Are Feedlot-based Performance Cattle Limiting Ecological Services for Rangeland Ecosystems in Northern Mixed-grass prairies?

Final report for OW15-026

Project Type: Professional + Producer
Funds awarded in 2015: $49,961.00
Projected End Date: 03/31/2018
Grant Recipient: Montana State University
Region: Western
State: Montana
Principal Investigator:
Dr. Emily Meccage
Montana State University
Expand All

Project Information


The goal of this project is to evaluate the impacts of steer frame-size and finishing time in a forage-based system. In January and November of 2016, the advisory committee met to discuss the project. The committee consists of 4 Montana ranchers who are located throughout the state. The goal of the January 2016 meeting was to outline the research project, the sampling methods, as well as discuss end-point expectations and questions. Committee members discussed the practicalities of the project, and whether or not the project was practical from a rancher standpoint. During the summer of 2016, an undergraduate student was hired to construct all fencing, as well as help put in place permanent water tanks. Four separate pastures were created, 2 which which enclosed the introduced Russian wildrye pasture that was established the previous year, and two that were larger, native range pastures.

In November 2016, the committee met once again to go out to the fields, observe the vegetation, fencing, and water availabilities, and again offer the insight. The fencing plan was changed based on some of the advice. Initial herbage mass samples were taken from all pastures, and the steers were allowed in to the Russian wildrye pastures on November 29, 2016. They stayed on the Russian wildrye pasture for about a month, and were then moved onto native range the end of December. Due to extremely cold temperatures for a prolonged period of time, steers were fed supplemental forage in order to prevent any health issues from occuring. Herbage mass, vegetation, and habitat sampling will continue to occur throughout the summer of 2017, until all animals are finished.

The group of “control” or feedlot-finished steers were put into the feedlot the same time as the forage-finished steers, and they are expected to finish sometime in May 2017.

Project Objectives:
  1. Assess the economic impacts of reducing frame size of cattle finished on forage.
  2. Assess the influence of cattle frame size on the capacity of conventional stocking rate models to protect or improve watersheds, wildlife habitat and rangeland condition.
  3. Develop new winter stocking rate model for native ranges that will protect/promote soil health, forage quality, watershed function and wildlife habitat persistence.


Click linked name(s) to expand


Participation Summary

Educational & Outreach Activities

5 Consultations
1 Curricula, factsheets or educational tools
1 Journal articles
3 On-farm demonstrations
1 Published press articles, newsletters
2 Tours
1 Workshop field days

Participation Summary

4 Farmers
Education/outreach description:

We are in the process of submitting a journal article to Rangelands, which will be disseminated among professionals and ranchers. We also plan on hosting a field day once all animals are processed and we receive the final products, in order to do a “taste test” among participating ranchers and department faculty. Results from this study will also be included in several presentations in the faculty outreach programs, which will be given around the state of Montana.

Project Outcomes

2 Grants received that built upon this project
4 New working collaborations
Project outcomes:



Emily Meccage, Lance McNew, Clayton Marlow, Danielle Peterson,

Department of Animal & Range Sciences, Montana State University, 103 Animal Biosciences Building, Bozeman, MT



Livestock grazing is the most wide-spread influence on native ecosystems in western North America with more than 70% of the western U.S. being grazed.  Livestock grazing provides food and fiber, as well as economic and social benefits to local communities, and protection to grassland ecosystems from conversion and loss.  However, poor management of rangelands, including chronic overgrazing, has deleterious effects on grassland ecosystems. Overgrazing reduces primary productivity, impedes plant growth and survival, alters plant species composition, increases soil erosion, and reduces ground water quality. In contrast, properly managed rangelands can improve water filtration and ground water recharge, protect soils and nutrient cycles, impede invasions of non-native plants, collect and store atmospheric carbon, produce habitat for wildlife, and provide a host of other ecosystem services.  Successful grazing systems maintain sustainable grassland ecosystems that provide viable revenue for producers, and successful grazing strategies will have multiple management and environmental goals, including economic livestock production, maintenance of plant community health, protection of water quality and quantity, reduction of soil erosion, as well as venues for dealing with climate change.

