMMs, which are based on maximum likelihood estimation, enable us to use observed goat movements as a proxy for the underlying states and subsequently deduce the spatial and temporal effects of transitioning between these behavioral states [10, 35, 50]. The observed movement processes of the step length and turn angle were used to estimate the hidden state processes. In this study, the pixel values of elevation, slope, solar radiation, VRM, and NDVI were extracted for all GPS locations of goats. The hidden state transition probability was linked to these environmental covariates and the specific time of day (from 8:00 am to 8:00 pm) during different seasons. All HMMs, assuming a von Mises distribution for turn angle and a gamma distribution for step length, were carried out using the moveHMM package in R version 4.2.3 [35].

To select an appropriate number of states, determine the initial parameters of step length and turn angle, and decide if a covariate should be included, the Akaike Information Criterion (AIC) was applied in the HMMs [35, 51, 52]. To select the appropriate number of movement states, we fitted one-state, two-state, and three-state models with no covariates (null models) for all goats and compared their AIC values. The two-state model with the lower AIC value was selected as the number of movement states (Supplementary Table S2). Based on observations, we found that Zhongwei goats hardly rest during daylight hours, and rumination primarily happens at night [53]. Therefore, the two-state model was selected, and the behavioral states within the movement behavior were classified as either foraging (low step lengths, varying turning angles) or travelling (long step lengths, small turning angles). Next, we determined the initial movement parameters by fitting each null model with 15 sets of randomly selected and reasonable step lengths and turning angles, achieved through an examination of histograms depicting the distribution of step lengths and turning angles. The set of parameters with the lowest AIC value of the model was applied to the null and covariate models [35].

The collinearity of the six covariates was examined, showing that all combinations produced correlation coefficients below 0.42 (Supplementary Fig. S2), indicating limited collinearity [54]. We first fitted univariate HMMs, modeling one covariate at a time and comparing its AIC score with the null model for each season. Second, we fitted multivariate HMMs using a forward stepwise procedure, i.e., iteratively adding covariates to a model based on their importance to the model’s performance, to include or exclude the retained covariates [51, 55, 56]. The minimum number of covariates was entered in the final model if the difference in the AIC value of the multivariate HMM was lower than 2 [51, 55]. Therefore, the final HMMs over the course of full year and individual seasons were fitted and selected, and the Viterbi algorithm was applied to estimate goat behavioral transition probabilities of covariates. The goodness-of-fit of the final models was validated using pseudo-residuals analysis. For a well-fitted model, the pseudo-residuals should exhibit an approximately normal distribution [35, 57]. To assess seasonal variation in behavior, a Least Significant Difference test was conducted in R to compare the movement time, speed, foraging time, and foraging percentage of goats at different climatic seasons.

Results

Behavioral states

The two-state HMMs were fitted to estimate the seasonal impact of environmental factors and specific times of day on the movement behavioral transitions of goats. After a forward covariate selection procedure, four covariates including specific time of day, slope, VRM, and NDVI, were selected as inputs to the final HMMs for summer and over the full course of a year (Table 1). The final HMMs for spring and autumn included three covariates: specific time of day, slope, and NDVI. In contrast, the covariates for winter included slope, VRM, and specific time of day (Table 1). The covariates representing elevation and solar radiation were excluded from all models (Table 1). The pseudo-residuals were checked to assess the goodness-of-fit of the HMMs in spring, summer, autumn, winter, and full year. The quantile-quantile plots for step lengths and turning angles were found to exhibit a normal distribution (Supplementary Fig. S3), indicating that these HMMs were well fit.

Table 1 Summary of the Akaike information criterion (AIC) and ΔAIC values for the null models, one covariate model, and multivariate models. The covariates include normalized difference vegetation index (NDVI), solar radiation, slope, elevation, specific time of day (ToD), and vector ruggedness measure (VRM). In step 1, singlecovariate hidden Markov models (HMMs) were conducted, and its AIC score was compared to that of the null model. The lower the ΔAIC, the higher the relative importance of the covariate for explaining the goat behavior states. In step 2, multivariate HMMs were developed by forward stepwise selection procedure, i.e., a procedure of iteratively adding covariates to a model based on their importance to the model’s performance. The minimum number of covariates was chosen as the final model if the difference in the AIC value of the multivariate HMM was lower than 2
Full size table
Fig. 2
figure 2

The spatial distribution of behavioral states for goats in full year (124 daily goat tracks), spring (37 daily goat tracks), summer (21 daily goat tracks), autumn (14 daily goat tracks), and winter (52 daily goat tracks) at a time interval of 30 s

Full size image
Fig. 3
figure 3

Histograms depicting a gamma distribution for step length and a von Mises distribution for turn angle in the foraging and travelling states across different seasons, as estimated by the Hidden Markov Models (HMMs). The parameter estimates for each state include mean step length (mean) with standard deviation (sd), mean turning angle direction (mu), and angle concentration (kappa) at a time interval of 30 s

Full size image

The movement of goats alternated between the foraging and travelling states based on the applied HMMs (Fig. 2), with some routes being used mainly for travelling. The density distribution patterns of step length and turn angle were similar in different seasons (Fig. 3). Across different seasons, foraging states had low step lengths (mean value varied from 5.41 to 7.68 m), a mean turning angle around zero (i.e. mu close to zero), and minimally concentrated turning angles (i.e. low kappa), indicating slow, tortuous, and undirected movements (Fig. 3). The travelling state showed a correlation with longer step lengths (mean value varied from 15.58 to 18.31 m), a mean turning angle around zero (i.e. mu close to zero), and highly concentrated turning angles (i.e. high kappa), indicating mainly fast and directed movements (Fig. 3). Based on the seasonal HMMs, the foraging step length was longer in autumn, while the travelling step length was shorter than that in other seasons, possibly because autumn is an estrus period and mating season. Breeding may have an impact on speed or agility of male goats, thereby affecting their behavioral classification.

