Spatial and temporal activity of cattle grazing in Mediterranean oak woodland

G Model

ARTICLE IN PRESS

APPLAN-4370; No. of Pages 9

Applied Animal Behaviour Science xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Applied Animal Behaviour Science journal homepage: www.elsevier.com/locate/applanim

Spatial and temporal activity of cattle grazing in Mediterranean oak woodland Iris Schoenbaum a,d,∗ , Jaime Kigel a , Eugene D. Ungar b , Amit Dolev c , Zalmen Henkin d a The Robert H. Smith Institute for Plant Sciences and Genetics in Agriculture, Faculty of Agricultural, Food and Environmental Sciences, Hebrew University of Jerusalem, Rehovot 76100, Israel b Department of Natural Resources, ARO − The Volcani Center, P.O. Box 6, Bet-Dagan 50250, Israel c MIGAL − Galilee Technological Center, Qiryat Shemona, P.O. Box 90000, Rosh Pina 12100, Israel d Beef Cattle Section, Newe-Ya’ar Research Center, Department of Natural Resources, Agricultural Research Organization, P.O. Box 1021, Ramat Yishay 30095, Israel

a r t i c l e

i n f o

Article history: Received 5 May 2016 Received in revised form 21 November 2016 Accepted 27 November 2016 Available online xxx Keywords: Cattle behavior GIS GPS Habitat preference Grazing management

a b s t r a c t We examined the temporal and spatial variation patterns of the grazing activity of free-ranging cattle in Mediterranean oak woodland in the Western Galilee, Israel, as affected by seasonal and management factors. The vegetation is dominated by scrub-oak woodland (Quercus calliprinos Webb.), interspersed with patches of semi-dwarf shrubs and herbaceous vegetation. High and moderate animal population densities of 0.55 and 0.33 cow ha−1 , respectively, were replicated twice. Cattle behavior was monitored with activity sensors on GPS collars, and pedometers, and spatial data were processed with Geographic Information System (GIS) tools. Overall, cattle devoted 9.7 ± 0.7 h/day to grazing, mostly in woodland areas, although they are natural herbaceous grazers. Behavior was associated with seasonal changes in biotic and abiotic factors. Preference for the woody vegetation types was detectable over the annual time scale but large seasonal differences in preference canceled out to a large extent when viewed at that time scale. Cattle under high density spent more time grazing and made more use of woody vegetation and steeper slopes. Thus, relatively high population densities may be required for landscape-oriented management. The present findings should contribute to rational management of cattle grazing in Mediterranean woodlands. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Mediterranean evergreen oak woodlands have been grazed by domestic ruminants, predominantly goats, for millennia. In addition to supporting animal production, grazing maintains an open woodland structure, reduces biomass and the risk of fire (Gutman et al., 2000), increases species diversity (Perevolotsky and Seligman, 1998), and facilitates recreational use of the landscape (Henkin, 2011). Over recent decades the woodlands of Israel have seen a massive decline in goat grazing, for a variety of socio-economic reasons; but there has been a modest revival of cattle grazing, introduced to increase meat production and, at least partially, to restore grazing as a means of woodland management. However, in contrast to its suitability for goats, which

∗ Corresponding author at: The Robert H. Smith Institute for Plant Sciences and Genetics in Agriculture, Faculty of Agricultural, Food and Environmental Sciences, Hebrew University of Jerusalem, Rehovot 76100, Israel. E-mail address: isi [email protected] (I. Schoenbaum).

are well adapted to browsing in Mediterranean woodlands, the woody vegetation is considered unsuitable for cattle, because of its poor nutritional quality (Perevolotsky et al., 1993; Papachristou et al., 2005). The low prevalence of cattle grazing in Mediterranean woodlands is due, presumably, to the limited penetrability of the typically dense woody vegetation, low productivity of the herbaceous vegetation, and poor accessibility of the extensive woodland areas that grow on steep terrain. Although woodland utilization is affected by cattle behavior, knowledge of the behavior patterns of cows in Mediterranean woodlands, and of the spatial and temporal distributions of their grazing is scarce. Acquisition of this knowledge should help to resolve potential conflicts of interest between herd owners, land-management agencies, and conservationists, and, furthermore, it is important for sustainable landscape management (Zuo and Miller-Goodman, 2003; Bailey, 2004). Spatial and temporal patterns of livestock behavior and utilization of the landscape are affected by abiotic factors such as slope, distance from water, and presence or absence of shade, as well as by biotic factors such as species composition of the vegetation, and forage quantity and quality (Arnold, 1981; Ganskopp, 2001;

http://dx.doi.org/10.1016/j.applanim.2016.11.015 0168-1591/© 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model APPLAN-4370; No. of Pages 9