The future of working rangelands depends on developing novel approaches and partnerships that enhance sustainability of grazing.  Recent societal shifts toward grass-fed beef coupled with high grain prices offers an opportunity to explore the economic and ecological effects of finishing cattle on native or improved forage.  The majority of expenses and environmental impact associated with finishing cattle for beef production are due feed production costs. Forage-finishing may reduce inputs for cattle production by eliminating transportation of both feed and livestock to feedlots and the air and water pollution linked to feedlot production.  Moreover, forage-finishing eliminates the cost of crop production, feed grinding, storage, transportation, and manure management.  Ecological benefits of forage finishing are numerous and include a reduced carbon footprint from the reduction of grain production and transportation, as well as  restoring areas now in grain production to native or improved permanent forage.  Nevertheless, historical selection toward large-framed cattle for feedlot finishing may preclude the ability for these animals  to finish on forage.  Furthermore, large-framed cattle have higher nutritional requirements leading to increased forage intake that may reduce long-term forage availability, diversity, and quality.  High residual feed intakes may forestall the applicability and economic viability of forage-finishing. Selecting cattle that are efficient at processing feed is important for reducing the amount of days it takes for them to reach finished weights, and it is estimated that a 10% increase in feed efficiency has the potential to increase profit by 43% (Fox et al., 2001).


The goal of this proposed collaboration was to evaluate whether shifting to smaller-framed sires, and reducing progeny frame size, enhances the ability of these smaller framed calves to finish on forage, ultimately reducing production inputs, and benefiting the rangeland ecosystem. The length of the grazing or finishing season may be reduced for offspring of small-framed sires if they are more efficient at utilizing forage than offspring from large-framed sires. Shortened finishing periods may lower long-term impacts to rangeland resources and improve bottom-up ecosystem function and habitat conditions for livestock and wildlife. The first part of our hypothesis was that steers from smaller framed bulls [low expected progeny differences (EPD) mature weights] would limit forage over-utilization and allow for more efficient growth of cattle on native and improved forage, effectively reducing or eliminating the need for supplementation and feedlot finishing. The second part of our hypothesis was that using smaller framed steers in grass finishing operations would have positive effects on rangeland ecosystems by improving soil health, maintaining or improving vegetation composition, structural heterogeneity and floristic diversity.


Study Site

The study site used to finish steers in this study consisted of a 410 ha stockade pasture on the Montana State University Red Bluff Research Ranch near Norris, Montana. Two water tanks were located within the stockade pasture to allow cattle access to water during winter and summer (Fig. 1). Since the 1980’s, this pasture has been used to over-winter mature cows at a stocking rate of 0.73 AUMs/ha. Soils in this pasture are relatively shallow, with upper soil horizon textures (0 – 15 cm) embracing sandy loams and loams. The southern 20% of the pasture has a slope of 0 – 4 %; 45% of the pasture has a gently sloping (4 – 15%) terrain with a southerly aspect, and the remaining 35% of the pasture has slopes in excess of 15% with a northerly aspect (Fig. 2).

Scattered limber pine (Pinus flexilis), Rocky Mountain juniper (Juniperus scopulorum) and antelope bitterbrush (Purshia tridentata) cover the steeper slopes (15 – 35%) in the stockade pasture, but several grassland habitat types form the dominate vegetation cover. Similar to the results of Mueggler and Stewart (1980), data collected in this study indicated that 0 – 4% slopes composed of sandy loam soils are dominated by the needleandthread/blue grama/western wheatgrass (Hesperostipa comata/Bouteloua gracilis/Pascopyrum smithii) habitat type.  As slopes increase to 4 – 15%, loamy sites are dominated by the Idaho fescue/bluebunch wheatgrass (Festuca idahoensis/Pseudoroegnaria spicata) habitat type. In areas where hay has been fed to wintering mother cows, smooth brome (Bromus inermis), Kentucky bluegrass (Poa pratensis) and western wheatgrass have replaced much of the Idaho fescue and bluebunch wheatgrass. The bluebunch wheatgrass/ Sandberg bluegrass/ needleandthread grass (Pseudoroegnaria spicata/ Poa sandbergii/ Hesperostipa comata) habitat type dominates the steeper (15 – 35%) north facing slopes and the antelope bitterbrush/ Idaho fescue (Purshia tridentata/ Festuca idahoensis) habitat type occupies the same slopes with a southerly aspect.



