Is the likelihood of waterfowl presence greater on conserved lands?

(USC Thesis Other)

Is the likelihood of waterfowl presence greater on conserved lands?

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

IS THE LIKELIHOOD OF WATERFOWL PRESENCE GREATER ON CONSERVED
LANDS?

by

Brian Vance Kearns

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(GEOGRAPHIC INFORMATION SCIENCE AND TECHNOLOGY)

May 2012

Copyright 2012 Brian Vance Kearns
ii

ACKNOWLEDGEMENTS

I thank the following parties for their involvement in and contribution to my Masters
thesis: Travis Longcore and my thesis committee for guiding these analyses and assisting
with theoretical and research concerns; Ducks Unlimited’s Great Lakes Atlantic Region
for funding and support of data collection that made this analysis possible; Kurt
Anderson of Ducks Unlimited for his work developing stopover data and theory
background which he was willing to share; Rob Paige of Ducks Unlimited for his work
on data management and clean-up of the telemetry dataset; John Coluccy, Dane Cramer,
and the biological staff of the Ducks Unlimited Great Lakes Atlantic Regional Office for
their suggestions and contributions in developing my research questions; and the
University of Southern California for providing the framework in which to execute this
project and the knowledge to do so.
iii

TABLE OF CONTENTS

Acknowledgements ii

List of Figures iv

Abstract v

Introduction 1

Methods 5
Table 1. Counts of categorization of telemetry for HCat and PAD 15

Results 16

Discussion 19

Conclusion 23

References 25

Appendix A: Python code for “Add Field” tool 28

iv

LIST OF FIGURES

Figure 1: Range map of A. rubripes 3

Figure 2: Workflow of data management 6

Figure 3: Satellite telemetry data and selection 7

Figure 4: Flight path and stopover event 9

Figure 5: Window of “Split Layer by Attributes” tool 11

Figure 6: Protected areas used by A. rubripes 12

Figure 7: Mean PAD values by stopover event 14

Figure 8: Count histograms of telemetry points, HCat 16

Figure 9: Count histograms of telemetry points, protected lands only 17

Figure 10: Distributions of mean PAD use 18

v

ABSTRACT

Waterfowl are one of our Nation’s most precious and abundant natural resources,
and preserving habitat well suited to their needs has long been a goal of private and
public entities alike. In this study, I focused on the American Black Duck (Anas
rubripes), a species seeing a large decline in numbers since the mid 20
th
century. Using a
satellite telemetry dataset collected by Ducks Unlimited during 2008 and 2009 in the
context of the Protected Areas Database of the United States (PAD), I addressed the land
use habits of A. rubripes to assess the efficacy of costly conservation efforts implemented
through conservation easements and the maintenance of wildlife refuges and management
areas. Most analyses were conducted at the stopover level, grouping telemetry points
within a 0.5 decimal degree diameter. By creating distributions and studying
correlations, this study finds that during wintering months A. rubripes registered more
telemetry points in PAD lands where hunting is allowed in-season; during migration,
lands outside of the PAD were more frequently used. This could be attributed to
waterfowl specific management practices creating prime habitat during wintering and
food needs being fulfilled by residual agricultural products during migration. This
suggests an increased importance of management efforts in wintering habitats. Climate
variables were also assessed to test reported influences of temperature and precipitation
on distribution and stopover behaviors, but study data did not demonstrate a correlation
between stopover length and temperature or precipitation at arrival and departure. A finer
scale geospatial analysis using more detailed information about hunting status and
protection level is recommended to further interpret available data.
1

INTRODUCTION

Land use has long been one of the key concerns for wildlife conservation efforts.
Gaining an understanding about habitat dynamics and how animals interact with their
respective biomes is one of our only pathways as researchers to implement measures that
will provide benefit for species in increasingly disturbed ecosystems. Wetlands in
particular are of conservation concern because they provide key habitat for migrating
waterfowl, fish, invertebrates, shorebirds, songbirds, and are even important as a resource
for humans (Kirby et al., 2008).
Though natural processes are certainly of interest, many wetland disturbances and
subsequent disturbances to waterfowl populations are caused by anthropogenic factors.
Whether through recreation, fertilization, or the conversion of land to agriculture,
wetlands face a great deal of disturbance pressure. Filoso & Palmer (2011) documented
the presence of excess nitrogen from fertilizer runoff in streams and also the tendency for
this nitrogen to accumulate in downstream waters, often wetlands or estuarine
ecosystems. Bennett (2011) presented a method for modeling the effects of recreation on
the wetland habitat of the Black-crowned Night Heron, showing disturbance of critical
breeding habitat. Naugle et al. (2001) also documented the negative effects of habitat
fragmentation and the related edge effects on breeding waterfowl populations in shallow
water wetlands. Beyond these factors, disturbances can dissuade animals from using
habitat during migration (Vegvari et al., 2011).
To counteract these disturbances, conservation projects and programs have been
developed to protect known sensitive areas or convert agricultural land back to wetland.
2

