Assessing grazing behavior and intensity of small ruminants in a Mediterranean rangeland

George Mitri, Karen Gebrael, Georgy Nasrallah, Maria Bou Nassar, Nour Masri¹, Dominique Choueiter¹ and Apostolos P Kyriazopoulos²

Institute of the Environment, University of Balamand - Kelhat, El Koura, Lebanon
george.mitri@balamand.edu.lb
¹ GEF funded project “Sustainable Land Management in the Qaraoun Catchment, Lebanon”, Ministry of Environment, United Nations Development Programme, Lebanon
² Department of Forestry and Management of the Environment and Natural Resources, Democritus University of Thrace, Orestiada, Greece

Abstract

Monitoring grazing behavior, livestock movement and grazing intensity is essential for developing and implementing sustainable rangeland management plans and avoid problems associated with over-grazing (i.e., land degradation). This work aimed at investigating the spatio-temporal characteristics of grazing behavior and intensity of small ruminants in order to be incorporated into rangelands management strategies for a specific area. The specific objectives were to 1) monitor daily movement of livestock, 2) analyze existing transhumance routes, 3) investigate the relationship of grazing behavior and intensity with seasonal variation and 4) evaluate the association between grazing intensity and exposure to land degradation. Those objectives were achieved by using remotely-sensed (RS) and Geographic Information System (GIS) data. Seasonal movements of small ruminants (i.e., goats and sheep) were monitored in lowlands and highlands and throughout transhumance. Collars mounted with a Global Positioning System were employed for use in continuous monitoring of animal movement. Daily trajectories of animals were recorded during grazing/transhumance, in addition to the daily traveled distance, total spent duration in rangelands and average movement speed in rangelands, in order to evaluate their seasonal variability. Eventually, the grazing intensity of selected herds was computed in order to assess its association with exposure to land degradation. As the daily traveled distance and the total spent duration in rangelands were significantly different in each season, the highest average grazing intensity was in forests (i.e., lower exposure to degradation) and the lowest grazing intensity was in croplands (i.e., highest exposure to degradation).

Keywords: GPS collars, livestock, rangeland management, seasonal movement

Introduction

Several drivers are changing global livestock systems. Factors such as growth in human population and rapid urbanization have substantially increased demand for livestock products (Di Virgilio et al 2019), while the growing public concern for the animal welfare (Szűcs et al 2012) could increase the extensive and semi-extensive systems that rely on grazing (Turner and Dwyer, 2007). As a result, many grasslands are considered worldwide under pressure and degraded due to overgrazing and misuse of lands (Staddon and Faghihinia 2021).

According to Darwish et al (2012), areas prone to desertification represent 57% of Lebanese national territory. While climate conditions combined with demographic pressure and mismanagement of land resources proliferate the severity and extent of land degradation (Salvati and Zitti 2009), the pressure on land resources is exacerbated by land use change (Masri et al 2002) and overgrazing of marginal lands (Darwish and Faour 2008), among others. Currently, small ruminant farming is the most important livestock sector in Lebanon, based on sheep (i.e., 350,000 head; 4,100 herds) and goats (i.e., 450,000 head; 5,850 Herds) (IFAD 2017). Moreover, farming contributes to 25% of milk production (Serhan and Mattar 2017). Sheep and goats have always been an integral part of the rural mosaic in Lebanon. They mainly graze in rangelands, thus rangeland contribution to animal feeding is very important, especially to transhumant systems that are commonly applied in the country. Grazing rights and practices vary among regions; public lands are leased for grazing. While some legislations regarding rangeland exist, they are not properly applied and regulations in place are often disregarded due to their complexity and financial implications. As a result, overgrazing is experienced in various sites across the country, adversely affecting the sustainability of the rangelands. In this context, “National Guidelines for the Management of Rangelands Outside Forests” have recently been produced to maintain and enhance the economic, social, and environmental values of rangelands in Lebanon (Choueiter and Kyriazopoulos 2019).