Steer Selection and Treatments

The Angus steers in this study originated from the Northern Agricultural Research Center (NARC; 48.5447, -109.6866), Havre, Montana (Fig. 3). Access to the resident maternal cow herd enabled us to utilize information from detailed cow history records to select four hundred cows, similar in both size and age (average cow size = 590-635 kg; average cow age = 6.5 years old), which were bred to produce the offspring used in this study. Two hundred cows were randomly selected from this herd and bred to a sire with low mature size EPD. The resulting bull calves were identified with the group name “SMALL.” The remaining two hundred cows were bred to a sire with high mature size EPD. The resulting bull calves were identified with the group name “LARGE.”  From these two groups of bull calves, 46 head of “SMALL” and 46 head of “LARGE” bull calves were randomly selected to participate in the forage finishing portion of this study. These bull calves were implanted with Ralgro at 4 weeks of age to maximize pasture utilization and improve weaning weights.

At approximately seven months of age, the steers were weaned and separated into groups according to sire EPD (SMALL and LARGE) on 20 October 2016. As anticipated, the LARGE steers were heavier than the SMALL steers at weaning (Fig. 4). Once separated from their mothers, the steers remained at the NARC in Havre for one month, where they were allowed to graze improved pasture until it was time for them to be transported to their respective locations to begin the feeding portion of this study. As expected, the pattern noted at weaning remained unchanged after one month of pre-conditioning, with LARGE steers remaining bigger than SMALL steers (Fig. 4). 

Half of the steers from each size group (23 from SMALL and 23 from LARGE) were moved into a feedlot at the NARC on 29 November 2016 for conventional feedlot finishing (80% concentrate (corn/barley), 12% straw, 3% oil and 5% supplement containing Tylosin and Monensin). The remaining steers (23 from SMALL and 23 from LARGE) were shipped to Red Bluff Research Station, Norris, Montana (45.5504, -111.6585) on 29 November 2016 (Fig. 3). Once on the ranch, the steers were moved to an improved pasture consisting of a two year old stand of Russian wildrye (Psathyrostachys juncea), split in half using electric fencing. Steers remained separated by treatment and were allowed to graze the Russian wildrye pasture until 3 January 2017. During 8 of the 33 days (29 November 2016 to 3 January 2017) on the introduced pasture, steers were fed hay due to severe weather conditions.  Hay was provided at a rate of 9 kg/ hd/ day during this period to maintain gains of approximately 0.18 to 0.36 kg/ head/ day.  After 3 January 2017, both groups of steers were turned into separate native range pastures consisting of the previously described community types. These grassland and shrub community types are common throughout much of southwestern Montana. On 3 May 2017, both groups were moved back to the improved pasture to enhance nutritional status (rate of gain) and to provide early spring rest for the native pastures. Steers remained separated by treatment and grazed this pasture until 6 June 2017 when they were again moved back onto native range for the summer months.  Because of less than average rainfall, it became apparent in late September that neither group would likely meet the target harvest weight of 544 kg before another winter began. To reduce competition between the test steers and the production cow herd at the Red Bluff Ranch, the project was terminated on 15 November 2017 when all 46 steers were shipped to market.