Various conservation organizations, both private and public, have been instrumental in
the implementation of these programs. As human development increases, providing
areas where animals can experience undisturbed habitat and resources becomes of greater
concern and must be an important factor in land management decision-making.
Documents such as the North American Waterfowl Management Plan (NAWMP)
indicate a growing need to provide habitat that caters to specific groups of organisms;
they also guide many large-scale conservation efforts by the private and public sectors
(Brasher et al., 2007). In relation, a recent study shows that stopover duration, a potential
indicator of the benefit an individual organism gets from a particular habitat, is indeed
positively correlated with foraging habitat quality specific to the forager (O'Neal et al.,
2012). Because many restoration efforts focus on improving carrying capacity at
particular key sites, a connection can be drawn between management practices and
observed species-level benefits.
Although wildlife is often the key concern when managing habitat, delivery of
conserved lands obviously must focus on selection and delineation of parcels where
greatest benefit will be seen. Land is therefore set aside through various programs that
work towards this goal. These programs include the Federal Wildlife Refuge System, as
well as United States Department of Agriculture programs such as the Conservation
Reserve Program (CRP), Wetland Reserve Program (WRP), Grassland Reserve Program
(GRP), and more. CRP, WRP, and GRP allow landowners to establish easements and “to
receive incentive payments for installing specific conservation practices that help protect
3

environmentally sensitive land, decrease erosion, restore wildlife habitat, and safeguard
ground and surface water” (USDA, 2012).
These private easements have become paramount for wildlife. Especially in areas
of the country with widespread agricultural development, they simultaneously preserve
habitat and attempt to make developed land mimic its natural state. Efforts such as these
help populations of threatened species stay robust. Unfortunately, despite these efforts,
many areas still see loss of native habitat at a rate that cannot be mitigated by the current
rate of easement creation (Gascoigne et al., 2011). Direction to where land conservation
efforts are or would be most effective would be very beneficial.
The species of concern in this study is the American Black Duck (Anas rubripes),
a dabbling duck common to the Atlantic and Mississippi flyways (Longcore et al., 2000).
Their wintering range is typically along the mid-Atlantic coast, while breeding and
nesting takes place in the northeastern provinces of Canada (see figure 1).

Figure 1. Range map of A. rubripes. Individuals migrate north to south; wintering
occurs in the southeastern United States, while breeding takes place as far north as
the Canadian province of Newfoundland and Labrador (Longcore et al., 2000).
4

A. rubripes and other dabbling ducks feed primarily on plants and macro invertebrates in
shallow water wetlands, a habitat type that is often protected through conservation
programs. In recent research efforts by Ducks Unlimited and other organizations,
numbers of A. rubripes were observed to have declined by about 60% in many wintering
areas since the 1950’s (DU, 2010).
Conroy et al. (2002) identified three specific factors affecting and regulating
populations of A. rubripes: loss in the quantity or quality of breeding habitat, loss in the
quantity or quality of wintering habitat, and harvest through hunting. Although drops in
A. rubripes populations can be partially attributed to habitat change, hunting and harvest
are also important. I chose to study A. rubripes because declining numbers in waterfowl
species are an important indicator of ecosystem health. Also, in an era where resources
are scarce, they must be allocated efficiently through careful direction based on
knowledge of wintering and migrating patterns on land.
Factors such as ecological carrying capacity, habitat availability, and level of
harvest are important for this species, and land conservation and wildlife management
likely affect these factors. This study seeks to evaluate the effectiveness of land
conservation efforts using spatial tools to observe whether migration events of A.
rubripes, namely stopovers and wintering, are affected by conservation projects.
Although researchers have assessed habitat variables such as food availability and
carrying capacity (which admittedly play a large role in initial management efforts, see
Plattner (2010)), effects of widespread hunting disturbance have not been extensively
explored in conjunction with land use of targeted species.
5

My hypothesis is that A. rubripes will prefer lands that are managed via land
protection efforts and experience a lower degree of hunting pressure, one of the observed
causes of A. rubipes population decline (Morton, 1998). Because of higher energetic
carrying capacity, protected habitats should also be preferable while the waterfowl are in
transit between wintering and breeding grounds (Brasher, 2007). Influence of
temperature and precipitation, two important climatic factors affecting migration
behaviors, on stopover time will also be investigated (Brook et al., 2009). Protected
lands are created with wildlife and habitat management in mind, and should therefore
represent as close to ideal habitat as is achievable through human efforts. This study
should begin to indicate whether costly conservation efforts are protecting habitats that
are preferentially used by a species at risk.
METHODS

This study investigates the distribution and land use of A. rubripes through data
cross tabulation and classification. A satellite telemetry dataset was pared down and
parsed out based on criteria relevant to the study, and points were then attributed with
values based on overlay of the Protected Areas Database (PAD), a USGS project to that
collects parcel information and attributes conservation easements and protected areas
throughout the United States. Count histogram and mean value range histograms were
then constructed to determine land use by birds during migration and wintering periods
separately. Figure 2 shows a generalized workflow for the data management and
tabulation portions of this analysis.
6

Figure 2. A generalized workflow, following data management steps through to
the point before distributions were created.