As overgrazing is one of the most important causes of land degradation and biodiversity loss, especially in rural and low-income areas (Kosmas et al 2015), the development of rangeland management strategies must be considered as priority, especially in those areas (Akasbi et al 2012). In this context, developing and implementing suitable and efficient land management plans must be accompanied by reliable scientific evidences (Staddon and Faghihinia 2021). A significant interaction between livestock and the environment was observed in different studies (Otte et al 2019; Steinfeld 1998). As such, monitoring/evaluating livestock mobility across different landscape units and during different seasons is a key component in developing and implementing sustainable grazing strategies and initiatives, by government and researchers, especially in agro-forest lands. More specifically, land management strategies that take into consideration seasonal variability have been proved to be efficient in preventing land degradation (Kleinebecker et al 2011). Such strategies are essential for the conservation and sustainable use of rangelands (Akasbi et al 2012). Furthermore, understanding grazing behavior and tracking transhumance routes of livestock are essentially needed for planning a more efficient resource use and sustained land productivity while determining possible impact on ecosystem functions (Milchunas and Lauenroth 1993). According to Matches (1992), moving, slowing down, speeding up, grazing, and resting are primary activities of grazing livestock and have a large impact on rangeland ecosystems. In general, some grazing behaviors contribute to increasing or decreasing heterogeneity of vegetation cover (Adler et al 2001), therefore possibly worsening land degradation.

Consequently, evaluating grazing intensity (i.e., cumulative effect of ruminants on rangelands during a specific duration) is a crucial key component while choosing appropriate land strategies (Akasbi et al 2012). However, there is lack of studies that tackle the effect of grazing intensity on rangelands, especially that high grazing intensity is linked to loss of soil carbon, nutrients and microbial activity (Staddon and Faghihinia 2021).

Yet, understanding how individual animals or groups of animals interact with their environment represents a very important aspect of research in animal ecology (Homburger et al 2014). However, researchers experience different challenges related to continuously observing tracks and behavior of animals over a relatively long period of time. In this context, traditional direct human observation is not efficient especially if needed observations must be continuously carried out over large space and long period of time, and thus it is a labor intensive procedure (Augustine and Derner 2013). In addition, the behavior of observed animals might be influenced by human presence (Homburger et al 2014). Some of the traditional livestock monitoring techniques relied on the number of animals in a specific location; however, these methods assumed that livestock were uniformly distributed in this area. As such, monitoring grazing pattern should be based on the spatial distribution of livestock in a particular area, in addition to the daily spent time and crossed distance in a specific rangeland (Gou et al 2019).

Recent studies used automated methods to monitor grazing behavior since those techniques proved to be cost-effective and required less time and fewer number of field workers (Gou et al 2019). In this context, advances in sensor tracking technology with the use of Global Positioning System (GPS) data allowed the extensive collection of animal movement data and GPS collars were widely used to monitor livestock and track animal’s movement, position and location (Bailey et al 2018).

In light of the above, the aim of this work was to investigate the spatio-temporal characteristics of grazing behavior and intensity of small ruminants in order to be incorporated into rangelands management strategies for a specific area. The specific objectives were to 1) monitor daily movement of livestock, 2) analyze existing transhumance routes, 3) investigate the relationship of grazing behavior and intensity with seasonal variation and 4) evaluate the association between grazing intensity and exposure to land degradation.

This work focused on the extensive use of remotely-sensed (RS) data and their analyses in a Geographic Information System (GIS). However, it is essential to note that site-based methods continue to be widely used today especially when assessing structural and compositional attributes of vegetation conditions. Therefore, future work shall consider integrating the two approaches not only for monitoring grazing behavior but also for mapping vegetation condition across different land cover categories.

Materials and methods

Study area description

The study area covered the administrative districts of Zahleh (418 km²), Rachaya (545 km²) and West Bekaa (445 km²) in the Bekaa Valley (MoE/UNDP/GEF 2018) of Lebanon (Figure 1). The country has a Mediterranean climate categorized by dry and hot summers, rainy and cool winters, dry to moderately dry fall and spring. The Bekaa Valley is characterized by a semi-arid to continental climate comprising recurrent drought events (Darwish et al 2012). Most of the Rachaya district is pasture, rangeland and bare land. Most residents of the West Bekaa district depend on agriculture.

