Thompson Rivers University
FINAL UREAP REPORT
September 2019
Plant Response to Grazing: Observing Changes in Community Structure and Individual
Characteristics
Written By:
Sarah Bayliff
Bachelor of Science, Major in General Biology
Supervised by: Dr Lauchlan Fraser and Dr. Wendy Gardner

1

1. Introduction
Plants, like all living organisms, are capable of responding to their environment by adapting and
evolving to better survive and reproduce. The phenomenon of this plant adaptation is known as
phenotypic plasticity, when a genotype is capable of expressing different phenotypes in
response to different environments (West-Eberhard 2008). Put more simply, phenotypic
plasticity refers to the ability of a plant to change its characteristics in response to
environmental variables, such as available nutrients, weather conditions, or disturbance.
Among grasslands throughout British Columbia, grazing is a common disturbance that can
affect the expression of plant characteristics and the structure of plant communities.
A plant’s typical response to grazing includes changing its characteristics to be tolerant of, or
resistant to, grazing (Wang et al. 2017). Characteristics known to show plasticity include size
and number of leaves and tillers (Wang et al. 2017), as well as plant height (Olson and
Wallander 1997). The direction and magnitude of which these characteristics will change
depends on several variables, including grazing intensity and frequency (Wang et al. 2017),
annual precipitation (Lohmann et al. 2017), and available light and nutrients (Hayashi et al.
2007). Typically, plants that have been subject to long term grazing will exhibit a lower number
of shoots or tillers, smaller and fewer leaves, and will grow closer to the ground. This change in
characteristics is an attempt to become grazing resistant or tolerant by lowering the likelihood
that they will be grazed and reducing the amount of lost plant material when grazing does
occur (Launchbaugh and Walker 2006).
The structure of plant communities is also subject to change when grazing occurs. As grazing
intensity changes, the functional groups and composition of the plant community also change
(Zhang et al. 2018). There are mixed opinions on the ideal grazing intensity to manage the
health of grasslands. Some studies argue that the occasional ‘resting’ of grasslands, meaning a
period where no grazing occurs, can benefit grassland health and productivity by increasing the
amount of palatable species, and reducing the amount of undesirable species (Zhang et al.
2018). Common between most papers, a high grazing intensity is thought to reduce grassland
health and forage availability by increasing the amount of weeds present, and lowering the
amount of grass species (Zhang et al. 2018, Pakeman RJ. 2004). However, it is also thought that
grazing at a proper intensity can increase grassland health and biodiversity by reducing the
amount of competitive and dominant species, allowing other species to better compete and
successfully grow (Zhang et al. 2018).
Managing the health of a grassland is not only important for the ecosystem, but should also be
a concern of cattle ranchers who have grazing rights to the land. Biodiversity is important for an
ecosystem to maintain function and productivity (Maestre et al. 2012), as well as increasing the
2

resistance of a grassland to grazing (Lyons and Hansel 2001). As biodiversity and grazing
resistance increases, so does the health and productivity of a grassland. With higher
productivity comes more available forage for cattle, increasing the profitability of a rancher’s
rangeland.
This experiment explored how plant characteristics and community structures change in
response to grazing. Plant characteristics focused on throughout this project were plant height,
number of tillers produced per plant, and average leaf area. Available biomass and vegetation
surveys were also examined. To compare the effects of grazing, three different treatments
were performed. An exclusion cage treatment was constructed to examine plants which were
not grazed for a single grazing event, but had been grazed in previous years. A grazed treatment
consisted of area that has been continuously grazed, including throughout the study months.
The final treatment involved a range reference area, where plants have not been subject to any
cattle grazing since 1920 (Government of British Columbia 2003). It was hypothesized that
grasses that have been exposed to long-term grazing, such as those in the cage and grazed
treatments, would show signs of grazing resistance. As the literature suggests, it was thought
that plants with higher grazing resistance would exhibit a lower plant height, fewer number of
tillers, and reduced leaf area. It was also hypothesized that grazed sites would have higher
measures of diversity as grazing reduces competition, allowing less-dominant species to
succeed.
2. Materials and Methods
2.1 Site Selection
Two study sites were selected for this project, one near Merritt and one in the Lac du Bois
grasslands. The Merritt site was located in the Drum Lake Pasture, approximately 3 kilometres
north of the Laurie Guichon Memorial Grasslands Interpretative Site. This site was also used by
Master’s students conducting grazing trials to try to control spotted knapweed and there were
three 50m x 50m enclosures already laid out. On May 8th, 2019, six exclusion cages were set up
around the existing enclosures (see Figure 1). The dimensions of the exclusion cages were
approximately 1.5m x 1.5m. All sampling at the Merritt site occurred on June 12th, 2019.