Data Sources

Duck location data were derived from a satellite telemetry study conducted by
Ducks Unlimited during 2008 and 2009. This study was conducted to determine the
effects of habitat changes on declining populations of A. rubripes. Female A. rubripes
were trapped and tagged with satellite transmitters in New Jersey and Delaware during
wintering, and were monitored over subsequent seasons. This collection of points served
as the raw data for this study (n = 33442) (see figure 3). Point locations were determined
via GPS with an accuracy of approximately 10 m, an accuracy that is sufficient for
regional analyses conducted in this study. Data ranged geographically from the state of
Virginia to the northern reaches of the Canadian province of Newfoundland and
Labrador. Birds were uniquely identified in data tables by the attribute field “PTT”,
7

which refers to the unique radio frequency transmitted by the collar. This allowed the
movements of individual birds to be analyzed independently, as discussed in more detail
in the coming sections.

(a) (b)

Figure 3. (a) All satellite telemetry locations collected from tagged A. rubripes
during 2008 and 2009. (b) Telemetry points used for this analysis, which exclude
points beyond 0.5 dd from the administrative boundary of the US.

Preparation of Telemetry Data
The data used for analyses were pared down and parsed out in several ways. First,
the ArcGIS “Select by Location” function was used with a 0.5 decimal degree search
distance around the polygon of the United States to produce a telemetry dataset
containing points only in the US (n = 10670). This was necessary due to the lack of a
PAD equivalent for Canada. The 0.5 decimal degree buffer was selected in response to
research showing this to be the constraining range of a stopover event (Afton, 2008).
8

This also allowed for selection of data points that were beyond the actual land boundary
of the United States, since these birds used lacustrine and estuarine habitats in addition to
terrestrial ones (see figure 3).
In addition to selecting individual telemetry points, analyses were conducted to
identify individual stopover events (n = 92). Points representing stopover events in this
dataset were initially determined by Ducks Unlimited biologist Kurt Anderson
(Anderson, 2009). Anderson defined stopover events as when two or more temporally
consecutive telemetry points for the same bird (PTT) were clustered within a 0.5 dd
diameter circular range of one another. Consequently, each stopover event had a varying
number of telemetry points associated with it, but always from the same PTT. Stopover
events were represented in Anderson’s data by a single (x, y) coordinate from the original
telemetry dataset. That was the first point in the temporal sequence of telemetry points
composing each stopover event (see figure 4). Anderson used this single representational
point to analyze variables such as stopover length, arrival and departure date, and general
spatial distribution; in the context of my study, these points were used chiefly as a link
between the stopover and telemetry datasets, which both contained information necessary
for this analysis.
The clustering method used in my analysis was slightly non-standard, as the distance
measure from the center of each cluster was not necessarily centered at the point
representing the stopover event because of the temporal nature of the designation.

9

(a) (b)

Figure 4. (a) An example of the migration route for bird with PTT 87847. Each
point represents a stopover event during migration. (b) The stopover point (first
temporal telemetry point) and telemetry points for the PTT code 78605 clustered
within a 0.5 dd diameter of the first point.

Therefore, it was necessary for me to establish clusters manually in order to ensure the
inclusion of all telemetry points within a stopover event. Each bird, represented by its
PTT, had one to several unique stopover events of varying length throughout the
monitoring period. For this analysis, all telemetry points contained within each single
stopover event were selected via the clustering method described above and attributed
with an additional unique identifier which consisted of the PTT of the bird conducting the
stopover concatenated with an integer value to delineate different stopover events within
a bird’s migration (see figure 4).
Next, since telemetry points were collected during different A. rubripes life
history periods, stopover events were separated into wintering (n = 37) and migration
10

(n = 55) categories. The time at which birds transitioned from wintering to migration was
originally determined by Anderson, and occurred when birds moved outside of the 0.5 dd
range where they wintered. Data was queried from the original stopover table in ArcMap
using the “Select by Attributes” utility. By exporting these data into two unique feature
classes, I was able to examine these two very different life history periods separately
during all further analyses.
An additional key geoprocessing task in this study was to separate the complete
dataset containing telemetry values for all birds in order to conduct data analysis for
individual birds. This was accomplished using a tool called “Split Layer by Attributes”
which was obtained from the Esri ArcGIS Geoprocessing Gallery (Patterson, 2011). The
process coded by this *.tbx file selected attributes from the telemetry and stopover event
feature classes by consecutive PTT values and exported them to separate shapefiles (see
figure 5). This created a telemetry and stopover event feature class for each unique PTT.
One pitfall of this tool for this particular study was its lack of support for the Esri
Geodatabase. This was remedied by exporting the shapefiles resulting from the tool into
folders, and then importing them into a File Geodatabase entitled “BlackDuck” which
was constructed for the storage of telemetry and stopover event points. Within this file
geodatabase I constructed separate feature datasets for telemetry points and stopover
events. This facilitated other geoprocessing tasks, which are discussed below.

11

Figure 5. An example of parameter inputs for the “Split Layer by Attributes” tool.
This tool was used to split stopover and telemetry feature classes by PTT value.