ARTICLE IN PRESS I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

2

Bailey, 2005). Mediterranean oak woodlands are characterized by a patchy structure comprising a fine-scale mosaic of vegetation types that can influence grazing activity and spatial distribution of the cattle. In turn, heterogeneous distribution of the cattle in the woodland may result in uneven utilization of the vegetation (Bailey et al., 2006; Ganskopp and Bohnert, 2009). This can result in excessive grazing pressure in specific areas, which then may suffer degradation, while leaving other areas underutilized and, possibly, then prone to progressive closure by the woody vegetation and the attendant reduction of accessibility, increased fire risk, and loss of biodiversity (Perevolotsky and Seligman, 1998). Understanding the factors that influence cattle distribution in woodlands is a critical step in devising management strategies to foster uniformity of grazing pressure. Spatial distribution and habitat selection of cattle have been studied in various rangelands around the world (Bailey et al., 2001; Schlecht et al., 2004; Kaufmann et al., 2013), including Israel (Henkin et al., 2012). However, little is known about grazing behavior in Mediterranean woodlands, where the logistical constraints on monitoring animals generally encountered in predominantly herbaceous landscapes are compounded by the fact that the animals cannot be easily observed. Such constraints are, at least partially, relieved by the use of animal collars that combine the Global Positioning System (GPS) with activity sensors to provide spatially and temporally continuous and accurate data collection regarding livestock grazing activities (Turner et al., 2000; Schlecht et al., 2004). Such data can be processed by a Geographic Information System (GIS) to relate animal location data to abiotic and biotic terrain factors. The objectives of the present study were: 1) to analyze the effects of animal population density and season on the diurnal activity pattern and the spatial distribution of grazing activity of free-ranging cattle in Mediterranean oak woodlands; and 2) to analyze the roles of abiotic and biotic landscape features, including topography, vegetation structure, and the locations of watering and supplementary feeding points, in shaping the spatial distribution of grazing activity. Understanding the relationships between cattle behavior and these factors will help to develop a management program that seeks to combine sustainable utilization of Mediterranean oak woodlands with improved herd productivity and performance. 2. Materials and methods 2.1. Study site The study was conducted at the Hatal Experimental Farm in Western Galilee, Israel (long. 35◦ 15 , lat. 33◦ 01 ), from October 2007 through November 2009. The site was described in detail by Henkin et al. (2005). Briefly, the site is at 400–500 m a.s.l. and consists of moderate to steep slopes of up to 40◦ . Limestone and dolomite rocks form 15–40% of the surface cover, and between them are pockets of terra rossa soil up to 40 cm deep. The dominant vegetation is scrub-oak woodland (Quercus calliprinos Webb), interspersed with batha vegetation comprising shrubs and dwarf shrubs, mainly Calicotome villosa (Poiret) Link and Sarcopoterium spinosum (L.) Spach. Herbaceous vegetation occurs as patches in open areas among the woody vegetation, and provides 3–4 months of high-quality forage during winter and spring. The climate is typically Mediterranean, with mild winters and hot dry summers. The hydrological year (and hence measurement of annual rainfall) commences on 1 October. The long-term average (±SD) annual rainfall is 796 ± 201 mm, most of which falls between November and March. Annual rainfall was 535 mm in 2007/8 (i.e. 1 Oct. 2007–30 Sept. 2008) and 610 mm in 2008/9, which are low values compared with the long-term average, and 797 mm in 2009/10. The average, minimum, and maxi-

mum ambient temperatures during the experimental periods were 19.5, 8.3, and 35.5 ◦ C, respectively, in spring; 27.8, 18.8, and 38.8 ◦ C, respectively, in summer; and 23.2, 13.4, and 37.5 ◦ C, respectively, in fall. 2.2. Experiment treatments, animals and grazing management Experiments were approved by the animal experimentation ethics committee of the Agricultural Research Organization. The experimental area occupied 212 ha, divided into four paddocks (East, West, North, South), each of 40–66 ha (Table 1; Supplementary Fig. S1). Treatments were two animal population densities − moderate and high relative to common practice, of 0.33 and 0.55 cow·ha−1 , respectively − which were replicated twice and randomly assigned to the four paddocks for the duration of the experiment. The paddock sizes were chosen to be representative of those found among commercial producers in the study region, and the number of animals per group was intended to be sufficiently large to yield robust results. The animals comprised the farm’s resident herd of 94 Baladi (Bos Taurus) × Hereford cows aged 3–12 years, of average body weight 502 ± 9 kg, and of fair-to-good body condition (at least 2.5 on scale of Edmonson et al., 1989) after calving. Cows were randomly allocated to paddocks for the duration of the experiment with the single constraint of being balanced by age. Calving occurred predominantly during November through March, and the calves were weaned in early June. The average annual calving and weaning rates were 85 ± 3% and 76 ± 3%, respectively. Following early-season grazing deferment, the cows entered the paddocks in spring (mid-March), grazed through fall (midNovember), in accordance with vegetation conditions, and averaged 260 grazing days per year. They had continuous access to water, and supplementary feed was provided during late summer and fall (May–November). During the rest of the year − midNovember through mid-March − the cattle were kept in holding paddocks outside the experimental area. 2.3. Monitoring cattle location and activity Cattle behavior was monitored in three seasons of the 2007/8 hydrological year: at the beginning of the rainy season in the fall (October–November), when the diet comprises mainly woody vegetation, oak acorns and supplementary feed; in the spring (March–April), when the diet comprises green herbaceous vegetation and woody vegetation; and in the summer (July–August), when the diet comprises dry herbaceous vegetation, woody vegetation and supplementary feed (Brosh et al., 2006a). Monitoring was similarly conducted in the spring and summer of the 2008/9 hydrological year, and in the fall of the 2009/10 hydrological year. Animal location and activity were monitored with Lotek 3300LR GPS collars with activity sensors (Lotek Engineering, Newmarket, ON, Canada) in conjunction with IceTag pedometers (IceRobotics Ltd, Edinburgh, Scotland, UK) (Supplementary Fig. S2). In each of the six monitoring seasons, the devices were deployed on eight randomly selected cows in one paddock and transferred weekly to eight animals in another paddock until the completion of one cycle of the four paddocks. The order in which the paddocks were monitored was randomized at each cycle. The weight, physiological status (empty, lactating, early pregnancy, late pregnancy) and body condition score (Edmonson et al., 1989) of every cow in the paddock were recorded; the last of data on a 1–5 scale (1 = low, 5 = high). These measurements were recorded either when the equipment was fitted or when it was removed. Based on Ungar et al. (2011), the collars were programmed to store latitude, longitude, and elevation, as well as associated data comprising date, time, ambient temperature, and motion sensor counts at 5-min intervals. The pedometers provided step counts,

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model

ARTICLE IN PRESS

APPLAN-4370; No. of Pages 9

I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

3

Table 1 Description of the experimental paddocks and treatments. Treatments were high and moderate animal population densities, replicated over two paddocks. Cover percentages were obtained by analysis of an aerial photograph with ERDAS Imagine software. These values assisted division of the area into three vegetation types. Paddock

North South West East

Area (ha)

40.5 65.5 45.0 60.5

Cover (%)

Treatment

Shrub

Tree

Rock

Soil

31.4 31.4 26.5 33.7

54.0 54.8 62.2 47.4

5.0 4.9 4.5 6.2

9.0 8.8 6.7 12.3

High High Moderate Moderate

Animal Density (cow ha−1 )

0.55 0.55 0.33 0.33

leg activity levels, and leg position (upright or lying) summed over 1-min intervals. Data from the pedometers and GPS collars, i.e., motion sensor counts and distance travelled since the previous measurement, were merged and used to predict activity for every 5-min interval. Activities were classified as grazing, walking (without grazing), or resting (while lying or standing). The prediction equations were based on classification-tree analysis developed by Ungar et al. (2011) which yielded an overall misclassification rate of 10%. The observations for that study were conducted at the Hatal Experimental Farm, with the collars used in the present study. Records from entire 24-h periods, starting at midnight, were extracted for data analysis. After correcting for occasional equipment failures, this dataset comprised 221,207 records, each representing a 5-min interval, taken from 164 cow-measurement periods (Supplementary Table S1). The structure of the dataset did not indicate that the woody vegetation impeded the ability of the GPS collars to determine animal location. The dataset used to analyze spatial distribution was based only on activities classified as grazing; it comprised 69,122 GPS locations taken from 146 individual cows.

