Relationship between terrain use and performance of beef cows grazing foothill rangeland 1

Relationship between terrain use and performance of beef cows grazing foothill rangeland 1 D. W. Bailey,* 2, D. D. Kress, D. C. Anderson*, D. L. Boss*, and E. T. Miller* *Northern Agricultural Research Center, Montana State University, Havre 59501 and Montana State University, Bozeman 59717 ABSTRACT: A study was conducted on foothills rangeland to determine whether grazing patterns differed among cow breeds and to determine whether there was a relationship between individual grazing patterns and performance of beef cows. Hereford (HH), Tarentaise (TT), ³ ₄ Hereford ¹ ₄ Tarentaise (3H1T), ¹ ₂ Hereford ¹ ₂ Tarentaise (HT), and ¹ ₄ Hereford ³ ₄ Tarentaise (1H3T) cows were observed during the summers of (n = 183) and (n = 159). Locations of individual cows were recorded two to three times per week during 1.5- to 2.5-h periods in the morning by observers on horseback. Statistical models included cow breed, age, and nursing status. Data from each year were analyzed separately. During and, nonlactating cows were located at greater (P < 0.05) vertical distances from water than lactating cows. In, nonlactating cows used steeper (P < 0.05) slopes than lactating cows. However, nonlactating cows did not travel as far horizontally from water (P < 0.05) as lactating cows in. Younger cows (3 yr) traveled further (P < 0.05) from water both vertically and horizontally than older cows (5+ yr) in, but not during. Tarentaise and 1H3T cows were observed at greater (P < 0.05) vertical distances from water than HH cows during both years of the study. During, TT and 1H3T cows used steeper (P < 0.05) slopes than HH cows. Using residual correlations, there were no consistent relationships between topographic aspects of individual grazing locations and cow weight, height, and body condition score. In, cows with earlier calving dates and correspondingly heavier calf weaning weights used areas that had greater vertical distances to water; however, in there were no relationships (P > 0.05) of calving date and weaning weight with cow location. During both years, pregnant and nonpregnant cows used terrain similarly (P > 0.05), which suggests that cow reproductive performance was not related to terrain use. Grazing patterns in foothills rangeland varied among cow breeds. Performance of cows that used more rugged topography was similar to cows using gentler terrain. Key Words: Cattle, Distribution, Grazing, Lactation, Performance, Weaning Weight 2001 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2001. 79:1883 1891 Introduction The distribution of grazing livestock is an important facet of proper rangeland management (Vallentine, 1990). If cattle are not dispersed evenly, preferred areas are overgrazed and other areas are underused. Developing water, salting, supplementing, herding, and fencing have been successfully used to modify livestock grazing distribution (Skovlin, 1957; Cook, 1966; Bailey and Welling, 1999). Another suggested approach for 1 This research was partially supported by the USDA Western Regional SARE program. This manuscript has been assigned Journal Series No. 2000-36, Montana Agric. Exp. Sta., Montana State Univ., Bozeman. 2 Correspondence: Star Route 36 Box 43 (phone: 406-265-6115; fax: 406-265-8288; E-mail: dbailey@montana.edu). Received May 4, 2000. Accepted April 9, 2001. improving livestock grazing distribution is to cull animals with undesirable grazing patterns and to select animals with desirable grazing patterns (Roath and Kruegar, 1982; Howery et al., 1996, ). Over time, the selected herd might use rugged and varying topography more evenly. Howery et al. (1996) found that some cows preferred upland areas in an Idaho mountain pasture whereas others spent more time in riparian areas. Livestock producers with extensive rangeland pastures could use breeds that travel further from water or use steeper slopes to improve uniformity of grazing. Herbel and Nelson (1966) found that Santa Gertrudis cows traveled further than Herefords. An important consideration when evaluating the feasibility of a selection program for grazing distribution is the relationship between animal performance and spatial grazing patterns. If cows that prefer gentle terrain near water raise calves with heavier weaning weights than cows using steeper terrain and areas farther from water, then a selection program for improved 1883

