PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of peerjLatest ArticlesFor AuthorsEditorial BoardPeerJPeerJ
 
PeerJ. 2016; 4: e2152.
Published online 2016 June 21. doi:  10.7717/peerj.2152
PMCID: PMC4924134

Towards evenly distributed grazing patterns: including social context in sheep management strategies

Academic Editor: Wendy Rauw

Abstract

Background. A large proportion of natural grasslands around the world is exposed to overgrazing resulting in land degradation and biodiversity loss. Although there is an increasing effort in the promotion of sustainable livestock management, rangeland degradation still occurs because animals’ foraging behaviour is highly selective at different spatial scales. The assessment of the ecological mechanisms modulating the spatial distribution of grazing and how to control it has critical implications for long term conservation of resources and the sustainability of livestock production. Considering the relevance of social interactions on animals’ space use patterns, our aim was to explore the potential effects of including animals’ social context into management strategies using domestic sheep grazing in rangelands as case study.

Methods. We used GPS data from 19 Merino sheep (approximately 10% of the flock) grazing on three different paddocks (with sizes from 80 to 1000 Ha) during a year, to estimate resource selection functions of sheep grazing in flocks of different levels of heterogeneity. We assessed the effects of sheep class (i.e., ewes, wethers, and hoggets), age, body condition and time since release on habitat selection patterns.

Results. We found that social rank was reflected on sheep habitat use, where dominant individuals (i.e., reproductive females) used more intensively the most preferred areas and low-ranked (i.e., yearlings) used less preferred areas. Our results showed that when sheep grazed on more heterogeneous flocks, grazing patterns were more evenly distributed at all the paddocks considered in this study. On the other hand, when high-ranked individuals were removed from the flock, low-ranked sheep shifted their selection patterns by increasing the use of the most preferred areas and strongly avoided to use less preferred sites (i.e., a highly selective grazing behaviour).

Discussion. Although homogenization and segregation of flocks by classes are common practices to increase flock productivity, we are proposing an alternative that employs behavioural interactions in heterogeneous flocks to generate more evenly distributed grazing patterns. This practice can be combined with other practices such as rotational grazing and guardian dogs (to decrease mortality levels that may be generated by sheep grazing on more risky habitats). This does not imply any modifications of livestock stocking rates and densities or any additional investments for labour and materials. Considering livestock behaviour is critical for the design of sustainable management practices that balance landscape conservation and livestock productivity.

Keywords: Selective grazing, Resource selection, Rangelands, Heterogeneous mixed flocks, Social rank, Sustainable management strategies, Overgrazing

Introduction

A large fraction of natural grasslands around the world is exposed to overgrazing and high stocking rates, leading to land degradation and biodiversity loss of an environment that besides holding key natural resources sustains socio-economically a vast proportion of human population (MA, 2006; FAO, 2009; Sayre et al., 2013). Even though there is an increasing effort in the promotion of sustainable range management (Morghan, Sheley & Svejcar, 2006; Briske et al., 2008; Boyd & Svejcar, 2009), rangeland degradation still occurs because grazing behaviour is highly selective at different spatial scales (Senft et al., 1987; Bailey et al., 1996). These heterogeneous patterns of space use often produce overgrazing at the most preferred areas while other areas remain under-grazed (Bailey, Dumont & WallisDevries, 1998; Pringle & Landsberg, 2004). Although herbivores’ movement decisions are related to the distribution of resources (e.g., Fortin, Fryxell & Pilote, 2002; Fryxell, Wilmshurst & Sinclair, 2004), for animals living in groups, social context can also be an important factor influencing their movement and patterns of space use (Scott, Provenza & Banner, 1995; Scott, Banner & Provenza, 1996; Couzin et al., 2005; Haydon et al., 2008). As a consequence, the movement patterns of gregarious herbivores can result in socially-generated disturbances that are somehow independent of resource distributions (Bailey et al., 2001; Seabloom & Reichman, 2001; Bailey, 2004).

The social environment of group-living herbivores has several benefits on individual fitness that are related to knowledge transmission from conspecifics’ experiences (Scott, Provenza & Banner, 1995; e.g., Scott, Banner & Provenza, 1996) and minimization of predation risk (Pinchak et al., 1991; Plumb & Dodd, 1993; Etzenhouser et al., 1998; Fortin et al., 2003; Fortin et al., 2005; Frair et al., 2005; Rubio et al., 2008; Fortin et al., 2009). Nevertheless, when resources are scarce and heterogeneously distributed across the landscape, animals have to balance the benefits of group living with the competition for resources (Childress & Lung, 2003; Veiberg et al., 2004; Fortin et al., 2009). This trade-off may contribute to the development of hierarchical structures, in which animals are part of a social rank that could determine their grazing and movement patterns (Thouless, 1990; Veiberg et al., 2004). Regardless of which factors determine an animal’s position in a social rank, once the hierarchy is established, higher-rank individuals have priority access to preferred areas while lower-rank individuals use less preferred areas in the landscape which otherwise would be underused (Petocz, 1973; Bøe, Berg & Andersen, 2006).

Our general aim was to explore the potential effects of including the animals’ social context into management strategies using domestic sheep (Ovis aries, Linnaeus 1758) grazing in Patagonian rangelands as case study. In particular, we wanted to explore which factors influenced the development of a social rank when sheep graze in heterogeneous flocks and if this social structure favours the development of more evenly distributed grazing patterns. Considering that animals could show different movement behaviours according to their age (e.g., Favre, Martin & Festa-Bianchet, 2008), sex (e.g., Michelena et al., 2005), nutritional condition and their energy requirements (e.g., Vervaecke, Roden & De Vries, 2005) we assessed their influence on sheep selection patterns.