Preparation of Land Characterization and Hunting Data
Telemetry data were analyzed in conjunction with land parcel data from the
Protected Areas Database (PAD), GIS data for which was obtained from the gap analysis
program of the United States Geologic Survey (USGS, 2011). This dataset includes a set
of land parcels that are officially designated as protected lands within the United States,
including private conservation easements, lands owned by non-profits or other entities,
and land managed under state and federal fish and wildlife programs such as wildlife
refuges or wildlife management areas.
For this study, land parcels on which A. rubripes telemetry locations occurred
were extracted from the PAD using an ArcGIS “Select by Location” function with a
“contains” parameter (see figure 6). This was executed for ease of data manipulation,
since the attributes of parcels without telemetry locations were not included in this study.
Parcel name, acreage, state, and other attributes were all included as data fields.
Consequently, land parcels included in the analysis were owned by federal, state, and
12

private entities (such as The Nature Conservancy) and private landowners with
conservation easements through state and federal programs (e.g. CRP, WRP, New Jersey
Green Acres Program, etc.). As mentioned, because a protected lands dataset for Canada
was not available, this study was limited to the telemetry points located within the United
States, and consequently includes mostly wintering and spring migration points.
To indicate protected land status in this analysis, an additional attribute was added
to the telemetry data to indicate whether the points were on PAD lands (PAD = 1) or not
on PAD lands (PAD = 0). This allowed for numeric analysis of PAD status as described
below.

Figure 6. Distribution of all PAD lands in the states where A. rubripes telemetry
points were present, and PAD parcels used by migrating and wintering A.
rubripes in 2008-2009; the selection display results from a “Select by Location”
operation using the telemetry feature class.

The level of hunting disturbance is also a variable of great interest, so I had to
determine the extent to which this was a factor for stopover and wintering events on
protected lands. To create a hunting status (HCat) attribute, I established a classification
13

system using an ordinal variable with three values: 0 = hunting is not permitted, 1 =
hunting is conditionally allowed (e.g. by limited draw permit or by private landowner
permission, assumed to both be of moderate impact), and 2 = hunting is allowed as
delineated by state regulations. Lands that were not included in the PAD were classified
as HCat 1 because hunting was a potential factor, but disturbance was assumed to be
moderate because of reduced hunter traffic. For the scope of this study, contacting land
owners for all telemetry points not accounted for by the PAD was not feasible so this
generalization was necessary.
Final Preparation of Feature Classes for Analysis
After telemetry points were associated with a particular stopover event, several
fields were added to all feature classes in the File Geodatabase using an original Python
script (see appendix). Fields containing values for HCat and PAD were added and
manually populated for all telemetry points.
Data Analysis
Once the datasets were fully populated with attributes from the spatial data, the
final analyses of the tabular data could begin. Mean values for HCat and PAD were
calculated for all telemetry points within each stopover event (see figure 7). Mean values
were determined via the “statistics” utility in the telemetry attribute table, and then
associated with the individual stopover event feature classes via a table join. These mean
values can be interpreted as the percentage of time that a bird spent in protected lands
within each stopover event.
14

Figure 7. A map showing the percentage of time spent in PAD lands for a subset
of stopover events from birds with various PTT values in the Chesapeake Bay
region. A great deal of variation is seen depending on location.

Next, distributions were constructed to explore mean PAD status for stopover
events and wintering. This was conducted for migration and wintering periods separately
to assess potentially different ecological needs. As described above, mean PAD values
were determined from all telemetry points associated with each stopover event. These
values were then plotted on a range histogram.
15

HCat values were analyzed slightly differently. This variable was strongly
correlated with PAD status because HCat 1 was very often an indicator of PAD 0 due to
category definitions (see table 1). This prompted the distribution of HCat values to be
assessed for only the data points where PAD = 1. A separate distribution was constructed
with telemetry points from both PAD values for comparison and to assess model
differences and biases when these categories were altered. Results are discussed in the
section following.
Table 1. Counts of categorization for HCat and PAD for each telemetry point in
wintering and migration periods. A value of 1 for HCat was assigned for private
lands where no other information was available. Consequently, PAD = 0 and
HCat = 1 are highly correlated.

PAD=0 PAD=1
Hcat= 0 3 386
Hcat=1 4562 1957
Hcat=2 3 2650

Climate Variables
Temperature and precipitation are also known to affect migratory behavior and
distribution of waterfowl species (Conroy et al., 2002). Therefore, historical weather data
was obtained from “weatherunderground.com”, which accesses historical National
Weather Service data to provide temperature and precipitation levels for a desired area
and temporal period. For the purposes of my analysis, climate variables were delineated
by ZIP code, as this was found to produce an analytic surface with a suitable scale. It
should be noted that although climate data were presented at this scale, values came from
weather stations that were most often spaced more distantly. Therefore, some reduction
16

in the level of precision and true spatial resolution should be considered with these data
and results. These data were continuous, and were kept as such for my analysis.
The first temporal point in each stopover event was overlaid with a ZIP code
administrative layer in ArcMap and attributed with the appropriate value. Then, data
were collected from “weatherunderground.com” for temperature and precipitation values
at arrival and departure. Appropriate attribute fields were added in feature classes
containing stopover events and then populated with these values to be used in regression.
Regressions were calculated using the JMP statistics program developed by SAS.
RESULTS
As a first indicator, I assessed a basic count of telemetry points that were
contained within the various hunting category values in all land areas (see figure 8) and
then only in land parcels contained within the Protected Areas Database (see figure 9).