Experimental Native Grazing Pastures

To evaluate production tradeoffs between grass finishing cattle and the impact to soils, range vegetation communities, and grassland bird habitat, the previously described historic winter range pasture at Red Bluff Ranch was subdivided into four experimental grazing paddocks to provide separate grazing opportunities for the 23 LARGE and 23 SMALL-framed cattle based on grazing season (Fig. 3).  Two pastures were grazed separately by each group in the winter/ spring season (3 January 2016 – May 3 2017) and the remaining pastures were grazed separately by each group during the summer/ fall season (6 June 2017 – 15 November 2017). The winter/spring season pastures used to graze the LARGE and SMALL-framed cattle were both 109 hectares in size with 62.3 total available AUM for the LARGE cattle and 53 AUM for the SMALL cattle.  The summer/fall season pastures varied in size with the LARGE-framed cattle pasture being 67 hectares and the SMALL-framed cattle pasture being 133 hectares.  However, steeper slopes (> 20°) in the larger pasture resulted in lower availability of forage. When the steep terrain was discounted, total available AUMs were similar between groups (LARGE-framed pasture = 49.5 AUM; SMALL-framed pasture = 50.5 AUM). Stocking rates in this study were considered moderate and were selected strategically to ensure maximum individual animal performance and allow for an observation of cattle size effect.

Cattle Performance

Steers were weighed on 20 October 2016 prior to being separated into the feedlot and pasture finishing treatments. Once moved to their respective locations, steer weights for both the SMALL and LARGE treatments were collected on six occasions and ADG was estimated for 5 intervals. Final feedlot steer weights were obtained on 13 June 2017, prior to shipping for harvest. Average final feedlot finished steer weight was 612 kg. Final pasture finished steer weights were obtained on 29 November 2016. Pasture finished steers did not reach the target finishing weight of 544 kg prior to winter, and due to Red Bluff Management implications, the project was terminated on 15 November 2017. The pasture finished steers were shipped to market weighing 489 kg on average. 

Linear mixed effects models in the R package lme4 (Bates et al. 2015, R Core Team 2017) were used to model effects of finishing system (feedlot v. forage), size class (SMALL v. LARGE), and animal age (days since birth) on ADG (kg d-1) of steers.  We computed ADG for each animal by calculating the weight gain in each period, then divided by the number of days in the period.  This resulted in multiple entries of ADG for each animal during the trial.  Because repeated measures of ADG for each animal are correlated, we used steer ID as a random effect in all models.  We used Akaike’s Information Criterion (AICc) to evaluate support for a candidate set of models representing a priori hypotheses regarding the effects of finishing system, frame size class, and steer age (Anderson and Burnham 2002).  Supported models with AICc values ≤ 2 from the best-fit model were considered parsimonious (Burnham et al. 2011). When a supported model differed from the top model by a single parameter, we considered the additional parameter to be uninformative and excluded this parameter from inclusion in the final model (Arnold 2010). Relative support of effects was evaluated by considering evidence ratios based on AICc weights (wi; Burnham and Anderson 2002). We report means and 95% confidence intervals (CI) of parameter estimates and ADG for steers in each finishing system and size class.  A similar analysis was conducted to evaluate the interactive effects of finish system, frame size, and days since birth on mass (kg) of steers during the study.

Structural Evaluation of Grassland Bird Habitat

To evaluate the ecological impacts of grazing between the two test groups (LARGE and SMALL), we conducted vegetation surveys along 20 randomly selected 100-m transects in June of 2016 (pre-grazing) and 2017 (post-grazing) within the four native pastures (4–5 transects per pasture).  In addition, we collected the same suite of vegetation measures at 100 randomly located line transects 20-m in length during just the post-grazing summer of 2017 (22–26 transects per pasture). All survey transects were separated by ≥ 200 m to minimize spatial autocorrelation.  Vegetation conditions were measured at subplots located every 10 m and 5 m along 100-m and 20-m transects, respectively.  At each subplot, visual obstruction reading (VOR) was measured from each of 4 cardinal directions at a distance of 2 m and height of 0.5 m (Robel et al., 1970). Vegetation coverage of native grass, exotic grass (predominantly cheatgrass [Bromus tectorum]), forbs, soil surface litter, and bare ground were estimated using the methods of Daubenmire (1959).  Heights (cm) of the nearest live plant of each type and detritus were measured at each subplot.