Material & Methods

Study area

Empirical data on sheep movement, body condition, and landscape traits were collected at Fortín Chacabuco Ranch (Supplemental Information 1). This farm is located in Los Lagos department (NW corner at 40°57′04.58″S, 71°08′44.14″W; SW corner at 40°59′45.27″S, 71°12′10.21″W; NE corner at 40°59′17.54″S, 71°05′52.75″W; SE corner at 41°03′09.17″S, 71°08′16.45″W) and belongs to the Pre-mountain range ecological area of Argentinean Patagonia (Cabrera, 1976). The landscape is characterized by mountain chains and hills, crossed by several rivers and water streams. The weather is cold, with mean annual temperatures of 10 °C and annual precipitations that range from 300 to 700 mm, concentrated during the cold season (i.e., May–August). Vegetation corresponds to the Sub-Andean district (León et al., 1998), composed by grasslands (dominated by Stipa speciosa var major in lower lands and by Festuca pallescens in upper lands), and wetlands of different extensions. Smaller wetlands are frequently associated to native and riparian forests.This ranch is characterized by high heterogeneity in sheep grazing patterns, with some areas intensively used and with degradation signs, and others under-used with high levels of litter accumulation from un-grazed senescent plants that decreases their productivity (Paramidani, Doffigny & Codesal, 2014).

Data collection and spatial data processing

Landscape data

In order to characterize the landscape where sheep grazed, we mapped the vegetation units of each paddock were livestock was kept during the sampling period. First, we identified and defined polygons of landscape units that we were able to classify using Google Earth imagery (Digital Globe Quickbird. Resolution: 15 m per pixel. 10-16-2013. Accessed: September 2014: Supplemental Information 1, Fig. 1); then we corroborated and defined each polygon’s limits by matching them with an unsupervised classification from a Landsat TM (2011) satellite image using ArcGIS (ESRI 2011. ArcGIS Desktop10.1. Redlands, Environmental Systems Research Institute). This classification was then validated through vegetation sampling that consisted in placing transects inside each polygon, in which we located a 1 m2 plot every 100 m. These points were selected previous to fieldwork on a Google Earth file containing the polygons of vegetation units. The number of transects depended on the size of each polygon (i.e., in wider polygons we placed two or more transects), and the length depended on the polygon’s length. In each plot we registered every plant species present, their percentage of cover, and the percentage of cover of litter and bare ground. The species were recognized in the field, and we took samples only from those specimens that could not be identified in the field (permit number 1339). By using a vegetation guides for this region (e.g., Bonvissuto & Somlo, 1995) we were able to identify eight main landscape units: central wetland, peripheral wetland, grassland, shrubland-grassland, native forest (dominated by Maytenus boaria, Austrocedrus chilensis, and Nothofagus spp.), riparian forest (associated with streams and water courses), high-lands (with rocky peaks which provide shadow during summer) and low production areas (which are eroded areas, with high proportions of bare ground and low levels of forage). Additionally, during those vegetation surveys we registered GPS locations of sheep carcasses and faeces from pumas and foxes to identify which vegetation units present higher predation risk (see Supplemental Information 1, Table S1.4).

For this study, we considered three paddocks, called Repunte-Bajo (RB), Frison-Guanaco (FG)and Side (S), which have sizes of 79, 944 and 146 Ha, respectively. This range of sizes among paddocks is representative for the paddocks used commonly in Patagonian rangelands for sheep production (see examples in Irisarri et al., 2012; Ormaechea & Peri, 2015). Information about the surface of each landscape unit, the amount of forage species, bare ground and predation risk levels of each paddock can be found in Supplemental Information 1. Central and peripheral wetlands were considered the most preferred landscape resources because they contain higher forage production. Native forests have relatively high quantities of forage and riparian forests are areas associated with semi-permanent water courses, but are closed areas with lower probabilities of predator detection (Risenhoover & Bailey, 1985; Kie, 1999) and high numbers of carcasses and predators’ faeces (Supplemental Information 1, Table S1.4). Grasslands and shrubland-grasslands are areas with moderate to low quantities of forage and predation risk signs. High-lands are presented only at Frison-Guanaco and are characterized by protruding rock walls that sheep use as shaded areas during summer and as wind refuges during winter (A di Virgilio, pers. obs., 2014). Finally, low production areas are sites of moderate-low elevation with the lowest quantities of forage observed in the landscape.

Sheep data

We selected 19 sheep from a flock of approximately 200 individuals, composed of 38% of ewes (i.e., reproductive females), 28% of hoggets (i.e., female and male yearlings), 31% of wethers (i.e., castrated males), and 3% of rams (reproductive males). In order to have a representative sample of these sheep classes (rams were kept apart from the flock during the sampling period), we selected 6 ewes, 7 hoggets, and 6 wethers. Sheep were equipped with GPS devices attached to collars (CatLog-B, Perthold Engineering, www.perthold.de; USA). Collared animals were marked for identification with a unique ID number and colour according to sheep class. The sampled animals were always the same individuals for all the sampling periods, and sheep wore the collars almost continuously for the whole duration of the experiment. The devices were programmed to acquire locations every 5 min, from September 2014 to September 2015.We used a Body Condition Score (BCS, Jefferies, 1961; Giraudo, 2010), as an indirect method to assess sheep fat reserves and estimated sheep’ age by their number of permanent teeth. Initial BCS of sheep and age were measured at the beginning of the sampling period. BCS ranged from 2.25 to 4 and ages ranged from 6 months to more than 4 years old (i.e., from 0 permanent teeth to 8 permanent teeth). We also measured BSC of all collared individuals every time the flock was translocated into the paddocks. We considered heterogeneous flocks (hereafter 3-class mixed flocks) as those composed of all sheep classes (i.e., ewes, hoggets and wethers) and less heterogeneous flocks (hereafter 2-class mixed flocks) as flocks with two sheep classes (e.g., ewes and hoggets, or wethers and hoggets). Because our research was conducted on non-regulated animals we did not required any ethical approval.

To assess if patterns of space use by sheep are more evenly distributed when animals graze in 3-class mixed flocks and to detect which individuals have priority access to most preferred areas, we selected movement data from the periods when all sheep classes grazed together in the same paddock (see Table 1). During our sampling period, the flock was divided according to sheep classes (see Borrelli et al., 1997; Golluscio, Deregibus & Paruelo, 1998) during lambing (i.e., when ewes give birth to lambs, approximately between October–November) and during weaning (i.e., when wethers and lambs were separated, approximately between January–February). To evaluate if sheep increase their selectivity under less heterogeneous mixed flocks, and if subordinates change their landscape use patterns when high-ranked individuals are removed from the flock, we used the following data sets: (i) Data from wethers and hoggets grazing when ewes were extracted from the flock during the lambing period (RB2C-mixed, see Table 1); and (ii) data from ewes and hoggets grazing when wethers were extracted from the flock during weaning (S2C-mixed, see Table 1). Both databases included GPS locations of individuals that experienced both flock types in the same paddock.