3

Figure 1: Grazing exclusion cage overlooking the town of Merritt
The Lac du Bois site consisted of four different site locations, each containing five exclusion
cages. These cages had been previously set up by a past student and have been in place since
the summer of 2017. Due to time constraints, only two of the five cages at each site were
sampled from, to give a total of eight cage samples. All sampling at the Lac du Bois site was
performed on June 19th, 2019 (see Figure 2).

Figure 2: View of Kamloops from one the Lac du Bois cage sites
4

Grazed sites were selected by placing a 1m x 1m quadrat frame five metres to the east of the
exclusion cage, measured by a transect tape. This was done for each exclusion cage, giving six
grazed samples from the Merritt site and eight grazed samples from the Lac du Bois site. In
Merritt, sampling was also done within a range reference site. To determine sites here, two
diagonal transects were run across the enclosure, with samples being taken every fifteen
metres, for a total of six samples.
2.2 Vegetation Surveys
For each cage, grazed, and reference site, a vegetation survey was completed to estimate
species abundance and density. Surveys were done within a 1m x1m quadrat frame. Plants
were identified to species and percent cover of the area was estimated for each species.
2.3 Biomass and Plant Characteristics
To attempt to see changes in species between the treatments, biomass and several plant
characteristics were measured. As the cage sites were also being used in other studies, these
measures had to be relatively small so as to not disturb the site too much. To select where to
sample from, a 50cm x 50cm quadrat frame was placed in the bottom left corner of the
previously placed 1m2 quadrat, then a 25cm x 25 cm quadrat frame was placed in the bottom
left corner of the 50cm2 quadrat (see Figure 3).

Figure 3: Quadrat setup for vegetation surveys, plant characteristics, and biomass samples
5

Biomass samples were taken from within the 25cm x 25cm quadrat by harvesting all available
vegetation, cutting it to ground level, and placing it in a labeled paper bag. Biomass samples
were taken from each site for a total of 18 samples from the Merritt site and 16 samples from
the Lac du Bois site. The biomass samples were then put in the drying oven for 48 hours at 65
degrees Celsius. Once dried, the samples were then weighed to give an idea of the available
forage and plant density between treatments.
Within the remaining area of the 50cm x 50cm quadrat, plant characteristics were measured.
Several of the dominant species (i.e. grasses and invasives) were selected and plant height, tiller
number, and leaf area were measured for each individual plant. Plant height was determined
by measuring from the plant base to the tallest point of the plant. Tiller number was only
measured for grasses and was determined by counting the total number of tillers, or stems, on
each individual grass. In order to measure leaf area, the entire plant was harvested and placed
in a plastic bag with moist paper towel in order to keep the leaves hydrated. Once back at the
lab, three leaves from each plant were selected to quantify leaf area. These leaves were taped
to a transparent folder cover and were then scanned to transfer their image onto the computer
(see Figure 4). Originally, for the Merritt plants, all leaves were analysed. However, it proved
too time consuming to transfer every single leaf so the protocol was changed for the Lac du
Bois site and only three leaves per plant were transferred. In order to select three leaves from
the Merritt plants, each leaf was numbered and a random number generator was used to select
three leaves to be analyzed for area. The scanned leaf images were analyzed by ImageJ, an
image processing program, to determine the area of each leaf. Once the areas of the three
leaves were determined, the average was taken of those values to give an average leaf area for
each individual plant.