Behavioral response to covariates

The importance of covariates for the movement behavioral transitions of goats varied in different seasons, as shown by the forward covariate selection in the multivariate HMMs (Table 1). During both the full year and summer, four covariates significantly affected the movement behavior of goats. For the full year, the importance of the covariates influencing movement behavior was ranked as follows: slope > specific time of day > VRM > NDVI, while in summer, the ranking shifted as follows: specific time of day > slope > VRM > NDVI, based on the improvement in model performance according to the AIC (Table 1). During spring, autumn, and winter, three covariates significantly affected the movement behavior of goats. For spring and autumn, the importance order was as follows: slope > specific time of day > NDVI, while for winter, the importance order was as follows: slope > VRM > specific time of day (Table 1). These results suggested that slope was generally a key factor influencing the movement behavior of goats. However, in summer, the specific time of day became the primary factor driving their movements.

Understanding is limited regarding balancing the importance of environmental factors and daily hours in affecting goat behavior. The abiotic factor of slope influenced movement behavior more than the biotic forage resource factor, represented by NDVI. Like many other animals, goats move in ways that balance energy acquisition from food intake and energy expenditure to find food, which are generally influenced by biotic and abiotic factors. The movement behavior identified here highlights that conserving energy on a day-to-day basis is more important than acquiring energy, especially in scenarios where resources are scarce or environmental conditions are harsh, such as in semi-arid mountain grassland areas. Elevation and solar radiation did not improve to enhance the predictive fit in any season, nor NDVI in winter or VRM in spring and autumn (Table 1). Elevation and solar radiation are probably not important drivers of their movement behavioral transition because Zhongwei goats are good at climbing and tolerant of hot thermal conditions. In conditions with low amounts of food resources and plants withering in winter months, food resources no longer influenced their behavioral decisions. Instead, in winter slope had the greatest impact on their behavior (Table 1), with goats focusing on conserving energy for survival.

Fig. 4
figure 4

The seasonal state probabilities of foraging and travelling movement behaviors of goats predicted by HMMs, as functions of slope, specific time of day, Vector Ruggedness Measure (VRM), and Normalized Difference Vegetation Index (NDVI) with 95% confidence intervals

Full size image

The probability of foraging increased with greater slope, specific time of day, and NDVI, but decreased with increasing VRM, whereas the probability of travelling showed the opposite pattern (Fig. 4). Specifically, the increase in foraging probability with NDVI was more obvious in spring (an increase from 0.44 to 0.89 within the NDVI range of 0.05–0.26) than in summer (an increase from 0.48 to 0.75 within the NDVI range of 0.05–0.32) and autumn (an increase from 0.56 to 0.75 within the NDVI range of 0.05–0.34) (Fig. 4). The early onset of vegetation growth in spring, with rapid changes in NDVI and increasing availability of high-quality forage, had the largest influence on the goat behavioral transition. In summer and autumn, the quality of the forage often decreases as it matures, leading to a moderate impact on the movement’s behavioral characteristics. In winter, the diminished availability of food supplies meant that NDVI failed to enhance the predictive fit and was therefore excluded from the HMM.

Behavioral time allocation

The behavioral time allocation varied in response to different seasons. The goats stayed significantly longer in the mountains in summer and autumn than in spring and winter (Fig. 5a). Such a pattern corresponds to longer daylight hours during summer and autumn, in contrast to winter and spring. The goats spent similar foraging time in different seasons, ranging from 4.80 to 5.15 hours (Fig. 5c), but the goats moved more slowly and spent a larger percentage of time for foraging in spring and winter than in summer and autumn (Fig. 5b, d). The percentage of time that the goats allocated toward foraging was negatively related to the available forage resources, i.e. goats spent a higher percentage of time foraging to meet their energy requirements in periods with limited forage resources.

Fig. 5
figure 5

The boxplot of seasonal movement of daily duration (a), average daily movement speed (b), daily foraging time (c), and percentage of daily foraging (d) of goats in the mountain derived from the one-year HMM. The black dashed line represents the average value of the full year. Different letters indicate seasonal significant differences at P = 0.05

Full size image

The goats’ hourly foraging percentage increased from 9:00 am to 8:00 pm across various seasons (Fig. 6), which meant that goats tended to move more slowly and undirected during the later hours of daylight. In the hotter seasons in summer and autumn, the foraging percentage was lower in the midday from 12:00 to 15:00, compared to the colder seasons in spring and winter (Fig. 6). Foraging more in late daylight hours could be linked to a decrease in the risk of ingesting gastrointestinal parasites from wet forage blades in the morning and an increase in the storage of energy for rumination during the night. In addition, a goat’s internal state could regulate their hourly activities, such that they are more energetic and quick-moving in the early morning and lethargic and slow-moving in the late afternoon.