Table 1
Description of datasets used for each analysis, including the paddock considered, the type of flock, the number of individuals included of each sheep class, and the period and its length in days.

In all cases, we used only diurnal locations when sheep are mainly grazing (Dudzinsky & Arnold, 1978). As a result, the total number of GPS locations to be included in the data analysis for Repunte-Bajo, Frison-Guanaco and Side paddocks were: 147,905, 208,275 and 28,976 respectively. The number of random points generated for statistical analyses (see below) was double of GPS locations for the three paddock, which represented a good balance between the quality of the availability estimation and computational demands (Northrup et al., 2013). The spatial data analysis was performed with R software (R Core Team, 2015). All information about spatial data analysis can be found in Supplemental Information 2.

Data analysis

To assess sheep resource selection patterns at landscape scale, we focused our analyses on availability vs. use of resources (e.g., Johnson et al., 2006; Beyer et al., 2010), where different resources are represented by the different vegetation units present in the landscape. This requires measuring the use of resources relative to their availability: a positive ratio indicates that animals are selecting a particular resource, a negative ratio suggests avoidance, and the magnitude of the ratio indicates the degree of the selection/avoidance. In particular, we estimated Resource Selection Functions (RSF, Manly et al., 2002), by fitting logistic regressions, assuming a Bernoulli distribution for the response variable and a logit link function. We used a Bayesian approach to fit a hierarchical logistic regression for each paddock allowing for individual differences in sheep responses and considering the lack of independence between the consecutive GPS locations (see Supplemental Information 3). The dependent variable for the RSF was the GPS locations (ones) combined with simulated random locations (zeroes) across the total surface of each paddock that approximate the availability of resources. The resources considered were landscape vegetation units: low production areas, high-lands, central wetlands, peripheral wetlands, grasslands, shrubland-grasslands, native forestsand riparian forests.

To assess the effect of social structure on the use of landscape resources, we fitted a hierarchical logistic model using time since release (as the number of days since the release in the paddock), sheep class, age and body condition score as co-variables. Continuous variables were centred and standardized. The model structure can be found in Supplemental Information 3. To evaluate if sub-dominant individuals shift their space use patterns when dominant sheep are removed from the flock, we fitted a hierarchical logistic model using as independent variable time since release, sheep type and flock type (i.e., 3-class vs. 2-class); see the model structure in Supplemental Information 3. All analyses were performed using R (R Core Team, 2015) and WinBUGS (Spiegelhalter et al., 1996). Details on parameters’ and hyper-parameters’ priors and the number of iterations for each Monte Carlo Markov Chain (MCMC) can be found in Supplemental Information 3. We checked for convergence and autocorrelation of the posterior distributions and estimated the 95% Highest Posterior Density (HPD) intervals for all parameters. To assess if resource selection patterns differed among sheep classes, we calculated the proportion of posterior distribution overlapping of the parameters for each landscape resource between (i) ewes-hoggets, (ii) ewes-wethers, and (iii) hoggets-wethers in each paddock. To evaluate the differences of resource selection under both types of flocks, we calculated the proportion of posterior distribution overlapping between the estimates of 3-class and 2-class mixed flocks for each resource in each paddock.

Results

We found that when sheep were maintained in more heterogeneous flocks they selected almost all vegetation units with similar intensities and strongly avoided low production areas or riparian forests (Fig. 1). Furthermore, we found that for the areas with higher forage production (i.e., central and peripheral wetlands) and for less preferred areas (i.e., low production sites and riparian forests), the patterns of resource selection differed among sheep classes. This resulted in ewes selecting more intensively the most productive areas and avoiding with more strength the less preferred areas; while hoggets and wethers selected landscape resources similarly (Fig. 1, Supplemental Information 4: Table S4.1, Supplemental Information 5). We did not detect effects of time since release, age or BCS on sheep selection patterns (RSF coefficients were close to zero and HPD-intervals included zero; see Supplemental Information 4, Table S4.2). These patterns of more evenly distributed use and differences among sheep classes were observed at the three paddocks evaluated (Figs. 1 and and4,4, Supplemental Information 5).

Figure 1
Resource Selection Functions mean estimates and 95% Highest Posterior Density intervals for the logistic model fitted to 3-class mixed flock for ewes (orange circles), hoggets (blue triangles) and wethers (red squares) grazing at the three paddocks.
Figure 4
Diurnal movement patterns of sheep under 3-category and 2-category mixed flocks at Repunte-Bajo paddock.

When sheep were kept in less heterogeneous flocks, we noticed that the outcome of the resource selection patterns depended on which sheep class was removed (Figs. 2 and and3).3). When wethers were removed, the selection patterns resulted similar to those in 3-class mixed flocks (Figs. 3A and and3B),3B), and neither ewes nor hoggets showed major changes in their landscape use. Nevertheless, when ewes were removed from the flock, it resulted in a more unevenly distributed space use, with higher selection of most preferred areas (i.e., central and peripheral wetlands) and higher avoidance of the rest of the landscape units (especially for low production areas and both types of forests). Moreover, hoggets were the sheep class that showed a marked switch in their resource selection patterns in comparison to wethers (Figs. 2A, ,2B2B and and4;4; Supplemental Information 5).

Figure 2
Resource Selection Function mean estimates and 95% Highest Posterior Density intervals for the logistic model fitted to data from hoggets (blue triangles) and wethers (red squares) grazing in 3-class (A) and 2-class (B) mixed flocks after ewes were removed ...
Figure 3
Resource Selection Function mean estimates and 95% Highest Posterior Density intervals for the logistic model fitted to data from hoggets (blue triangles) and ewes (orange circles) when grazing in a 3-class mixed flock (A) and in a 2-class mixed flock ...