6

A)

B)

C)

D)

Figure 4: Scanned images of leaves, A) Idaho Fescue from Merritt reference area, B) Spotted
Knapweed from Merritt cage treatment, C) Bluebunch Wheatgrass from Lac du Bois cage
treatment, D) June Grass from Lac du Bois cage treatment

7

2.4 Statistical Analysis
Due to low sample size, plant characteristics were grouped to functional groups. The Merritt
site compared groups of grasses and invasives between the treatments. The Lac du Bois site
only compared grasses. Comparisons were made between Merritt treatments, between Lac du
Bois treatments, and also between Merritt and Lac du Bois sites. All data was tested for the
assumptions of fitting a normal distribution and having equal variances. The majority of the
data met these assumptions and thus underwent parametric tests. To compare between the
three Merritt site treatments, ANOVAs were used. If a p-value of less than 0.05 was returned,
subsequent two-sample t-tests were performed to find where the difference between
treatments was occurring. The Lac du Bois site only had two treatments which were compared
using two-sample t-tests.
Several groups of data did not meet the assumptions of parametric tests. Merritt biomass, Lac
du Bois tiller number, and Lac du Bois biomass data sets were able to be transformed using the
log10+1 function. Once the data was transformed, it fit the assumptions of parametric tests and
were then able to undergo ANOVA and t-test statistical tests. Merritt tiller number was unable
to be transformed and had to be analyzed using non-parametric tests. Non-parametric tests
included Kruskal-Wallis and Mann-Whitney U tests.
The values from the vegetation surveys were analyzed using the Shannon-Wiener Diversity
index, the Simpson Diversity index, and counts of species richness. Once calculated, these data
sets were also tested for normality and equal variances. Once confirmed that the data met the
assumptions of parametric tests, ANOVAs and t-tests were performed for these three measures
of diversity to test for different measures of diversity between treatments.
3. Results
3.1 Merritt site results
Grasses in the reference area showed several differences from grasses in the cage and grazed
treatments. Reference grasses were significantly taller (p=<0.001) and showed a larger leaf area
(p=0.0456) than grasses that had been in the grazed treatment. Reference area grasses also
appeared to have significantly larger numbers of tillers than grasses in the cage (p=<0.05) and
grazed treatments (p=<0.05) (see Figure 5). Grasses inside the cages were found to be
significantly taller than grasses that had been grazed (p=<0.001).

8

Figure 5: Comparison of average number of tillers on grasses between treatments at the
Merritt site
When looking at the invasive functional group, which contained only spotted knapweed, plants
in the grazed treatment were significantly shorter than knapweed plants in the cage (p=0.0013)
or reference treatments (p=0.0009), see Figure 6.

40

Average Height (cm)

35
30
25
20

*

15
10
5
0

Cage

Grazed

Reference

Treatment

Figure 6: Comparison of average height of spotted knapweed plants between treatments at the
Merritt site

9

Upon analysis of the diversity indices, the Shannon-Weiner diversity index gave the grazed sites
a significantly higher value of biodiversity than the reference sites (p=0.036). Similarly, the
reference site also showed a significantly lower species richness than the cage (p=0.0045) and
grazed (p=0.0015) treatments (see Figure 7).