(a) (b)

Figure 8. Counts of Telemetry points by HCat in (a) wintering and (b) migration.
These distributions contain telemetry values from PAD = 0 and PAD = 1 areas,
and are used to demonstrate HCat variable correlation with PAD.

17

(a) (b)

Figure 9. Telemetry points representing HCat values where PAD = 1 in (a)
wintering and (b) migration.

Despite a lack of hunting pressures during spring migration, hunting category
values were assessed for both wintering and migration periods due to the potential effects
of management practices associated with hunting. A. rubripes land use in areas that
either allowed hunting outright or via permitted harvest remained relatively similar and
high during wintering in protected lands. Conversely, lands where open hunting took
place in-season were strongly preferred during migration when birds were observed in
protected lands. This suggests a higher use of lands that have either open or permitted
hunting; in the case of migration, HCat 2 values, or open hunting, are almost universally
used given the analysis of only lands within the Protected Areas Database. One can also
observe that A. rubripes has an increased use of lands with lower hunting pressures
during wintering when hunting season is open, while choosing open hunting lands during
spring migration when hunting pressure is not a factor. Finally, it can be seen that when
PAD = 0 lands are removed from the hunting category analysis, private non-protected
lands cease to bias trends of land use. This implies the need for more detail regarding
private lands.
18

Next, I examined the distribution of mean values for PAD per stopover event.
This was accomplished by averaging the PAD values for each telemetry point associated
with a particular stopover or wintering event. Mean PAD land use values for wintering
were relatively evenly spread, while PAD use during migration seems to be skewed
towards the “0” category, or land not included in the PAD (see figure 10).

(a)

(b)

Figure 10. (a) Distribution of mean PAD values for wintering habitat. (b)
Distribution of mean PAD values for stopovers during migration.

19

This suggests that the use of protected areas by wintering birds was more strongly
favored than that of birds that were in migration, which were observed more frequently
outside of protected land parcels.
Lastly, I assessed climate variables to see whether climatic factors had an effect
on the observed length of stopover events. For this, I conducted linear regressions for
arrival temperature, departure temperature, arrival precipitation, and departure
precipitation against stopover length (n = 55). A small selection of values was omitted
due to lack of available climate data. Although some potential correlation was observed,
adjusted r-squared values were too low to suggest any type of compelling implication;
therefore, these results were not explored further.
DISCUSSION
The relative distributions of A. rubripes telemetry points between wintering and
migration life history periods prove to be one of the more interesting results of this study.
A. rubripes were indeed observed more often on land with permit-limited hunting than in
other areas, and while within protected land parcels, open and permit-limited hunting
areas were used more than areas where hunting was prohibited. This suggests that even
during hunting season when disturbance is direct and present, birds are still choosing
managed areas where hunting is allowed over non-managed areas or hunting-prohibited
areas.
There are a few implications of these results. The fewest observations are found
in hunting-prohibited areas because the area covered by protected lands with this
designation is comparatively very small. Permitted hunting areas and private lands show
20

the highest count of telemetry observations since they also include all areas not accounted
for by the Protected Areas Database. With the removal of biasing points outside of the
PAD, however, we see more even usage of open and permit-limited hunting areas.
Although the level of hunting disturbance is likely higher in these areas, the level of
management for creating habitat specific to the needs of waterfowl is likely higher to
boost population sustainability for harvest and species well-being.
For instance, St. Clair Flats Wildlife Area in Michigan possesses many water
control structures and converted agricultural land areas where water level is specifically
tailored to appropriate dabbling duck feeding depths (MDNR, 2002). This may well
draw in wintering birds that require a reliable winter food source. In another case, land
modifications, such as marsh terracing, were found to increase the numbers of wintering
water birds in an estuarine ecosystem, and even prompted increased count volumes in
certain species around anthropogenic features such as pathways or towns (O'Connell and
Nyman, 2011). Although it cannot be guaranteed, one could assume that conservation
efforts involving water control structures and other infrastructure are likely to be more
intensive in protected areas. These findings suggest that land management benefits might
outweigh hunting disturbances in wintering habitat in terms of habitat choice, and have
positive effects outside of hunting season during migration.
Although distributions of utilized PAD lands remained somewhat similar during
migration, this temporal period has a unique set of conclusions to be drawn from it given
a significantly lower use of PAD parcels. The mean PAD values’ distribution for
telemetry points during stopover events confirm that non-protected areas seem to be used
21

