2009 SOUTHEASTERN NATURALIST 8(3):437–444
Mud Track Plots: An Economical, Noninvasive Mammal
Survey Technique
Ross R. Conover1,* and Eric T. Linder2
Abstract - The active component (e.g., bait or scent lures) common to mammal survey
techniques can bias detectability (e.g., aversion or attraction) and reduce their efficacy
for spatially explicit, habitat-use research. We overcame this obstacle with 1-m2 mud
track plots, which use natural ground substrates to sample mammalian species’ occupancy
with minimal bias of natural movements. Performance of this method was based
on criteria that included implementation effort, cost, species detection, proportion of
plots to capture tracks, and species identification by track capture. Mud track plots
were quickly established, maintained, and monitored with minimal effort and cost.
We recorded tracks in 85% of plots over 8 nights, documented all visually confirmed
medium and large-sized mammals (>800 g), and captured identifiable (>98%) tracks
on a row-crop farm in the Mississippi Alluvial Valley. Drawbacks included limited application
for small mammals, potential ambiguity of using track imprints for species
identification, and weather-dependent, spatio-temporal restrictions. Mud track plots
are an inexpensive, simple approach to noninvasive mammal detection.
Introduction
Occupancy information provides insight on a variety of spatio-temporal
habitat-use patterns and ecological effects (Vojta 2005). Several noninvasive
(i.e., does not require capture or handling) survey techniques have been
successfully implemented to obtain occupancy data for varied taxa and research
objectives (Connors et al. 2005, Mooney 2002, Savidge and Seibert
1988), with some common methods including camera traps, track plates,
snowtracking, scent stations, and scat surveys (Gompper et al. 2006). The
versatility and low-cost of track-plate methods has resulted in their use to
evaluate area species-richness (Silveira et al. 2003), predator activity (Connors
et al. 2005, Kuehl and Clark 2002), general habitat-use (Fecske et al.
2002), relative abundance (Mooney 2002), and species occupancy (Mowat
et al. 2000, Smith et al. 2007). The efficacy of track plates depends on artificial components (e.g., bait, scent, boxed enclosures, or media-tracking
platforms) to enhance mammal attraction and track identification (Foresman
and Pearson 1998). However, the potential of these components to invoke
attraction or aversion responses may also limit track-plate applicability for
research that requires information on natural mammal movements.
Track-capture methods that detect mammals without infl uencing their
natural movements can elucidate habitat-use patterns at smaller spatial scales
1Department of Natural Resource Ecology and Management, Iowa State University,
339 Science II, Ames, IA 50011. 2Department of Biological Sciences, University of
Texas – Brownsville, 80 Fort Brown, Brownsville, TX 78520. *Corresponding author
- melospiza77@yahoo.com.
438 Southeastern Naturalist Vol. 8, No. 3
more accurately than techniques that incorporate artificial components. For
example, some carnivores have exhibited wariness to enter track-plate stations
despite deployment of bait and scent lures (Gompper et al. 2006, Vanak
and Gompper 2006). Although track-plate efficacy differs relative to platform
and media types (Belant 2003, Wemmer et al. 1996), Procyon lotor (Raccoon)
demonstrated reluctance to step on artificial platforms (Wolf et al. 2003)
and the lack of platforms altogether has increased visitation rates for foxes
(Sargeant et al. 2003). Snowtracking evades these issues and captures tracks
from natural mammal movements by avoiding artificial components; however;
its use is restricted to snow-covered regions (Gompper et al. 2006). Sand
or raked-soil track plots are plausible alternatives because they capture tracks
more naturally (Bider 1968). However, their ability to register identifiable
mammal tracks may be unreliable (Glen and Dickman 2003), without mineral
oil (Kuehl and Clark 2002, Renfrew et al. 2005), and albeit untested, some
mammals may exhibit wariness of sand where it represents a foreign edaphic
substrate. Hence, a technique that detects mammal presence with minimal bias
should avoid using artificial or foreign components whenever possible.
We propose mud track plots, which use natural ground substrates in an area
cleared of debris to capture tracks, as an inexpensive technique to detect mammals.