2.4. GIS mapping The following GIS thematic layers were created with ArcView 10.0 software (Supplementary Fig. S3):

1. Cow activity: This layer contains GPS locations associated with grazing activity. 2. Vegetation type: A polygon layer was produced by analyzing an aerial photograph with ERDAS Imagine software to distinguish between trees, shrubs, rocks, and bare ground; it was validated in the field using a GPS device (Supplementary Fig. S4). Three main vegetation types were designated according to tree cover (Table 1): (i) dense woodland (>80% coverage); (ii) open woodland (50–80% coverage); and (iii) garrigue (<50% coverage). Each polygon was inspected visually and ascribed an accessibility score based on the density of the woody vegetation cover and how easily a person could traverse the polygon on foot. The scores were: 0–easily traversed due to the absence of woody vegetation; 1–woody vegetation has a small impact on traversability; 2–shrubs and trees force a tortuous path but the polygon can be traversed easily; 3–shrubs and trees are sufficiently close to each other to make it somewhat difficult to traverse the polygon; 4–density of woody vegetation makes it difficult to traverse the polygon; 5–extremely difficult or impossible to traverse the polygon. 3. Topography: An elevation contour layer at 5-m resolution was produced from data provided by the national mapping agency (Survey of Israel), and raster layers for slope and aspect were derived from the elevation layer with the Spatial Analyst tool of the ArcView 10.0 software. 4. Management: This layer contains locations of watering troughs, supplementary feed bins, and paddock fence lines.

Number of cows

22 37 15 20

Vegetation type (%) Garrigue

Open woodland

Dense woodland

35.9 43.2 37.6 56.5

46.1 41.8 44.6 29.8

17.9 15.1 17.8 13.6

2.5. Data analysis 2.5.1. Statistical data analysis The diurnal activity pattern, i.e., the periods of time devoted to grazing, resting, or walking, during the course of a 24-h period, was calculated for each combination of cow, treatment, and season. Daily grazing time was subjected to analysis of variance (ANOVA) in which factors in the model were: season (spring, summer, fall), animal population density (high, moderate), cow parameters (age and physiological status), seasonal cycle (1, 2) as a random factor and paddock (east, west, north, south) as a random effect nested in animal population density. Significance of interactions between these factors was tested, and random interaction terms that did not improve model fit were removed. The Shapiro-Wilk test and the Bartlett test were applied to verify normal distribution and homogeneity of variance (P < 0.05), respectively. Following ANOVA, Student’s t-test was applied at P < 0.05 to detect significant differences within seasonal cycles. The coefficient of variation (CV) of the proportions of time spent grazing in each hour of the day was calculated and used to measure differences between the seasons in the diurnal pattern of grazing activity; it was calculated as the average of the proportion of each hour devoted to grazing, divided by its standard deviation. Spatial distribution analyses were based on animal locations at which activity was classified as grazing. For each cow these locations were linked to the various GIS layers (vegetation, topography, and management data), and the distance of each location from management factors (water, feed, fences) was calculated. Buffer zones of 30-m radius around the water troughs and supplementary feed stations were excluded from these calculations. ANOVA was applied to the values of accessibility of animal locations, and those of their slope and distance from management factors (water, feed, fences), based on the same factors. Statistical analyses were carried out with the JMP software, version 7.1 (SAS Institute, Cary, NC, USA).

2.5.2. Preference for vegetation type Preferences for or avoidance of vegetation types were assessed according to Jacobs’ Selectivity Index (JSI) (Jacobs, 1974; Henkin et al., 2012). The index was calculated at the animal level as:

JSIi = (ri −pi )/(ri + pi −2ri pi )

in which ri = proportion of all grazing locations in a given paddock that were located in vegetation type i; pi = proportion of paddock area that contained vegetation type i (1, 2, or 3 represents dense woodland, open woodland or garrigue, respectively). The JSI can range from −1 to +1, with JSI > 0 indicating preference, JSI < 0 indicating avoidance, and values at or near zero indicating random selection. The ANOVA model defined above (2.5.1) was used to analyze this index for each vegetation type, and a one-sample t-test was used to determine whether JSI differed from zero.

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model

ARTICLE IN PRESS

APPLAN-4370; No. of Pages 9

I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

4

Seasonal cycle 2 35

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

35

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

35

30 25 20 15 10 5 0

30 25 20 15 10 5 0

30 25 20 15 10 5 0

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

35

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

35

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

35

30 25 20 15 10 5 0

30 25 20 15 10 5 0

30 25 20 15 10 5 0 0 2 4 6 8 10 12 14 16 18 20 22

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

0 2 4 6 8 10 12 14 16 18 20 22

Proportion time Fall

1

Temperature (ᵒC)

Summer

Spring

Season

Hour in the day

Hour in the day

Fig. 1. Proportions of time allocated by cattle to grazing (blue), resting (green) and walking (red) activities in each hour of the day, as affected by seasonal cycle and season. Also shown are ambient temperature (black line) and daylight duration (white horizontal bar). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3. Results 3.1. Cattle activity The overall average times that cows spent grazing, resting, and walking were, respectively, 9.7 ± 0.4, 13.7 ± 0.4, and 0.6 ± 0.1 h/day. This allocation of time was influenced by seasonal conditions (Table 2): in spring, when high-quality herbaceous vegetation was available and most of the cows were lactating (Supplementary Table S1), the cattle spent more time grazing than they did in fall and summer: 11.7 ± 0.4, 9.0 ± 0.4, and 8.5 ± 0.3 h/day, respectively (P = 0.009). These values are averaged over two seasonal cycles and two animal population densities. These seasonal differences also were reflected in the daily activity patterns of the cattle (Fig. 1), which responded to seasonal variations in day length and temperature. During summer and fall, when temperatures were relatively high, most of the grazing was conducted during early morning (05:00–08:00 h) and late afternoon (15:00–19:00 h), thus limiting total grazing time. The CV of the hourly proportions of time spent grazing was higher in summer