1884 Bailey et al. grazing distribution may not be practical. Researchers have spent little, if any, time evaluating the relationship between where cows graze in rugged topography and their corresponding performance. The objectives of this study were to determine whether there were differences among cattle with Hereford, Tarentaise, and Hereford Tarentaise breeding and to examine the phenotypic relationships between where cows graze on foothills rangeland and their performance. Study Site Materials and Methods The study was conducted in the Bear s Paw Mountains 30 km south of Havre, Montana. Cattle were observed in three pastures: AI, Back, and Hay. Topographic relief in the AI pasture (254 ha) varies from 1,164 to 1,305 m with slopes of 0 to 53%. In the Back pasture (337 ha), relief varies from 1,184 to 1,402 m with slopes of 4 to 55%. In the Hay pasture (134 ha), relief varies from 1,201 to 1,323 m with slopes of 0 to 49%. In all three pastures, lower elevations with gentle slopes were dominated by Kentucky bluegrass (Poa pratensis L.), and very steep slopes (> 40%) were dominated by rough fescue (Festuca scabrella Torr.). Kentucky bluegrass, rough fescue, bluebunch wheatgrass (Pseudoregnaria spicata [Pursh] A Love), and Idaho fescue (Festuca idahoensis Elmer) were dominant in the majority of areas in each pasture. In all pastures, less than 15% of the pasture contained any trees such as Ponderosa pine (Pinus ponderosa Laws) and aspen (Populus tremuloides Michx.). The 15-yr average annual precipitation at the study site is 41 cm. Grass production was roughly 1,600 and 1,500 kg/ha in and, respectively. Grasses compose 80 to 85% of the total vegetative standing crop during most years. Cattle Cattle used in this study resulted from a crossbreeding experiment (Kress et al. 1996) with Hereford and Tarentaise cattle. Hereford, Hereford Tarentaise reciprocal cross, and Tarentaise sires were mated to Hereford, Hereford Tarentaise reciprocal cross, and Tarentaise females to produce five cow types: Hereford (HH), ³ ₄ Hereford ¹ ₄ Tarentaise (3H1T), ¹ ₂ Hereford ¹ ₂ Tarentaise (HT), ¹ ₄ Hereford ³ ₄ Tarentaise (1H3T), and Tarentaise (TT). Cattle were familiar with the study area and had annually grazed in study pastures sometime during the grazing season (late May to early January) since birth. The same animals were used in both and grazing seasons, with the exception of 23 cows that either died or were culled during the winter of. During, 161 lactating cows and 22 nonlactating cows were observed. During, 151 lactating cows and 9 nonlactating cows were observed. Cow age varied from 3 to 9 yr, but most of the cows Table 1. Distribution of cows by age and lactation status during and Trait Age, yr 3 55 4 57 53 5 35 47 6 18 29 7 12 16 8 6 9 9 6 Lactation status Lactating 161 151 Nonlactating 22 9 were young (Table 1). Cows began calving in mid-march and ended in late April (45-d breeding season). All cows included in the study grazed together in the same pastures. From early June to mid-july cattle grazed the Hay pasture in and the AI pasture in. Cattle grazed the Back pasture from mid-july to early September in both and. In the Back pasture during, an additional 40 Hereford Tarentaise cows grazed with the study cattle, but they were sold in June and were excluded from the analyses. During, 54 crossbred 2-yr-old cows grazed with the study animals (daughters of study animals sired by Angus, Charolais, Piedmontese, and Salers bulls). These animals were excluded from the analyses because they were only observed for 1 yr and their breeding was different from that of the study animals. Cows may use areas of pastures similar to those used by their dams (Howery et al., ). If the animals used in the study contained both dams and daughter, the study potentially could be confounded. However, almost all of the dams of the cows used in this study had been culled prior to. Any data collected from either dams or daughters of the cows used in this study were excluded from the analyses. Cattle Performance Weight, hip height, and body condition scores (BCS) were recorded for all cows at weaning (October 1). Body condition scores were the average of two observers using a 1-to-9 scale (1 = emaciated and 9 = obese). In addition, cows were rectally palpated to determine pregnancy status. Calves were weighed at weaning (October 1). Weaning weights were adjusted to 205 d using birth-to-weaning ADG. Grazing Location Observations In this study, slope and horizontal and vertical distance to water characterized where cattle were observed. These abiotic factors are primary determinants of grazing distribution patterns observed at larger scales (Senft et al., 1987; Bailey et al., 1996). Areas

Terrain use and cow performance 1885 that are located on steeper slopes typically receive less use than those on gentle slopes (Mueggler, 1965), and locations that are located at further (horizontal) distances from water also receive less use than areas near water (Valentine, 1947). Roath and Kruegar (1982) suggested that vertical distance to water was one of the most important determinants of livestock grazing patterns in mountain rangeland. Topographic maps of each pasture were subdivided into 1- to 7-ha units based on slope, elevation, aspect, and distance to water. Observers were trained to recognize the boundaries of all subunits within each pasture. Two to four observers on horseback rode pastures during a 1.5- to 2.5-h period each day and attempted to record the location and activity of every cow in the pasture. Observers rode close enough to each cow to observe her identification