Fig. 6
figure 6

The seasonal foraging percentage at a specific time of day from the one-year HMM

Full size image

Discussion

Goats demonstrate highly flexible and adaptable in harsh mountainous terrain and changing environmental features. Goat responses to food resources are complex and context-dependent and can be influenced by the spatial and temporal dynamics of various plant species [8], as well as their height and density [58], and forage quantity and quality [21]. Many studies have suggested that forage availability is the major determinant of the grazing behavior of goats, often have overlooked other factors such as terrain [59,60,61]. However, our results based on the HMMs show that slope and specific time of day influence movement behavior more than forage resource in the mountain region (Table 1). Hence, the results highlight the importance of considering the combined effect of environmental factors and daily behavioral rhythms on the movement characteristics and behavioral transitions of livestock.

Effect of biotic factors on movement behavior

The specific movement behavioral transition of goats was influenced by the seasonal dynamics of food resources (Figs. 4 and 5). The goats moved more sinuously and slowly and foraged more in patches with abundant forage resources in spring, summer, autumn, and full year (Fig. 4). Some authors [8, 62] believe that herbivores reduce speed and increase forage in areas of abundant food resources because they spend more time feeding to obtain more energy. As a comparative example, the movement speed of goats was observed to decrease as forage abundance increased in winter for a dry grassland area of the northern Negev region of Israel [21]. Schlecht et al. [63] found that the amount of forage encountered along the itineraries of free-ranging goats was higher than the average available forage. In contrast, increases in the speed of goats in response to available resources have been reported during summer in a Mongolian grassland [22], in summer in south Texas [23], and in spring in the dry grasslands of the northern Negev region of Israel [21]. Shimada et al. [22] reported that goats do not have to spend time selecting forage under vegetation-rich conditions. Svoray et al. [21] contend that the change in the opposite response of the speed of goats to forage availability between spring and winter may be due to the increase in biomass availability in spring. When taken together, these studies demonstrate that goat movement exhibits plasticity and adaptability based on their speed response to forage quantity or palatable species, highlighting the importance of additional and focused studies such as the one presented herein.

The period of high-quality forage significantly influences the behavioral transitions of goats. The impact of NDVI on the probability of foraging behavior was largest in spring and lowest in winter (Table 1; Fig. 4), which suggests that the early growth stage of plants in spring, characterized by high forage quality [16, 27, 64], exerts the strongest influence on goat behavior. In summer and autumn, as forage matures, its quality tends to diminish [65], which moderately affects the movement and behavioral patterns of goats. The stage of plant withering in winter represents the period of low food resources. Hence, NDVI has the weakest impact on goat behavior during winter. Instead, slope played the most significant role in influencing their behavior in winter (Table 1), as goats prioritize energy conservation rather than energy acquisition for survival. The early onset of vegetation growth, with rapid changes in NDVI leading to a shorter period of availability of high-quality forage, will have a significant influence on the animal population dynamics and life histories in alpine environments [64]. There is a “green wave” hypothesis that migrating animals should track high-quality forage at the leading edge of spring green-up to get more benefits [27, 66]. Behavioral transition for non-migrating livestock is also strongly affected by the period of high-quality forage.

Goats increased their allocation of foraging percentage in winter and spring when the food resources were low, and daylight hours were short (Fig. 5). Considering the phenomenon of “summer fullness, autumn fat, spring and winter fatigue” in goats within the semi-arid grassland studied herein, it is reasonable to assume that they need to dedicate a higher percentage of foraging behavior to meet their energy requirements in the period of low forage availability. Safari et al. [67] and Nyamangara et al. [68] also found that the proportion of foraging time increased during seasons of low forage availability in a semi-arid area. Conversely, goats dedicated a higher percentage of time to foraging in Mediterranean forest rangelands during periods with higher food availability, due to the effects of a high abundance of preferred shrubs and herbaceous plants [59]. Kronberg and Malechek [59] found that the foraging time allocation of goats was positively related to available forage resources and inversely related to dietary crude protein content. Overall, the foraging time allocation of goats is complex and context-dependent, relating to interactions between movement behavior and dynamic abiotic and biotic features. Understanding the allocation of time to various behaviors during different seasons is contingent upon variations in both the availability and quality of forage.

Effect of abiotic factors on movement behavior of goats

Some studies have concentrated on the movement pattern of goats in relation to abiotic factors [17, 18], but few have explored how abiotic factors influence movement characteristics and behavior. Our results showed that the movement behavior of goats was poorly predicted by elevation and solar radiation, while slope and VRM proved to be crucial variables for explaining their movement characteristics in a semi-arid mountain grassland (Table 1). Zhongwei goats are good at climbing and tolerant of hot thermal conditions [29], which partially explains why their movement behavior was unaffected by elevation and solar radiation. The goats exhibited a higher probability of foraging behavior in regions characterized by greater slopes irrespective of seasons (Fig. 3; Table 1). As ascending and descending movements require time and energy [69, 70], goats tend to favor movement along gentle slopes [15]. Goats use steep slopes to forage, expending more energy in order to gain additional energy from feeding. Some wild mountain goats specifically inhabit rugged terrain, as it facilitates escape from predators [71,72,73]. The domestic goats exhibited a higher probability of foraging behavior in regions characterized by lower VRM (Fig. 3; Table 1). VRM measured terrain ruggedness independently of slope [44, 45]. In this study, areas with high VRM values were primarily located in valleys, and the domestic goats tended to exhibit more travelling behavior within these valley regions.