Discussion

This work provides information about a possible mechanism underlying the generation of more evenly distributed grazing patterns, which is related to ecological and behavioural aspects of livestock and can be employed as a tool for the development of sustainable livestock management strategies. Our outcomes showed that maintaining flocks of sheep from different classes resulted in a more homogeneous distribution of grazing, independently of the paddock in which sheep grazed. On the other hand, we observed that this evenly distributed grazing patterns could be easily overridden by common management practices that tend to maintain more uniform groups to increase flock productivity (Arnold & Pahl, 1974; Stenseth, 1983; Lynch et al., 1989; Borrelli et al., 1997; Borrelli & Oliva, 2001), such as flock segregation according to nutritional requirements or culling, among others. Finally, we noticed a shift from an evenly distributed grazing pattern toward a more selective one only when dominant individuals were removed from the flock.

Although there is evidence that food preferences, weather, forage distribution and predators influence herbivores grazing patterns (Scott, Provenza & Banner, 1995; Prache, Gordon & Rook, 1998), we propose that social interactions are also important factors influencing landscape use and ultimately population dynamics (Morales et al., 2010). This is particularly relevant for gregarious animals like livestock and we believe that it should be considered into management strategies in addition to the rest of the factors. Commonly, social interactions among group members result in dominance structures or social ranks (Lynch, Hinch & Adams, 1992; Erhard et al., 2004) that are reflected in their spatial distribution when grazing (see Lynch, Hinch & Adams, 1992; Sibbald, Shellard & Smart, 2000). We observed that ewes selected with more intensity the most preferred areas in comparison to hoggets and wethers that showed higher selectivity towards less preferred areas when sheep were maintained in more heterogeneous flocks. This pattern could be indicating the development of a well established social rank, in which ewes are dominant and have preferential access to resources, as was previously observed in other studies with sheep (Lynch, Wood-Gush & Davies, 1985; Eccles & Shackleton, 1986; Festa-Bianchet, 1991; Færevik, Andersen & Bøe, 2005), and hoggets and wethers occupied lower positions in the social rank.

Generally, to avoid aggressive interactions, subordinate individuals use the resources after dominants and are forced to feed on less preferred areas (Eccles & Shackleton, 1986; Prache, Gordon & Rook, 1998) while high-ranked animals are generally unaffected by the proximity of subordinates (e.g., Bennett, Finch & Holmes, 1985). This would mean that low-ranked individuals are more limited on their choices when high-ranked individuals are present. Under this context, we would expect that, when these dominant individuals are removed from the flock, lower ranked individuals respond by showing a switch in their selection patterns toward the most preferred areas. Although some studies focused on dominance levels of wethers (Stolba et al., 1990; Keeling, 2001) there is not a conclusive statement about their position in social ranks. In this work, we did not observed any switch in the landscape use patterns of hoggets and ewes when wethers were extracted from the flock. Moreover, when ewes were removed, only hoggets showed a significant shift in their selectivity, strongly selecting most preferred areas and avoiding less preferred areas. These results imply that reproductive females are upper positioned in the social rank, maybe due to their higher energetic demands (Keeling, 2001); wethers are more or less independent of this social rank maybe associated to their relatively constant energetic demands generated by castration, and immature individuals are at the lowest levels of this ranking.

Although there is evidence that age and body condition can be factors influencing animals’ resource selection and their social interactions (Festa-Bianchet, 1991; Côté, 2000; Veiberg et al., 2004; Vervaecke, Roden & De Vries, 2005), we did not find evidence about the potential effects of these factors on the landscape use of our sheep. This lack of evidence may be due to the fact that we did not have enough variation in age or body condition score among sampled individuals, or because we would need a bigger sample to assess those effects. Studies that found an effect of age and body condition on sheep social rank were conducted on homogeneous flocks composed mostly of one sheep class and under pen or small paddock conditions instead of rangelands. Nevertheless, the lack of effect of body condition could have important implications on livestock productivity, because if sub-dominant individuals would be severely food restricted under more heterogeneous flocks, we would expect a strong association between body condition score and the use of less preferred areas and it was not the case in our results.

Unevenly distributed grazing, besides generating degradation of over-used areas, decreases the productivity of under-grazed parts of the landscape mainly because of the accumulation of senescent material (Belsky, 1986). Thus, if grazing distribution could be controlled, we could consider it as a potential tool for natural grasslands conservation and recovery (see Weber et al., 1998; Adler, Raff & Lauenroth, 2001). Even though the stocking rates were reduced when sheep shifted from a 3-class to a 2-class mixed flock, if our findings would have been related to the number of animals in a flock, we would expect to find similar results if we would have removed ewes as we would have removed wethers, because all sheep classes represented similar proportions of the flock. Nevertheless, we observed a marked switch in grazing patterns, only after ewes were removed. For this reason, we propose that social interactions among flock members is the underlying mechanism that generates more evenly distributed grazing patterns and the disruption of this social rank generates highly selective grazing patterns. Considering that livestock resource selection patterns affect not only animals’ nutritional levels (Stricklin, Zhou & Gonyou, 1995) but also landscape dynamics (Adler, Raff & Lauenroth, 2001), our findings have important management implications and provide information that could be used to decrease the negative impacts of sheep grazing and maintain the long-term productivity of natural grasslands.

Conclusions

Several management tools have been proposed to homogenize livestock grazing patterns, such as water developments (Porath et al., 2002), fencing (Bailey, Walker & Rittenhouse, 1990), strategic supplement placement (Bailey et al., 2001), herding (Rhodes & Marlow, 1997), and stocking density increasing (Savory, 1983). However, most of these tools require relatively high investments (e.g., Bailey, 2004) or resulted controversial (Briske et al., 2013; Cibils et al., 2014; Teague, 2014). We are aware that the adoption of a novel management strategy strongly depends on ecological and economical constraints, and that social and cultural factors play a key role in this process (Borrelli et al., 1997; Golluscio, Deregibus & Paruelo, 1998). In this work, we are suggesting to land managers and livestock producers an alternative for flock management that employs animals’ behavioural interactions without the need for reducing livestock rates, modifying stocking densities or additional investments in labour and materials. Moreover, managing more heterogeneous mixed flocks can be combined with other strategies such as rotational grazing and the use of guardian dogs (to decrease the mortality levels that may be generated by sheep grazing on more risky habitats) to increase livestock productivity and landscape conservation.