Using means and standard deviations, we calculated the difference between pre- and post-grazing measures at each sampling transect for pastures in each grazing treatment. Because our data did not conform to assumptions of normality, we used a Wilcoxon Rank-Sums test to compare pre- and post-grazing differences in vegetation conditions between pastures grazed by SMALL and LARGE cattle.  Treatment-level differences in pre- and post-grazing differences were considered statistically significant at α = 0.05. 

Soils Analyses

Nine permanently monumented sites (RB3-RB60) were used to track total soil organic matter (TSOM) within the four native range pastures from April 2016 until October 2017 (Fig. 5).  This sampling timeline bracketed the grazing period, with samples being taken December 2016 through Oct 2017.  A soil sample was collected at each site during March/April (spring), June/July (summer) and October/November (fall). Soil samples were taken to the BART Nutrition Center where they were oven-dried, individually ground, weighed and then ashed at 375o C for 16 hours (Robertson, 2011; Hoogsteen et al., 2015). Initial minus final sample weight was assumed to represent TSOM; the precursor to soil organic carbon. In addition to the soil samples, four samples of above-ground production or biomass were collected using a stratified pattern at each soil collection site during the respective collection periods. Biomass was separated into standing dead, surface litter and green or actively growing material. Each fraction was weighed separately to test the concept that green actively growing material produced the standing dead and standing dead provided the surface litter that would eventually become TSOM.

One pasture was grazed for each treatment group x season (SMALL steers – winter/spring; LARGE steers- winter/spring; SMALL steers- summer/fall and LARGE steers- summer/fall), providing implication for statistical analyses. For this reason, a one way ANOVA was used to compare TSOM across dates rather than among pastures.  Paired t-tests were used to compare TSOM and plant residue (surface litter) between spring, summer, and fall grazed pastures. Sites within pastures were considered independent samples.



Structural Evaluation of Grassland Bird Habitat

Average cheatgrass coverages did not differ between pre- and post-grazing periods for either group, however, variability in cheatgrass cover was greater during the post-grazing period in pastures grazed by the SMALL groupfinal-figures-WSARE.  Despite observing apparent pre- and post-grazing differences in several vegetation measurements (Table 1), we found no statistical evidence for differences in vegetation conditions among grazing treatments (P > 0.05); differences in average pre- and post-grazing vegetation conditions were similar for pastures grazed by SMALL and LARGE steers (Table 1). Similarly, variability in vegetation conditions assessed by standard deviation at the transect level were similar among LARGE and SMALL groups (Table 1).

Table 1. The mean differences in vegetation measures means and standard deviations before and after grazing and their standard deviations are depicted here.  Wilcoxon tests were used to analyze for significance using an alpha of 0.05.  No vegetation metric was significantly different before or after grazing for either large or small framed cattle (all p-values > 0.05).

Mean Difference in Vegetation Measures Before and After Grazing


                      Cattle Frame Size

Vegetation Metric





mean VOR (cm)

– 2.85 ± 0.10

– 2.42 ± 0.87



sd VOR

– 0.38 ± 0.35

– 0.06 ± 0.80



mean % native grass

– 92.00 ± 45.33

– 98.48 ± 65.13



sd % native grass

9.18 ± 7.23

3.53 ± 4.32



mean native grass height (cm)

– 9.45 ± 10.97

– 4.51 ± 11.51



sd native grass height (cm)

3.04 ± 2.83

2.86 ± 5.60



mean % cheat grass

6.31 ± 7.70

8.93 ± 9.91



sd % cheat grass

5.89 ± 6.19

7.81 ± 10.06



mean cheat grass height (cm)

0.66 ± 6.84

3.22 ± 6.35



Sd cheat grass height (cm)

