The Implications of Varying Levels of Cattle Grazing

Cattle grazing has been a large source of controversy, especially on the fescue grasslands of Cypress Hills International Park, encompassing an endangered habitat. In an attempt to better understand the effects of grazing on diversities in the area, a study was undertaken using three representative sites; ungrazed, moderately grazed, and highly grazed. We assessed arthropod and plant abundance and diversity using sweep nets, traps, point frame analysis, and collection of plant biomass. Unfortunately, there was no significant difference in the three treatments for arthropod abundance or diversity, seemingly due to poor weather. However, plant abundance and diversity followed a significant trend with intermediate grazing being the most abundant diverse. In conclusion, it appears that moderate levels of grazing may be beneficial if implemented properly.

Introduction

Cypress Hills Interprovincial Park lies on the border of Alberta and Saskatchewan in Canada. The area comprises two subregions, a montane natural subregion as well as a mixed grassland natural subregion, composed mostly of fescue grasslands. In recent years, it has been a hot spot for research endeavors due to the unique processes that occur in the area and the opportunity to study the incredible diversity. The fescue grasslands in this area are particularly of importance because they are an endangered habitat. Additionally, this area is not prone to fire suppression. Many studies suspect that this is due to road fragmentation of habitat, human activity, the encroachment of forests, introduction of invasive species, and the grazing activities of domestic and non-domestic ungulates. Due to this, there has been a large controversy on the effect of grazing on this historical site and the potential that it may increasingly reduced native species and productivity of the area. Our research group in particular was concerned with the effects of grazing on plant and arthropod abundance and diversity. Three representative sites were chosen, an ungrazed exclosure, an intermediately grazed field, and a highly grazed area where cattle are primarily released prior to grazing beginning. We predicted that the highly grazed site would have the lowest abundance and diversity of not only plants but also arthropods. Due to a large portion of cattle being in the area, competition for feed is high and therefore it follows that cattle become less picky about their foliage choice, opting to eat what is available. In addition to competition, cattle are also known for trampling the ground, which results in more compacted soils. Consequently this leads to reduced soil moisture content due to the inability of soils to absorb and hold water. Finally, erosion is also likely due to reduced plant cover and drier soils overall. Conversely, the ungrazed grazed site was predicted to have the greatest abundance of plants and arthropods due to the lack of competition for resources within the arthropods, and a lack of suppressing factor of grazing for the plants. Finally, we predicted that the intermediate site would likely show the greatest diversity of both plants and arthropods due to the lack of overcrowding, and the suppression of aggressively dominant plants.

Methods

Location

This study took place in the fescue grasslands on the Alberta side of Cypress Hills Interprovincial Park. It encompassed three different representative sites; ungrazed, intermediately grazed, and highly grazed, all considered as different treatments. The ungrazed site was an exclosure where cattle were not permitted to enter, and although it is suspected to be ten plus years old, the exact date it was built is unknown. Prior to sampling, eight random positions in a 50x50m grid were selected for each of the three sites using a table of random numbers to act as replicates.

Arthropod Sampling

Arthropod sampling was undertaken in two ways, primarily using a large mesh sweep net, to collect larger flying insects, and secondarily with pan and pitfall traps, to collect ground-dwellers and flying insects respectively. Net sampling was completed first to avoid disturbing any arthropods present prior to collection of other data. From the origin point of each replicate, the net was swung ten times on alternating sides in time with steps for a total of ten paces, making sure to brush the foliage in the area with each sweep. This was completed in each cardinal direction (south, east, north, and west) for a total of forty sweeps per replicate at each site. Following net sampling, pitfall and pan traps were set up. Both contained water and dish soap to drown any arthropods who fell into the traps. Pit fall traps were set up using red Solo cups, in which the tops were flush with the ground and packed with soil to ensure no arthropods would fall into the cracks rather than the cup. A tile was then placed overhead, supported by nails to ensure that potential rainfall would not overfill the cups. Pan traps were aluminum and painted yellow, as many flowering plants in the area display a similar colour. They were placed on top of the soil and weighed down by rocks to ensure they would not move due to weather disturbances. Any rocks placed in the traps were below the water line so as not to allow any arthropods a potential to escape once they had fallen in. In total, 3 pitfall traps and 3 pan-traps were set up per replicate, approximately a meter away from, and surrounding the origin. Traps were collected three days later and all six traps from each replicate were combined, filtered of water, and preserved in ethanol until they could be counted. Trap and net data were pooled following identification. Identification and counting of preserved species occurred based on a previously established arthropod library of the area.