number from a plastic ear tag or a firebrand on the animal s hip. Observers recorded the activity and the pasture unit in which the animal was located. The goal of these observations was to obtain a scan sample of individual animal locations of the entire herd. Ideally, scan samples should be instantaneous (Lehner, 1979). However, individually identifying and observing over 160 animals instantaneously on extensive foothills rangeland pastures was not feasible. Overall, observers recorded the locations of 87% of the animals in the herd during an observation period (Table 2). Cattle were observed at least 15 times in each pasture (two to three times per week). Most of the observations were collected during the early morning, just after sunrise (Table 2). Location of cattle during the early morning hours is a good predictor of where cattle spend most of their time grazing (Low et al., 1981; Bailey et al., 1990). Morning observations began between 0530 and 0630 and ended before 0945. During the morning observations most of the cattle were actively grazing. Afternoon observations were usually collected between 1600 and 1930. Most cattle were near water and were not grazing during the afternoon observations. Table 2. Number of observation periods and mean number of times cows were observed in study pastures Year and pasture Timing a Observation periods Observations/cow Hay am 13 11.1 All 18 14.8 Back am 13 10.0 All 15 13.4 AI am 10 8 All 17 15.2 Back am 12 10.4 All 17 14.9 a Most observations were collected in the early morning (am). The combination of morning and afternoon observation is identified as All. Average slope, elevation, and distance to water (horizontal and vertical) were determined for each pasture subunit. Slopes were determined from topographic maps and verified with inclinometers. Horizontal distance to water was measured directly from topographic maps. For most subunits, vertical distance to water was determined by subtracting subunit elevation from the elevation to the nearest water. After leaving water, cattle had to climb up and down slopes (over a ridge) to reach some subunits. For these subunits, vertical distance to water was calculated as the difference in elevation from the lowest ridge crossing and the nearest water plus the difference in elevation between the ridge crossing and the subunit. For each cow, all location data collected in a pasture during a grazing season were pooled and used to determine the average slope, horizontal distance to water, and vertical distance to water of observed cow locations. These mean values were weighted averages of slopes and distances to water (horizontal and vertical) of the subunits where each cow was observed. For example, if a cow was observed four times on a subunit with a slope of 10% and six times on a subunit with a slope of 20%, average slope use would be 16%. The purpose for calculating these averages was to characterize the type of terrain at which cows were observed. Thus, cows using gentler terrain near water would have lower average slope use and smaller distances (horizontal and vertical) to water than cows using steeper terrain far from water. Although data were collected during both morning and afternoon periods, only data from morning observations are presented. Values from all observations (morning and afternoon) were highly correlated (r > 0.7). Using only morning observational data, mean values from each of the two pastures grazed during a season ( or ) were then averaged to give values of the average slope use, horizontal distance to water, and vertical distance to water for each cow for the year. The purpose for averaging data from the two pastures was to provide a more accurate representation of the grazing location of individual cows during the majority of the grazing season. Animal responses evaluated from data collected at weaning (October) should be more closely related to grazing locations evaluated over a 12- wk period (both pastures) than over a 6-wk period (one pasture). Data from each pasture are also presented to provide a measure of the variability among pastures. Statistical Analyses Location data from each grazing season were analyzed separately. Three dependent variables were used to characterize average terrain use of each cow: mean slope use, mean observed horizontal distance from water, and mean observed vertical distance from water. These values presented in these analyses were calculated from morning observations. Other dependent variables analyzed with the location data were cow