Diurnal behavior of goats

The diurnal movement behavior of goats was rhythmic in the mountain regions. The percentage of time spent on foraging increased from the morning to the afternoon across different seasons (Fig. 6). Foraging had the highest probability of occurring in late daylight hours in all four seasons (Fig. 4). Similar results have been observed for castrated male goats in silvopasture [74, 75], in semi-arid regions in India [76], in a heathland area [61], and in a Mediterranean forest rangeland [59]. Karki et al. [77] found that the possible dew on vegetation in the morning hindered the foraging of goats. Reducing wet forage could aid goats in decreasing the risk of ingesting gastrointestinal parasites, which tend to ascend onto wet forage blades and descend closer to the ground as the forage dry up [78]. Additionally, it was observed that the rumination of Zhongwei goats mostly occurred at night [53]. Increased foraging activity in the afternoon, followed by rumination at night, helps compensate for reduced grazing opportunities during nighttime. This pattern ensures adequate daytime feeding and supports efficient rest, digestion, and nutrient absorption at night. In addition, many studies have found goats to be reluctant to graze in the middle of the day [61, 79], which was also showed in our study, especially in summer (Fig. 6), because part of the time was used for rumination during the hottest hours in the daytime [53]. Moreover, a goat’s inherent physiological state may shape their daily activities, resulting in them being more energetic and fast-moving in the early morning, yet weary and slow-moving by the late afternoon.

Management implications

Grazing by domesticated livestock is the most extensive land-use activity globally [1]. However, livestock overgrazing is a widespread and significant driver of grassland degradation [80, 81], particularly in northern China, where 90% of grasslands have suffered from overgrazing and subsequent degradation [82]. Recent advancements in real-time GPS tracking and other sensor technologies have given rise to the emerging field of precision livestock management [83], which provides information on where, when, why, and how livestock move, offers valuable insights for improving livestock management, promoting environmental sustainability, and maximizing economic benefits [84]. Managers can ensure that livestock grazes more evenly across the landscape, and don’t graze too much time in one area to prevent overgrazing, such as prohibiting overexploitation of the areas of entrance pass and exit pass by periodically changing these places through intervention [85]. Implementing adaptive multi-paddock grazing management protocols, which include maintaining short grazing periods and sufficient litter and plant cover, helps protect the soil and promotes rapid plant regrowth [12]. For example, adding supplementary feed for goats during winter and early spring can sustain the nutritional needs of the herd, and reducing herd size through selling goats at the end of the growing season is another strategy that supports both ecological health and economic optimization [85]. In addition, managers can implement time-specific grazing schedules that align with the natural behaviors of goats. By allowing livestock access to more foraging resources during peak activity hours, particularly in the afternoon when foraging is most active, grazing efficiency can be maximized. During the hottest parts of the day, especially in summer, goats tend to ruminate. Providing shaded resting areas or shelters during these times can help them manage heat stress and improve digestion.

Conclusions

Understanding livestock movement is crucial for unraveling how livestock engage with their environment at different spatial and temporal scales, which offers valuable insights for both herd and rangeland management [84]. HMMs applied for fine-scale analysis of movement behavior were found to be useful for understanding the movement behavioral transitions of domestic goats in response to environmental factors and specific times of day. Slope plays a vital role in explaining their movement characteristics, while elevation and solar radiation provide limited information on movement behavior. Goat movement behavior was influenced more by slope and a specific time of day than by the forage resource, as represented by NDVI. The impact of NDVI on behavioral transitions was large in spring but insignificant in winter, when the limited availability of food resources did not affect goat behavior. The likelihood of foraging behavior increased with slope, NDVI, and specific time of day (from 8:00 am to 8:00 pm), while it decreased with VRM. These results are critical for predicting how goats perceive and utilize their environment under increased soil degradation, biodiversity loss, and demand for animal products. In addition, behavioral transition is complex and context-dependent, related to interactions between movement behavior and dynamic abiotic and biotic features. The space–time–action system links the behavioral state of goats and landscape characteristics, which provides new information for understanding the movement behavior of goats. Future research should focus on developing a precision livestock management system employing real-time tracking, remote sensing imagery, and robust statistical analysis models to integratelivestock behavior with grassland conditions, which allows managers to enhance livestock well-being and promote environmental sustainability.

Data availability

Data is provided via Figshare at: https://doi.org/10.6084/m9.figshare.28784639.v1.

Abbreviations

AIC:

Akaike Information Criterion

DEM:

Digital Elevation Model

GPS:

Global Positioning System

HMM:

Hidden Markov Model

NDVI:

Normalized Difference Vegetation Index

References

  1. Food and Agriculture Organization of the United Nations (FAO): The State of Food and Agriculture: Leveraging food systems for inclusive rural transformation. In, Rome. Italy; 2017.

  2. Maestre FT, Le Bagousse-Pinguet Y, Delgado-Baquerizo M, Eldridge DJ, Saiz H, Berdugo M, Gozalo B, Ochoa V, Guirado E, García-Gómez M, et al. Grazing and ecosystem service delivery in global drylands. Science. 2022;378(6622):915–20.

  3. Herrero M, Havlík P, Valin H, Notenbaert A, Rufino MC, Thornton PK, Blümmel M, Weiss F, Grace D, Obersteiner M. Biomass use, production, feed efficiencies, and greenhouse gas emissions from global livestock systems. Proc Natl Acad Sci. 2013;110(52):20888–93.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Herrero M, Thornton PK, Gerber P, Reid RS. Livestock, livelihoods and the environment: Understanding the trade-offs. Curr Opin Environ Sustain. 2009;1(2):111–20.