Plant Sampling

Plant species abundance and diversity were assessed using point frame analysis. From the origin at each replicate, a 0. 5m long point frame was used twice for a total of 1m and 20 pinpoints along the transect in a singular direction, and always the same direction for each replicate. Every time the pin hit a plant, it was identified and counted. Plants may have multiple hits in one given pin point. Counting continued until the pin hit either the ground or litter. This data was used to calculate percent cover. Since a single pin point could hit more than one plant, it was possible to achieve percent covers exceeding 100%. In addition to plant diversity assessments, estimates of plant organic material and litter biomass for the area were taken by laying a transect of 0. 2x0. 5m at the end of the point frame. Plant matter and litter was separated and removed from the grid before being dried at 60°C upon returning to the lab. Finally, a 10x10m area was marked in the south-east corner of the initially marked larger area for each treatment site to assess plant species diversity. Abundance was not taken into account, only if the species was present or not. However, since there was only 1 replicate completed per site, statistical and error analysis were unable to be completed.

Statistical Analysis

Data was statistically assessed using single factor ANOVAs to compare means in which there were three treatments (ungrazed, intermediate grazing, and high grazing). Significance was assessed based on a p<0. 05 level. If there was a significant difference between the three treatments, assessment was followed up with a Student’s two-tailed t-test assuming equal variance to compare between two treatments. Calculations and assessments of data were done using Excel. All error bars and error estimates were done using the standard error of the mean (SE).

Results

Arthropod Sampling Although no pan or pitfall traps were disturbed by cattle activity, there was a significant amount of rain on the day of arthropod collection which may have had an effect on estimates either due to reduced arthropod presence, or the washing away of some already captured samples, especially concerning the pan traps. However, all data collected was still considered since this was consistent across all three treatments. There was no significant difference in mean abundance of distinct arthropod species present between the three sites (ANOVA, F2, 21 = 2. 51, p=0. 105). The ungrazed, intermediately grazed, and highly grazed treatments showed mean abundance of species present to be 32. 9(±3. 0), 27. 0(±1. 7), and 26. 1(±2. 1) respectively (fig. 1). Similarly, there was no significant difference in the mean number of individuals captured at each site (ANOVA, F2, 21=2. 77, p=0. 086). The ungrazed, intermediately grazed, and highly grazed treatments showed mean number of individuals captured to be 185. 0(±23. 5), 156. 1(±24. 4), and 111. 1(±18. 7) respectively. Furthermore, there was no significant difference between mean arthropod species richness, as calculated by the Mehinick’s index (ANOVA, F2, 21 = 0. 35, p=0. 712). The ungrazed, intermediately grazed, and highly grazed treatments showed a species richness index of 2. 48(±0. 22), 2. 32(±0. 27), and 2. 62(±0. 27) respectively (fig. 3). Finally, there was no significant difference in genus diversity, as calculated by the Shannon Weiner index of diversity (ANOVA, F2, 21=3. 22, p=0. 0599). The ungrazed, intermediately grazed, and highly grazed treatments showed a diversity index of 1. 13(±0. 05), 0. 927(±0. 07), and 1. 67(±0. 07) respectively (fig. 4). Plant Sampling There was a significant difference in the percent coverage of plants, and therefore abundance, between the ungrazed, intermediately grazed, and highly grazed treatments (ANOVA, F2, 21=43. 25, p=0. 000000036). Furthermore, all treatments were found to be significantly different from each other. The intermediately grazed treatment had significantly more plant cover than either the ungrazed treatment (two tailed t-test, t14=2. 18, p=0. 047) as well as the highly grazed treatment (two tailed t-test, t14=10. 96, p=0. 000000029). Additionally, the ungrazed treatment had significantly more plant coverage than the highly grazed treatment (two tailed t-test, t14=6. 68, p=0. 00001). The mean percent coverage in the ungrazed, intermediate, and highly grazed treatments were 118. 1% (±12. 8%), 154. 4% (±10. 6%), and 26. 9% (±4. 7%) respectively. There was a significant difference between the litter biomass across the three treatments (ANOVA, F2, 21=7. 75, p=0. 003) as well as the plant organic matter biomass between the three treatments (ANOVA, F2, 21=18. 0, p=0. 00005). Once this was established, a Student’s two-way t-test was completed to compare significant differences between two of the three treatments. It was shown that plant biomass was significantly less in the highly grazed treatment compared to the intermediate treatment (two sample t-test, t14=-2. 73, p=0. 016), as well as the ungrazed treatment (two sample t-test, t14=-7. 11, p=0. 000005). Similarly, there was significantly less plant biomass in the intermediate treatment compared to the ungrazed treatment (two sample t-test, t14=-4. 05, p=0. 0012). The mean plant biomass of the ungrazed, intermediately grazed, and highly grazed were 6. 63g(±1. 4g), 13. 85g(±2. 1g), and 26. 43g(±2. 3g) respectively (fig. 6, blue bars). In terms of the litter biomass, a similar pattern emerged. The highly grazed treatment litter biomass was significantly less compared to the intermediately grazed (two sample t-test, t14=-4. 44, p=0. 0006), and the highly grazed (two sample t-test, t14=-3. 42, p=0. 0041) treatments. However, there is no significant difference between intermediately grazed and ungrazed treatments in terms of litter biomass (two sample t-test, t14=-1. 686, p=0. 114). The mean litter biomass of the ungrazed, intermediately grazed, and highly grazed were 1. 59g(±0. 40g), 19. 16g(±4. 0g), and 38. 55g(±10. 8g) respectively (fig. 6, orange bars).