Table 2 Mean (±SE) time (hours per day) allocated to the three activities (grazing, resting, walking) according to seasonal cycle, season and animal population density (AD). Significant differences (according to Student’s test at ␣ = 0.05) between grazing times within seasonal cycle, based on a model including both season and animal population density, are indicated by different letters: upper case for cycle 1; lower case for cycle 2. Seasonal cycle

Season

1

High Moderate Summer High Moderate High Fall Moderate High Spring Moderate Summer High Moderate High Fall Moderate

2

Spring

AD

Time in activity (h/day) Graze

Rest

Walk

11.9 ± 0.3 A 11.5 ± 0.6 A 9.4 ± 0.3 B 8.2 ± 0.4 C 9.2 ± 0.3 BCE 8.9 ± 0.5 BCE 12.3 ± 0.3 a 11.3 ± 0.3 b 7.9 ± 0.5 d 8.4 ± 0.5 cd 9.4 ± 0.3 c 8.4 ± 0.3 cd

11.7 ± 0.3 12.0 ± 0.6 13.9 ± 0.3 15.1 ± 0.4 14.3 ± 0.4 14.6 ± 0.5 11.5 ± 0.3 12.3 ± 0.3 15.4 ± 0.5 14.9 ± 0.5 14.0 ± 0.3 14.9 ± 0.3

0.5 ± 0.1 0.5 ± 0.1 0.7 ± 0.1 0.7 ± 0.0 0.5 ± 0.0 0.5 ± 0.1 0.3 ± 0.0 0.4 ± 0.0 0.6 ± 0.0 0.7 ± 0.0 0.5 ± 0.1 0.7 ± 0.1

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model APPLAN-4370; No. of Pages 9

ARTICLE IN PRESS I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

5

Fig. 2. Spatial distribution of grazing activity in the experimental paddocks according to seasonal cycle and season. Each point represents a 5-min period during which cow activity was classified as grazing.

and fall than it was in spring − 0.46, 0.59, and 0.56 in spring, summer, and fall, respectively (P = 0.003)–indicating wider variation in the level of grazing activity over the course of the day. In contrast, in spring, when temperatures were moderate and available forage was of higher quality, grazing was more evenly distributed through the day, with 50% of the daily grazing activity occurring around midday. In all seasons some grazing activity was detected also around midnight. Also animal population density influenced the amount of time allocated to grazing, but this effect depended on season and seasonal cycle (Table 2). Relative to those at the moderate population density, cattle at the high density grazed 72 ± 18 min/day longer in the summer of the relatively dry first seasonal cycle, and 59 ± 18 min/day longer during the spring of the second seasonal cycle − a cycle that followed a rainy winter; P = 0.04 for the combined effect of seasonal cycle × season × population density. However, no significant differences were detected in the other seasons. 3.2. Spatial distribution of the cattle in relation to vegetation types The spatial distribution of grazing in the various paddocks was not uniform, and was influenced by both vegetation type and sea-

son (Fig. 2). In spring, the cows spent more time grazing in the shrubby garrigue, which has the highest cover of herbaceous vegetation. In contrast, in summer and fall, when temperatures were highest and herbaceous forage was scarce, the cows spent most of their time in the dense woodland. The proportions of the day in which the cows grazed in the garrigue were 47.0, 28.2, and 21.8% during spring, summer and fall, respectively. This changing preference for vegetation types was reflected in Jacobs’ Selectivity Index (JSI), which showed the garrigue being preferred or neutral in the spring and avoided in the summer and fall, with values of 0.09 ± 0.04, −0.32 ± 0.04, and −0.47 ± 0.04 in spring, summer, and fall, respectively (P < 0.0001; Fig. 3, Supplementary Table S2). In contrast, during fall the cows spent relatively more time in the dense woodland and avoided the garrigue, with JSI = 0.36 ± 0.05 and −0.28 ± 0.06, respectively (P = 0.05). In the open woodland there were moderate seasonal fluctuations, with slight preference in summer and fall, and avoidance or neutral in spring (P = 0.003). Animal population density affected preferences for the various vegetation types. Under high density, the cows generally preferred to graze in the open woodland (P = 0.02; Fig. 3, Supplementary Table S2), whereas under moderate density, during the spring they showed preference or indifference for the garrigue and avoided the dense woodland (season × animal density interaction significant at

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model

ARTICLE IN PRESS

APPLAN-4370; No. of Pages 9

I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

6

Moderate

*** ***

0.6 0.4 0.2

*** ***

***

*** ***

BC B

C BC

b

c

c c

-0.2 -0.4 -0.6 -0.8 Summer

Fall

Spring

Summer

0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8

***

***

A C

ABC

BC

Spring

Summer

**

C

Fall

*** **

A

ab b

Spring

***

A C C

Spring

BC

B

Summer Seasonal cycle 1

a a

a a

Summer

Fall

***

abc

Spring

b

** ***

ab

a

bc

Summer

b

***

Fall

Seasonal cycle 2

Fig. 3. Effects of season and animal population density (high versus moderate) on grazing preferences, as expressed by Jacobs’ selectivity index, for the three main vegetation types: garrigue, dense woodland, open woodland. The data are index values ± SE. The index can range from −1 to +1, with positive values indicating preference and negative values indicating avoidance. Significant differences between means within seasonal cycles are indicated by different letters: capital and lowercase letters for the first and second seasonal cycle, respectively (Student’s t-test ␣ = 0.05). Significance of difference from zero according to one-sample t-test is shown above each set of axes (* P < 0.05, ** P < 0.01, *** P < 0.001).