Furthermore, this approach to track capture avoids use of bait, foreign
substrates, or platforms, thus reducing potential sources of movement bias.
The performance of mud track plots was evaluated based on criteria that includes
1) implementation effort, 2) cost, 3) species detection, 4) proportion of
plots to capture tracks in ≤3 days, and 5) species identification (Foresman and
Pearson 1998). We predict that properly employed mud track plots will detect
mammal occupancy under natural movement conditions.
Methods
Study site
We conducted this study intermittently from May–July, 2004 on a soybean
(Glycine sp.) and cotton (Gossypium sp.) row-crop production farm in
Sunfl ower County, MS, located in the Mississippi Alluvial Valley (MAV)
physiographic region. The study area (33°20'N, 90°40'W; approximately
615 ha) was representative of the MAV landscape, being dominated by large
(89.6 ± 15.7 ha), row-crop agricultural fields fragmented by wooded fencerows
and linear riparian zones (i.e., streams, rivers, drainage ditches) with
negligible topographic relief. Mud track plots were established immediately
adjacent to 10-m wide, herbaceous strips juxtaposed to a wooded fencerow
with an embedded drainage ditch (Conover et al. 2007). The ground substrate
in the MAV was hard when dry, though softened considerably with
moderate rainfall. Soil associations were mostly Dundee silt loam or Forestdale
silt loam, which are stratified alluvium soils of fine to coarse texture
with poor to moderate drainage (Powell et al. 1952). Daily precipitation
measurements were obtained from the nearby (<25 km) US Department of
Agriculture weather station at Beasley Lake, MS.
2009 R.R. Conover and E.T. Linder 439
Mud track plots
Our mud track plots were a 1-m2 region cleared of all vegetation and litter,
such that only bare ground was exposed. We evaluated implementation effort
and cost to establish, maintain, and monitor these plots based on equipment
and manual labor requirements. We established mud track plots by first removing
standing vegetation in a pre-measured plot, then scraping the dry ground
with the fl at side of a hand shovel to eliminate residual vegetation and grade
the surface to maximize track detail. We wore gloves and rubber boots during
plot preparation to minimize anthropogenic cues that may infl uence mammal
behavior. Ensuing litter piles were removed to avoid obstructing the natural
travel pathway of mammals; however, surrounding vegetation remained intact.
We established 46 mud track plots during early May of 2004. Plots were
re-scraped between data collection periods and otherwise every ten days to
prevent track recounts and suppress vegetative growth to ensure the capture
of clean track imprints under appropriate conditions. The 46 mud track plots
were established in a paired design such that plots were in 23 adjacent sites at
50-m intervals, and located on opposite edges of 10-m wide herbaceous habitat
strips. The ability of ground substrates to capture track detail depended on
a suitable combination of precipitation and drying time. We determined mud
track plot viability on post-rainfall mornings by the presence of identifiable
mammal tracks or using finger imprints to test plot texture. As precipitation
can vary locally in the MAV, all mud track plots were located within 3.5 km to
maximize track plot viability in isolated precipitation events.
Track capture
We collected track data during four distinct periods (1–3 nights per period)
that followed precipitation events of >1.0 cm. We evaluated the speciesdetection
criteria by comparing medium- and large-sized mammals (>800 g)
detected by track plots with mammals confirmed to be present through opportunistic
sightings on the farm from May–July over three years (2002–2004).
We measured the proportion of plots that captured tracks per tracking period
to evaluate potential mammal wariness toward mud track plots. Other studies
commonly recorded an average latency-to-first-detection of >3 days for mustelids
and as such, we expected to capture few tracks in <2 days if mammals
exhibit wariness (Foresman and Pearson 1998, Zielinski 1995). This effect is
particularly relevant for mammals with previously documented wariness (e.g.,
Raccoons, foxes; Sargeant et al. 2003, Wolf et al. 2003). As we combined our
data per tracking period, we report proportion of viable plots to capture tracks
for 1-, 2-, and 3-day tracking periods.