Supplemental Information

10.7717/peerj.2152/supp-1

Supplemental Information 1

Landscape description:
10.7717/peerj.2152/supp-2

Supplemental Information 2

Spatial data processing:
10.7717/peerj.2152/supp-3

Supplemental Information 3

Model structure:
10.7717/peerj.2152/supp-4

Supplemental Information 4

Model output:
10.7717/peerj.2152/supp-5

Supplemental Information 5

Movement paths and GPS locations:

Acknowledgments

We thank the Nature Conservancy and S Gary for allowing us to conduct field work at Fortín Chacabuco ranch; to F Montenegro and N Rodríguez for all the assistance and invaluable help and to G Iglesias and J Beltrán for their support. Also, we thank ML Villar for all her help and knowledge shared about sheep and livestock management; and A Sáez and ML Villar for all their comments that improved the previous versions of this manuscript.

Funding Statement

This study was funded by PIP 112 201101 00058 from CONICET (to JMM). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

The following grant information was disclosed by the authors:

CONICET PIP 112 201101 00058.

Additional Information and Declarations

Competing Interests

The authors declare there are no competing interests.

Author Contributions

Agustina di Virgilio conceived and designed the experiments, performed the experiments, analyzed the data, wrote the paper, prepared figures and/or tables, reviewed drafts of the paper.

Juan Manuel Morales conceived and designed the experiments, contributed reagents/materials/analysis tools, reviewed drafts of the paper.

Animal Ethics

The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers):

Our research was conducted on non-regulated animals. The legislation of the site where we conducted our work did not required any ethical approval. Moreover, there is not a specific institution that regulates the manipulation of domestic animals. Note also, that the manipulation of the sheep included fitting the GPS collars to them and checking their body condition. There was no use of drugs or phlebotomy involved in the project. Nevertheless, we have an agreement letter (in Spanish and a brief summary translated in English) from the ranch owner and the administrator, allowing us to conduct our project with their sheep.

Field Study Permissions

The following information was supplied relating to field study approvals (i.e., approving body and any reference numbers):

The vegetation data used in this work is shared with another project and researchers, for which we have applied for permission to National Parks Administration (Permit number 1399). The permit (in Spanish) and an English summary is attached in this submission.

Data Availability

The following information was supplied regarding data availability:

The data can be freely downloaded by following these steps: (1) Clicking on the projects map (https://www.movebank.org/panel_embedded_movebank_webapp) and searching the project (by key words such as Ovis aries Patagonia or by the entire project name: Ovis aries di Virgilio Patagonia); (2) selecting the study (by clicking on the green point representing our study); (3) opening the study page (by clicking on “Open in study page”); and (4) clicking on “Download” tab. After reading the study license terms, the data can be freely downloaded in different file formats.