P = 0.009). This trend was marginally stronger in the wetter, second seasonal cycle (seasonal cycle × population density interaction significant at P = 0.015 for dense woodland; P = 0.12 for garrigue). The degree to which animals displayed preference for or avoidance of the various vegetation types was weaker when viewed on a year-round basis (JSI range −0.27 to 0.16; Fig. 4) than when viewed seasonally, but the seasonal levels of preference and avoidance did not cancel each other completely. Notably, there was a consistent degree of avoidance of the garrigue, on the one hand, and a trend for preference shared by the two woody vegetation types, on the other hand. In four cases the preference/avoidance was stronger under the high population density condition: the JSI became significantly different from zero (Fig. 4). 3.3. Spatial distribution of the cattle in relation to terrain and management factors Terrain conditions affected the spatial distribution of the cows (Table 3; Supplementary Table S3). In the first seasonal cycle the cattle preferred to graze in the more accessible terrain (P = 0.03) and closer to the fences (P < 0.0001). This preference for the more accessible terrain was highly significant (P < 0.0001) in spring. In addition, the cows preferred (P = 0.06) to graze on more moder-

***

*

**

* Indicate difference from zero one A BC AB sample tC b a

0 -0.2

test (*P<0.1, **P<0.05,***P<0.01)

-0.4 -0.6 -0.8

Open Open Dense Garrigue Dense woodland woodland woodland woodland

Seasonal cycle11

c d

Fall

0.2

Garrigue

Seasonal cycle 2

*** ***

**

*** ***

0.4

***

A

Seasonal cycle 1

0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6 -0.8

Fall

Seasonal cycle 2

Seasonal cycle 1

Open woodland Jacobs selectivity index

c

0.0

Spring

Dense woodland Jacobs selectivity index

*** **

a A A

High Moderate

High 0.6

Jacob selectivity index

Garrigue Jacobs selectivity index

High

Seasonal cycle 2

Fig. 4. Effects of animal population density (high versus moderate) and vegetation type (garrigue, dense woodland, open woodland) on grazing preferences, as expressed by Jacobs’ selectivity index, on a year-round basis. The data are index values ± SE. The index can range from −1 to +1, with positive values indicating preference and negative values indicating avoidance. Significant differences between means within vegetation type are indicated by different letters: capital letters for dense woodland and lowercase letters for garrigue (Student’s t-test ␣ = 0.05). Significance of difference from zero according to one sample t-test is shown above the set of axes (* P < 0.05, ** P < 0.01, *** P < 0.001).

ate slopes during the spring than during the fall and summer. The cattle under high population density grazed more on the steeper slopes during the first summer than did those under moderate density during the second spring: 17.9 and 9.7%, respectively (seasonal cycle × season × population density interaction significant at P < 0.0001). In general, under moderate population density grazing tended to be in the more accessible areas, particularly in the spring (season × population density interaction P = 0.09). The cattle tended to graze closer to the water troughs in fall than in spring (502 ± 27 and 751 ± 23 m, respectively; P = 0.07), and closest, at 455 ± 5 m (P = 0.002), in the first fall. Similarly, in the first seasonal cycle they grazed closer to the supplementary feeding stations during fall than during summer: 435 ± 40 and 671 ± 35 m, respectively (seasonal cycle × season interaction significant at P = 0.02). Animal population density affected grazing distances from fences, water troughs, and supplementary feed stations (Table 3; Supplementary Table S3). In all seasons cattle under high density grazed 25 ± 1.5 m further from the fences than those under moderate density (P = 0.05), and this difference was particularly pronounced in fall (season × population density interaction significant at P = 0.02). The high-density cows also grazed 110 ± 12 m further from the water troughs than the moderate-density ones, although there was a significant (P < 0.0001) season × population density interaction. Furthermore, the former cows grazed 163 ± 12 m further away from the supplementary feed stations than the latter, and this effect was stronger in the dryer first seasonal cycle (seasonal cycle × population density interaction significant at P = 0.02). 4. Discussion 4.1. Utilization of woody mediterranean landscapes by cattle Free-ranging ruminants face a variety of foraging decisions, including site selection within a given landscape, plant selection within a given site, and allocation of time to foraging (Dumont and Gordon, 2003; Marquardt et al., 2010). These decisions are influenced by the availability and quality of vegetation, terrain factors, and environmental factors (Bailey et al., 1996; Moser et al., 2006; Kaufmann et al., 2013). In the present study we examined the foraging decisions of cattle grazing in Mediterranean oak woodland, as reflected in their activity and spatial distribution patterns.

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model

ARTICLE IN PRESS

APPLAN-4370; No. of Pages 9

I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

7

Table 3 Various measures of the spatial distribution of cows according to seasonal cycle, season, and animal population density (AD). Measures are accessibility (scale 1–5), slope (◦ ), and distances (m) from water trough, supplementary feeder, and fences. Values are mean ± SE. Significant differences (according to Student’s test at ␣ = 0.05) within seasonal cycle, based on a model including both season and animal population density, are indicated by different letters: upper case for cycle 1; lower case for cycle 2. Seasonal cycle

Season

AD

1

Spring

High Moderate High Moderate High Moderate High Moderate High Moderate High Moderate

Summer Fall 2

Slope (◦ )

Accessibility (1–5)

Spring Summer Fall

Distance (m) from Water

3.1 ± 0.1 2.8 ± 0.1 3.3 ± 0.1 2.9 ± 0.1 3.4 ± 0.1 2.9 ± 0.1 3.4 ± 0.0 2.6 ± 0.1 3.4 ± 0.1 3.0 ± 0.2 3.4 ± 0.1 3.1 ± 0.2

BC E AB D A CD a c a b a b

13.8 ± 0.7 12.8 ± 1.5 17.9 ± 0.5 14.5 ± 0.8 12 ± 0.3 15.3 ± 1.5 14.2 ± 0.5 9.7 ± 0.8 14.7 ± 0.8 15.8 ± 1.4 14.4 ± 0.7 15.7 ± 1.9