Part of the monitoring was conducted in low pasture lands located outside the initial study area (i.e., south Lebanon/low pasture lands) to account for transhumance routes with connections to the study area.

Figure 1. Study area location

Sampling methodology

Prior to fieldwork, a questionnaire was administered to the selected herd owner/herder in order to confirm willingness to participate in the study. Data such as type of herd, number of animals, current herd location, summer/winter rangeland and transhumance means/date/destination were collected.

A two-phase sampling approach was adopted, in order to cover the seasonal movements of livestock (i.e., goats and sheep) as follows:

· Phase 1 consisted of monitoring the mobility of animals on low pasture lands, then on the transhumance routes (i.e., from low pasture lands to highlands) and eventually in highlands.

· Phase 2 consisted of monitoring the mobility of animals only in the highlands.

Information about herds such as owner/herder contact information, number/type of animals and location were obtained from a survey-based herders’ database produced in 2019 by the “Sustainable Land Management in the Qaraoun Catchment” project. A total of 138 animals (i.e. 32 herds) were tracked, 37 sheep and 101 goats, organized in three groups, namely A, B and C (Table 1). Given a population of 126,014 animals (i.e., goats and sheep), the number of sampled animals involved a 7% margin of error and a 90% confidence level.

The sampling approach considered tracking the movement of 5 animals (i.e., 1 male and 4 females) in each herd. It was not possible to sample a higher number of animals due to timeline and logistical restrictions.

Recording transhumance routes and grazing behavior

Transhumance routes included all autumn and spring transitions from the highlands to the winter shelter and return. Based on the results of the conducted surveys, transhumance herds were classified into two categories (i.e., by truck or on foot). Herds with confirmed “on foot” transhumance were selected for monitoring transhumance routes, which was conducted in May/June 2019 (i.e., Group B, phase 1).

As for grazing behavior, animals were monitored in the lowlands during April/May 2019 (i.e., Group B, phase 1). In addition, the grazing behavior was assessed in the highlands during May/June 2019 (i.e., Group B, phase 1) and from August through December 2019 (i.e., Group A, B and C, phase 2).

Monitoring animal mobility was done using GPS collars (Photo 1.) designed to transmit animal location via a General Packet Radio Service (GPRS). The number of days recorded was 4 consecutive days for each herd.

During each field visit, an additional questionnaire was addressed to the herd owner/herder to obtain more detailed spatial information such as location of milk collection points and water drinking points. Accordingly, milk collection and water drinking points were identified with the help of herders using high spatial resolution imagery.

Photo 1. Use of GPS collars for monitoring animal mobility (source: Institute of the Environment, University of Balamand)

Analyzing spatial patterns of the grazing herds

The GPS devices recorded geographical position (i.e., latitude and longitude) every 5 minutes. This time interval was considered adequate to record in sufficient detail the distances/trajectories travelled daily, the total time spent daily in the rangeland, the proportion of time spent in each vegetation type or on degraded lands (Mitri et al 2019), the movement speed of livestock, the milk collection points and the water drinking points.

The GPS locations for each animal were converted into trajectories (i.e., the sequence of location points and segments connecting them), which were used to calculate “grazing duration” (h d^-1), the time elapsed between the first and last position recorded. This variable indicated the time spent in the rangeland. The distance travelled daily (km d ^-1) was calculated by summing each segment length corrected for the altitudinal gradient between the initial and final position. Moreover, “movement speed” (km h^-1) was calculated by dividing “total distance” by “grazing duration”. The “total time spent in each vegetation type” (h) for each trajectory was identified by overlapping the points recorded on each trajectory with land cover/ land use maps of the study area. Other datasets employed in this work for analyzing the recorded GPS data included the most recent land cover/land use map of Lebanon and a 10 x 10 m digital elevation model (DEM).

