Elliott W.R. Parsons
Online Supporting Information
Parsons, Elliott, J.L. Maron, T.E. Martin: Elk herbivory alters small mammal assemblages in high elevation drainages
Appendix S1. Studies that quantified the effects of large herbivores on small mammal community richness and diversity.
Reference
Large herbivores excluded
Delta richness
Delta diversity
(inside exclosures) (inside exclosures)
Moulton 1978
Cattle
Decrease
Not determined
Medin & Clary 1989
Cattle
Increase
Increase
Schulz & Leininger 1991
Cattle
No effect
No effect
Rosenstock 1996
Cattle
Increase
Not determined
Jones & Longland 1999
Cattle
No effect
Not determined
Eccard et al. 2000
Cattle
Increase
Not determined
Moser & Witmer 2000
Cattle and elk
Increase
Increase
Chapman & Ribic 2002
Cattle
Increase
Increase
Beever & Brussard 2004
Feral horses
No effect
No effect
Giuliano & Homyack 2004 Cattle
Increase
Not determined
Hoffmann & Zeller 2005
Cattle
Increase
Increase
Valone & Sauter 2005
Cattle
Increase
Not determined
1
Elliott W.R. Parsons
Mathis et al. 2006
Cattle
No effect
Saetnan & Skarpe 2006
Greater kudu, red hartebeest, impala, warthog Increase
Not determined
Yarnell et al. 2007
18 African ungulates, (page 390).
No effect
No effect
No effect
Literature Cited, Appendix S1
Chapman, E.W. & Ribic, C.A. (2002) The impact of buffer strips and stream-side grazing on small mammals in southwestern
Wisconsin. Agriculture, Ecosystems and Environment, 88, 49-59.
Beever, E.A. & Brussard, P.F. (2004) Community-and landscape-level responses of reptiles and small mammals to feral-horse grazing
in the Great Basin. Journal of Arid Environments, 59, 271-297.
Eccard, J.A., Walther, R.B. & Milton, S.J. (2000) How livestock grazing affects vegetation structures and small mammal distribution
in the semi-arid Karoo. Journal of Arid Environments, 46, 103-106.
Giuliano, W.M. & Homyack, J.D. (2004) Short-term grazing exclusion effects on riparian small mammal communities. Journal of
Range Management, 57, 346-350.
Hoffmann, A. & Zeller, U. (2005) Influence of variations in land use intensity on species diversity and abundance of small mammals
in the Nama Karoo, Namibia. Belgian Journal of Zoology, 135 (supplement), 91-96.
Jones, A.L. & Longland, W.S. (1999) Effects of Cattle Grazing on Salt Desert Rodent Communities. American Midland Naturalist,
141, 1-11.
2
Elliott W.R. Parsons
Mathis, V.L., Whitford, W.G., Kay, F.R., Alkon, P.U. (2006) Effects of grazing and shrub removal on small mammal populations in
southern New Mexico, USA. Journal of Arid Environments, 66, 76-86.
Medin, D.E. & Clary, W.P. (1989) Small mammal populations in a grazed and ungrazed riparian habitat in Nevada. United States
Department of Agriculture: Forest Service Intermountain Research Station research paper, INT-413, 1-6.
Moser, B.W. & Witmer, G.W. (2000) The effects of elk and cattle foraging on the vegetation, birds, and small mammals of the Bridge
Creek Wildlife Area, Oregon. International Biodeterioration & Biodegradation, 45, 151-157.
Moulton, M. (1978) Small mammal associations in grazed versus ungrazed cottonwood riparian woodland in eastern Colorado.
Lowland river and stream habitat in Colorado: a symposium, Greely, Colorado, 133-139.
Rosenstock, S.S. (1996) Shrub-grassland small mammal and vegetation responses to rest from grazing. Journal of Range
Management, 49, 199-203.
Saetnan, E.R. & Skarpe, C. (2006) The effect of ungulate grazing on a small mammal community in southeastern Botswana. African
Zoology, 41, 9-16.
Schulz, T.T. & Leininger, W.C. (1991) Nongame wildlife communities in grazed and ungrazed montane riparian sites. Great Basin
Naturalist, 51, 286-292.
Valone, T.J. & Sauter, P. (2005) Effects of long-term cattle exclosure on vegetation and rodents at a desertified arid grassland site.
Journal of Arid Environments, 61, 161-170.
3
Elliott W.R. Parsons
Yarnell, R.W., Scott, D.M., Chimimba, C.T. & Metcalfe, D.J. (2007) Untangling the roles of fire, grazing and rainfall on small
mammal communities in grassland ecosystems. Oecologia, 154, 387-402.