Tracks were not recorded when plots received precipitation at night to
prevent raindrops from effacing track detail. To minimize observer bias, one
individual trained on track identification monitored all mud track plots on
post-rainfall mornings. Pad measurements and photographs were initially
used to confirm species identifications from track imprints (Murie 1954). We
report track data as the number of detections/track plot-night for all mammal
species. Track plot-night is defined as the number of viable mud track plots
440 Southeastern Naturalist Vol. 8, No. 3
per night for all nights in a tracking period. If a track plot was unviable, it
was removed from the track plot sample.
Results
One person established 46-mud track plots in two half-days, with individual
plots being established in approximately 3–15 minutes/plot, dependent
on the amount of vegetation removal required. Maintenance and monitoring
commenced either post-track capture or every ten days, and temporal effort
was approximately 1–5 minutes/plot, dependent on track detail and not
including travel time. Costs were limited to the manual labor required to
establish, maintain, and monitor track plots of a desired sample, as well as
necessary materials (i.e., hand shovel, work gloves, boots).
Throughout the study, mud track plots received enough precipitation to
collect data on at least four distinct occasions (8 total nights), including 28
May (1.07-cm precipitation on previous day), 04–06 June (4.1-cm precipitation
on previous five days), 29–30 June (17.20-cm precipitation on previous
five days), and 16–17 July (3.4-cm precipitation on previous day) 2004.
However, we did not monitor plots during every potential opportunity (i.e.,
precipitation event) from May–July. The edaphic conditions at our study site
were such that 1.07 cm of precipitation was sufficient to capture identifiable
tracks of medium- and large-sized mammals for one night. A two-night data
collection period (29–30 June) that was preceded by 17.20-cm of precipitation
over five days rendered four of forty-six track plots (8.7%) unviable
from standing water. Mud track plots captured identifiable tracks between
1–3 nights per data collection period, which depended on plot texture relative
to total precipitation and drying time (e.g., cloud cover).
Species detection was considered robust because track plots detected all
medium- and large-sized mammals previously confirmed on the study farm
(Table 1). Odocoileus virginianus (White-tailed Deer), Dasypus novemcinctus
(Nine-banded Armadillo), Raccoon, and Canis latrans (Coyote) (30, 19, 10,
and 10%, respectively) were the most frequently detected species (Table 1).
Eighty-five percent of mud track plots successfully recorded tracks at least
once across the study. Using data combined across tracking periods, the proportion
of viable plots that captured tracks included 15% in one night (28 May),
78% in three nights (04–06 June), 81% in two nights (29–30 June), and 74% in
two nights (16–17 July). Mud track plots captured a total of 104 tracks for at
least 11 mammal species (Table 1). Medium- and large-sized mammals produced
descriptive track impressions. We identified all medium- and large-sized
mammal tracks except one (100 and 98%, respectively), which was speculated
to be a large Canis familiaris L. (Domestic Dog). We failed to distinguish
tracks between Sylvilagus fl oridanus Allen (Eastern Cottontail) and Sylvilagus
aquaticus Bachman (Swamp Rabbit) or Sciurus carolinensis Gmelin (Gray
Squirrel) and Sciurus niger Linnaeus (Fox Squirrel), thus recorded them to genus.
Small mammals only represented 5% of visible tracks, of which none were
identified to species (Table 1).
2009 R.R. Conover and E.T. Linder 441
Discussion
We evaluated mud track plot efficacy using implementation effort, cost,
species detection, proportion of plots to capture tracks, and species identification
(Foresman and Pearson 1998). Implementation effort for this technique
was minimal during establishment, maintenance, and monitoring aspects
of application. Initial establishment of 46 plots only required two half-days
by one person. Although establishment effort will vary with geography and
edaphic conditions, this technique has potential for application of large
samples with a relatively nominal time requirement. Maintenance and monitoring
effort depends on plot texture, surface damage, vegetative growth, and
observer track-identification skills. We maintained plots after they had dried
to prevent accidental recounts during subsequent tracking periods; however,
preliminary plot re-scraping prior to drying (after recording track data) facilitated
easier smoothing of dried plots. Many plots required no maintenance
between precipitation events from lack of surface damage or vegetative
growth. As mud track plots required minimal equipment, the primary costs
will be manual labor, which is relative to the desired plot sample.