1886 weight, height, and BCS at weaning. Only the performance data that were associated with the year of location observations ( or ) were used in the analyses. Including information from other years would improve the characterization of performance for older cows. However, the majority of the cows were either 3 or 4 yr of age at the beginning of the study (Table 1), and additional performance data are limited. Including all performance data would have complicated the analysis because the variability records among cows would have to be included in the analyses. Most importantly, behavioral processes during grazing are hierarchical in time and space (Senft et al., 1987; Bailey et al. 1996), and observed behaviors should be examined at the temporal and spatial scales where responses are most likely to be measured (O Neill et al., 1986). In this study, observed grazing patterns during the grazing season corresponded most closely to the weights and other performance information collected at the end of the grazing season. The five cow breed types (HH, 3H1T, HT, 1H3T, and TT) were compared using a GLM model of SAS (SAS Inst. Inc., Cary, NC) that included cow breed type, cow age, and lactation status (lactating or nonlactating). In, cow age was separated into 3, 4, and greater than 4 yr of age. During, cow age categories were changed to 4, 5, and greater than 5 yr to ensure that both and analyses were similar. Mean separations among cow ages and breed types were conducted using Tukey s studentized range test (SAS Inst. Inc.). Residuals from this model were saved for all dependent variables to calculate residual correlations between grazing location and cow traits at weaning. The residual correlations allowed us to examine the phenotypic relationships between terrain use and performance traits after the model adjusted for the fixed effects (cow breed, age, and lactation status). A similar analysis was conducted using grazing location data and calf performance. The model included cow breed type, cow age, and breed of calf s sire (Angus, Charolais, Piedmontese, and Salers). Dependent variables included season-long values for terrain use (mean observed slope, horizontal distance to water, and vertical distance to water), calving date and performance of the calf (actual weaning weight and 205-d age adjusted weaning weight). Residuals from this model were also saved and used to calculate residual correlations between grazing location data from the cow and performance of their calves. Evaluating the relationship between fall pregnancy status was not possible using residual correlations because pregnancy status is categorical data (pregnant or nonpregnant). Terrain use of pregnant and nonpregnant cows was compared using a model that contained cow breed, cow age, lactation status, and fall pregnancy status. Differences in observed grazing patterns between pregnant and nonpregnant cows in the fall would suggest a relationship between reproductive performance and terrain use. Bailey et al. Performance Results Nonlactating cows were heavier and in better body condition (P < 0.001) in the fall than lactating cows. In, nonlactating cows were 606 ± 14 kg with a BCS of 6.5 ± 0.2, whereas lactating cows were 564 ± 5kg with a BCS of 5.1 ± 0.1. During, nonlactating cows were 637 ± 41 kg with a BCS of 7.2 + 0.2, whereas lactating cows were 589 ± 4 kg with a BCS of 5.5 ± 0.1. Weights and heights of younger cows were less (P < 0.05) than those of older cows (Table 3). Body condition scores of 4-yr-old cows were greater (P < 0.001) than those of cows of other ages in. In, BCS of cows of all ages were similar (P = 0.2). Age of cow did not affect (P = 0.6) calving date, but calves from younger cows weighed less (P < 0.05) than calves from older cows (Table 3). In the fall, Tarentaise cows weighed less and had a lower BCS (P < 0.05) than cows with greater proportions of Hereford breeding (Table 4). During, cow breed did not affect actual (P = 0.6) and 205-d age adjusted (P = 0.9) calf weaning weights. Again in, cow breed did not affect actual (P = 0.16) and 205-d age adjusted weaning weights (P = 0.06). Breed of the calf s sire did not affect calf weaning weight in (P = 0.4) or in (P = 0.1). Grazing Location. Nonlactating cows traveled farther (P = 0.04) from water vertically but less horizontally (P = 0.05) than did lactating cows (Table 5). Lactation status did not affect (P = 0.14) use of slopes during. Cow age was important (P < 0.01) for horizontal and vertical distance to water during. Younger cows (3 yr) traveled farther from water and used higher elevations than 5-yr-old and older cows (Table 6). Use of slopes was similar (P = 0.13) among cow breed types during (Table 7). However, horizontal and vertical distance traveled to water differed (P < 0.01) among cow breeds. Hereford Tarentaise cows were farther (P < 0.05) horizontally from water than Hereford cows, and TT, 1H3T, and HT cows were farther (P < 0.05) vertically from water than Hereford cows. Cows that were pregnant at weaning did not use steeper slopes (P = 0.08), travel farther horizontally from water (P = 0.3), or travel farther vertically from water (P = 0.9) than cows that were not pregnant at weaning (Table 8).. Nonlactating cows used steeper slopes (P = 0.02) and were farther vertically from water (P = 0.006) than nonlactating cows (Table 5). Lactation status did not affect (P = 0.9) horizontal distance to water. Cow age (4, 5, and 6+ yr) did not affect slope use (P = 0.4), horizontal distance traveled to water (P = 0.7), or vertical distance traveled to water (P = 0.2). Cow breed types differed (P < 0.005) in their use of slopes and vertical distance traveled to water (Table 7). However, hori-