The multi-temporal recorded measurements were investigated in order to assess the connection between mobility behavior and seasonal variations. In this context, an independent t-test (i.e., normal data) and a Mann-Whitney U test (i.e., non-normal data) were used to determine the statistical difference in mobility behavior between goats and sheep. Furthermore, the non-parametric Kruskal Wallis test (i.e., non-normal data) test was performed using SPSS to evaluate the relationship between movement behavior and seasonal differences.

Evaluating grazing intensity

The grazing intensity of each herd was computed in order to check its association with the exposure to land degradation of grazed sites. Accordingly, rectangular grids were generated around every grazing location using GIS. The grid cell area was 1 ha (i.e., 100 m * 100m). The following formula was used in order to calculate the grazing intensity:

Where,

GI _herd (goats/sheep ha ^-1): grazing intensity of the herd in each grid cell (i). It is worth mentioning that one male (i.e., sheep or goat) was chosen from each herd in order to calculate the grazing intensity;

n: number of records within each grid cell i, obtained through the intersection between each grid cell and the recorded readings, using QGIS, as shown in Figure 2;

N _herd: total number of animals in the herd;

R: GPS recording interval (h) (i.e., 0.083 h in this study);

D: daily grazing period (h);

GP: total number of recorded days (i.e., 4 days in this study); and

S: grid cell area (i.e. 1 ha in this study)

Figure 2. Example of intersection between each grid cell and the recorded readings (source: Institute of the Environment, University of Balamand)

The computed multi-temporal grazing intensities were also investigated in order to evaluate the link between grazing intensities and seasonal variations. Accordingly, the non-parametric Kruskal Wallis test (i.e., non-normal data) test was performed using SPSS.

Subsequently, 3 herds were selected (i.e., one in Zahleh, one in Rachaya and one in West Bekaa) in order to check the link between grazing intensity and land degradation. For this purpose, the intersection between each grid cell (i.e., having a specific grazing intensity) and its degradation score (Mitri et al 2019) was performed using GIS. In this work, lands were categorized into least, moderate and highly prone to degradation, according to their degradation score (Mitri et al 2019). In this context, the variability in degradation scores and grazing intensities across land cover categories (i.e., grassland, cropland and forest) was assessed through the non-parametric Kruskal Wallis test (i.e., non-normal data). Furthermore, the variability in grazing intensities across degradation score categories was tested using non-parametric Kruskal Wallis test (i.e., non-normal data). In addition, the correlation between degradation score and grazing intensity was assessed through Spearman correlation (i.e., relationship between data was not linear).

In principle, utilization and residual vegetation field-based data, in combination with the previously described monitoring information, can be used to improve the understanding of spatial and temporal patterns of livestock use, grazing intensity, and possible causes of changes in rangeland attributes. Yet, additional work is needed in the future to collect further data in the field and accordingly, guide adjustments to management strategies.

Results and discussion

Spatial patterns of the grazing herds

Daily trajectories as a function of different land cover/land use categories (i.e., cropland, forest, grassland, urban land and others) were determined. Almost all milk collection locations coincided with existing water points. Herders installed water tanks as drinking points at the entrance of each corral (i.e., also milk collection points).

Animal mobility statistics from phases 1 and 2 were computed and segregated by animal species (i.e., goats and sheep). The statistical analysis didn’t show any significant difference between goats and sheep movement behavior (p value > 0.05). Accordingly, the remaining analysis was completed by combining data for both sheep and goats.

Travelled distance

The average distance traveled by day for both sheep and goats varied between 5.23 km/day and 9.16 km/day. The highest average travelled distances occurred in October and May and the lowest ones were in June and November (Figure 3). The statistical difference between months in term of travelled distance was statistically significant (p value = 0.048 < 0.05 – Kruskal Wallis Test). This indicated that the distance travelled by herds was affected by seasonality. In fact, Lebanon is characterized by hot/ dry summers extended from June to September and cool/rainy weather from November until March (RCCC 2021). This mainly explains having the lowest traveled distances in June (i.e., mobility affected by high temperature) and in November (i.e., mobility affected by low temperature/rain). Additionally, the scarcity of forage in dry season may affect the behavior of ruminant animals and consequently the travelled distance is reduced (Lamidi and Ologbose 2014).