Average Species Richness

12
10
8

*

6
4
2
0

Cage

Grazed

Reference

Treatment

Figure 7: Comparison of species richness between treatments at the Merritt site
3.2 Lac du Bois site results
There were no significant differences between treatments for any variables at the Lac du Bois
site.
3.3 Comparison between Merritt and Lac du Bois sites
Lac du Bois grasses in the cage treatment were significantly higher than grasses in the Merritt
cage treatment (p=0.009). Grazed grasses in Lac du Bois were also significantly taller than
grazed grasses in Merritt (p=0.0002), see Figure 8. Lac du Bois grasses in cages showed a
significantly lower number of tillers than grasses in the Merritt cages (p=<0.05). When
comparing leaf area, Lac du Bois grasses had a significantly higher value than Merritt grasses for
the cage (p=<0.001) and grazed treatments (p=<0.001). Lac du Bois grasses from the cage and
grazed treatments were also compared to grasses within the Merritt reference site. Leaf area
was found to be significantly lower in reference area grasses than both Lac du Bois cage
(p=<0.001) and grazed grasses (p=<0.001). There were no other significant differences for any
other variables between the Lac du Bois grasses and the Merritt reference area grasses.

10

Average Height

80
70
60
50
40
30
20
10
0

*

Merritt - Cage

Lac du Bois - Cage

Site

A)
Average Height

70
60
50
40
30
20
10
0

B)

*

Merritt - Grazed

Lac du Bois - Grazed

Site

Figure 8: Comparison of average height of grass species between sites, A) Cage treatments, B)
Grazed treatments
Diversity index values were also compared between Merritt and Lac du Bois sites. The ShannonWiener H-values given showed that grazed Merritt sites having significantly higher diversity
than grazed Lac du Bois sites (p=0.008). The Simpson index also showed grazed Merritt sites as
more diverse than grazed Lac du Bois sites (p=0.0103). Lastly, species richness for grazed
Merritt sites was significantly greater than that of grazed Lac du Bois sites (p=0.014).
4. Discussion
4.1 Plant Height
At the Merritt site, grass height was found to be significantly lower in grazed plants when
compared to cage and reference plants. While it is thought that plant species respond to
grazing by lowering their height and growing lower to the ground (Launchbaugh and Walker
2006), this is likely not the case here as there was no difference between grasses in the cages
and grasses in the reference area. What is the more likely explanation for this is herbivory. The
plants in the grazed treatment were subject to grazing. When cattle graze, they generally

11

decrease the height of the plant as defoliation occurs. For this reason, the difference in grass
height between treatments is not an important finding.
The invasive functional group however, did produce interesting results when comparing heights
between treatments. Spotted knapweed plants had a significantly lower height in the grazed
treatment than in other treatments. While this is also likely due to defoliation, this result has
more significance than the grass functional group. Shorter knapweed plants in grazed area
suggests the cows have eaten this invasive plant. This supports the use of cattle grazing as a
successful way of controlling the growth and spread of this competitive invasive species using
targeted grazing practices.
4.2 Number of Tillers
Merritt grasses in the reference area appeared to have a greater number of tillers than grasses
in the cage or grazed treatment. Grasses in the cage and grazed treatments have been
previously grazed and should show grazing resistance, so this result is consistent with the
hypothesis that plants with a higher grazing resistance will lower the number of tillers
produced. The number of tillers a grass produces is a very plastic trait that varies with grazing
intensity (Wang et al. 2017). Previous studies suggest that many grass characteristics tend to
reduce in magnitude in response to disturbance (Li et al. 2015). For example, reducing the
number of tillers produced in response to grazing. These studies’ findings coincide with the
results found in this experiment. As the reference area grasses have been enclosed for nearly
one hundred years, they have not been subject to high grazing pressure and haven’t had to
change any characteristics as a response to grazing disturbance. The grasses within the cage
and grazed areas however, have been subject to grazing and may have produced phenotypic
changes in order to adapt to continuous grazing pressure, thus lowering the number of tillers
produced.
4.3 Leaf Area
Leaf area is another plant characteristic that is subject to change in response to disturbance
(Wang et al. 2017). At the Merritt site, grasses in the reference area had a significantly larger
average leaf area than grasses in the cage and grazed treatments. Again, this is possibly due to
the miniaturization of plant traits that occurs in response to grazing. The grasses within the
cage and grazed treatments have been exposed to grazing and perhaps have adapted to be
more grazing resistant by lowering the area of their leaves. This is unlike the reference area
grasses which have not been grazed and have not had the need to adapt to such disturbances.
4.4 Diversity Indices
Grazing can alter ecosystem structure and change the composition of plant communities (Zhang
et al. 2018). Upon analysis of the three Merritt treatments, it appeared as though the Merritt
reference area had a lower measure of diversity than the cage and grazed treatments. This
result supports the hypothesis that grazing changes community structure, and can be beneficial
12