Terrain use and cow performance 1887 Table 3. Effect of cow age on weight, height, and body condition score (BCS) of cows and actual weaning weight and 205-d age-adjusted weaning weight of their calves during the and grazing seasons Calf Calf 205-d Cow Calving wean wt, wean wt, age, yr n a Wt, kg Hip height, cm BCS b n c date d kg kg 3 55 525 ± 7 x 131 ± 0.5 x 5.2 ± 0.1 x 42 89 ± 2 x 206 ± 4 x 224 ± 4 x 4 57 568 ± 8 y 132 ± 0.5 xy 5.5 ± 0.1 y 49 86 ± 2 x 230 ± 4 y 246 ± 4 y > 4 71 604 ± 5 z 133 ± 0.4 y 5.3 ± 0.1 xy 70 87 ± 1 x 236 ± 4 y 254 ± 4 y 4 53 577 ± 8 x 132 ± 0.5 x 5.8 ± 0.1 x 42 94 ± 2 x 227 ± 4 x 247 ± 3 x 5 47 575 ± 6 x 131 ± 0.6 x 5.5 ± 0.1 x 47 93 ± 2 x 243 ± 4 y 263 ± 3 y > 5 60 619 ± 6 y 134 ± 0.4 y 5.6 ± 0.1 x 60 90 ± 2 x 249 ± 4 y 266 ± 3 y a Number of cows in each age class for cow weight, hip height, and BCS. b Body condition score based on a 1-to-9 scale (1 = emaciated and 9 = obese). c Number of cows in each age class for calving date, calf weaning weight, and age-adjusted calf weaning weight. d Julian date. x,y,z Means ± SE within the same column and within the same year ( or ) without common superscripts differ (P < 0.05). zontal distance traveled to water was similar (P = 0.4) for all cow breed types. Cattle with more Tarentaise breeding (TT and 1H3T) used steeper slopes and climbed higher from water (P < 0.05) than Herefords (HH). Cows that were pregnant at weaning did not use steeper slopes (P = 0.5), travel farther horizontally from water (P = 0.8), or travel farther vertically from water (P = 0.5) than did cows that were not pregnant at weaning (Table 8). Residual Correlations Between Grazing Locations and Performance. During and, the weight, height, and BCS of the cows were not correlated with location of grazing (Table 9). Calving date was negatively correlated to slope use (r = 0.18, P = 0.02) and vertical distance to water (r = 0.19, P = 0.02) during. Cows that used steeper slopes and were farther vertically from water calved earlier in the calving season. In, calving date was not related to any measure Table 4. Differences among cow breeds in weight, hip height, body condition score (BCS), calving date, actual calf weaning weight, and 205-d age-adjusted weaning weight of the calf during the and grazing seasons Calf 205-d Cow Calving Calf wean wean wt, breed a n b Wt, kg Hip height, cm BCS c n d date e wt, kg kg HH 43 559 ± 8 xy 130 ± 0.6 x 5.5 ± 0.1 x 41 89 ± 2 x 212 ± 4 x 231 ± 4 x 3H1T 40 581 ± 10 yz 132 ± 0.6 x 5.5 ± 0.1 x 35 83 ± 1 x 232 ± 4 x 247 ± 5 x HT 34 593 ± 11 z 134 ± 0.7 y 5.5 ± 0.2 x 27 88 ± 2 x 232 ± 6 x 251 ± 6 x 1H3T 31 581 ± 8 yz 133 ± 0.7 y 5.3 ± 0.1 x 28 87 ± 2 x 236 ± 8 x 253 ± 8 x TT 35 533 ± 10 x 132 ± 0.6 xy 4.7 ± 0.1 y 30 89 ± 2 x 224 ± 5 x 243 ± 5 x HH 37 599 ± 8 x 131 ± 0.6 x 6.1 ± 0.1 x 34 95 ± 2 x 230 ± 5 x 251 ± 4 x 3H1T 36 602 ± 10 x 132 ± 0.6 x 5.7 ± 0.1 xy 34 90 ± 2 x 241 ± 4 x 257 ± 4 x HT 30 606 ± 10 x 133 ± 0.7 x 5.6 ± 0.1 y 28 93 ± 2 x 240 ± 5 x 261 ± 4 x 1H3T 25 597 ± 8 x 133 ± 0.7 x 5.5 ± 0.1 yz 25 93 ± 3 x 244 ± 6 x 264 ± 5 x TT 32 557 ± 9 y 132 ± 0.9 x 5.1 ± 0.1 z 28 90 ± 2 x 252 ± 6 x 269 ± 5 x a Cow breed types were Hereford (HH), ³ ₄ Hereford ¹ ₄ Tarentaise (3H1T), ¹ ₂ Hereford ¹ ₂ Tarentaise (HT), ¹ ₄ Tarentaise ³ ₄ Hereford (1H3T), and Tarentaise (TT). b Number of cows in each breed class for cow weight, hip height, and body condition score. c Body condition score based on a 1-to-9 scale (1 = emaciated and 9 = obese). d Number of cows in each breed class for calving date, calf weaning weight, and age-adjusted calf weaning weight. e Julian date. x,y,z Means ± SE within the same column and within the same year ( or ) without common superscripts differ (P < 0.05).

1888 Bailey et al. Table 5. Effect of lactation status on average terrain use during and Indicator of terrain use Lactating Nonlactating P-value Slope, % 18.4 ± 0.2 19.7 ± 0.5 0.14 Horizontal distance to water, m 445 ± 5 440 ± 13 0.05 Vertical distance to water, m 49 ± 1 58 ± 2 0.04 n 161 22 Slope, % 19.0 ± 0.2 21.0 ± 0.4 0.02 Horizontal distance to water, m 836 ± 9 828 ± 38 0.9 Vertical distance to water, m 62 ± 1 70 ± 3 0.01 n 151 9 of terrain use (Table 9). Actual weaning weight and 205-d adjusted weaning weight were positively related to vertical distance from water during. Cows that were further vertically from water had calves with heavier actual (r = 0.21, P = 0.01) and adjusted weaning weights (r = 0.16, P = 0.05). In, actual weaning weight was not related to any measure of terrain use (Table 9). The direction of the relationship between 205-d age-adjusted weaning weight and slope use was negative, but the correlation was not significant (r = 0.16, P = 0.06). Horizontal (r = 0.10, P = 0.2) and vertical (r = 0.02, P = 0.8) distance to water were not related to 205-d age-adjusted weaning weight. Lactation Status Discussion As