It is by no means obvious that such a landscape is suitable for cattle production. Firstly, cattle are considered to be natural herbaceous grazers (Papachristou et al., 2005) but the landscape presents only limited availability of herbaceous biomass. Secondly, to the extent that cattle consume woody vegetation, its nutritional quality is poor (Perevolotsky, 1994; Henkin et al., 2005). Nevertheless, we found that the cattle spent 66% of their total grazing time, which averaged 9.7 ± 0.7 h/day, among the woody vegetation, in either open or dense woodland, which together account for 56% of the total area. The daily grazing times that were measured in the various seasons and for both population densities (Table 2) were comparable with those reported for herbaceous rangeland systems (Di Marco and Aello, 2001; Zuo and Miller-Goodman, 2004; Henkin et al., 2012). Furthermore, analysis of the daily activity pattern showed that most of the grazing was concentrated into the early morning and late afternoon, as found in other woody systems (Roath and Krueger, 1982; Parsons et al., 2003) and in rangelands dominated by herbaceous vegetation (Celaya et al., 2008; Henkin et al., 2012). In addition, nocturnal foraging activity was recorded around midnight, as found in other studies (Brosh et al., 2006b; Ganskopp and Bohnert, 2006; Henkin et al., 2012). In all, our present findings suggest that the behavior of the cattle and their daily allocation of time to grazing in the Mediterranean oak woodland were quite similar to those observed in grasslands, despite the large differences in the structure and composition of the vegetation. 4.2. Seasonal effects on grazing activity Various studies have found contrasting responses of cattle to reductions in the quantity and quality of forage; responses that include increases (Arnold, 1981; Di Marco and Aello, 2001), and decreases (Schlecht et al., 2006; Henkin et al., 2012) in daily grazing time. As shown by Brosh et al. (2006a) at our study site, the Mediterranean woodland provided high-quality herbaceous forage during 3–4 months in late winter and spring, whereas during summer and fall the cattle depended on the woody vegetation and supplementary feed. Consequently, the time devoted to grazing activity varied between seasons: in spring the high-quality herbaceous vegetation was relatively abundant and, as most of the cows were lactating, nutrient requirements were relatively high (Aharoni et al., 2004). Furthermore, consumption beyond immediate requirements builds energy reserves that can be used when high-quality forage becomes a limiting factor (Brosh et al., 2006b). Together these factors resulted in a springtime increase of up to 3 h in daily grazing time compared with the summer and fall. In the dry seasons, cattle consumed woody vegetation, supplementary feed, and also acorns when available in fall. These low-quality, woody diet components also are rich in secondary

C E A BC D B a b a a a a

796 ± 60 706 ± 69 797 ± 26 742 ± 25 554 ± 40 356 ± 59 651 ± 53 845 ± 46 754 ± 16 676 ± 41 636 ± 66 504 ± 22

Feeder B A BC C D E b a ab b bc c

Fence

790 ± 26 552 ± 72 534 ± 37 335 ± 60

A B B C

741 ± 17 633 ± 47 621 ± 65 491 ± 22

a b b c

78 ± 6 52 ± 6 86 ± 4 75 ± 3 88 ± 4 45 ± 3 89 ± 6 71 ± 4 101 ± 5 75 ± 3 102 ± 6 72 ± 5

AB C A B A C a b a b a b

compounds such as tannins, that limit food intake (Rogosic et al., 2008) and therefore have a strong effect on grazing time. Moreover, the supplementary feed that compensates for protein deficiency during the dry season can contribute to a reduction in grazing time. The decline in grazing time in summer and fall was also reflected in the daily pattern of grazing activity, which was affected by length of daylight and by ambient temperature. In the summer, grazing started earlier and also was more restricted around mid-day (11:00–15:00 h) when temperatures were high (average 32.2 ◦ C), compared with those in spring (average 24.4 ◦ C). A temperature of 25 ◦ C is considered to be the threshold temperature above which heat stress causes a decrease in food consumption (Hahn, 1999; Taweel et al., 2006). 4.3. Seasonal effects on spatial distribution of grazing activity The type, quality, and quantity of forage play central roles in determining the spatial distribution of cattle (Kie and Boroski, 1996; Ganskopp and Bohnert, 2006; Ganskopp and Bohnert, 2009). Therefore, changes in forage characteristics with the passage of the seasons form a main driver of changes in the spatial distribution of cattle (Parsons et al., 2003). Because the present study was conducted in Mediterranean woodland characterized by a wide range of vegetation types, grazing distribution varied throughout the year. These vegetation types vary in species composition and structure and thereby influence forage choices; also, they vary in micro-environmental factors, such as shade. Because herbaceous vegetation is generally found in more open areas, during spring the cattle increased the proportion of their grazing time in the garrigue vegetation from 22% in fall to 50% in spring, and avoided the dense woodland, as shown by Jacobs’ selectivity index. In spring, forage is more available and is of better nutritional quality, temperatures are moderate, and most of the cows are suckling their calves. In contrast, during summer and fall, when amount and quality of forage are relatively low and ambient temperatures are high, cows prefer to graze vegetation with greater cover of shading trees. This provides them with both food and shade and, in fall, also acorns. The fact that cattle prefer different types of woodland vegetation in different seasons should improve year-round utilization of the available vegetation (Putfarken et al., 2008), and thereby contribute to sustainable management of the woodland. According to optimal foraging theory, herbivores are expected to reduce their energy expenditure while maximizing energy benefits (MacArthur and Pianka, 1966). Thus, they will move to where there is less difficult topography, choosing relatively moderate slopes and easier trails (Ganskopp et al., 2000). Because in spring the Mediterranean woodland contains abundant high-quality herbaceous forage, the cattle preferred to graze in less difficult terrain,

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model APPLAN-4370; No. of Pages 9