by increasing biodiversity (Zhang et al. 2018). As grazing occurs, competition is lowered
between different plant species. As defoliation of dominant species occurs, opportunity is given
to less dominant species to successfully grow, increasing the biodiversity of the area (Zhang et
al. 2018).
4.5 Comparing Study Sites
The Merritt site supported the hypothesis that grazed sites would have higher biodiversity and
lower measures of plant characteristics. However, no significant results were found from the
Lac du Bois site. To investigate this finding, all variables were compared between the Lac du
Bois and Merritt sites. Grasses at the Lac du Bois cage and grazed sites were significantly taller
than grasses at the Merritt site. This result suggested a lower grazing intensity in Lac du Bois as
they have not been defoliated as much and do not show signs of reducing their plant height in
response to grazing. Overall, the Lac du Bois grasses were quite similar to the grasses in the
Merritt reference area, the only significant difference between the two being a larger leaf area
in the Lac du Bois grasses. These results suggested that the Lac du Bois grasses, like the Merritt
reference area, have not been subject to intense grazing and thus have not undergone
phenotypic plasticity. The Merritt site, which has been subject to grazing, shows higher diversity
measures and plants with grazing resistant characteristics. From these results, it appeared as
though the Merritt site was more resistant to grazing than the Lac du Bois site. The reason for
no significant results between Lac du Bois treatments was likely due to low frequency, low
intensity grazing, which does not produce drastic phenotypic changes in grasses, as the
response of plants is dependent upon grazing intensity (Zhang et al. 2018).
5. Conclusion
The results of this experiment suggested that in response to grazing, grasses will miniaturize
their characteristics to become more resistant to grazing. In this experiment, the traits that
showed reduction were leaf area and the number of tillers produced. While grazing can alter
the structure of individuals, it can also cause changes at the community level. As shown at the
Merritt cage and grazed sites, it appeared as though proper grazing practices can increase
biodiversity, enhancing the health of the grassland. Overall, grassland ecosystems change
depending on grazing intensity. With proper grazing management practices, grasslands can
become rich in biodiversity while providing sufficient forage for range use.
6. Potential revisions for future studies
If this study were to be repeated, it should take place over a time period longer than four
months. This four-month study was too short to observe phenotypic plasticity in response to
grazing between the cage and grazed plants as there was only a single sampling period. Ideally,
grasses would be measured before grazing, after grazing, and the previous year when they’ve
had the time to possibly adapt to the grazing disturbance.