    Article  Google Scholar 

  5. Schick RS, Loarie SR, Colchero F, Best BD, Boustany A, Conde DA, Halpin PN, Joppa LN, McClellan CM, Clark JS. Understanding movement data and movement processes: current and emerging directions. Ecol Lett. 2008;11(12):1338–50.

    Article  PubMed  Google Scholar 

  6. Morales JM, Haydon DT, Frair J, Holsinger KE, Fryxell JM. Extracting more out of relocation data: Building movement models as mixtures of random walks. Ecol 2004, 85(9):2436–45.

  7. Patterson TA, Thomas L, Wilcox C, Ovaskainen O, Matthiopoulos J. State-space models of individual animal movement. Trends Ecol Evol. 2008;23(2):87–94.

    Article  PubMed  Google Scholar 

  8. Fryxell JM, Hazell M, Börger L, Dalziel BD, Haydon DT, Morales JM, McIntosh T, Rosatte RC. Multiple movement modes by large herbivores at multiple Spatiotemporal scales. Proc Natl Acad Sci. 2008;105(49):19114.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Nathan R, Monk CT, Arlinghaus R, Adam T, Alós J, Assaf M, Baktoft H, Beardsworth CE, Bertram MG, Bijleveld AI. Big-data approaches lead to an increased Understanding of the ecology of animal movement. Science. 2022;375(6582):eabg1780.

    Article  CAS  PubMed  Google Scholar 

  10. McClintock BT, Langrock R, Gimenez O, Cam E, Borchers DL, Glennie R, Patterson TA. Uncovering ecological state dynamics with hidden Markov models. Ecol Lett. 2020;23(12):1878–903.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Li BV, Jiang B. Responses of forest structure, functions, and biodiversity to livestock disturbances: A global meta-analysis. Glob Chang Biol. 2021;27(19):4745–57.

    Article  CAS  PubMed  Google Scholar 

  12. Teague R, Kreuter U. Managing grazing to restore soil health, ecosystem function, and ecosystem services. Front Sustain Food Syst. 2020;4:534187.

    Article  Google Scholar 

  13. Fetzel T, Havlik P, Herrero M, Erb K-H. Seasonality constraints to livestock grazing intensity. Glob Chang Biol. 2017;23(4):1636–47.

    Article  PubMed  Google Scholar 

  14. Jiang K, Teuling AJ, Chen X, Huang N, Wang J, Zhang Z, Gao R, Men J, Zhang Z, Wu Y, et al. Global land degradation hotspots based on multiple methods and indicators. Ecol Indic. 2024;158:111462.

    Article  Google Scholar 

  15. Ganskopp D, Cruz R, Johnson DE. Least-effort pathways? A GIS analysis of livestock trails in rugged terrain. Appl Anim Behav Sci. 2000;68(3):179–90.

    Article  CAS  PubMed  Google Scholar 

  16. Cheng H, Jin B-C, Luo K, Zhang X-L, Pei J-Y, Zhang Y-H, Han L-Q, Tang J-Q, Li F-M, Sun G-J, et al. Seasonal resource selection of free-ranging Zhongwei goats in the semi-arid grassland. Animal. 2023;17(10):100972.

    Article  PubMed  Google Scholar 

  17. Rivero MJ, Grau-Campanario P, Mullan S, Held SDE, Stokes JE, Lee MRF, Cardenas LM. Factors affecting site use preference of grazing cattle studied from 2000 to 2020 through GPS tracking: a review. Sensors. 2021;21(8):2696.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Owen-Smith N. Effects of Temporal variability in resources on foraging behaviour. Dordrecht: Springer Netherlands; 2008.

    Book  Google Scholar 

  19. White KS, Gregovich DP. Mountain goat resource selection in relation to mining-related disturbance. Wildl Biol. 2017;2017(1):wlb00277.

    Article  Google Scholar 

  20. Hebblewhite M, Merrill E, McDermid G. A multi-scale test of the forage maturation hypothesis in a partially migratory ungulate population. Ecol Monogr. 2008;78(2):141–66.

    Article  Google Scholar 

  21. Svoray T, Shafran-Nathan R, Ungar ED, Arnon A, Perevolotsky A. Integrating GPS technologies in dynamic spatio-temporal models to monitor grazing habits in dry rangelands. Taylor and Francis: Leiden;; 2009.

  22. Shimada S, Yokoyama H, Toyoda H, Sekiyama A, Aosier B, Yokohama M. GIS analysis of grazing selection by goats in Tov Province, Mongolia. J Arid Land Stud. 2009;19(1):29–32.

    Google Scholar 

  23. Etzenhouser MJ, Owens MK, Spalinger DE, Murden SB. Foraging behavior of browsing ruminants in a heterogeneous landscape. Landsc Ecol. 1998;13(1):55–64.

    Article  Google Scholar 

  24. Chebli Y, El Otmani S, Hornick J-L, Keli A, Bindelle J, Cabaraux J-F, Chentouf M. Forage availability and quality, and feeding behaviour of Indigenous goats grazing in a mediterranean silvopastoral system. Ruminants. 2022;2(1):74–89.

    Article  Google Scholar 

  25. Ngwa AT, Pone DK, Mafeni JM. Feed selection and dietary preferences of forage by small ruminants grazing natural pastures in the Sahelian zone of Cameroon. Anim Feed Sci Technol. 2001;88(3):253–66.