Literature review methods:
The literature review was conducted using the database: Web of Knowledge/Web of Science. The search terms were Topic =
(herbivory) AND Topic = (small mammal) OR Topic = (exclosure). Citation database was: Science Citation Index Expanded, 1899present. Time-span included “all years” until the date 8/22/12. The search returned 1,538 titles and abstracts. Of these, 25 were
related to the effects of large herbivores on small mammal abundance, density, diversity, or richness. A review of the literature cited
sections from these 25 studies yielded another 14 studies on the same topic. From these 39 studies, 15 (38%) quantified the effects of
large herbivores on small mammal richness (listed in literature cited above). Furthermore, of these 15 studies, 53% also quantified
ungulate impacts on small mammal diversity.
4
Elliott W.R. Parsons
Appendix S2. The locations and names of study drainages.
Drainage classification and location
Drainage Pair
Drainage Name
Treatment
Closest ridge
UTM location
Elevation (m)
1
14
Exclosure
Buck Springs
486213 E, 3806519 N
2367
1
8
Control
Buck Springs
486213 E, 3806519 N
2340
2
E4
Exclosure
Dane
484684 E, 3808369 N
2366
2
E3
Control
Dane
484684 E, 3808369 N
2397
3
12
Exclosure
McClintock
483153 E, 3808372 N
2356
3
11
Control
McClintock
483149 E, 3806524 N
2389
The snowmelt drainages used in this study flow into steeper northward-draining canyons and are paralleled by ridges. The closest
ridge to each study drainage pair is listed above. All UTM locations are in Zone 12N.
5
Elliott W.R. Parsons
Appendix S3. Detailed description of methods for calculating paired differences in relative
abundance.
We calculated the relative abundance of all species (in all drainages and for all primary
trapping sessions) as the number of unique individuals captured divided by sampling effort
(Equation 1) according to Slade and Blair (2000).
𝐸𝑞𝑢𝑎𝑡𝑖𝑜𝑛 1: 𝑅𝑒𝑙𝑎𝑡𝑖𝑣𝑒 𝑎𝑏𝑢𝑛𝑑𝑎𝑛𝑐𝑒 =
𝑢𝑛𝑖𝑞𝑢𝑒 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙𝑠
𝑠𝑎𝑚𝑝𝑙𝑖𝑛𝑔 𝑒𝑓𝑓𝑜𝑟𝑡
𝐸𝑞𝑢𝑎𝑡𝑖𝑜𝑛 2: 𝑆𝑎𝑚𝑝𝑙𝑖𝑛𝑔 𝑒𝑓𝑓𝑜𝑟𝑡 = (𝑡𝑟𝑎𝑝𝑠 ∗ 𝑖𝑛𝑡𝑒𝑟𝑣𝑎𝑙𝑠) − (𝑠𝑝𝑟𝑢𝑛𝑔 𝑡𝑟𝑎𝑝𝑠 ∗ 0.5)
We calculated sampling effort as shown by Equation 2 (Beauvais and Buskirk 1999).
Because traps were checked twice a day for three mornings and four evenings (from 2006 to
2009), all animals had the potential of being caught a maximum of seven times (i.e. intervals = 7)
during a trapping session. During 2004 and 2005, however, traps were only checked once a day
(and thus animals could only be caught a maximum of four times during each trapping session)
so we used four trap intervals in our calculations for sampling effort for these two years. We
subtracted sprung traps (closed traps without animals in them) from traps*intervals because these
traps were not available to catch animals. However, we multiplied sprung traps by 0.5 before
subtraction because the best estimate of when traps close (and thus became unavailable) is half
way between intervals (Beauvais and Buskirk 1999). This method corrects for trap saturation
and variation in trap-springing by site or time and provides a more accurate measure of sampling
effort than simply multiplying traps by trapping intervals (Beauvais and Buskirk 1999).
Sampling effort was important to quantify at our field site because physical disturbances,
6
Elliott W.R. Parsons
including direct sun, wind, rain, and hail, as well as animals (mostly black bears), caused our
traps to close.
We first determined yearly estimates of relative abundance for all drainages by separately
averaging the relative abundance for each species for May, June, and July. We then calculated
paired-drainage differences by subtracting the non-fenced control drainage estimate for a species
from the paired exclosure estimate. By examining changes in paired differences over time (as
opposed to simply comparing exclosure and control averages), we were able to control for
variation in relative abundance between drainage-pairs and across the three trapping sessions.
Finally, we multiplied all relative abundance estimates by the total number of trap occasions
(intervals*total available traps for 2006-2009) to convert the relative abundance estimates to
estimates of the total number of individual animals on a trapping grid.