All mammal species documented on the study farm over 3 summers (2002–
2004) were detected by mud track plots, excepting small mammals (e.g., mice),
sciurids, and lagomorphs due to difficulty in distinguishing species from
tracks. Our three-year, mammal inventory was relatively comprehensive as determined
by repeat sightings of every species. We acknowledge this species list
may be incomplete as it was a product of opportunistic observation and not an
intensive, systematic survey. However, our study farm was also conducive for
visual documentation from a paucity of vegetative cover and mammal diversity.
Furthermore, we did not detect any species on our mud track plots that had
Table 1. Mammal tracks recorded using mud track plots to survey mammal species’ occurrence
on edges of herbaceous strip habitats in the Mississippi Alluvial Valley during 2004. TD = total
detections; May= 28 May; June1 = 4–6 June; June2 = 29–30 June; July = 16–17 July.
Mammal species TD May %A June1 % June2 % July %
White-tailed Deer 31 1 0.022 9 0.065 9 0.107 12 0.130
(Odocoileus virginianus Zimmermann)
Nine-banded Armadillo 20 2 0.043 5 0.036 9 0.107 4 0.043
(Dasypus novemcinctus Linnaeus)
Raccoon (Procyon lotor Linnaeus) 10 1 0.022 2 0.014 5 0.060 2 0.022
Coyote (Canis latrans Say) 10 1 0.022 5 0.036 3 0.036 1 0.011
Red Fox (Vulpes vulpes Linnaeus) 7 0 0.000 1 0.007 2 0.024 4 0.043
Bobcat (Lynx rufus Schreber) 5 2 0.043 0 0.000 2 0.024 1 0.011
Opossum (Didelphis virginiana Kerr) 5 0 0.000 2 0.014 1 0.012 2 0.022
Squirrel (Sciurus spp.) 4 0 0.000 3 0.022 1 0.012 0 0.000
Feral Cat (Felis silvestris Schreber) 3 0 0.000 2 0.014 1 0.012 0 0.000
Striped Skunk 2 0 0.000 1 0.007 0 0.000 1 0.011
(Mephitis mephitis Schreber)
Rabbit (Sylvilagus spp.) 1 0 0.000 0 0.000 1 0.012 0 0.000
Small mammals 6 2 2 0 2
AProportion of detections per species relative to number of viable mud track plot-nights in data
collection period.
442 Southeastern Naturalist Vol. 8, No. 3
not been previously identified in the study area. Maximum species detection
in other studies may be facilitated by appropriate track plot placement (e.g.,
natural corridors, diverse habitats). The success with which mud track plots
quantified species richness for medium- and large-sized mammals supports
previous findings of track census accuracy (Silveira et al. 2003).
Our mud track plots captured mammal tracks on the first night of adequate
substrate texture. First night detections included Raccoons and second-night
detections included Red Foxes, which are both species known to exhibit wariness
of other track capture methods (Sargeant et al. 2003, Wolf et al. 2003).
We speculate that failure of some plots to capture tracks during the first night
may have resulted from lack of mammal activity in that particular area. The
immediacy of mud track plot detection viability supports the idea that mammals
exhibit reduced wariness toward track plots of natural substrates than
track-capture techniques with foreign components (Zielinski 1995). The majority
(74–81%) of plots captured tracks during two and three-night periods,
which may be associated with the duration of rainfall (e.g., 5 days) prior to
track capture, although this is untested. We are unaware of any studies that
report aversion of wild mammals to step on mud substrates. Furthermore, the
ability of these plots to capture tracks immediately (i.e., first night) suggests
equal detectability across tracking-period length, and therefore, data would
be comparable in either a three-night period or three 1-night tracking periods.
Hence, we predict applicability of mud track plots for study areas and seasons
with broad ranges of precipitation events.
Mud track plots captured identifiable tracks of medium- and large-sized
mammals, yielding information on species occurrence. Track plots were
viable after a minimum of 1.0 cm precipitation, depending on plot texture.