    Google Scholar 

  26. Chebli Y, Otmani SE, Chentouf M, Hornick J-L, Bindelle J, Cabaraux J-F. Foraging behavior of goats browsing in Southern mediterranean forest rangeland. Animals. 2020;10(2):196.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Wade C, Trotter MG, Bailey DW. Small ruminant landscape distribution: A literature review. Small Ruminant Res. 2023;223:106966.

    Article  Google Scholar 

  28. Nair MRR, Sejian V, Silpa MV, Fonsêca VFC, de Melo Costa CC, Devaraj C, Krishnan G, Bagath M, Nameer PO, Bhatta R. Goat as the ideal climate-resilient animal model in tropical environment: revisiting advantages over other livestock species. Int J Biometeorol. 2021;65(12):2229–40.

    Article  PubMed  Google Scholar 

  29. Zhang D. The ecological environment and characteristics of Zhongwei goat. J Grass Livest. 1995;3:5–7.

    CAS  Google Scholar 

  30. Remelgado R, Wegmann M, Safi K. rsmove—An R package to Bridge remote sensing and movement ecology. Methods Ecol Evol. 2019;10(8):1212–21.

    Article  Google Scholar 

  31. Hebblewhite M, Haydon DT. Distinguishing technology from biology: a critical review of the use of GPS telemetry data in ecology. Philos Trans R Soc Lond B Biol Sci. 2010;365(1550):2303–12.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Neumann W, Martinuzzi S, Estes AB, Pidgeon AM, Dettki H, Ericsson G, Radeloff VC. Opportunities for the application of advanced remotely-sensed data in ecological studies of terrestrial animal movement. Mov Ecol. 2015;3:8.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Bartumeus F, da Luz MGE, Viswanathan GM, Catalan J. Animal search strategies: a quantitative random-walk analysis. Ecology. 2005;86(11):3078–87.

    Article  Google Scholar 

  34. Patterson TA, Basson M, Bravington MV, Gunn JS. Classifying movement behaviour in relation to environmental conditions using hidden Markov models. J Anim Ecol. 2009;78(6):1113–23.

    Article  PubMed  Google Scholar 

  35. Michelot T, Langrock R, Patterson TA. MoveHMM: an R package for the statistical modelling of animal movement data using hidden Markov models. Methods Ecol Evol. 2016;7(11):1308–15.

    Article  Google Scholar 

  36. Bartumeus F, Catalan J, Viswanathan GM, Raposo EP, da Luz MGE. The influence of turning angles on the success of non-oriented animal searches. J Theor Biol. 2008;252(1):43–55.

    Article  CAS  PubMed  Google Scholar 

  37. Wang L, Shao Ma, Wang Q, Gale WJ. Historical changes in the environment of the Chinese loess plateau. Environ Sci Policy. 2006;9(7):675–84.

    Article  Google Scholar 

  38. Wang X, Zhang W, Sun Y, Han G. Yuzhong County annals. Lanzhou, China: GanSu People’s Publishing House; 2001.

    Google Scholar 

  39. Jin B, Sun G, Zhang Y, Zou M, Ni X, Luo K, Zhang X, Cheng H, Li F, Wu XB. Livestock tracks transform resource distribution on terracette landscapes of the loess plateau. Ecosphere. 2016;7(4):e01337.

    Article  Google Scholar 

  40. Jin B, Cheng H, Sun G, Li F, Wu XB. Multi-parallel structure and a generalized conceptual model of livestock track network. Catena 2022, 216:106380.

  41. Boyce MS, Pitt J, Northrup JM, Morehouse AT, Knopff KH, Cristescu B, Stenhouse GB. Temporal autocorrelation functions for movement rates from global positioning system radiotelemetry data. Philos Trans R Soc Lond B Biol Sci. 2010;365(1550):2213.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Turchin P. Ecology of animal movement. In.: Princeton, NJ: Princeton University Press; 1998.

  43. Gupte PR, Beardsworth CE, Spiegel O, Lourie E, Toledo S, Nathan R, Bijleveld AI. A guide to pre-processing high‐throughput animal tracking data. J Anim Ecol. 2022;91(2):287–307.

    Article  PubMed  Google Scholar 

  44. Hobson R. Surface roughness in topography: quantitative approach. In: Spatial analysis in geomorphology. Routledge; 2019: 221–246.

  45. Sappington JM, Longshore KM, Thompson DB. Quantifying landscape ruggedness for animal habitat analysis: A case study using Bighorn sheep in the Mojave desert. J Wildl Manage. 2007;71(5):1419–26.

    Article  Google Scholar 

  46. Pettorelli N, Ryan S, Mueller T, Bunnefeld N, Jędrzejewska B, Lima M, Kausrud K. The normalized difference vegetation index (NDVI): unforeseen successes in animal ecology. Clim Res. 2011;46(1):15–27.

    Article  Google Scholar 

  47. Exelis Visual Information Solutions. Environment for Visualizing Images (ENVI). Version 5.0. 2013.

  48. ESRI: ArcGIS Desktop. Release 10.6. In. Redlands, CA: Environmental Systems Research Institute: Release;; 2018.

    Google Scholar 

  49. China Meteorological Administration: Meteorological industry standard of the People’s Republic of China: division of climatic season. In, Beijing. China: China Meteorological Press; 2012.

  50. Mor B, Garhwal S, Kumar A. A systematic review of hidden Markov models and their applications. Arch Comput Method E. 2021;28(3):1429–48.