more frequently during migration than during the wintering period (see figure 10). The
literature does not seem to suggest a particular reason why this pattern would vary, other
than to say that migrating waterfowl in various regions tend to favor areas with better
habitat conditions and greater nutrient reserves to replenish large amounts of expended
energy (Lok et al., 2011). Another connected factor might be residual grains left in
agricultural lands that likely would not be protected under conservation programs while
still under harvest. Especially as migration locations approach the Great Plains, this
becomes a much more important variable and potential management concern (Foster et
al., 2010, Sherfy et al., 2011). I also conjecture that the lower use of managed lands may
simply be due to necessity during long migration journeys; protected lands are not evenly
distributed geographically and given similar habitat conditions, protected and non-
protected land will be used during the course of migration.
The lack of correlation between climatic factors and stopover length is somewhat
surprising. Many studies in the past have found that climate seems to, at the very least,
affect distribution of dabbling duck species (Brook et al., 2009, Schummer et al., 2010).
Therefore one would expect that migratory behaviors might be affected as well,
especially in months with more extreme weather conditions. The most likely cause for
this study’s lack of correlation would seem to be attempting to extrapolate two discrete
values over extensive periods of time. For example: assessing mean temperature on the
day of arrival of a 30-day stopover is likely a gross generalization for examining actual
dynamics. Instead, better results might be observed analyzing average snow cover over
the duration of the stopover since this has been shown to affect A. rubripes survey
22

studies taken during spring migration as the birds fly further north towards their nesting
grounds (Chaulk and Turner, 2007). Furthermore, I would suggest that rather than taking
temperature and precipitation data readings from the beginning and end of the stopover, a
more continuous spatio-temporal analysis be employed to attempt to correlate patterns in
temperature fluctuations with the initiation of bird movements. Although the first
inclination might be to use average temperature over a stopover event, this method has
the difficulty of accounting for large differences in stopover length given that this dataset
contained values between 4 and 1000 hours. This variable is still relatively poorly
studied owing to the difficulty of establishing long term data sets, so these are potentially
very good areas for future research.
Lastly, the correlation and resulting bias occurring between HCat and PAD
categories in this study should be remedied in future research efforts. The strong
connection between HCat = 1 and PAD = 0 resulted from a lack of information about
hunting conditions on non-protected lands. This could be addressed by conducting more
detailed research on lands not contained within the PAD. Including additional categories
for HCat that represented hunting practices on private lands where A. rubripes were
observed would help to refine the analysis for non-protected lands. Unfortunately, given
the time frame and resources of this study, extensive landowner research could not be
conducted.
In addition, many National Wildlife Refuges or Wildlife Management Areas have
hunting sites delineated within the boundaries of the larger property. One could
hypothesize that sub-sites within management areas where hunting pressures were lower
23

might foster more usage by A. rubripes and similar species. This information would, if
collected, provide an excellent way of assessing the effect of fine scale intermediate
hunting disturbances on migrating and wintering waterfowl.
CONCLUSIONS
Although hunting is one of the principal disturbances associated with A. rubripes
population decline, wintering habitat where hunting is allowed is used extensively by
ducks. This is demonstrated by a large proportion of telemetry observations occurring in
protected and managed lands where hunting is a prevalent recreational activity.
My chief research question of whether or not protected areas in and of themselves
have an effect on waterfowl migration seems to have several answers, and those answers
are largely based on temporal factors that require further investigation. During wintering,
birds seem to more often choose lands where hunting is allowed in-season at varying
degrees, potentially due to greater land manipulation and management to create habitat
best suited to the wildlife of concern in these areas. This could also be attributed to the
establishment of these sites in areas with high inherent habitat quality. During migration,
telemetry points occurred largely outside protected areas of any kind. These data suggest
that either habitat requirements are different during migration (assuming all of the best
habitats are protected) or that not all of the best habitats are protected (assuming habitat
preferences are unchanged during migration). When ducks were recorded in protected
areas during migration, they were predominantly found in sites where hunting was openly
permitted when in season.
24

Further research in this area would be productively directed towards climate-
related variables and further exploration of stopover length. These concepts are
somewhat nebulous and a bit difficult to manage in a spatio-temporal sense, especially in
the vein of determining an effective model for tracking weather changes over the course
of a stopover event. It is quite probable that dynamics of temperature, precipitation, and
snow-cover over the course of a stopover event, rather than just two discrete points,
would be more descriptive of resource allocation, food availability, and carrying
capacity; further investigation is certainly warranted.
At a more basic level, this study would benefit greatly from more ancillary data
development. Contacting individual land owners where birds landed during stopover
events and discovering their hunting practices would be a great addition. This would aid
in seeing the effects of hunting on birds while they move, specifically during fall
migration (were these points to be collected by future surveys). For PAD lands, analysis
using not only the outlines of the wildlife areas in question but also digitizing hunting
sites within said areas could give different results; perhaps birds avoid areas within
managed sites where hunting is allowed in favor of less disturbed areas. Providing
greater resolution for HCat and PAD categories and developing more robust datasets
would grant better results, but the descriptive analyses conducted here suggest that these
behaviors warrant further research.