13

Another area of concern was the biomass samples. No significant differences were found for
any biomass samples for this experiment. This was likely because the samples were so small. If
this study were to be done again, larger biomass samples should be taken at an attempt to
properly quantify the sites.
This study would also benefit from an increased sample size so plant characteristics could be
compared down to species level between treatments to give more accurate results.
References
Government of British Columbia. 2003. Range Reference Areas of the Southern Interior Forest
Region- Drum Lake RRA. [Internet]. Cited September 11th, 2019. Available from:
https://www.for.gov.bc.ca/rsi/range/RRA/cascades/drumlake.htm
Hayashi M, Fujita N, Yamauchi A. 2007. Theory of grazing optimization in which herbivory
improves photosynthetic ability. 248(2):367-376. [Internet]. Cited September 11th, 2019.
Available from: https://www.sciencedirect.com/science/article/pii/S0022519307002494
Launchbaugh K and Walker JW. 2006. Targeted grazing—a new paradigm for livestock
management. Targeted grazing: a natural approach to vegetation management and landscape
enhancement. American Sheep Industry Association. 2006: 1-56. [Internet]. [Cited September
11th 2019]. Available from: http://sccd.org/wp-content/uploads/2015/08/Section-1-principlesand-overview.pdf
Lohmann D, Guo T, Tietjen B. 2017. Zooming in on coarse plant functional types-simulated
response of savanna vegetation composition in response to aridity and grazing. 11(2):161-173.
[Internet]. Cited September 11th, 2019. Available from: https://link-springercom.ezproxy.tru.ca/article/10.1007%2Fs12080-017-0356-x
Li X, et al. 2015. Contrasting Effects of Long-Term Grazing and Clipping on Plant Morphological
Plasticity: Evidence from a Rhizomatous Grass.10(10):1-19. [Internet]. Cited September 11th,
2019. Available from:
https://eds.a.ebscohost.com/eds/pdfviewer/pdfviewer?vid=3&sid=9c449f9a-9adc-4a4b-b85e348a56b3a8f8%40sdc-v-sessmgr01
Lyons RK and Hansel CW. 2001. Grazing and Browsing: How Plants are Affected. AgriLife
Extension Texan A&M System. [Internet]. [Cited September 11th 2019]. Available from:
http://oaktrust.library.tamu.edu/bitstream/handle/1969.1/87088/pdf_1523.pd
Maestre FT, Quero JL, Gotelli NJ, Escudero A, Ochoa V, Delgado-Baquerizo M, Garcia-Gomez M,
Bowker MA, Soliveres S, Escolar C, et al. 2012. Plant species richness and ecosystem
multifunctionality in global drylands. NCBI. [Internet]. [Cited February 2nd 2019]. Available

14

from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3558739/ DOI:
10.1126/science.1215442
Olson B and Wallander RT. 1997. Biomass and carbohydrates of spotted knapweed and Idaho
fescue after repeated grazing. 50(4):409-412. [Internet]. Cited September 11th, 2019. Available
from: https://www-jstororg.ezproxy.tru.ca/stable/4003308?seq=1#metadata_info_tab_contents
Pakeman RJ. 2004. Consistency of plant species and traits to grazing along a productivity
gradient: a multi-site analysis. 92(5):893-905. [Internet]. Cited September 11th, 2019. Available
from: https://www-jstororg.ezproxy.tru.ca/stable/3599387?seq=1#metadata_info_tab_contents
Wang D, Du J, Zhang B, Ba L, Hodgkinson KC. 2017. Grazing Intensity and Phenotypic Plasticity in
the Clonal Grass Leymus chinensis. 70(6):740-747. [Internet]. Cited September 11th, 2019.
Available from: https://bioone-org.ezproxy.tru.ca/journals/rangeland-ecology-andmanagement/volume-70/issue-6/j.rama.2017.06.011/Grazing-Intensity-and-PhenotypicPlasticity-in-the-Clonal-Grass-iLeymus/10.1016/j.rama.2017.06.011.full
West-Eberhard MJ. 2008. Encyclopedia of Ecology. Phenotypic Plasticity. [Internet]. Cited
September 11th, 2019. Available from: https://www.sciencedirect.com/topics/agricultural-andbiological-sciences/phenotypic-plasticity
Zhang C, Dong Q, Chu H, Shi J, Li S, Wang Y, Yang X. 2018. Grassland Community Composition
Response to Grazing Intensity Under Different Grazing Regimes. 71(2):196-204. [Internet]. Cited
September 11th, 2019. Available from: https://bioone-org.ezproxy.tru.ca/journals/rangelandecology-and-management/volume-71/issue-2/j.rama.2017.09.007/Grassland-CommunityComposition-Response-to-Grazing-Intensity-Under-DifferentGrazing/10.1016/j.rama.2017.09.007.full

15