    Article  Google Scholar 

  51. van Beest FM, Mews S, Elkenkamp S, Schuhmann P, Tsolak D, Wobbe T, Bartolino V, Bastardie F, Dietz R, von Dorrien C, et al. Classifying grey seal behaviour in relation to environmental variability and commercial fishing activity - a multivariate hidden Markov model. Sci Rep. 2019;9(1):5642.

    Article  PubMed  PubMed Central  Google Scholar 

  52. Akaike H. Information theory and an extension of the maximum likelihood principle. Selected papers of Hirotugu Akaike. New York: Springer New York; 1998. pp. 199–213.

    Chapter  Google Scholar 

  53. Luo Y, Zhao Y, Wang J. Study on main behavior of Zhongwei goats in spring. J Gansu Agr Univ. 1994;29(1):20–6.

    Google Scholar 

  54. Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carré G, Marquéz JRG, Gruber B, Lafourcade B, Leitão PJ, et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography. 2013;36(1):27–46.

    Article  Google Scholar 

  55. Ferreira E, Valerio F, Medinas D, Fernandes N, Craveiro J, Costa P, Silva JP, Carrapato C, Mira A, Santos S. Assessing behaviour States of a forest carnivore in a road-dominated landscape using hidden Markov models. Nat Conserv. 2022;47:155–75.

    Article  Google Scholar 

  56. Beumer LT, Pohle J, Schmidt NM, Chimienti M, Desforges J-P, Hansen LH, Langrock R, Pedersen SH, Stelvig M, van Beest FM. An application of upscaled optimal foraging theory using hidden Markov modelling: year-round behavioural variation in a large Arctic herbivore. Mov Ecol. 2020;8(1):25.

    Article  PubMed  PubMed Central  Google Scholar 

  57. Zucchini W, MacDonald IL. Hidden Markov models for time series: an introduction using R. New York: Chapman and Hall/CRC; 2009.

    Book  Google Scholar 

  58. Laca EA, Distel RA, Criggs TC, Demment MW. Effects of canopy structure on patch depression by grazers. Volume 75. Ecology; 1994. pp. 706–16. 3.

  59. Chebli Y, El Otmani S, Hornick J-L, Keli A, Bindelle J, Chentouf M, Cabaraux J-F. Using GPS collars and sensors to investigate the grazing behavior and energy balance of goats browsing in a mediterranean forest rangeland. Sensors. 2022;22(3):781.

    Article  PubMed  PubMed Central  Google Scholar 

  60. Zampaligré N, Schlecht E. Livestock foraging behaviour on different land use classes along the semi-arid to sub-humid agro-ecological gradient in West Africa. Environ Dev Sustain. 2018;20(2):731–48.

    Article  Google Scholar 

  61. Ferreira LMM, Celaya R, Benavides R, Jáuregui BM, García U, Sofia Santos A, Rosa García R, Rodrigues MAM, Osoro K. Foraging behaviour of domestic herbivore species grazing on heathlands associated with improved pasture areas. Livest Sci. 2013;155(2):373–83.

    Article  Google Scholar 

  62. Bailey DW, Gross JE, Laca EA, Rittenhouse LR, Coughenour MB, Swift DM, Sims PL. Mechanisms that result in large herbivore grazing distribution patterns. J Range Manage. 1996;49(5):386–400.

    Article  Google Scholar 

  63. Schlecht E, Hiernaux P, Kadaouré I, Hülsebusch C, Mahler F. A spatio-temporal analysis of forage availability and grazing and excretion behaviour of herded and free grazing cattle, sheep and goats in Western Niger. Agric Ecosyst Environ. 2006;113(1):226–42.

    Article  Google Scholar 

  64. Pettorelli N, Pelletier F, Hardenberg Av, Festa-Bianchet M, Côté SD. Early onset of vegetation growth vs. rapid green-up: impacts on juvenile mountain ungulates. Ecology. 2007;88(2):381–90.

    Article  PubMed  Google Scholar 

  65. Nelson CJ, Moser LE. Plant Factors Affecting Forage Quality. In: Forage Quality, Evaluation, and Utilization. 1994: 115–154.

  66. Van der Graaf A, Stahl J, Klimkowska A, Bakker JP, Drent RH. Surfing on a green wave-how plant growth drives spring migration in the Barnacle Goose Branta leucopsis. Ardea 2006, 94(3):567.

  67. Safari J, Mushi DE, Kifaro GC, Mtenga LA, Eik LO. Seasonal variation in chemical composition of native forages, grazing behaviour and some blood metabolites of small East African goats in a semi-arid area of Tanzania. Anim Feed Sci Technol. 2011;164(1):62–70.

    Article  CAS  Google Scholar 

  68. Nyamangara ME, Ndlovu LR. Feeding behaviour, feed intake, chemical and botanical composition of the diet of Indigenous goats Raised on natural vegetation in a semi-arid region of Zimbabwe. J Agric Sci. 1995;124(3):455–61.

    Article  Google Scholar 

  69. Taylor CR, Shkolnik A, Dmi’el R, Baharav D, Borut A. Running in cheetahs, gazelles, and goats: energy cost and limb configuration. Am J Physiol. 1974;227(4):848–50.

    Article  CAS  PubMed  Google Scholar 

  70. Lachica M, Prieto C, Aguilera JF. The energy costs of walking on the level and on negative and positive slopes in the granadina goat (Capra hircus). Brit J Nutr. 1997;77(1):73–81.