25

REFERENCES

AFTON, A. D. 2008. Chronology and rates of migratory movements, migration
corridors, and habitats used throughout the annual cycle by female Lesser Scaup
radio-marked on Pool 19 of the Mississippi River. Baton Rouge, LA: Louisiana
State University.

ANDERSON, K. A., BOWMAN, J.B., COLUCCY, J.M., & YERKES, T. 2009. Spring
Migration Ecology of American Black Ducks Determined by Satellite Telemetry.
Master of Science, University of Delware.

BENNETT, V. J., FERNANDEZ-JURICIC, E., ZOLLNER, P. A., BEARD, M. J.,
WESTPHAL, L. & FISHER, C. L. L. 2011. Modelling the responses of wildlife
to human disturbance: An evaluation of alternative management scenarios for
black-crowned night-herons. Ecological Modelling, 222, 2770-2779.

BRASHER, M. G., STECKEL, J. D. & GATES, R. J. 2007. Energetic carrying capacity
of actively and passively managed wetlands for migrating ducks in Ohio. Journal
of Wildlife Management, 71, 2532-2541.

BROOK, R. W., ROSS, R. K., ABRAHAM, K. F., FRONCZAK, D. L. & DAVIES, J. C.
2009. Evidence for Black Duck Winter Distribution Change. Journal of Wildlife
Management, 73, 98-103.

CHAULK, K. G. & TURNER, B. 2007. The timing of waterfowl arrival and dispersion
during spring migration in Labrador. Northeastern Naturalist, 14, 375-386.

CONROY, M. J., MILLER, M. W. & HINES, J. E. 2002. Identification and synthetic
modeling of factors affecting American black duck populations. Wildlife
Monographs, 1-64.

DU. 2010. Black Duck Study [Online]. Memphis, TN: Ducks Unlimited. Available:
http://www.ducks.org/conservation/black-duck-study/black-duck-study [Accessed
March 18 2012].

FILOSO, S. & PALMER, M. A. 2011. Assessing stream restoration effectiveness at
reducing nitrogen export to downstream waters. Ecological Applications, 21,
1989-2006.

FOSTER, M. A., GRAY, M. J. & KAMINSKI, R. M. 2010. Agricultural Seed Biomass
for Migrating and Wintering Waterfowl in the Southeastern United States.
Journal of Wildlife Management, 74, 489-495.

26

GASCOIGNE, W. R., HOAG, D., KOONTZ, L., TANGEN, B. A., SHAFFER, T. L. &
GLEASON, R. A. 2011. Valuing ecosystem and economic services across land-
use scenarios in the Prairie Pothole Region of the Dakotas, USA. Ecological
Economics, 70, 1715-1725.

KIRBY, J. S., STATTERSFIELD, A. J., BUTCHART, S. H. M., EVANS, M. I.,
GRIMMETT, R. F. A., JONES, V. R., O'SULLIVAN, J., TUCKER, G. M. &
NEWTON, I. 2008. Key conservation issues for migratory land- and waterbird
species on the world's major flyways. Bird Conservation International, 18, S49-
S73.

LOK, E. K., ESLER, D., TAKEKAWA, J. Y., DE LA CRUZ, S. W., BOYD, W. S.,
NYSEWANDER, D. R., EVENSON, J. R. & WARD, D. H. 2011. Stopover
Habitats of Spring Migrating Surf Scoters in Southeast Alaska. Journal of Wildlife
Management, 75, 92-100.

LONGCORE, J. R., MCAULEY, D.G, HEPP, G.R., RHYMER, J.M. 2000. Anas
rubripes: American Black Duck. Birds of North America.

MDNR 2002. St. John's Marsh Wildlife Area: Special Use and Hunting Rules. In:
RESOURCES, M. D. N. R. (ed.). Clay Township.

NAUGLE, D. E., JOHNSON, R. R., ESTEY, M. E. & HIGGINS, K. F. 2001. A
landscape approach to conserving wetland bird habitat in the prairie pothole
region of eastern South Dakota. Wetlands, 21, 1-17.

O'CONNELL, J. L. & NYMAN, J. A. 2011. Effects of Marsh Pond Terracing on Coastal
Wintering Waterbirds Before and After Hurricane Rita. Environmental
Management, 48, 975-984.

O'NEAL, B. J., STAFFORD, J. D. & LARKIN, R. P. 2012. Stopover duration of fall-
migrating dabbling ducks. Journal of Wildlife Management, 76, 285-293.

PATTERSON, D. 2011. Split Layer By Attributes. In: ESRI (ed.) ArcGIS Resource
Center.

PLATTNER, D. M., EICHHOLZ, M. W. & YERKES, T. 2010. Food Resources for
Wintering and Spring Staging Black Ducks. Journal of Wildlife Management, 74,
1554-1558.

SCHUMMER, M. L., KAMINSKI, R. M., RAEDEKE, A. H. & GRABER, D. A. 2010.
Weather-Related Indices of Autumn-Winter Dabbling Duck Abundance in Middle
North America. Journal of Wildlife Management, 74, 94-101.
27

SHERFY, M. H., ANTEAU, M. J. & BISHOP, A. A. 2011. Agricultural Practices and
Residual Corn During Spring Crane and Waterfowl Migration in Nebraska.
Journal of Wildlife Management, 75, 995-1003.