References

Adler, Raff & Lauenroth (2001) Adler P, Raff D, Lauenroth W. The effect of grazing on the spatial heterogeneity of vegetation. Oecologia. 2001;128:465–479. doi: 10.1007/s004420100737. [Cross Ref]
Arnold & Pahl (1974) Arnold G, Pahl P. Some aspects of social behaviour in domestic sheep. Animal Behaviour. 1974;22:592–600. doi: 10.1016/S0003-3472(74)80004-7. [Cross Ref]
Bailey (2004) Bailey D. Management strategies for optimal grazing distribution and use of arid rangelands. Journal of Animal Science. 2004;82:E147–E153. [PubMed]
Bailey, Dumont & WallisDevries (1998) Bailey DW, Dumont B, WallisDeVries MF. Utilization of heterogeneous grasslands by domestic herbivores: theory to management. Annales de Zootechnie. 1998;47:321–333.
Bailey et al. (1996) Bailey DW, Gross JE, Laca EA, Rittenhouse LR, Coughenour MB, Swift DM, Sims PL. Mechanisms that result in large herbivore grazing distribution patterns. Journal of Range Management. 1996;49:386–400. doi: 10.2307/4002919. [Cross Ref]
Bailey et al. (2001) Bailey D, Kress D, Anderson D, Boss D, Miller E. Relationship between terrain use and performance of beef cows grazing foothill rangeland. Journal of Animal Science. 2001;79:1883–1891. [PubMed]
Bailey, Walker & Rittenhouse (1990) Bailey D, Walker J, Rittenhouse L. Sequential analysis of cattle location: day-to-day movement patterns. Applied Animal Behaviour Science. 1990;25:137–148.
Belsky (1986) Belsky A. Does herbivory benefit plants? a review of the evidence. American Naturalist. 1986:870–892.
Bennett, Finch & Holmes (1985) Bennett I, Finch VA, Holmes C. Time spent in shade and its relationship with physiological factors of thermoregulation in three breeds of cattle. Applied Animal Behaviour Science. 1985;13:227–236. doi: 10.1016/0168-1591(85)90046-2. [Cross Ref]
Beyer et al. (2010) Beyer HL, Haydon DT, Morales JM, Frair JL, Hebblewhite M, Mitchell M, Matthiopoulos J. The interpretation of habitat preference metrics under use–availability designs. Philosophical Transactions of the Royal Society of London B: Biological Sciences. 2010;365:2245–2254. doi: 10.1098/rstb.2010.0083. [PMC free article] [PubMed] [Cross Ref]
Bonvissuto & Somlo (1995) Bonvissuto G, Somlo R. Guías de condición para los principales tipos de campo de dos áreas ecológicas de Patagonia–Argentina. I. Precordillera. II. Sierras y Mesetas. En Seminario-taller: Producción, nutrición y utilización de pastizales. Grupo regional Patagónico de Ecosistemas de pastoreo. In: Somlo R, Becker G, editors. FAO.-UNESCO/MAB.-INTA. Rome: FAO; 1995.
Borrelli & Oliva (2001) Borrelli P, Oliva G. Ganadería ovinasustentable en la Patagonia austral: tecnología de manejo extensivo. Rio Gallegos: INTA-UNPA-CAP; 2001. 270 pp.
Borrelli et al. (1997) Borrelli P, Oliva G, Williams M, Gonzalez L, Rial P, Montes L. Santa Cruz yTierra del Fuego. Proyecto Prodeser. Buenos Aires: INTA-GTZ; 1997. Sistema regional de soporte de decisiones. 134 p.
Boyd & Svejcar (2009) Boyd CS, Svejcar TJ. Managing complex problems in rangeland ecosystems. Rangeland ecology and management. 2009;62:491–499. doi: 10.2111/08-194.1. [Cross Ref]
Briske et al. (2013) Briske DD, Bestelmeyer BT, Brown JR, Fuhlendorf SD, Polley HW. The savory method can not green deserts or reverse climate change: a response to the allan savory ted video. Rangelands. 2013;35:72–74. doi: 10.2111/RANGELANDS-D-13-00044.1. [Cross Ref]
Briske et al. (2008) Briske D, Bestelmeyer B, Stringham T, Shaver P. Recommendations for development of resilience-based state-and-transition models. Rangeland Ecology and Management. 2008;61:359–367. doi: 10.2111/07-051.1. [Cross Ref]
Cabrera (1976) Cabrera AL. Regiones fitogeográficas argentinas. Second edition Enciclopedia Argentina de Agricultura y Jardinería; Buenos Aires: 1976.
Childress & Lung (2003) Childress M, Lung M. Predation risk, gender, and the group size effect: does elk vigilance depend upon the behavior of conspecifics? Animal Behaviour. 2003;66:389–398. doi: 10.1006/anbe.2003.2217. [Cross Ref]
Cibils et al. (2014) Cibils AF, Lira Fernández RJ, Oliva GE, Escobar JM. Is holistic management really saving patagonian rangelands from degradation? a response to teague. Rangelands. 2014;36:26–27.
Côté (2000) Côté SD. Dominance hierarchies in female mountain goats: stability, aggressiveness and determinants of rank. Behaviour. 2000;137:1541–1566. doi: 10.1163/156853900502718. [Cross Ref]
Couzin et al. (2005) Couzin ID, Krause J, Franks NR, Levin SA. Effective leadership and decision-making in animal groups on the move. Nature. 2005;433:513–516. doi: 10.1038/nature03236. [PubMed] [Cross Ref]
Dudzinsky & Arnold (1978) Dudzinsky M, Arnold G. Ethology of free-rangingdomestic animals. Elsevier Scientific Publishers; New York: 1978.
Bøe, Berg & Andersen (2006) Bøe KE, Berg S, Andersen IL. Resting behaviour and displacements in ewes—effects of reduced lying space and pen shape. Applied Animal Behaviour Science. 2006;98:249–259. doi: 10.1016/j.applanim.2005.10.001. [Cross Ref]
Eccles & Shackleton (1986) Eccles T, Shackleton D. Correlates and consequences of social status in female bighorn sheep. Animal Behaviour. 1986;34:1392–1401. doi: 10.1016/S0003-3472(86)80210-X. [Cross Ref]
Erhard et al. (2004) Erhard HW, Fàbrega E, Stanworth G, Elston DA. Assessing dominance in sheep in a competitive situation: level of motivation and test duration. Applied Animal Behaviour Science. 2004;85:277–292. doi: 10.1016/j.applanim.2003.09.013. [Cross Ref]
Etzenhouser et al. (1998) Etzenhouser MJ, Owens MK, Spalinger DE, Murden SB. Foraging behavior of browsing ruminants in a heterogeneous landscape. Landscape Ecology. 1998;13:55–64. doi: 10.1023/A:1007947405749. [Cross Ref]
Færevik, Andersen & Bøe (2005) Færevik G, Andersen IL, Bøe KE. Preferences of sheep for different types of pen flooring. Applied Animal Behaviour Science. 2005;90:265–276. doi: 10.1016/j.applanim.2004.08.010. [Cross Ref]