expected, lactating cows were thinner and weighed less than nonlactating cows in the fall. Energy requirements of nonlactating cows are lower than those of lactating cows (NRC, 1996), and they readily gained weight during the summer grazing season. Most of the observations in this study suggested that nonlactating cows used higher terrain and steeper slopes than lactating cows. Higher water requirements (NRC, 1996) and the presence of a young calf may limit the distance lactating cows will travel. However, in, cows with calves traveled farther horizontally from water than cows without calves. This apparent contradiction may be at least partially explained by the topographic feature of the pasture. Some of the more gentle terrain with lower elevations is located a large horizontal distance from water. Roath and Kruegar (1982) suggested that vertical distance from water was more important than horizontal distance from water. Land managers have been aware of the restrictions in travel of cows nursing calves. Skovlin (1965) suggested that cow-calf pairs were more suited to moderate terrain and that yearlings were more suited to rugged terrain. Cow Age Younger cows weighed less in the fall and hip heights were lower than those of older cows. Weaning weights of calves from younger cows were also lower than those of calves from older cows. Grazing locations differed among cow age classes during, but not during. Young cows (3 yr) used steeper slopes and climbed farther up slopes. Separate analyses by pastures show that the differences among age classes occurred in the Hay pasture, which was grazed early in the grazing season (June and early July). Later in the season in the Back pasture, differences among age classes were not apparent. Mosley (1999) suggested that subordinate animals might be displaced and forced to use suboptimal habitat. Younger animals may have lower social ranking than older animals and correspondingly may be forced to use the rougher terrain. In our study, social Table 6. Differences in terrain use among cows of differing ages during and Horizontal distance Vertical distance Cow age, yr n Slope, % to water, m to water, m 3 55 19.1 ± 0.3 a 461 ± 7 a 53 ± 1 a 4 57 18.6 ± 0.4 a 443 ± 9 ab 50 ± 2 ab 5+ 71 18.2 ± 0.3 a 433 ± 8 b 48 ± 1 b 4 53 19.3 ± 0.3 a 826 ± 15 a 62 ± 1 a 5 47 19.2 ± 0.4 a 847 ± 14 a 64 ± 1 a 6+ 60 18.8 ± 0.3 a 836 ± 14 a 63 ± 1 a a,b Means within the same column and same year without common superscript differ (P < 0.05).

Terrain use and cow performance 1889 Table 7. Effect of cow breed on the mean slope, horizontal distance to water, and vertical distance from water that cows were observed during the and grazing seasons a ; mean values are given for each grazing season and for each pasture within a grazing season Behavior Year and pasture HH 3H1T HT 1H3T TT Slope, % Mean 18.4 ± 0.4 y 17.9 ± 0.4 y 18.6 ± 0.4 y 19.3 ± 0.4 y 18.9 ± 0.4 y Hay 15.5 ± 0.5 y 15.1 ± 0.5 y 16.0 ± 0.7 y 15.5 ± 0.5 y 15.9 ± 0.4 y Back 21.2 ± 0.6 y 20.8 ± 0.6 y 21.2 ± 0.7 y 23.0 ± 0.7 y 21.9 ± 0.7 y n 43 40 34 31 35 Mean 18.1 ± 0.4 y 18.9 ± 0.4 yz 19.2 ± 0.3 yz 19.7 ± 0.5 z 19.9 ± 0.4 z AI 14.3 ± 0.5 y 15.1 ± 0.5 y 15.0 ± 0.5 y 14.9 ± 0.6 y 15.5 ± 0.5 y Back 21.8 ± 0.5 y 22.7 ± 0.5 yz 23.5 ± 0.5 yz 24.4 ± 0.8 z 24.3 ± 0.6 z n 37 36 30 25 32 Horizontal distance to water, m Mean 422 ± 9 y 454 ± 10 yz 461 ± 7 z 438 ± 11 yz 452 ± 9 yz Hay 390 ± 10 y 410 ± 9 yz 421 ± 11 z 417 ± 11 yz 420 ± 10 yz Back 454 ± 18 y 499 ± 20 y 501 ± 25 y 456 ± 17 y 485 ± 16 y n 43 40 34 31 35 Mean 808 ± 21 y 832 ± 20 y 853 ± 17 y 845 ± 17 y 849 ± 17 y AI 1046 ± 35 y 1059 ± 30 y 1123 ± 22 y 1140 ± 43 y 1119 ± 30 y Back 570 ± 22 y 606 ± 23 y 582 ± 26 y 551 ± 23 y 578 ± 19 y n 37 36 30 25 32 Vertical distance to water, m Mean 44 ± 2 y 49 ± 2 yz 52 ± 2 z 53 ± 2 z 54 ± 2 z Hay 34 ± 2 y 38 ± 2 y 39 ± 3 y 37 ± 2 y 41 ± 2 y Back 53 ± 2 y 61 ± 3 yz 64 ± 3 yz 68 ± 3 z 68 ± 3 z n 43 40 34 31 35 Mean 58 ± 1 y 62 ± 2 yz 63 ± 2 yz 65 ± 2 z 67 ± 2 z AI 49 ± 1 y 49 ± 2 y 51 ± 2 y 52 ± 2 y 52 ± 2 y Back 67 ± 2 y 76 ± 3 yz 76 ± 3 yz 79 ± 3 z 82 ± 2 z n 37 36 30 25 32 a Cattle breeds were Hereford (HH), ³ ₄ Hereford ¹ ₄ Tarentaise (3H1T), ¹ ₂ Hereford ¹ ₂ Tarentaise (HT), ¹ ₄ Hereford ³ ₄ Tarentaise (1H3T), and Tarentaise (TT). y,z Means within the same row without common superscripts differ (P < 0.05). interactions may have been less important as the cows aged. Howery et al. () reported that peers influenced where yearling heifers grazed. When animals were older, they tended to graze where their dam or foster dam grazed, suggesting an effect of early learning on where older cattle graze. Additional study may be required to understand the effect of age on grazing dis- tribution patterns, especially as cows age from yearlings to mature animals. Cow Breed Tarentaise and 1H3T cows climbed higher above water than Hereford cows during and and used Table 8. Terrain use of cows that were pregnant or nonpregnant at weaning Distance to water Pregnancy status n Slope, % Horizontal, m Vertical, m Pregnant 157 18.4 ± 0.2 446 ± 5 50 ± 1 Nonpregnant 26 19.6 ± 0.5 437 ± 9 51 ± 2 P-value 0.08 0.3 0.9 Pregnant 150 19.0 ± 0.2 835 ± 9 63 ± 1 Nonpregnant 10 19.1 ± 0.9 856 ± 31 62 ± 4 P-value 0.5 0.4 0.5