ARTICLE IN PRESS I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

8

thereby investing less energy in searching for feed. In contrast, in summer and fall, when there was a shortage of forage, the cows grazed in areas with more restricted accessibility and with steeper slopes, which were not exploited earlier in the year. Therefore, in the latter seasons the cattle increased their foraging range in search of available forage (Kie and Boroski, 1996), as manifested in their grazing at greater distances from the paddock fences (Owens et al., 1991). However, because in summer and, especially, during fall there is a lack of forage and a greater need of water, the cows grazed about 250 m closer to the water trough and supplementary feeding station. A similar trend has been found in other studies, in which cattle grazed closer to the water source as temperatures increased at the end of the dry season (Parsons et al., 2003; Delcurto et al., 2005). Thus, judicious placement of the water and feed stations can increase uniformity of grazing and improve utilization of the range. 4.4. Effect of animal population density on grazing activity and spatial distribution Effects of animal population density on grazing activity in Mediterranean woodland scarcely have been studied. In the present study we found that in some seasons the cattle in the highdensity paddocks grazed about 1 h/day longer than those in the moderate-density ones. Previous studies showed that forage availability decreased with increasing animal population density (Brosh et al., 2006b) and that diet quality was lower (Cornelissen and Vulink, 2015). We suggest that in the woodlands, under the higherdensity condition, additional grazing activity served to compensate for lower availability of high-quality forage. Cows in the moderate-density paddocks did not have to extend their home range in searching for forage to the same extent as those in the high-density one. In spring, grazing was concentrated in the most accessible terrains and on the more moderate slopes, but in the high-density paddock the cattle showed greater utilization of the woody vegetation types and extended their foraging range, as reflected in their grazing further from fences, water troughs, and supplementary feed. This finding has management implications with regard to effects on the landscape: the degree to which animals will penetrate and exploit the more densely wooded vegetation formations depends on animal population density, and a high density may be required to achieve goals such as reduced fire hazard. A key question, beyond our present scope, concerns the consequences in terms of animal production, health and welfare. 5. Conclusions The GPS and GIS methodologies that we applied serve as a reliable tool for monitoring cattle behavior during the various seasons, even in a dense woodland formation. Because of seasonal changes in vegetation production and quality, we observed differential behavior of the cattle: although they are natural herbaceous grazers, they utilized and even preferred the woody vegetation in summer and fall, when the quality and quantity of the herbaceous vegetation were low. Although the shift away from the herbaceousrich garrigue and towards the two woody vegetation types was still detectable on the annual time scale, the shifting seasonal preference patterns balanced out to a large extent. Animal population density influenced cattle behavior, but this effect was season- and year-dependent. Broadly, at the higher population density, cattle allocated more time to grazing, and showed stronger preferences for the woody vegetation and the less favorable areas. This should be considered in management planning. The present study provides basic data essential for optimal management planning that takes the landscape and season into account. As a general rule, sustainable management of cattle grazing in

woodland must be based on the various ecological and landscape factors, in addition to productivity factors associated with grazing as a source of livelihood for herders. Such a management plan would lead to optimal utilization of woodland landscapes without damaging them, while increasing their productivity and the performance of the cattle herds. In this way, it should be possible to minimize the conflicts between herders and organizations engaging in nature conservation, and to enable application of cattle grazing as an efficient, multi-purpose management tool for managing open landscapes. Acknowledgements The authors thank the herd owners, Yigal and Chaim Haike, for their exceptional cooperation in conducting this study; Yehuda Yehuda for technical assistance in conducting the field work; and Hillary Voet for statistical consultations. The study was funded by grants from the Israel Rangeland Advisory Board, Northern R&D, and the Jewish National Fund. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.applanim.2016. 11.015. References Aharoni, Y., Brosh, A., Orlov, A., Shargal, E., Gutman, A., 2004. Measurements of energy balance of grazing beef cows on Mediterranean pasture, the effects of stocking rate and season − 1. Digesta kinetics: faecal output and digestible dry matter intake. Livest. Prod. Sci. 90, 89–100. Arnold, G.W., 1981. Grazing behaviour. In: F.H.W, M. (Ed.), Grazing Animals. , pp. 79–104. Bailey, D.W., Gross, J.E., Laca, E.A., Rittenhouse, L.R., Coughenour, M.B., Swift, D.M., Sims, P.L., 1996. Mechanisms that result in large herbivore grazing distribution patterns. J. Range Manage. 49, 386–400. Bailey, D.W., Welling, G.R., Miller, E.T., 2001. Cattle use of foothills rangeland near dehydrated molasses supplement. J. Range Manage. 54, 338–347. Bailey, D.W., VanWagoner, H.C., Weinmeister, R., 2006. Individual animal selection has the potential to improve uniformity of grazing on foothill rangeland. Rangeland Ecol. Manage. 59, 351–358. Bailey, D.W., 2004. Management strategies for optimal grazing distribution and use of arid rangelands. J. Anim. Sci. 82 (E-Suppl), E147–153. Bailey, D.W., 2005. Identification and creation of optimum habitat conditions for livestock. Rangeland Ecol. Manage. 58, 109–118. Brosh, A., Henkin, Z., Orlov, A., Aharoni, Y., 2006a. Diet composition and energy balance of cows grazing on Mediterranean woodland. Livest. Sci. 102, 11–22. Brosh, A., Henkin, Z., Ungar, E.D., Dolev, A., Orlov, A., Yehuda, Y., Aharoni, Y., 2006b. Energy cost of cows’ grazing activity: use of the heart rate method and the Global Positioning System for direct field estimation. J. Anim. Sci. 84, 1951–1967. Celaya, R., Benavides, R., Garcia, U., Ferreira, L.M.M., Ferre, I., Martinez, A., Ortega-Mora, L.M., Osoro, K., 2008. Grazing behaviour and performance of lactating suckler cows, ewes and goats on partially improved heathlands. Animal 2, 1818–1831. Cornelissen, P., Vulink, J.T., 2015. Density-dependent diet selection and body condition of cattle and horses in heterogeneous landscapes. Appl. Anim. Behav. Sci. 163, 28–38. Delcurto, T., Porath, M., Parsons, C.T., Morrison, J.A., 2005. Management strategies for sustainable beef cattle grazing on forested rangelands in the Pacific northwest. Rangeland Ecol. Manage. 58, 119–127. Di Marco, O.N., Aello, M.S., 2001. Energy expenditure due to forage intake and walking of grazing cattle. Arquivo Brasileiro De Medicina Veterinaria E Zootecnia 53, 105–110. Dumont, B., Gordon, I.J., 2003. Diet selection and intake within sites and across landscapes. In: ‘t Mannetje, L., R.-A, L., Sandoval-Castro, C.A., Ku-Vera, J.C. (Eds.), The VI International Symposium on the Nutrition of Herbivores. Universidad Autónoma de Yucatán, Mérida, Yucatán, Mexico, pp. 175–194. Edmonson, A.J., Lean, I.J., Weaver, L.D., Farver, T., Webster, G., 1989. A body condition scoring chart for Holstein dairey-cows. J. Dairy Sci. 72, 68–78. Ganskopp, D., Bohnert, D., 2006. Do pasture-scale nutritional patterns affect cattle distribution on rangelands? Rangeland Ecol. Manage. 59, 189–196. Ganskopp, D.C., Bohnert, D.W., 2009. Landscape nutritional patterns and cattle distribution in rangeland pastures. Appl. Anim. Behav. Sci. 116, 110–119. Ganskopp, D., Cruz, R., Johnson, D.E., 2000. Least-effort pathways?: a GIS analysis of livestock trails in rugged terrain. Appl. Anim. Behav. Sci. 68, 179–190.