    Article  CAS  PubMed  Google Scholar 

  71. Weterings MJA, Moonen S, Prins HHT, van Wieren SE, van Langevelde F. Food quality and quantity are more important in explaining foraging of an intermediate-sized mammalian herbivore than predation risk or competition. Ecol Evol. 2018;8(16):8419–32.

    Article  PubMed  PubMed Central  Google Scholar 

  72. Gross JE, Kneeland MC, Reed DF, Reich RM. Gis-based habitat models for mountain goats. J Mammal. 2002;83(1):218–28.

    Article  Google Scholar 

  73. Lowrey B, Garrott RA, Miyasaki HM, Fralick G, Dewey SR. Seasonal resource selection by introduced mountain goats in the southwest Greater Yellowstone Area. Ecosphere 2017, 8(4):e01769.

  74. Bhattrai S, Karki U, Poudel S, McElhenney W, Paneru B. Diurnal behavior and distribution patterns of Kiko wethers and Katahdin Rams in woodlands with different vegetation heights during fall. Agroforest Syst. 2020;94(5):1809–23.

    Article  Google Scholar 

  75. Poudel S, Karki U, Karki Y, Tillman A. Diurnal behavior of Kiko wethers in southern-pine silvopastures planted with warm-season forages. Small Ruminant Res. 2019;175:1–6.

    Article  Google Scholar 

  76. Solanki G. Grazing behaviour and foraging strategy of goats in semi-arid region in India. Trop Ecol. 2000;41(2):155–60.

    Google Scholar 

  77. Karki U, Karki Y, Khatri R, Tillman A. Diurnal behavior and distribution patterns of Kiko wethers in southern-pine silvopastures during the cool-season grazing period. Agroforest Syst. 2019;93(1):267–77.

    Article  Google Scholar 

  78. Zvinorova PI, Halimani TE, Muchadeyi FC, Matika O, Riggio V, Dzama K. Prevalence and risk factors of Gastrointestinal parasitic infections in goats in low-input low-output farming systems in Zimbabwe. Small Ruminant Res. 2016;143:75–83.

    Article  CAS  Google Scholar 

  79. Bojkovski D, Štuhec I, Kompan D, Zupan M. The behavior of sheep and goats co-grazing on pasture with different types of vegetation in the karst region1. J Anim Sci. 2014;92(6):2752–8.

    Article  CAS  PubMed  Google Scholar 

  80. Lai L, Kumar S. A global meta-analysis of livestock grazing impacts on soil properties. PLoS ONE. 2020;15(8):e0236638.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  81. Eldridge DJ. M Delgado-Baquerizo 2017 Continental-scale impacts of livestock grazing on ecosystem supporting and regulating services. Land Degrad Dev 28 4 1473–81.

    Article  Google Scholar 

  82. Kemp DR, Guodong H, Xiangyang H, Michalk DL, Fujiang H, Jianping W, Yingjun Z. Innovative grassland management systems for environmental and livelihood benefits. Proc Natl Acad Sci. 2013;110(21):8369–74.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  83. Bailey DW, Trotter MG, Tobin C, Thomas MG. Opportunities to apply precision livestock management on rangelands. Front Sustain Food Syst. 2021;5:611915.

    Article  Google Scholar 

  84. Zobel G, Neave HW, Webster J. Understanding natural behavior to improve dairy goat (Capra hircus) management systems. Transl Anim Sci. 2018;3(1):212–24.

    Article  PubMed  PubMed Central  Google Scholar 

  85. Cheng H, Jin B, Luo K, Pei J, Zhang X, Zhang Y, Tang J, Yang Q, Sun G. Vegetation response to goats grazing intensity in semiarid hilly grassland of the loess plateau, Lanzhou, China. Sustainability. 2021;13(6):3569.

    Article  Google Scholar 

Download references

Acknowledgements

We thank the local shepherd Jie-Long Jin for helping us to collect GPS tracking data with his goat herd and sharing his knowledge with us.

Funding

The work was partially supported by grants to BJ from the National Natural Science Foundation of China (32360286), and grants to HC from Henan Normal University (QD2021093).

Author information

Authors and Affiliations

  1. School of Tourism, Henan Normal University, Xinxiang, 453007, China

    Hua Cheng

  2. Climate and Livability Initiative, Division of Biological and Environmental Sciences and Engineering, King Abdullah of Science and Technology (KAUST), Thuwal, 23955-6900, Saudi Arabia

    Hua Cheng, Kasper Johansen & Matthew F. McCabe

  3. College of Animal Science, Guizhou University, Guiyang, 550025, China

    Baocheng Jin

  4. State Key Laboratory of Grassland Agro-ecosystems, College of Ecology, Lanzhou University, Lanzhou, Gansu, 730000, China

    Guojun Sun

Authors
  1. Hua Cheng
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Kasper Johansen
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Baocheng Jin
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Guojun Sun
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Matthew F. McCabe
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

G.S., H.C., and B.J. developed the research idea. H.C. and B.J. collected the GPS data, with the help from G.S. H.C. performed the data analyses. H.C. wrote the manuscript with critical input from K.J., B.J., and M.M.

Corresponding author

Correspondence to Hua Cheng.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Cheng, H., Johansen, K., Jin, B. et al. Seasonal movement behavior of domestic goats in response to environmental variability and time of day using Hidden Markov Models. Mov Ecol 13, 28 (2025). https://doi.org/10.1186/s40462-025-00557-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s40462-025-00557-2

Keywords

Movement Ecology

ISSN: 2051-3933