FAO (2009) FAO (Food and Agriculture Organization of the United Nations) Review of evidence on drylands pastoral systems and climate change. In: Neely C, Bunning S, Wilkes A, editors. Land and water discussion paper 8, Land Tenure and Management Unit (NRLA) FAO, Land Water division; Rome: 2009.
Favre, Martin & Festa-Bianchet (2008) Favre M, Martin JGA, Festa-Bianchet M. Determinantsand life-history consequences of social dominance in bighorn ewes. Animal Behaviour. 2008;76:1373–1380. doi: 10.1016/j.anbehav.2008.07.003. [Cross Ref]
Festa-Bianchet (1991) Festa-Bianchet M. The social system of bighorn sheep: grouping patterns, kinship and female dominance rank. Animal Behaviour. 1991;42:71–82. doi: 10.1016/S0003-3472(05)80607-4. [Cross Ref]
Fortin et al. (2005) Fortin D, Beyer HL, Boyce MS, Smith DW, Duchesne T, Mao JS. Wolves influence elk movements: behavior shapes a trophic cascade in Yellowstone National Park. Ecology. 2005;86:1320–1330. doi: 10.1890/04-0953. [Cross Ref]
Fortin et al. (2009) Fortin D, Fortin M-E, Beyer HL, Duchesne T, Courant S, Dancose K. Group-size-mediated habitat selection and group fusion-fission dynamics of bison under predation risk. Ecology. 2009;90:2480–2490. doi: 10.1890/08-0345.1. [PubMed] [Cross Ref]
Fortin et al. (2003) Fortin D, Fryxell JM, O’brodovich L, Frandsen D. Foraging ecology of bison at the landscape and plant community levels: the applicability of energy maximization principles. Oecologia. 2003;134:219–227. doi: 10.1007/s00442-002-1112-4. [PubMed] [Cross Ref]
Fortin, Fryxell & Pilote (2002) Fortin D, Fryxell JM, Pilote R. The temporal scale of foraging decisions in bison. Ecology. 2002;83:970–982. doi: 10.1890/0012-9658(2002)083[0970:TTSOFD]2.0.CO;2. [Cross Ref]
Frair et al. (2005) Frair JL, Merrill EH, Visscher DR, Fortin D, Beyer HL, Morales JM. Scales of movement by elk (Cervus elaphus) in response to heterogeneity in forage resources and predation risk. Landscape Ecology. 2005;20:273–287. doi: 10.1007/s10980-005-2075-8. [Cross Ref]
Fryxell, Wilmshurst & Sinclair (2004) Fryxell JM, Wilmshurst JF, Sinclair AR. Predictivemodels of movement by serengeti grazers. Ecology. 2004;85:2429–2435. doi: 10.1890/04-0147. [Cross Ref]
Giraudo (2010) Giraudo CG. El empleo de la condición corporal como indicador del estado nutricional de los ovinos. Presencia. 2010;54:32–35.
Golluscio, Deregibus & Paruelo (1998) Golluscio R, Deregibus V, Paruelo JM. Sustainabilityand range management in the patagonian steppes. Ecología Austral. 1998;8:265–284.
Haydon et al. (2008) Haydon DT, Morales JM, Yott A, Jenkins DA, Rosatte R, Fryxell JM. Socially informed random walks: incorporating group dynamics into models of population spread and growth. Proceedings of the Royal Society of London B: Biological Sciences. 2008;275:1101–1109. doi: 10.1098/rspb.2007.1688. [PMC free article] [PubMed] [Cross Ref]
Irisarri et al. (2012) Irisarri JGN, Oesterheld M, Paruelo JM, Texeira MA. Patterns and controls of above-ground net primary production in meadows of patagonia. A remote sensing approach. Journal of Vegetation Science. 2012;23:114–126. doi: 10.1111/j.1654-1103.2011.01326.x. [Cross Ref]
Jefferies (1961) Jefferies B. Body condition scoring and its use in management. Tasmanian Journal of Agriculture. 1961;32:19–21.
Johnson et al. (2006) Johnson CJ, Nielsen SE, Merril EH, McDonald TL, Boyce MS. Resource selection functions based on use–availability data: theoretical motivation and evaluation methods. Journal of Wildlife Management. 2006;70:347–357. doi: 10.2193/0022-541X(2006)70[347:RSFBOU]2.0.CO;2. [Cross Ref]
Keeling (2001) Keeling J. Social behavior in farm animals. CABI Publishing; Wallingford: 2001.
Kie (1999) Kie JG. Optimal foraging and risk of predation: effects on behavior and social structure in ungulates. Journal of Mammalogy. 1999;80:1114–1129.
León et al. (1998) León RJ, Bran D, Collantes M, Paruelo JM, Soriano A. Grandes unidades de vegetación de la patagonia extra andina. Ecología Austral. 1998;8:125–144.
Lynch, Hinch & Adams (1992) Lynch JJ, Hinch G, Adams D. The behaviour of sheep: biological principles and implications for production. East Melburne: CSIRO publications; 1992. 237 pp.
Lynch et al. (1989) Lynch J, Hinch G, Bouissou M, Elwin R, Green G, Davies H. Social organization in young merino and merino×border leicester ewes. Applied Animal Behaviour Science. 1989;22:49–63. doi: 10.1016/0168-1591(89)90079-8. [Cross Ref]
Lynch, Wood-Gush & Davies (1985) Lynch J, Wood-Gush D, Davies H. Aggression and nearest neighbours in a flock of scottish blackface ewes. Biology of Behaviour. 1985;10:215–225.
MA (2006) MA (Millennium Ecosystem Assessment) Millennium ecosystem assessment synthesis reports. Island Press; Washington: 2006. Available at http://maweb.org/en/Synthesis.aspx.
Manly et al. (2002) Manly BFJ, McDonald LL, Thomas DL, McDonald TL, Erickson WP, editors. Resource selection by animals: statistical analysis and design for field studies. 2nd edition Boston: Kluwer; 2002.
Michelena et al. (2005) Michelena P, Henric K, Angibault J, Gautrais J, Lapeyronie P, Porter RH, Deneubourg J, Bon R. An experimental study of social attraction and spacing between sexes in sheep. Journal of Experimental Biology. 2005;208:4419–4426. doi: 10.1242/jeb.01909. [PubMed] [Cross Ref]
Morales et al. (2010) Morales JM, Moorcroft PR, Matthiopoulos J, Frair JL, Kie JG, Powell RA, Merrill EH, Haydon DT. Building the bridge between animal movement and population dynamics. Philosophical Transactions of the Royal Society B. 2010;365:2289–2301. doi: 10.1098/rstb.2010.0082. [PMC free article] [PubMed] [Cross Ref]
Morghan, Sheley & Svejcar (2006) Morghan KJR, Sheley RL, Svejcar TJ. Successful adaptive management—the integration of research and management. Rangeland Ecology and Management. 2006;59:216–219. doi: 10.2111/05-079R1.1. [Cross Ref]
Northrup et al. (2013) Northrup JM, Hooten MB, Anderson Jr CR, Wittemyer G. Practical guidance on characterizing availability in resource selection functions under a use-availability design. Ecology. 2013;94:1456–1463. doi: 10.1890/12-1688.1. [PubMed] [Cross Ref]