1890 Bailey et al. Table 9. Residual correlations between indicators of terrain use and cow performance during and a Distance to water Trait Slope (%) Horizontal, m Vertical, m Cow weight, kg 0.03 0.08 0.03 (0.6) (0.3) (0.7) Cow height, cm 0.04 0.09 0.04 (0.6) (0.2) (0.6) Cow BCS 0.04 0.06 0.04 (0.6) (0.4) (0.5) Calving date 0.18 0.05 0.19 (0.02) (0.6) (0.02) Actual weaning wt 0.13 0.04 0.21 of calf, kg (0.09) (0.7) (0.01) 205-d age-adjusted 0.08 0.02 0.16 weaning wt of calf, kg (0.3) (0.8) (0.05) Cow weight, kg 0.02 0.07 0.12 (0.8) (0.4) (0.1) Cow height, cm 0.03 0.14 0.09 (0.7) (0.08) (0.2) Cow BCS 0.06 0.04 0.10 (0.4) (0.6) (0.2) Calving date 0.11 0.09 0.05 (0.2) (0.3) (0.6) Actual weaning 0.06 0.12 0.04 wt of calf, kg (0.5) (0.1) (0.6) 205-d age-adjusted 0.16 0.10 0.02 weaning wt of calf, kg (0.06) (0.2) (0.8) a Residual correlations are presented and the associated P-values are given below in parentheses. steeper slopes in. These differences were most apparent in the Back pasture based on analyses of individual pastures. In the other pastures (Hay and AI), differences among cow breeds for terrain use were not statistically significant. However, breed means followed the same trends in the Hay and AI pastures (Table 7). The Back pasture is larger and has more topographic relief than the other pastures, which may have provided more opportunity for any inherent difference among breed types to be expressed. Havstad and Doornbos (1987) reported breed differences in distance traveled of cattle with Hereford, Angus, and Simmental breeding within the same pastures used in this study. However, differences among breeds were not consistent between years. For example, Hereford Simmental traveled less than other breeds evaluated during 1984, but in 1985 this breed type traveled more than the other breeds. Herbel and Nelson (1966) found that Santa Gertrudis cattle traveled farther than Hereford cattle in extensive rangeland pastures in New Mexico. The authors suggested that Santa Gertrudis cattle may have traveled farther because the Brahman influence in their breeding made them more adapted to warm temperatures and allowed them to travel farther from water and shade. In this study, Tarentaise cattle may have been more adapted to grazing in steep, rugged terrain. Tarentaise cattle originated in the French Alps, and Herefords were developed in less rugged terrain near Hereford, England. In extensive, rugged rangeland pastures, managers may be able to improve uniformity of grazing by selecting cattle breeds that were developed in mountainous terrain. Relationships Between Grazing Location and Performance No relationships were observed between measurements of terrain where a cow grazed (slope and horizontal and vertical distance to water) and the cow s height, weight, or BCS. Selecting cows based on weight, height, or BCS should not affect grazing distribution patterns as long as cows remain within the moderate values observed in this study. Likewise, any selection or culling programs based on grazing distribution patterns should have minimal, if any, effects on cow size or body condition. However, genetic correlations among cow size and terrain use could potentially differ from the phenotypic relationships observed in this study (Falconer, 1960). Relationships between terrain use and calf weaning weight were also inconsistent. In, cows that used higher elevations calved earlier and had heavier unadjusted and age-adjusted weaning weights. However, in the Back pasture during terrain use was not related to weaning weight. The inconsistent results from and between terrain use and calf weaning weight suggest that there is no consistent phenotypic relationship between these traits. Results from this study demonstrate that culling cows that prefer gentle terrain near water should not adversely affect calf weaning weights. Cows using steeper and more rugged terrain seem similar in size and raise calves with weaning weights similar to those of calves from cows using gentler terrain near water. Cow reproductive performance does not seem to be related to terrain use. Nonpregnant cows and pregnant cows traveled similar distances from water (horizontally and vertically) during the fall at weaning. There were no statistically significant differences in use of slopes between pregnant and nonpregnant cows. Culling cows based on terrain use should not adversely affect pregnancy rates. Implications Uniformity of grazing in rugged rangeland may be improved by selecting breeds originating from mountainous terrain (Tarentaise) rather than breeds developed in gentler topography (Hereford). Tarentaise cows used steeper slopes and climbed to higher elevations above water than Herefords. Terrain was not consistently related to the weight, height, and body condition of the cow or to calving date and calf weaning weight.

