Carnivores in the Farmland Matrix: Presence Depends on Avocado Orchards and Other Neighboring Land U - PowerPoint PPT Presentation

1 / 1
About This Presentation
Title:

Carnivores in the Farmland Matrix: Presence Depends on Avocado Orchards and Other Neighboring Land U

Description:

Carnivores in the Farmland Matrix: Presence Depends on Avocado Orchards and Other Neighboring Land U – PowerPoint PPT presentation

Number of Views:67
Avg rating:3.0/5.0
Slides: 2
Provided by: tnog
Category:

less

Transcript and Presenter's Notes

Title: Carnivores in the Farmland Matrix: Presence Depends on Avocado Orchards and Other Neighboring Land U


1
Carnivores in the Farmland Matrix Presence
Depends on Avocado Orchards and Other Neighboring
Land Use Types
Theresa Nogeire, Frank W. Davis, John L.
Orrock Donald Bren School of
Environmental Science and Management, University
of California, Santa Barbara, CA 93106-5131
Program in Evolution, Ecology, and Population
Biology, Washington University at St. Louis
ABSTRACT
RESULTS
Conservation of wide-ranging species is
increasingly difficult as land throughout the
world is increasingly dominated by human
activities. Researchers have shown that
remnants of native habitat embedded within the
agricultural matrix can support mammalian
carnivores. Farmland can be useful for movement,
human-dominated lands can be incorporated into
home ranges, and, especially for omnivores like
coyotes and bears, crops can provide food
subsidies. Human-dominated landscapes can no
longer be regarded simply as an unfriendly matrix
surrounding islands of native habitat, but
rather, as a special type of habitat for some
species. Avocado orchards are an important land
use in southern California and avocados are a
food resource for many omnivores, including
coyote, bear, and gray fox. The orchards are
also used by bobcats. But agricultural
landscapes present dangers for these animals
including exposure to roads and vehicles and
exposure to poisons such as rodenticides. Using
remotely triggered cameras and scat surveys, we
surveyed carnivore activity in several avocado
orchards in Santa Barbara and Ventura Counties.
We also recorded indices of rodent abundance.
Using digital land use/land cover data we
characterized the extent and configuration of
natural, avocado orchard, other agricultural, and
urban or built-up lands around surveyed sites.
There is more carnivore activity in agricultural
landscapes than in wildlands, and most of this
activity appears to be concentrated at the edge
of orchards. The amount of avocado orchard
within a home-range-sized buffer of the
camera-trapping location is predictive of the
probability of encountering a black bear, bobcat,
coyote, and gray fox. The amount of urban land,
natural vegetation, and other agricultural lands
does not significantly affect the probability of
encountering a black bear, bobcat or gray fox.
Other agriculture does help predict coyote
occurrence, but urban and natural lands do not.
Rodent abundance and diversity are lower in
orchards than in native vegetation. Preliminary
results suggest a negative relationship between
rodents and carnivores.
The amount of avocado orchard in the landscape at
one or both scales is predictive of the presence
of black bear, coyote, bobcat, and gray fox, but
the amount of urban or built-up land and natural
vegetation does not have a significant effect at
either scales. The amount of other agriculture
was also significant in the coyote models (see
table below).
It is unlikely that any fox has a circular home
range centered on one of our camera locations.
To test the theory that a carnivore has a home
range in the most natural place possible that
still includes the camera location within the
home range radius, we compared models with data
from the camera sites to data from the most
natural site within the home range (see table
below). Adding these variables improved the
model nearly 2-fold.
OBJECTIVES
There were also significant relationships between
species (see table to left). These could be
caused by interactions between species or
covariates. For example, I would like to examine
the effect of water sources.
Assess the influence of land use on the
probability of detecting mammalian carnivores in
Santa Barbara and Ventura Counties. Focus on
lands surrounding avocado orchards, where high
numbers of carnivores have been
reported. Evaluate the relationship between
carnivores and rodents in avocado
orchards. Develop predictive models that can be
tested with further camera-trapping.
METHODS
Mammal surveys Probability of encountering
species was tabulated from camera trapping over
the past year in Santa Barbara and Ventura
Counties. Orchards were selected for variety of