Ormaechea & Peri (2015) Ormaechea S, Peri P. Landscape heterogeneity influences on sheep habits under extensive grazing management in southern patagonia. Livestock Research for Rural Development. 2015;27:1–6.
Paramidani, Doffigny & Codesal (2014) Paramidani M, Doffigny C, Codesal P. Estudio inicial de Pastizales Ea. “Fortín Chacabuco”. Buenos Aires: Ovis XXI S.A; 2014. pp. 28.
Petocz (1973) Petocz RG. The effect of snow cover on the social behavior of bighorn rams and mountain goats. Canadian Journal of Zoology. 1973;51:987–993. doi: 10.1139/z73-143. [Cross Ref]
Pinchak et al. (1991) Pinchak WE, Smith MA, Hart RH, Waggoner Jr JW. Beef cattle distribution patterns on foothill range. Journalof Range Management. 1991;44:267–275. doi: 10.2307/4002956. [Cross Ref]
Plumb & Dodd (1993) Plumb GE, Dodd JL. Foraging ecology of bison and cattleon a mixed prairie: implications for natural area management. Ecological Applications. 1993;3:631–643. doi: 10.2307/1942096. [Cross Ref]
Porath et al. (2002) Porath M, Momont P, DelCurto T, Rimbey N, Tanaka JA, McInnis M. Offstream water and trace mineral salt asmanagement strategies for improved cattle distribution. Journal of Animal Science. 2002;80:346–356. [PubMed]
Prache, Gordon & Rook (1998) Prache S, Gordon IJ, Rook AJ. Foraging behaviour anddiet selection in domestic herbivores. Annales de Zootechnie. 1998;47:335–345. doi: 10.1051/animres:19980502. [Cross Ref]
Pringle & Landsberg (2004) Pringle HJ, Landsberg J. Predicting the distribution of livestock grazing pressure in rangelands. Austral Ecology. 2004;29:31–39. doi: 10.1111/j.1442-9993.2004.01363.x. [Cross Ref]
R Core Team (2015) R Core Team R: a language and environment for statistical computing. R Foundation for Statistical Computing; Vienna: 2015.
Rhodes & Marlow (1997) Rhodes B, Marlow C. Abstracts, Society for Range Management Annual Meeting; 1997. Intensive range riding for improving livestock distribution. pp. 21.
Risenhoover & Bailey (1985) Risenhoover KL, Bailey JA. Foraging ecology of mountain sheep: implications for habitat management. The Journal of Wildlife Management. 1985;49:797–804. doi: 10.2307/3801714. [Cross Ref]
Rubio et al. (2008) Rubio CMB, Cibils AF, Endecott RL, Petersen MK, Boykin KG. Pinon–juniper woodland use by cattle in relation to weather and animal reproductive state. Rangeland Ecology and Management. 2008;61:394–404. doi: 10.2111/07-056.1. [Cross Ref]
Savory (1983) Savory A. The savory grazing method or holistic resource management. Rangelands. 1983:155–159.
Sayre et al. (2013) Sayre NF, McAllister RR, Bestelmeyer BT, Moritz M, Turner MD. Earth stewardship of rangelands: coping with ecological, economic, and political marginality. Frontiers in Ecology and the Environment. 2013;11:348–354. doi: 10.1890/120333. [Cross Ref]
Scott, Banner & Provenza (1996) Scott CB, Banner RE, Provenza FD. Observations of sheep foraging in familiar and unfamiliar environments: familiarity with the environment influences diet selection. Applied Animal Behaviour Science. 1996;49:165–171. doi: 10.1016/0168-1591(96)01040-4. [Cross Ref]
Scott, Provenza & Banner (1995) Scott CB, Provenza FD, Banner RE. Dietary habits and social interactions affect choice of feeding location by sheep. Applied Animal Behaviour Science. 1995;45:225–237. doi: 10.1016/0168-1591(95)00605-R. [Cross Ref]
Seabloom & Reichman (2001) Seabloom EW, Reichman OJ. Simulation models of the interactions between herbivore foraging strategies, social behavior, and plant community dynamics. The American Naturalist. 2001;157:76–96. doi: 10.1086/317009. [PubMed] [Cross Ref]
Senft et al. (1987) Senft R, Coughenour M, Bailey D, Rittenhouse L, Sala O, Swift D. Large herbivore foraging and ecological hierarchies. BioScience. 1987;37:789–799. doi: 10.2307/1310545. [Cross Ref]
Sibbald, Shellard & Smart (2000) Sibbald A, Shellard L, Smart T. Effects of space allowance on the grazing behaviour and spacing of sheep. Applied Animal Behaviour Science. 2000;70:49–62. doi: 10.1016/S0168-1591(00)00145-3. [PubMed] [Cross Ref]
Spiegelhalter et al. (1996) Spiegelhalter D, Thomas A, Best N, Gilks W. BUGS0.5: Bayesian inference using Gibbs sampling manual (version ii) MRC Biostatistics Unit, Institute of Public Health; Cambridge: 1996.
Stenseth (1983) Stenseth NC. Causes and consequences of dispersal in smallmammals. In: Swingland JR, Greenwood P, editors. The ecology of animal movement. Oxford University Press; Oxford: 1983. pp. 63–101.
Stolba et al. (1990) Stolba A, Hinch G, Lynch J, Adams D, Munro R, Davies H. Social organization of merino sheep of different ages, sex and family structure. Applied Animal Behaviour Science. 1990;27:337–349. doi: 10.1016/0168-1591(90)90129-2. [Cross Ref]
Stricklin, Zhou & Gonyou (1995) Stricklin W, Zhou J, Gonyou H. Selfish animats and robot ethology: using artificial animals to investigate social and spatial behavior. Applied Animal Behaviour Science. 1995;44:187–203. doi: 10.1016/0168-1591(95)00613-W. [Cross Ref]
Teague (2014) Teague R. Deficiencies in the Briske et al. rebuttal of the savory method. Rangelands. 2014;36:37–38.
Thouless (1990) Thouless C. Feeding competition between grazing red deer hinds. Animal Behaviour. 1990;40:105–111. doi: 10.1016/S0003-3472(05)80669-4. [Cross Ref]
Veiberg et al. (2004) Veiberg V, Loe LE, Mysterud A, Langvatn R, Stenseth NC. Social rank, feeding and winter weight loss in red deer: any evidence of interference competition? Oecologia. 2004;138:135–142. doi: 10.1007/s00442-003-1399-9. [PubMed] [Cross Ref]
Vervaecke, Roden & De Vries (2005) Vervaecke H, Roden C, De Vries H. Dominance, fatness and fitness in female american bison, bison bison. Animal Behaviour. 2005;70:763–770. doi: 10.1016/j.anbehav.2004.12.018. [Cross Ref]
Weber et al. (1998) Weber GE, Jeltsch F, Van Rooyen N, Milton SJ. Simulated long-term vegetation response to grazing heterogeneity in semi-arid rangelands. Journal of Applied Ecology. 1998;35:687–699. doi: 10.1046/j.1365-2664.1998.355341.x. [Cross Ref]

Articles from PeerJ are provided here courtesy of PeerJ, Inc