Figure 3. Seasonal variation of travelled distance per day

Spent duration in rangelands

The average spent duration in rangelands per day for both sheep and goats varied between 9.20 h/day and 14 h/day (Figure 4). The highest spent duration was in June and August and the lowest spent durations were in November and December. The statistical difference between months in term of travelled spent duration was statistically significant (p value = 0.04 < 0.05 – Kruskal Walllis Test). This was mainly attributed to longer days during summer, where herds started grazing at sunrise and continued until sunset (Garcia-Gonzalez et al 1990).

Figure 4. Seasonal variation of spent duration in rangelands

Speed in rangelands

The average speed in rangelands for both sheep and goats varied between 0.37 km/h in June and 0.86 km/h in October (Figure 5). The statistical difference between months in term of speed was not statistically significant (p value = 0.33 > 0.05 – Kruskal Walllis Test), meaning that seasonality did not necessarily affect the speed of small ruminants speed.

Figure 5. Seasonal variation of speed in rangelands

Transhumance routes

The calculated total distance travelled during transhumance was 51 km involving a movement for total duration of 33 hr. All identified transhumance routes crossed grasslands, forests and croplands. Grasslands and forests represented the main vegetation types for livestock grazing during transhumance. Herders were found to avoid cropland areas throughout transhumance. This could be mostly related to avoid conflict with landowners and reduce the risk of damaging existing crops (Ntassiou et al 2019).

In extended monitoring tasks, variability in growing conditions, affected by both climate and weather, provides important context for interpreting transhumance data. More specifically, precipitation and temperature influence the start and end of the growing season as well as the quantity and quality of forage (Jansen et al 2021) which in turn affect transhumance routes in their spatial and temporal characteristics.

Seasonal variation of grazing intensity

The average grazing intensity in rangelands for both sheep and goats varied between 9animal/ha in October and 30 animals/ha in May (Figure 6). The statistical difference between months in term of grazing intensity was not statistically significant (p value = 0.79 – Kruskal Walllis Test), meaning that seasonality did not necessarily affect grazing intensity. This finding was in accordance with results reported by Akasbi et al (2012), where the computed grazing intensities showed small variation between seasons.

Figure 6. Seasonal variation of grazing intensity

In general, grazing intensity could be affected by the abundance and types of plant species in a certain location (Akasbi et al 2012). To this regard, Castro and Fernández Núñez (2016) reported that grazing of small ruminants is affected by the forage availability and the applied grazing management. Generally, the decisions of the shepherds on managing daily grazing were based on the availability of forage resources in order to maximize the productivity of animals (Bonanno et al 2008; Akasbi et al 2012).

Grazing intensity and land degradation

The difference in the three land cover categories, namely grassland, cropland and forest, in term of degradation score was also investigated. The results showed a significant statistical difference in proneness to land degradation among the 3 different categories (p value = 0.0001 < 0.05). Table 2 showed that the highest average degradation score was in croplands. This could be mostly explained by adopting unstainable farming practices.

Furthermore, the relationship between degradation score and grazing intensity was investigated. Accordingly, the grazing intensity was found to be statistically different (p value = 0.00013 < 0.05) in the 3 degradation categories (i.e., least, moderately and highly prone to degradation). Table 3 showed that the highest average grazing intensity was in areas that were least prone to degradation.

The correlation between the degradation score and grazing intensity was further evaluated. The results showed a negative significant correlation (p value = 1.41E-26 < 0.05, correlation coefficient = - 0.26). Also, the variation between land cover categories (i.e., grassland, cropland and forest), in term of grazing intensities was investigated (Photo 2). The results showed a significant statistical difference in grazing intensity among the 3 land cover categories (p value = 3.47 E-9).