Duration of track-plot viability depended on amount of precipitation and
drying time (e.g., cloud cover). This variability is exemplified by a 14.2-cm
precipitation event (over 5 days) that only facilitated two viable nights of track
capture before desiccation occurred from extreme heat, whereas a 4.1-cm
precipitation event (over 5 days) followed by mild temperatures and greater
cloud cover captured tracks for three-nights. The intense heat in the MAV
may reduce duration of track plot viability from desiccation relative to other
areas. Furthermore, we also recorded a one-day 3.4-cm precipitation event
that may have permitted more than two-nights of track capture, but the data
collection ended prematurely from additional heavy rainfall. Alternatively,
we acknowledge the potential value of mild precipitation during track capture
periods to maintain plot saturation, although caution should be used to avoid
track imprint effacement. Furthermore, study areas with more frequent precipitation
may allow more consistent track plot employment. While the lucid
detail of most pad impressions often permitted quick, accurate track identifi-
cation (Murie 1954), there is potential for difficulty in identifying species by
tracks to limit the applicability of this technique, particularly in areas where
species with similar tracks co-occur (Zielinski and Truex 1995). Although the
majority of mammals on our study farm had relatively distinguishable tracks,
the possibility of misidentification exists between species with similar tracks
(e.g., Red Fox and young Coyote).
2009 R.R. Conover and E.T. Linder 443
Additional limitations associated with mud track plots include 1) difficulty to detect and/or identify tracks of small mammals, 2) temporal
unpredictability from precipitation dependence, and 3) spatial restrictions
from isolated precipitation events. The under-representation of small mammals
was not surprising, as difficulties in capturing and identifying small
mammal tracks have been previously reported (Wemmer et al. 1996). Although
the dependence of mud track plots on precipitation may preclude
interpreting resultant data as absolute values, they remain viable estimates for
relative comparisons. The precipitation dependence of mud track plots may be
circumvented using manual water saturation; however, this remains untested.
Future studies should investigate applying mud track plots in baited applications
as potential replacements to tracking platforms.
According to Foresman and Pearson’s (1998) criteria to evaluate technique
efficacy, mud track plots performed effectively to detect occupancy of mediumand
large-sized mammals. The success of this technique is attributed to their
minimal cost and effort, high proportion of plots that captured tracks in 1–3
days, having recorded all visually confirmed mammal species in the study area,
and capturing identifiable (≥98%) track imprints. Given the variety of mammalian
survey approaches, technique selection should be based on research
objectives and method applicability (Foresman and Pearson 1998, Mowat et al.
2000, Wolf et al. 2003). We advocate the use of mud track plots for spatially explicit
habitat-use investigations of medium- and large-sized mammals to detect
presence-absence with minimal bias of natural movements.
Acknowledgments
We thank B. Leopold and W. Clark for their reviews of a previous manuscript
version, Delta Wildlife for logistical support, and the Jones’ Planting Company for
land access. This manuscript was also improved by reviews from J.L. Belant and two
anonymous reviewers.
Literature Cited
Belant, J.L. 2003. Comparison of 3 tracking mediums for detecting forest carnivores.
Wildlife Society Bulletin 31:744–747.
Bider, J.R. 1968. Animal activity in uncontrolled terrestrial communities as determined
by a sand-transect technique. Ecological Monographs 38:269–307.
Connors, M.J., E.M. Schauber, A. Forbes, C.G. Jones, B.J. Goodwin, and R.S. Ostfeld.
2005. Use of track plates to quantify predation risk at small spatial scales.
Journal of Mammalogy 86:991–996.
Conover, R.R., L.W. Burger, Jr., and E.T. Linder. 2007. Winter avian community
and sparrow response to field border width. Journal of Wildlife Management
71:1917–1923.
Fecske, D.M., J.A. Jenks, and V.J. Smith. 2002. Field evaluation of a habitat-relation
model for the American Marten. Wildlife Society Bulletin 30:775–782.
Foresman, K.R., and D.E. Pearson. 1998. Comparison of proposed survey procedures
for detection of forest carnivores. Journal of Wildlife Management 62:1217–1226.