Discussion

Effects of Grazing on Arthropods

One surprising outcome of this experiment is that there was no distinct difference or pattern between treatments regarding the abundance or diversity of arthropods. This did not support our predictions, as we expected that the intermediate treatment would produce the highest species richness, and the ungrazed site would produce the largest abundance of individuals. Unfortunately, there was no significant difference in mean species abundance, mean number of individuals captured, mean species richness, or mean order diversity between the three treatments. Though one might be inclined to believe that the data suggests grazing has no impact on arthropod abundance or diversity, this may be misleading. In a study done on mites, they found that the abundance and diversity of these arthropods was significantly different between different grazing intensities. Similarly, data from previous years undertaking the same study would not support our data from this year. We believe that due to the cold weather and large amount of rain, these two factors in conjunction may have had confounding effects on our study concerning arthropods. In general, arthropods have reduced activity levels in colder weather since their metabolism decreases and locomotion becomes restricted. This may have not been a huge factor during the day we used the sweep net but may have impacted the number of arthropods who fell into traps, which is sufficient to alter the data. Additionally, the amount of rain may have washed out some of the arthropods who were previously drowned, especially where the aluminum pan traps were concerned. There is very little to no evidence to support the data we collected in this study.

Effects of Grazing on Plants

The effects of grazing on plants followed our predictions with few exceptions. As expected of litter and plant biomass calculations, both significantly increased with decreasing grazing pressure Simply put, the cattle were consuming less and therefore more plant biomass from the previous year was left to settle. Additionally, cattle are not hindering the current plants from growing to the same extent. The only exception to this is that there was no significant difference between litter biomass of the intermediate and ungrazed areas. This suggests that there may have been similar plant biomass in the selected areas the year prior to this study.

Overall, the intermediate site had significantly more percent coverage than both the highly grazed and ungrazed site. Though the specific reason for this is not known, it is likely due to another factor we did not take into account in this study such as the effects of cattle grazing on soil quality or perhaps the specific species composition. Based on this data, we can conclude that although high levels of grazing have evidently detrimental effects on plant diversity and abundance, a moderate level of grazing may indeed be beneficial to some extent. Our data is supported by a study in which they assessed high impact areas of cattle and found that not only was soil moisture reduced, but so was the quality of it which they believe led to a reduced species diversity. Another study found similar results in that lightly grazed areas suffered no long term consequences other than minorly reduced recovery rate.

Ideally, allowing time for recovery between grazing periods would be beneficial as well, since the timing of grazing is also a factor which should be taken into account. Furthermore, the scope of this study was only designed to consider two trophic levels of this system; plants and arthropods. If possible, it may be beneficial to better understand how other trophic levels, such as predators of wild ungulates, may play a part in the greater picture of the development of the fescue grasslands. Similarly, it may also be beneficial to consider how abiotic factors play a role as well, such as fire suppression.