Photo 2. Variety of grazing intensities across the different land cover categories

Table 4. indicated that the highest average grazing intensity was in forests (i.e., least prone to degradation) and the lowest grazing intensity was in croplands (i.e., moderately prone to degradation). This could be mostly caused by limited opportunities of grazing in croplands especially when not managed for allowing sustainable farming practices. It is well documented that livestock, specifically goats, consume more woody species in comparison to herbaceous ones (El Aich et al 2007; Manousidis et al 2016). This could explain the high grazing intensity in forests, compared to grasslands and croplands, therefore risking overgrazing of limited forest resources if no rangeland management plans were put in place. Furthermore, grazing intensity is affected by weather, due to its relationship on plant growth and livestock behavior (Smith et al 2016). In this context, addressing the climate variability challenge provides an important context for interpreting grazing intensity.

In future work, this assessment could be further complemented by estimates about how much forage has been removed or what remains across larger spatial and temporal extents. However, failure to accurately account for biodiversity, variability in biomass and growing conditions can introduce bias to measurements, thus adding complexity to data interpretation (Jansen et al 2021). In addition, it is important to note that commonly used in-field methods have been shown to be subjective (Halstead et al 2000) and costly to collect (Caughlan and Oakley 2001).

Conclusions

This study provided direct observation of daily livestock movement and transhumance routes, evaluated grazing intensities, investigated interlinkages between grazing behavior and seasonal variation, and assessed relationship between grazing intensities and land degradation. The following was mainly concluded:

• The distance travelled by herds was affected by seasonality.
  
  
• The highest spent grazing duration was in June and August and the lowest spent durations were in November and December, however the speed of animal movement was not significantly affected by seasonality.
  
  
• Grasslands and forests represented the main vegetation types for livestock grazing during transhumance as herders tended to avoid cropland areas.
  
  
• Failure to account for the influence of climatic conditions on rangelands can lead to an incomplete understanding of temporal and spatial variabilities in transhumance.
  
  
• The average grazing intensity in rangelands for both sheep and goats varied between 9animal/ha (i.e., in October) and 30 animals/ha (i.e., in May), yet seasonality did not necessarily affect grazing intensity.
  
  
• The highest average grazing intensity was in areas categorized as least prone to degradation (i.e., forests), therefore providing needed forage for animals.
  
  
• Highly prone to degradation croplands provided little grazing opportunities.

Overall, allowing sustainable farming practices taking into account grazing opportunities and restoring degraded grassland are expected to improve rangeland conditions, therefore providing more opportunities for sustainable rangeland management across the different land cover categories (e.g., cropland, forests and grassland).

Future research involves 1) expanding the number of sampled animals for improved analysis of statistical significance and 2) track the animals for a longer period of time to account for improved monitoring of monthly variability in their movements in function of reproduction cycles, 3) consider the inclusion of adaptive actions in response to climate variability challenges and 4) combine and analyze the results of this study with field data collected on the floristic composition of the study area therefore addressing limitations of not accounting for field data in this work. Overall, a better knowledge of the flora in combination with direct observation of animal behavior is expected to advance knowledge in the sustainable management of rangeland.

Acknowledgements

This work was supported by the “Sustainable Land Management in the Qaraoun Catchment” project funded by the Global Environment Facility (the GEF), implemented by the Lebanese Ministry of Environment and the United Nations Development Programme (UNDP). Also, the authors would like to thank Engineer Ali Ibrahim for assisting in data collection and mounting of collars.

References

Adler P B, Raff D A and Lauenroth W K 2001 The effect of grazing on the spatial heterogeneity of vegetation. Oecologia , 128, 465-479, DOI: 10.1007/s004420100737

Akasbi Z, Oldeland J, Dengler J and Finckh M 2012 Analysis of GPS trajectories to assess goat grazing pattern and intensity in Southern Marocco. The Rangeland Journal, 34, 415-427, DOI: 10.1071/RJ12036

Augustine D G and Derner J D 2013 Assessing herbivore foraging behavior with GPS collars in a semiarid grassland. Sensors, 13, 3711-3723, DOI: 10.3390/s130303711

Bailey D W, Trotter M G, Knight, C W, Thomas M G 2018 Use of GPS tracking collars and accelerometers for rangeland livestock production research. Translational Animal Science, 2, 81-88, https://doi.org/10.1093/tas/txx006

Bonanno A, Fedele V and Di Grigoli A 2008 Grazing management of dairy goats on Mediterranean herbaceous pastures. In ‘Dairy Goats Feeding and Nutrition’. (Eds Cannas, A and Pulina, G pp. 189-220). Wallingford, UK: CAB International.