USDA. 2012. Conservation Programs [Online]. United States Department of
Agriculture. Available:
http://www.fsa.usda.gov/FSA/webapp?area=home&subject=copr&topic=landing
[Accessed March 18 2012].

USGS. 2011. Protected Areas Database of the United States [Online]. Available:
http://gapanalysis.usgs.gov/padus/data/ [Accessed February 25 2012].

VEGVARI, Z., BARTA, Z., MUSTAKALLIO, P. & SZEKELY, T. 2011. Consistent
avoidance of human disturbance over large geographical distances by a migratory
bird. Biology Letters, 7, 814-817.

28

APPENDIX: PYTHON CODE FOR “ADD FIELD” TOOL

import arcpy
#set environment
arcpy.env.workspace = "C:/Users/Brian/Desktop/Black Duck Data/Black Duck
Data/Black Duck 2.gdb/Stopover"
arcpy.env.overwriteOutput = True
#create list of datasets in blackduckspoints.gdb
fcList = arcpy.ListFeatureClasses()
#create a loop to add fields
for fc in fcList:
arcpy.AddField_management(fc, "Arrival_Date", "Text", "", "", "8", "ArrDate",
"NULLABLE", "")

For use in other settings, the workspace would need to be modified to suit the
individual computer being used. Additionally, parameters of the field should be adjusted
to fit the need in the final “for” loop.

Abstract (if available)

Abstract Waterfowl are one of our Nation’s most precious and abundant natural resources, and preserving habitat well suited to their needs has long been a goal of private and public entities alike. In this study, I focused on the American Black Duck (Anas rubripes), a species seeing a large decline in numbers since the mid 20th century. Using a satellite telemetry dataset collected by Ducks Unlimited during 2008 and 2009 in the context of the Protected Areas Database of the United States (PAD), I addressed the land use habits of A. rubripes to assess the efficacy of costly conservation efforts implemented through conservation easements and the maintenance of wildlife refuges and management areas. Most analyses were conducted at the stopover level, grouping telemetry points within a 0.5 decimal degree diameter. By creating distributions and studying correlations, this study finds that during wintering months A. rubripes registered more telemetry points in PAD lands where hunting is allowed in-season

Conceptually similar

PDF

Modeling open space acquisition in Boulder, Colorado

PDF

The geography of voter power in the U.S. electoral college from 1900-2012

PDF

Spatial distribution of the Nile crocodile (Crocodylus niloticus) in the Mariarano River system, Northwestern Madagascar

PDF

Effect of spatial patterns on sampling design performance in a vegetation map accuracy assessment

PDF

Investigation and analysis of land use/land cover to determine the impact of policy developments in cities

PDF

Geographic information systems eelgrass (Zostera marina) habitat restoration suitability model, Long Island Sound, USA - a ‘sound-wide’ model

PDF

Testing LANDIS-II to stochastically model spatially abstract vegetation trends in the contiguous United States

PDF

Exploring urban change using historical maps: the industrialization of Long Island City (LIC), New York

PDF

Spatiotemporal visualization and analysis as a policy support tool: a case study of the economic geography of tobacco farming in the Philippines

PDF

Investigating electoral college reform: geography's impact on elections, and how maps influence our perception of election outcomes

PDF

Analysis of future land use conflict with volcanic hazard zones: Mount Rainier, Washington

PDF

Utilizing GIS and remote sensing to determine sheep grazing patterns for best practices in land management protocols

PDF

Questioning the cause of calamity: using remotely sensed data to assess successive fire events

PDF

Modeling burn probability: a Maxent approach to estimating California's wildfire potential

PDF

Using pattern oriented modeling to design and validate spatial models: a case study in agent-based modeling

PDF

A comparison of GLM, GAM, and GWR modeling of fish distribution and abundance in Lake Ontario

PDF

Reducing maternal mortality by improving medical facility accessibility: a methodology demonstrated for the Democratic Republic of the Congo

PDF

Using Landsat and a Bayesian hard classifier to study forest change in the Salmon Creek Watershed area from 1972–2013

PDF

Integrating land survey data into measurement-based GIS: an assessment of challenges and practical solutions for surveyors in Texas

PDF

The role of GIS in asset management: integration at the Otay Water District

Asset Metadata

Creator Kearns, Brian Vance (author)

Core Title Is the likelihood of waterfowl presence greater on conserved lands?

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 04/18/2012

Defense Date 03/24/2012

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag conservation,GIS,land easements,Land use,migration,OAI-PMH Harvest,waterfowl

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Longcore, Travis R. (committee chair), Kemp, Karen K. (committee member), Vos, Robert O. (committee member)

Creator Email bkearns@usc.edu,brian.v.kearns@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-7713

Unique identifier UC1113219

Identifier usctheses-c3-7713 (legacy record id)

Legacy Identifier etd-KearnsBria-619.pdf

Dmrecord 7713

Document Type Thesis

Rights Kearns, Brian Vance

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA