Download nest webs and path modeling methods

CAVITY EXCAVATION AND ENLARGEMENT AS MECHANISMS
FOR INDIRECT INTERACTIONS IN AN AVIAN COMMUNITY
Lori A. Blanc, Department of Biology, Virginia Tech University, Blacksburg, VA 24061, E-mail: [email protected]
Jeffrey R. Walters, Department of Biology, Virginia Tech University, Blacksburg, VA 24061, E-mail: [email protected]
INTRODUCTION
EXPERIMENTAL RESULTS
Mean Change in # Detections Per Plot .
Direct and indirect species interactions within ecological communities may play a strong role in
influencing or maintaining community structure. Complex community interactions pose a
major challenge to predicting ecosystem responses to environmental change because
predictive frameworks require identification of mechanisms by which community interactions
arise. Cavity-nesting communities are well suited for mechanistic studies of species
interactions because 1) cavity-nesters interact through the creation of and competition for
cavity nest-sites and 2) cavity availability is relatively easy to measure and manipulate. In this
study, we use a cavity ‘nest-web’ and path analysis as a predictive framework for identifying
potential indirect species interactions between the federally endangered Red-cockaded
Woodpecker and other cavity-nesting birds in northwest Florida. We then use experimental
manipulation of cavity availability to test predictions about potential species interactions.
Model fit: X2 = 5.8, P=0.33; RMSEA = 0.04
A concurrent experimental study enabled us to test relationships shown in our path model. We
experimentally blocked 2 pathways within the model, including a) that between enlarged Redcockaded Woodpecker cavities and Northern Flicker presence and b) that between enlarged
Red-cockaded Woodpecker cavities and large SCNs. Experimental manipulations were
designed to prevent the cavity enlargement behavior of the Northern Flicker and remove
enlarged Red-cockaded Woodpecker cavities from the research plots.
METHODS: We randomly selected 8 plots for experimental manipulation along with 8 control
plots. Plots within both groups contained a comparable range of enlarged Red-cockaded
Woodpecker cavities and snag availability. Following the 2003 field season, we installed metal
restrictor plates on all inactive and enlarged Red-cockaded Woodpecker cavities within
experimental plots (n=135 cavities, Fig. 4). Metal restrictor plates are commonly used to repair
and prevent enlargement of Red-cockaded Woodpecker cavities by other woodpecker species.
Experimental manipulation was complete in January 2005. The experimental portion of this
study consists of 2003 (pre-treatment) and 2005 (post-treatment) data. To control for annual
variation, we used the amount of change in abundance and number of nests found in each plot
between 2003 and 2005 for analysis. For each plot we subtracted pre-treatment from posttreatment values and used the difference for analysis. We used a Mann Whitney U test to
detect significant differences in the amount of change between treatment and control plots.
EXPERIMENTAL MANIPULATION
Figure 2. A nest web diagram depicting cavity creation and
use within the cavity-nesting bird community at Eglin Air
Force Base. N indicates the number of nests found for a
species; E indicates the number of nest cavities excavated by
that species. Links between the Secondary Cavity Nester
(SCN) level and the Primary Excavator (PCE) level represent
the proportion of SCN nests found cavities excavated by the
indicated PCE. Links between the PCE and Tree level show
the proportion of PCE nests found in that tree resource.
Photo Credit: Kevin Rose
Figure 3. Path diagram depicting the indirect relationship between the RCW
cavities, pine snags, Northern Flicker and large SCNs, including the
southeastern American Kestrel and Eastern Screech-Owl. Arrows reflect
hypotheses about causal relationships between variables; associated
numbers indicate the path coefficient (i.e. direct effect of independent variable
on adjacent dependent variable). The model explained 35% of variation (R2)
in large SCN nests and had a strong fit to observed data. When modified to
include only snags suitable for nesting (dbh > 20cm), model fit improved
(X2 = 5.0, P=0.41; RMSEA = 0.008). In addition, the path between pine snags
and Northern Flicker became significant (P=0.001).
0.5
Control
Restrict
00
-0.5
 No difference between the change in Northern Flicker (NOFL)
abundance in experimental and control plots (U = 70.5, P = 0.42).
 Decrease in American Kestrel (AMKE) abundance in experimental
plots, but not significantly more than control (U = 71, P = 0.40).
-1
 Marginally significant increase in Red-cockaded Woodpecker
(RCW) abundance in experimental plots (U = 54.5, P = 0.08).
-1.5
AMKE
NOFL
RCW
(n=16)
.
PART 2: EXPERIMENTAL STUDY
135 cavities restricted
(8 treatment plots / 8 control)
Figure 4. Metal restrictor plates -- a cavity management technique commonly used for management of the
federally endangered Red-cockaded Woodpecker. We used restrictor plates to experimentally reduce the number
of enlarged Red-cockaded Woodpecker cavities and prevent Red-cockaded Woodpecker cavity enlargement.
Mean Change in # Nests Found Per Plot
Mean Change in # Nests Found Per Plot .
The federally endangered Red-cockaded Woodpecker, endemic to fire-maintained
southern pine forests, is the only species in North America that excavates cavities in
living pine. Sap wells excavated around the cavity help prevent nestling predation
by rat snakes, which can climb the bark of living pine. Other woodpecker species,
such as the Northern Flicker and Pileated Woodpecker, often enlarge these cavities,
subsequently creating nesting habitat for large, secondary-cavity nesters (e.g.,
Wood Ducks, Eastern Screech-Owls, American Kestrels, Fox Squirrels). Over 27
vertebrate species are known to used Red-cockaded Woodpecker cavities.
Figure 1. The study was conducted at Eglin Air Force Base in the Florida panhandle.
RESULTS
1
1
A dead tree, referred to as a ‘snag’. Dead wood is
critical for cavity-excavation by most woodpecker
species. There are many species that cannot
excavate, but require cavities for nesting. These
‘secondary cavity nesters’ (SCN) rely on
woodpeckers to excavate cavities.
METHODS: From 2002-2005, we monitored abundance and nests of cavity-nesters on 36
48ha plots at Eglin Air Force Base (Fig. 1). We constructed a nest web using nest cavities for
which we could identify the excavator (Fig. 2). We then developed a path diagram to depict an
indirect relationship identified in the nest web (Fig. 3). Observational units in the model were
research plots from 2003-2005 (n=92), excluding plots which had experimental manipulations
from a concurrent study on large cavity availability (n=16, see Part 2). We developed our path
model using AMOS v.5.0 and assessed model fit with chi-square analysis (significant at P >
0.05) and root mean square error approximation (RMSEA, significant at P < 0.05).
Bars reflect mean change in relative abundance from 2003 to 2005,
including StdErr.
1.5
-2
Study Site
PART 1: NEST WEBS AND PATH MODELING
2
NOFL
Bars reflect mean change in the number of Northern Flicker (NOFL)
nests found from 2003 to 2005, including StdErr. Asterisk indicates
significant difference (P<0.05) between control and experimental plots.
0.5
*
00
Control
Restrict
-0.5
-1
Nests in RCW Cavities
Nests in Snags
1
(n=16)
SCN
0.5
00
*
Control
Restrict
-0.5
Nests in RCW Cavities
Nests in Snags
 Significant increase in NOFL nests found in snags in experimental
plots (U = 48.5, P = 0.01), indicating a switch to the use of snags.
Bars reflect the mean change in the number of large SCN nests found
from 2003 to 2005, including standard error. Asterisk indicates
significant difference (P<0.05) between control and experimental plots.
 Number of large SCN nests found in Red-cockaded Woodpecker
(RCW) cavities decreased significantly more in experimental plots
than in controls (U = 89.5, P = 0.01).
 No change in number of large SCN nests found in snags
(U = 64.5, P = 0.38).
-1
-1.5
 Decrease in number of NOFL nests found in RCW cavities in
experimental plots, although not significantly more than in controls
(U = 78, P = 0.11).
(n=16)
The application of metal restrictor plates caused a decline in large SCN nests found in Redcockaded Woodpecker cavities and a switch to the use of snags by the Northern Flicker.
Large pine snags apparently buffered the impact of metal restrictor plates. The use of snags
for nesting requires either the ability to find existing cavities in snags or excavate new cavities.
The Northern Flicker, capable of cavity-excavation, had greater flexibility than SCNs in
switching to the use of available snags when Red-cockaded Woodpecker cavities were
restricted and this flexibility was reflected in our experimental results. These results indicate
that Red-cockaded Woodpecker cavity management can indirectly impact large SCNs.
CONCLUSIONS
In summary, we identified an indirect interaction between the Red-cockaded
Woodpecker and large SCNs, mediated by the Northern Flicker and used path analysis to
show how cavity excavation and enlargement can drive these relationships. Then, through
experimental manipulation, we a) confirmed the role of these 2 mechanisms in creating habitat
for large SCNs and b) demonstrated that metal restrictor plates can disrupt this indirect
relationship by affecting Northern Flicker cavity-enlargement behavior. Recent studies indicate
that indirect interactions based on modification of traits (e.g., behavior), such as that shown
here, may be highly important and widespread phenomena within ecological communities.
This study also provides an empirical example of how, in ecological communities where
complex species interactions occur, single-species management can have indirect impacts on
non-target species. In some cases, indirect effects may conflict with ecosystem management
goals and subsequently, have broad implications for conservation management. Indeed, it has
been proposed that biodiversity conservation include protection of critical interspecific
interactions. Finally, this research demonstrates how a study of basic ecological principles can
inform conservation management and concommitally, how the application of a management
technique can aid in the study of basic ecological principles. Our findings highlight the need to
develop a better understanding of how mechanisms underlying species interactions can
influence community structure and subsequently cause unexpected community responses to
environmental change.
FUNDING AND FIELD ASSISTANCE
Funding for this study was provided by D.O.D. Threatened & Endangered Species funds from Eglin Air Force Base, a National Science Foundation DDIG, and
grants from Sigma Xi, Florida Ornithological Society, WPI International, PEO International and the Virginia Tech Graduate Research Development Program.
Many thanks to the RCW Research Team, Jackson Guard personnel at Eglin AFB and seasonal field crews for their valuable assistance and support.