Castro M and Fernández Núñez E 2016 Seasonal grazing of goats and sheep on Mediterranean mountain rangelands of northeast Portugal. Livestock Research for Rural Development, 28 (5), http://www.lrrd.org/lrrd28/5/cast28091.html

Caughlan L and Oakley K 2001 Cost considerations for long-term ecological monitoring, Ecol. Indic., 1 (2), pp. 123-134.

Choueiter D and Kyriazopoulos A P 2019 National Guidelines for rangeland management. Report LBN/CO/IC/18/18 United Nations Development Programme. UNDP Lebanon Country Office, Beirut, Lebanon, p 211.

Darwish T and Faour G 2008 Rangeland degradation in two watersheds of Lebanon. Lebanese Science Journal, 1, 71-80, http://lsj.cnrs.edu.lb/2008/06/01/t-darwish-and-g-faour/

Darwish T, Zdruli P, Saliba R., Awad M, Shaban A and Faour G 2012 Vulnerability to Desertification in Lebanon Based on Geo-information and Socioeconomic Conditions. Journal of Environmental Science and Engineering, 1, 851-864, https://www.academia.edu/11300499/Vulnerability_to_Desertification_in_Lebanon_Based_on_Geo_information_and_Socioeconomic_Conditions

Di Virgilio A, Lambertucci S A and Morales, J M 2019 Sustainable grazing management in rangelands: Over a century searching for a silver bullet. Agriculture, Ecosystem and Environment, 283, 106561, https://doi.org/10.1016/j.agee.2019.05.020

El Aich, A, El Assouli N, Fathi A, Morand-Fehr P and Bourbouze A 2007 Ingestive behavior of goats grazing in the Southern Argan (Argania spinosa) forest of Morocco. Small Ruminant Research, 70, 248-256, DOI: 10.1016/j.smallrumres.2006.03.011

Garcia-Gonzalez R, Hidalgo R and Montserrat C, 1990 Patterns of livestock use in time and space in the summer ranges of the Western Pyrenees: A case study in the Aragón valley. Mountain Research and Development, 10, 241-255, https://www.jstor.org/stable/3673604?seq=1#metadata_info_tab_contents

Gou X, Tsunekawa A, Peng F, Zhao X, Li Y and Lian J 2019 Method for Classifying Behavior of Livestock on Fenced Temperate Rangeland in Northern China. Sensors, 19 (23), 5334, doi: 10.3390/s19235334

Halstead L, Howery L, Ruyle G 2000 Comparison of 3 Techniques for Monitoring Use of Western Wheatgrass.” Journal of Range Management 53, no. 5 (2000): 499–505. https://doi.org/10.2307/4003650.

Homburger H, Schneider M K, Hilfiker S and Lüscher A 2014 Inferring Behavioral States of Grazing Livestock from High-Frequency Position Data Alone. PLoS ONE, 9(12), e114522. https://doi.org/10.1371/journal.pone.0114522

IFAD 2017 Harmonised Actions for Livestock Enhanced Production and Processing (HALEPP). Detailed design report. Project No. 2000001478, https://webapps.ifad.org/members/lapse-of-time/docs/english/EB-2017-LOT-P-11-Project-Design-Report.pdf

Jansen V, Traynor A, Karl J, Lepak N, Sprinkle J, 2021 Monitoring grazing use: Strategies for leveraging technology and adapting to variability, Rangelands, https://doi.org/10.1016/j.rala.2021.07.005.