Terrain use and cow performance 1891 Therefore, culling cows that graze gentle terrain near water to improve uniformity of grazing should have no immediate adverse impact on herd productivity. However, genetic correlations among these traits must be determined before the effects of long-term selection for grazing distribution can be evaluated. The absence of adverse phenotypic relationships between terrain use and cow performance and the apparent difference among breeds suggest that continued evaluation of selection programs for more uniform grazing distribution is warranted. Literature Cited Bailey, D. W., J. E. Gross, E. A. Laca, L. R. Rittenhouse, M. B. Coughenour, D. M. Swift, and P. L. Sims. 1996. Mechanisms that result in large herbivore grazing distribution patterns. J. Range Manage. 49:386 400. Bailey, D. W., J. W. Walker, and L. R. Rittenhouse. 1990. Sequential analysis of cattle location: Day-to-day movement patterns. Appl. Anim. Behav. Sci. 25:137 148. Bailey, D. W., and G. R. Welling. 1999. Modification of cattle grazing distribution with dehydrated molasses supplement. J. Range Manage. 52:575 582. Cook, C. 1966. Factors affecting utilization of mountain slopes by cattle. J. Range Manage. 19:200 204. Falconer, D. S. 1960. Introduction to Quantitative Genetics. Ronald Press, New York. Havstad, K. M., and D. E. Doornbos. 1987. Effect of biological type on grazing behavior and energy intake. In: Proc. Grazing Livestock Nutrition Conf., Jackson, WY. pp 9 15. Herbel, C. H., and A. B. Nelson. 1966. Activities of Hereford and Santa Gertrudis cattle on a southern New Mexico range. J. Range Manage. 19:173 181. Howery, L. D., F. D. Provenza, and R. E. Banner.. Social and environmental factors influence cattle distribution. Appl. Anim. Behav. Sci. 55:231 244. Howery L. D., F. D. Provenza, R. E. Banner, and C. B. Scott. 1996. Differences in home range and habitat use among individuals in a cattle herd. Appl. Anim. Behav. Sci. 49:305 320. Kress, D. D., D. E. Doornbos, D. C. Anderson, and K. C. Davis. 1996. Genetic components for milk production of Tarentaise, Hereford, and Tarentaise Hereford cows. J. Anim. Sci. 74:2344 2348. Lehner, P. N. 1979. Handbook of Ethological Methods. Garland STMP, New York. Low, W. A., R. L. Tweedie, C. B. H. Edwards, R. M. Hodder, K. W. J. Malafant, and R. B. Cunnigham. 1981. The influence of environment on daily maintenance behavior of free-ranging Shorthorn cows in central Australia. 1. General introduction and descriptive analysis of day-long activities. Appl. Anim. Ethol. 7:11 26. Mosley, J. C. 1999. Influence of social dominance on habitat selection by free-ranging ungulates. In: K. L. Launchbaugh, J. C. Mosley, and K. D. Sanders (ed.) Grazing Behavior of Livestock and Wildlife. pp 109 1118. Idaho Forest and Range Exp. Sta. Bull. No. 70, Moscow. Mueggler, W. F. 1965. Cattle distribution on steep slopes. J. Range Manage. 18:255 257. NRC. 1996. Nutrient Requirements of Beef Cattle. 7th ed. National Academy Press, Washington, DC. O Neill, R. V., D. L. DeAngelis, J. B. Waide, and T. F. H. Allen. 1986. A Hierarchical Concept of Ecosystems. Princeton University Press, Princeton, NJ. Roath, L. R., and W. C. Krueger. 1982. Cattle grazing and behavior on a forested range. J. Range Manage. 35:332 338. Senft, R. L., M. B. Coughenour, D.W. Bailey, L. R. Rittenhouse, O. E. Sala, and D. M. Swift. 1987. Large herbivore foraging and ecological hierarchies. Bioscience 37:789 799. Skovlin, J. M. 1957. Range riding the key to range management. J. Range Manage. 10:269 271. Skovlin, J. M. 1965. Improving Cattle Distribution on Western Mountain Rangelands. USDA Farm Bull. 2212, Washington, DC. Valentine, K. A. 1947. Distance from water as a factor in grazing capacity of rangeland. J. For. 45:749 754. Vallentine, J. F. 1990. Grazing Management. Academic Press, San Diego, CA.