Glen, A.S., and C.R. Dickman. 2003. Monitoring bait removal in vertebrate pest control:
A comparison using track identification and remote photography. Wildlife
Research 30:29–33.
444 Southeastern Naturalist Vol. 8, No. 3
Gompper, M.E., R.W. Kays, J.C. Ray, S.D. Lapoint, D.A. Bogan, and J.R. Cryan.
2006. A comparison of noninvasive techniques to survey carnivore communities in
northeastern North America. Wildlife Society Bulletin 34:1142–1151.
Kuehl, A.K., and W.R. Clark. 2002. Predator activity related to landscape features in
northern Iowa. Journal of Wildlife Management 66:1224–1234.
Mooney, K.A. 2002. Quantifying avian habitat use in forests using track-plates. Journal
of Field Ornithology 73:392–398.
Mowat, G., C. Shurgot, and K.G. Poole. 2000. Using track plates and remote cameras
to detect Marten and Short-tailed Weasels in coastal cedar hemlock forests. Northwestern
Naturalist 81:113–121.
Murie, O.J. 1954. A Field Guide to Animal Tracks. Second Edition. Houghton Miffl in,
Boston, MA.
Powell, J.C., W.E. Keenan, W.A. Cole, L.C. Murphree, D.A. Yost, J.J. Pitts, and R.H.
Wells. 1952. Soil survey of Sunfl ower County, Mississippi. US Department of Agriculture
Soil Conservation Service. Mississippi Agricultural Experiment Station,
Mississippi State, MS.
Renfrew, R.B., C.A. Ribic, J.L. Nack, and E.K. Bollinger. 2005. Edge avoidance
by nesting grassland birds: A futile strategy in a fragmented landscape. Auk
122:618–636.
Sargeant, G.A., D.H. Johnson, and W.E. Berg. 2003. Sampling designs for carnivore
scent-station surveys. Journal of Wildlife Management 67:289–298.
Savidge, J.A., and T.F. Seibert. 1988. An infrared trigger and camera to identify predators
at artificial nests. Journal of Wildlife Management 52:291–294.
Silveira, L., A.T.A. Jacomo and J.A.F. Diniz-Filho. 2003. Camera trap, line-transect
census, and track surveys: A comparative evaluation. Biological Conservation
114:351–355.
Smith, J.B., J.A. Jenks, and R.W. Klaver. 2007. Evaluating detection probabilities for
American Marten in the Black Hills, South Dakota. Journal of Wildlife Management
71:2412–2416.
Vanak, A.T., and M.E. Gompper. 2007. Effectiveness of non-invasive techniques for
surveying activity and habitat use of the Indian Fox, Vulpes bengalensis, in southern
India. Wildlife Biology 13:219–224.
Vojta, C.D. 2005. Old dog, new tricks: Innovations with presence-absence information.
Journal of Wildlife Management 69:845–848.
Wemmer, C., T.H. Kunz, G. Lundie-Jenkins, and W.J. McShea. 1996. Mammalian
sign. Pp. 157–176, In D.E. Wilson, F.R. Cole, J.D. Nichols, R. Rudran, and M.S.
Foster (Eds.). Measuring and Monitoring Biological Diversity: Standard Methods
for Mammals. Smithsonian Institution Press, Washington, DC.
Wolf, K.N., F. Elvinger, and J.L. Pilcicki. 2003. Infrared-triggered photography and
tracking plates to monitor oral rabies vaccine bait contact by Raccoons in culverts.
Wildlife Society Bulletin 31:387–391.
Zielinski, W.J. 1995. Track plates. Pp. 67–86, In W.J. Zielinski and T.E. Kucera (Eds.).
American Marten, Fisher, Lynx, and Wolverine: Survey Methods for their Detection.
US Department of Agriculture, Forest Service, Pacific Southwest Research
Station, Albany, CA. General Technical Report PSW-GTR-157.
Zielinski, W.J., and R.L. Truex. 1995. Distinguishing tracks of Marten and Fisher at
track-plate stations. Journal of Wildlife Management 59:571–579.