9.18 ± 7.23

3.53 ± 4.32



mean % forb

0.94 ± 20.23

1.59 ± 7.92



sd % forb

– 0.27 ± 7.44

0.56 ± 11.46



mean forb height (cm)

– 0.06 ± 16.91

– 3.23 ± 7.86



sd forb height (cm)

– 2.90 ± 5.48

– 3.99 ± 6.13



mean % shrub

0.63 ± 2.34

5.16 ± 9.53



sd % shrub

1.98 ± 7.39

9.37 ± 17.30



mean shrub height (cm)

– 0.15 ± 2.35

4.06 ± 6.07



sd shrub height (cm)

– 0.47 ± 7.42

8.55 ± 11.70



mean % detritus

19.41 ± 19.09

23.86 ± 21.13



sd % detritus

7.53 ± 8.63

5.19 ± 11.02



mean % bare ground

11.03 ± 23.36

3.36 ± 14.24



sd % bare ground

5.72 ± 15.35

6.07 ± 11.11





Soils Analyses

While initial analyses indicated that TSOM changed dramatically across seasons (SPR 2017 and early SUM 2017 had the lowest amount of TSOM), there was no appreciable increase or decrease (P > 0.10) in TSOM at the conclusion of the grazing study (Fig. 6). By late October 2017 TSOM levels had recovered to values similar to those recorded at the beginning of the trial in April 2016. The lower values (P < 0.011) in spring and early summer 2017 could represent mineralization of TSOM from winter into early summer. Importantly, these analyses included data from two, 60-year-old grazing exclosures. There was no difference (P = 0.771) in June/July 2017 TSOM levels between the rested winter grazed pastures and the 60-year-old grazing exclosures. Analysis of variance of TSOM by site across all sampling dates indicated no difference (P = 0.35) among grazed and ungrazed and protected sites (Fig. 7).

Surface litter was the most likely plant residue to be incorporated into TSOM during the grazing trial. At the beginning of the study (April 2016) surface litter differed (P = 0.08) across all 4 pastures, and at the conclusion of the study (October 2017) the differences had become even more pronounced (P = 0.001; Fig. 8 and 9). However, these differences were not related to either grazing or season of grazing (P = 0.350).  Instead these values seem reflective of the landscape. These results indicate that the stocking rate used in this study did not influence the amount of residual litter that could be incorporated into the soil profile. 

Cattle performance

We analyzed weights (mass; kg) from 92 steers from 5 March 2016 to 30 November 2017.  Average daily gain of the feedlot group increased by 0.7% each day from birth to harvest (Fig. 10), and steers in this group were ready for harvest at 444 days of age.  In contrast, ADG for the forage finish group declined by an average of 0.2% each day from birth to harvest (Fig. 10), and were not harvested until 623 days of age.  An effect of frame size on ADG was not supported by the data (report p value).  Body weight increased with days since birth for both finish groups; however, the rate of increase was 48% greater for the feedlot group (Fig. 11).  The effect of frame size on final body weight was small but significant (P = 0.04) and this difference in body weight between the SMALL and LARGE groups was more pronounced in the forage-finished group, with SMALL gaining more weight during the study than LARGE (Fig. 11, 12).  Average weight at harvest for feedlot and forage-finished groups were 621 kg (range: 535–724 kg) and 465 kg (range 379–517 kg), respectively (Fig. 12).


Conclusively, vegetation conditions including the presence of native grasses, cheat grass, forbs, shrubs, detritus, and bare ground after grazing did not differ between the SMALL and LARGE treatments after being grazed. Visual obstruction readings were also similar for pastures being grazed by the SMALL and LARGE treatments. However, variation in VOR for pastures being grazed by small-framed steers was lower when compared to VOR in pastures being grazed by large-framed steers, indicating that small-framed steers tended to graze more uniformly than large-framed steers. Higher VOR variation associated with pastures being grazed by large-framed steers might be explained by the increase of within transect variability of native grass cover and height following grazing, which was not observed in pastures being grazed by small-framed steers. Although there appeared to be variation differences in VOR and native grass cover when grazed by steers of different sizes, total VOR and total native grass cover did not differ between large- and small-framed cattle, indicating that the variation differences might have been due to pasture non-uniformity, and thus sampling variability. Variation in VOR and native grass cover most likely reflects baseline differences in pastures, rather than the effects of steer frame size, and is most likely due to under-sampling of vegetation metrics for pre- and post-grazing analysis. Similarity in pre-and post-grazing vegetation conditions suggests that grazing smaller frame size steers does not benefit the rangeland ecosystem more than grazing large-frame size steers.