Kleinebecker T, Weber H and Hotzel N 2011 Effects of grazing on seasonal variation of aboveground biomass quality in calcareous grasslands. Plant Ecology, 212 (9), 15663-1576, https://www.jstor.org/stable/41508838?seq=1#metadata_info_tab_contents

Kosmas C, Detsis V, Karamesouti M, Kounalaki K, Vassiliou P and Salvati L 2015 Exploring long-term impact of grazing management on land degradation in the socio-ecological system of Asteroussia Mountains, Greece. Land, 4(3), 541-559, https://doi.org/10.3390/land4030541

Lamidi A A and Ologbose F I 2014 Dry season feeds and feeding: a threat to sustainable ruminant animal production in Nigeria. Journal of Agriculture and Social Research, 14, 17-30, https://www.ajol.info/index.php/jasr/article/view/166692#:~:text=Conclusively%20dry%20season%20feeds%20and,of%20strategy%20recommend%20in%20the

Manousidis T, Kyriazopoulos, A P, Parissi, Z M, Abraham, E M, Korakis, G and Abas, Z 2016 Grazing behavior, forage selection and diet composition of goats in a Mediterranean woody rangeland. Small Ruminant. Research, 145, 142-153, http://dx.doi.org/10.1016/j.smallrumres.2016.11.007

Masri T, Khawlie M and Faour G 2002 Land cover change over the last 40 years in Lebanon. Lebanese Science Journal, 2, 17-28.

Matches A 1992 Plant-response to grazing - a review. Journal of Production Agriculture , 5, 1-7, https://doi.org/10.2134/jpa1992.0001

Milchunas D G and Lauenroth W K 1993 Quantitative effects of grazing on vegetation and soils over a global range of environments. Ecological Monographs, 63, 327- 366, https://doi.org/10.2307/2937150

Mitri G, Nasrallah G, Gebrael K, Bou Nassar M, Abou Dagher M, Nader M, Masri N and Choueiter D 2019 Assessing land degradation and identifying potential sustainable land management practices at the subnational level in Lebanon. Environmental Monitoring and Assessment, 191(9), 567, https://doi.org/10.1007/s10661-019-7739-y

MoE/UNDP/GEF 2018 Sustainable Land Management in the Qaraoun Catchment, Lebanon. Beirut, Lebanon. Retrieved December 1, 2020 from https://www.thegef.org/sites/default/files/project_documents/12-01-14_Project_Document_PAD.pdf

Ntassiou K and Doukas I D 2019 Recording and mapping traditional transhumance routes in the South-Western Macedonia, Greece. GeoJournal, 84(1), 161-181, https://link.springer.com/article/10.1007/s10708-018-9857- 4

Otte J, Pica-Ciamarra U and Morzaria S 2019 A comparative overview of the livestock-environment interactions in Asia and Sub-Saharan Africa. Frontiers in Veterinary Science, 6, 37, doi: 10.3389/fvets.2019.00037

RCCC 2021 Country level, climate fact sheet, Lebanon. The Red Cross Red Crescent Climate Centre, Retrieved November 20, 2021 from https://www.climatecentre.org/wp-content/uploads/RCCC-ICRC-Country-profiles-Lebanon.pdf

Salvati L and Zitti M 2009 Assessing the impact of ecological and economic factors on land degradation vulnerability through multiway analysis. Ecological Indicators, 9, 357-363, DOI: 10.1016/j.ecolind.2008.04.001

Serhan M and Mattar J 2017 “The Goat Dairy Sector in Lebanon”, in: Goat science. In Tech Open Science.

Smith L, Ryle G, Maynard Jet al. 2015 . Principles of obtaining and interpreting utilization data on southwest rangelands. 2016:1-14.

Staddon P L and Faghihinia M 2021 Grazing intensity is key to global grassland carbon sequestration potential. Sustainable Environment, 7(1), https://doi.org/10.1080/27658511.2021.1895474

Steinfeld H 1998 Livestock and their interaction with the environment: An overview. BSAP Occasional Publication, 21, 67-76.

Szűcs E, Geers R, Jezierski T, Sossidou E N and Broom D M 2012 Animal Welfare in Different Human Cultures, Traditions and Religious Faiths. Asian-Australasian Journal of Animal Sciences, 25 (11), 1499-1506, doi: 10.5713/ajas.2012.r.02

Turner S P and Dwyer C M 2007 Welfare assessment in extensive animal production systems: challenges and opportunities. Animal welfare, 16, 189-192.