Beauvais, G.P. & Buskirk, S.W. (1999) Modifying estimates of sampling effort to account for
sprung traps. Wildlife Society Bulletin, 27, 39-43.
Slade, N.A. & Blair, S.M. (2000) An empirical test of using counts of individuals captured as
indices of population size. Journal of Mammalogy, 81, 1035-1045.
7
Elliott W.R. Parsons
Appendix S4.
In summer 2008, the cover of deciduous litter was measured on the experimental drainages. Cover was measured by randomly
sampling litter in 65 of the sampling sub-plots on the experimental drainages (see Martin 1998, 2007 for details regarding vegetation
sampling design). Within 1 m of the center-point of each selected sampling sub-plot, a 50 cm2 quadrat was randomly placed and all
deciduous litter was collected, dried to constant weight and weighed. Also within an 11 m radius circle around each center-point, all
stems of deciduous trees > 2.5 cm diameter breast height (dbh) were counted (in 2008, 75% of the deciduous trees were aspen and
maple combined). There was a significant correlation between deciduous litter cover (mass) and the number of deciduous stems
(Pearson’s r = 0.53, (2-tailed) P < 0.001).
Martin, T.E. (1998) Are microhabitat preferences of coexisting species under selection and adaptive? Ecology, 79, 656-670.
Martin, T.E. (2007) Climate correlates of 20 years of change in birds, plants and trophic interactions in a high elevation riparian
system. Ecology, 88, 367-380.
8
Elliott W.R. Parsons
Appendix S5. A comparison of the coefficients of determination for linear models and log-transformed models for explaining
changes in paired differences over time.
Linear model
Delta R2
Log model
F-value
P-value
Error df
R-squared
F-value
P-value
Error df
R-squared
Lin-Log
Total abundance
12.11
0.004
14
0.54
6.40
0.02
14
0.38
0.16
Total mice
6.95
0.02
14
0.4
5.56
0.03
14
0.36
0.04
Voles
6.94
0.02
14
0.44
5.68
0.03
14
0.40
0.04
We initially used ANCOVA with paired differences in relative abundance as the dependent variable, year as a covariate, and
plot-pair as a fixed factor to test whether changes in paired differences over time increased linearly across years for all three drainage
pairs. For total abundance, total mice, and voles we found that there was no significant drainage-pair*year interaction and all slopes
were positive, indicating that paired differences were positively associated with year for all three drainage pairs. However, this
relationship may not have been linear; for example, paired differences may have increased rapidly following rapid initial changes in
vegetation inside the exclosures compared to the controls, and the magnitude of the increase may have lessened with time (i.e. the
relationship may have been curvilinear).
To test whether this relationship was curvilinear, we first added a positive whole number equal to the absolute value of the
lowest paired difference for all values within each species in order to avoid negative values so we could log-transform them. Then we
9
Elliott W.R. Parsons
performed the same ANCOVA as above and examined the coefficient of determination for both the linear and log-transformed
models. In these three groups the log-transformed models performed worse (lower R2 values) than the linear models.
10
Elliott W.R. Parsons
Appendix S6. Small mammal trapping summary results: 2004 to 2009. The two species of Peromyscus were only successfully
differentiated in 2006. Between 2006 and 2009 we captured 1,505 individuals of Peromyscus maniculatus and 416 individuals of
Peromyscus boylii. This table does not include shrews (Sorex merriami) or Northern pocket gophers (Thomomys talpoides) because
we did not catch them frequently enough to include them in the analyses.
Genus
# individuals
% of total
# of captures
% of total
Captures/individual
Peromyscus
2566
61.79
6947
68.48
2.71
Tamias
751
18.08
1467
14.46
1.95
Neotoma
342
8.24
758
7.47
2.22
Microtus
222
5.35
418
4.12
1.88
Spermophilus
165
3.97
356
3.51
2.16
Tamiasciurus
107
2.58
198
1.95
1.85
11
Elliott W.R. Parsons
Appendix S7. We found a significant positive association between paired differences in shrub
cover (x-axis) and paired differences in the relative abundance of mice (y-axis).
20
Relative abundance
(exclosure - control)
10
0
-10
-20
-30
-40
-50
-10
-8
-6
-4
-2
0
2
4
6
Percent shrub cover (exclosure - control)
12
8
Elliott W.R. Parsons
Appendix S8. Differences in understory vegetation between exclosures and non-fenced controls
in 2009.
A
B
C
D
Five years after construction of the fences (i.e. 2009) there was noticeably higher cover of A)
maple, B) aspen, C) New Mexico locust, and D) grass cover inside (right) versus outside (left) of
the ungulate exclosures.
13