As expected, ADG and final steer weight was higher for feedlot finished steers than for forage finished steers. Additionally, feedlot finished steers reached harvest weight 179 days sooner than forage finished steers. Average daily gain was not influenced by steer frame size in either the feedlot or forage finished settings; however, final body weight was higher for SMALL-framed cattle in the feedlot finished group. For the forage finished steers, there was only about a 6 kg difference in body weight between SMALL and LARGE, indicating that shifting to small-framed cattle in a forage finished system does not enhance their ability to finish on forage, and does not decrease the length of time necessary for forage finishing.

This study indicates that moderate winter and summer grazing on foothills rangeland in southwestern Montana does not appear to impact TSOM or grassland bird habitat in either a negative or positive manner.  It is possible that heavier stocking rates may produce the anticipated change in soil health or grassland bird habitat, but such rates may be counter to other ecological output. The stocking rate in this study was strategically implemented to maximize individual steer performance. Had we selected a heavier stocking rate, we may have observed a different outcome for rangeland conditions such as vegetation cover, grassland bird habitat, VOR, and soil health.

Based on the results of this research it is apparent that high performance sires are not necessary if the objective is to finish cattle on grass. It is also apparent that in order to finish steers on grass, improved pasture or supplementation is necessary to ensure animals reach final finishing weight in a timely manner. Originally, steers in this study were meant to graze improved pasture during the fall, beginning in August. However, due to drought conditions, there was not enough regrowth to support these steers this early in the fall after the improved pasture had already been grazed earlier that year in the spring.  








Literature Cited

Anderson, D. R., and K.P Burnham. 2002. Avoiding pitfalls when using information-theoretic methods. The Journal of Wildlife Management 912-918.


Arnold, T. W. 2010. Uninformative parameters and model selection using Akaike’s Information Criterion. The Journal of Wildlife Management 74:1175-1178.


Bates, D., M. Maechler, B. Bolker, and S. Walker. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67:1-48.<doi:10.18637/jss.v067.i01>.

Burnham, K. P., D.R. Anderson, and K.P Huyvaert. 2011. AIC model selection and multimodel inference in behavioral ecology: some background, observations, and comparisons. Behavioral Ecology and Sociobiology 65:23-35.


Daubenmire, R. F. 1959. Plants and environment. A text book of plant autecology. John Wiley and Sons Inc.


Fox, D.G., L.O. Tedeschi, and P.J. Guiroy. 2001. Determining feed intake and feed efficiency of individual cattle fed in groups. Proceedings of the Beef Improvement Federation 33rd Annual Research Symposium Meeting  33:80-98.

Hoogsteen, M. J. J., E.A. Lantinga, E.J. Bakker, J.C.J Groot, and P.A. Tittonell. 2015. Estimating soil organic carbon through loss on ignition: effects of ignition conditions and structural water loss. European Journal of Soil Science 66:320-328.


Mueggler, W. F., and W.L Stewart. 1980. Grassland and shrubland habitat types of western Montana. Intermountain Forest and Range Experiment Station. Ogden, UT, USA: USDA Forest Service Intermountain Res. Sta. Gen. Tech. Rpt. INT-66 p. 10-74.


Robel, R. J., J.N. Briggs, A.D. Dayton, and L.C. Hulbert. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management 295-297.


Robertson, S. 2011. Direct estimation of organic matter by loss on ignition: methods. SFU Soil Science Lab, Burnaby 11p.


Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.