surrounding land use size, having approximately
average size, and landowner cooperation. Control
sites with only natural vegetation were trapped
at the Sedgwick Reserve and Gaviota State Park.
Remote-triggered cameras (Stealthcam) were set up
along dirt roads or trails in all sites, near
signs of carnivore activity or at trail junctions
when possible. A small amount of carnivore lure
was placed in front of the trap to encourage
animals to stop for long enough to be
photographed. Rodent trapping was done in a
subset of sites, using Sherman live traps.
Trapping webs were used and sites were trapped
for four consecutive nights.
A particularly interesting set of relationships
between species are the relationships between
rodents and carnivores. Are the avocado orchards
subsidizing larger then normal rodent
populations, which are attracting the carnivores?
We think not first, because colleagues have
documented coyotes, bears, foxes, and other
carnivores eating avocados (Borchert, Davis and
Kreitler, 2007). Second, we found very low
numbers of rodents in orchards compared to some
natural sites (see figure below). There is a
negative relationship between the number of
rodents and the number of predators in our sites
(r2.27). This could mean that there are not
many rodents in the orchards and therefore the
predators are eating avocados, that the predators
are keeping the rodent populations low, or that
the rodents in the orchards are more active
during the day (when we did not trap) to avoid
the nocturnal predators.
Landscape data Using data from the Farmland
Mapping and Monitoring Program, the California
Avocado Commission, and aerial photographs, we
outlined the following land use types in the
neighborhood of the camera sites Urban or
built-up, avocado orchard, other agriculture, and
natural vegetation. We performed neighborhood
analyses of the amount of each land use type at
two scales. The circular neighborhoods were
defined by a radius of 469m (mimicking an average
fox home range size, as reported in Crooks 2002)
and 968m (mimicking an average bobcat home range
size). Reported data are the average value for
each land use (present1, absent0) within the
neighborhood. These data were then entered into
a stepwise regression (JMP 7.0, SAS Institute
2007) to find significant predictors for each
carnivore species.
REFERENCES
CONCLUSIONS
The amount of avocados in the neighborhood
partially explains the probability of
encountering bobcat, black bear, coyote, and gray
fox. The amount of natural vegetation has not
proved to be a significant predictor for any
carnivore species. There is a negative
relationship between carnivores and rodents.
There are more rodents and fewer carnivores in
natural, control sites, and fewer rodents and
more carnivores in agricultural sites. There are
correlations between several of the carnivore
species. These could be caused by interactions
between species or covariates.
Borchert, M., F. Davis and J. Kreitler. 2007.
Carnivore use of an avocado orchard in southern
California. U.S.A. California Fish and Game
Journal. Crooks, K.R. 2002. Relative
sensitivities to mammalian carnivores to habitat
fragmentation. Conservation Biology 16(2)
488-502. Farmland Mapping and Monitoring Program
(FMMP) 2007. State of California Department of
Conservation. Hilty, J.A., C. Brooks, E. Heaton
and A.M. Merenlender. 2006. Forecasting the
effects of land-use change on native and
non-native mammalian predator distributions.
Biodiversity and Conservation 15(9) 2853-2871.
FURTHER ANALYSES AND FUTURE RESEARCH
ACKNOWLEDGMENTS
Using the predictive models developed in this
analysis, we would like to test the models by
camera-trapping at more sites. We would also
like to add one more scale to the analysis. We
are just getting started on a survey of avocado
growers throughout southern California and Baja
California, Mexico. The survey is designed to
gather knowledge from the owners and managers of
orchards throughout the growing region our
experience has been that growers have a pretty
good idea of what species are on their land.
While this data is more subjective then
camera-trapping, it will allow us to survey
dozens or even hundreds of orchards, and to
compare the species occurrence with management
practices, attitudes, and landscape variables.
Wed like to acknowledge Patrick Jantz, Theresa
Rusca, Lydia Ries, Oliver Soong, Jason Kreitler,
and California State Parks for help in the field.
Also all the orchard owners and managers who
spent time with us, especially Jay Ruskey.
Finally, thanks to the Bren School of
Environmental Science and Management and NSF
Grant DEB-0444217 for funding assistance.
Please address correspondence to
tnogeire_at_bren.ucsb.edu
Write a Comment
User Comments (0)
About PowerShow.com