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015

G Model APPLAN-4370; No. of Pages 9

ARTICLE IN PRESS I. Schoenbaum et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx

Ganskopp, D., 2001. Manipulating cattle distribution with salt and water in large arid-land pastures: a GPS/GIS assessment. Appl. Anim. Behav. Sci. 73, 251–262. Gutman, M., Henkin, Z., Holzer, Z., Noy-Meir, I., Seligman, N.G., 2000. A case study of beef-cattle grazing in a Mediterranean-type woodland. Agrofor. Syst. 48, 119–140. Hahn, G.L., 1999. Dynamic responses of cattle to thermal heat loads. J. Anim. Sci. 77, 10–20. Henkin, Z., Gutman, M., Aharon, H., Perevolotsky, A., Ungar, E.D., Seligman, N.G., 2005. Suitability of Mediterranean oak woodland for beef herd husbandry. Agric. Ecosyst. Environ. 109, 255–261. Henkin, Z., Ungar, E.D., Dolev, A., 2012. Foraging behaviour of beef cattle in the hilly terrain of a Mediterranean grassland. Rangeland J. 34, 163–172. Henkin, Z., 2011. Cattle grazing and vegetation managemant for multiple use of mediterranean shrubland in Israel. Israel J. Ecol. Evol. 57, 43–51. Jacobs, J., 1974. Quantitative measurement of food selection- modification of forage ratio and ivlevs electivity index. Oecologia 14, 413–417. Kaufmann, J., Bork, E.W., Blenis, P.V., Alexander, M.J., 2013. Cattle habitat selection and associated habitat characteristics under free-range grazing within heterogeneous Montane rangelands of Alberta. Appl. Anim. Behav. Sci. 146, 1–10. Kie, J.G., Boroski, B.B., 1996. Cattle distribution, habitats, and diets in the Sierra Nevada of California. J. Range Manage. 49, 482–488. MacArthur, R.H., Pianka, E.R., 1966. On optimal use of a patchy environment. Am. Nat. 100, 603-+. Marquardt, S., Beck, S.G., Encinas, F.D., Alzerreca, A.H., Kreuzer, M., Mayer, A.C., 2010. Plant species selection by free-ranging cattle in southern Bolivian tropical montane forests. J. Trop. Ecol. 26, 583–593. Moser, B., Schutz, M., Hindenlang, K.E., 2006. Importance of alternative food resources for browsing by roe deer on deciduous trees: the role of food availability and species quality. For. Ecol. Manage. 226, 248–255. Owens, M.K., Launchbaugh, K.L., Holloway, J.W., 1991. Pasture characteristics affecting spatial distribution of utilization by cattle in mixed brush communities. J. Range Manage. 44, 118–123. Papachristou, T.G., Platis, P.D., Nastis, A.S., 2005. Foraging behaviour of cattle and goats in oak forest stands of varying coppicing age in Northern Greece. Small Rumin.t Res. 59, 181–189. Parsons, C.T., Momont, P.A., DelCurto, T., McInnis, M., Porath, M.L., 2003. Cattle distribution patterns and vegetation use in mountain riparian areas. J. Range Manage. 56, 334–341.

9

Perevolotsky, A., Seligman, N.G., 1998. Role of grazing in Mediterranean rangeland ecosystems − Inversion of a paradigm. Bioscience 48, 1007–1017. Perevolotsky, A., Brosh, A., Ehrlich, O., Gutman, M., Henkin, Z., Holzer, Z., 1993. Nutritional − value of common oak (Quercus Calliprinos) browse as fooder for goats − experimental results in ecological prespective. Small Rumin. Res. 11, 95–106. Perevolotsky, A., 1994. Tannins in mediterranean woodland species − lack of response to browsing and thinning. Oikos 71, 333–340. Putfarken, D., Dengler, J., Lehmann, S., Haerdtle, W., 2008. Site use of grazing cattle and sheep in a large-scale pasture landscape: a GPS/GIS assessment. Appl. Anim. Behav. Sci. 111, 54–67. Roath, L.R., Krueger, W.C., 1982. Cattle grazing and behavior on a forested range. J. Range Manage. 35, 332–338. Rogosic, J., Estell, R.E., Ivankovic, S., Kezic, J., Razov, J., 2008. Potential mechanisms to increase shrub intake and performance of small ruminants in mediterranean shrubby ecosystems. Small Rumin. Res. 74, 1–15. Schlecht, E., Hulsebusch, C., Mahler, F., Becker, K., 2004. The use of differentially corrected global positioning system to monitor activities of cattle at pasture. Appl. Anim. Behav. Sci. 85, 185–202. Schlecht, E., Hiernaux, P., Kadaoure, I., Hulsebusch, C., Mahler, F., 2006. 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. 113, 226–242. Taweel, H.Z., Tas, B.M., Smit, H.J., Tamminga, S., Elgersma, A., 2006. A note on eating behaviour of dairy cows at different stocking systems- diurnal rhythm and effects of ambient temperature. Appl. Anim. Behav. Sci. 98, 315–322. Turner, L.W., Udal, M.C., Larson, B.T., Shearer, S.A., 2000. Monitoring cattle behavior and pasture use with GPS and GIS. Can. J. Anim. Sci. 80, 405–413. Ungar, E.D., Schoenbaum, I., Henkin, Z., Dolev, A., Yehuda, Y., Brosh, A., 2011. Inference of the activity timeline of cattle foraging on a mediterranean Woodland using GPS and pedometry. Sensors 11, 362–383. Zuo, H.T., Miller-Goodman, M.S., 2003. An index for description of landscape use by cattle. J. Range Manage. 56, 146–151. Zuo, H.T., Miller-Goodman, M.S., 2004. Landscape use by cattle affected by pasture developments and season. J. Range Manage. 57, 426–434.

Please cite this article in press as: Schoenbaum, I., et al., Spatial and temporal activity of cattle grazing in Mediterranean oak woodland. Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.11.015