AN ABSTRACT OF THE THESIS OF
Kristin N. Jones for the degree of Master of Science in Forest Ecosystems and Society
presented on November 20, 2015
Title: Combined Effects of Intensive Forest Management and Microclimate on
Reproduction in a Cavity-Nesting Songbird
Abstract approved: ________________________________________________________
James W. Rivers
Matthew G. Betts
ABSTRACT
Future scenarios of global climate change rely on large-scale climate envelope
models that do not account for local climatic conditions to which organisms most closely
respond. Shifts in species distributions and phenology driven by climate change are welldocumented, yet we lack a strong understanding of how climate change will influence the
demographic rates of animal populations, which directly determine the likelihood of
species persistence. Land cover change, an agent of global change that is increasing in
extent and intensity to meet the needs of a growing human population, can combine with
climate change to stress animal populations by altering microclimatic conditions, as well
as changing the availability of resources such as food and cover from predators. Thus,
identifying the individual and combined effects of climate and land management
practices is essential to accurately predict the responses of animal populations to global
change.
Intensive forest management results in land cover change by suppressing the
growth of competing plant species in favor of commercial species and altering the
abundance and composition of forest vegetation. In turn, changes in the abundance and
composition of forest vegetation can modify local air temperatures. I hypothesized that
herbicide application would alter the thermal environment for early-seral forest
organisms (Chapter 2). If this were the case, then air temperature could be expected to
increase in magnitude and variability along a gradient of herbicide treatment intensity. To
test this hypothesis, I used iButton dataloggers to monitor air temperature at 160 nest
boxes on 20 intensively managed early-seral forest stands (10.4 - 18.9) in the northern
Oregon Coast Range, USA representing a gradient in intensive forest management (i.e.,
no-spray control, light, moderate and intensive herbicide application). I also measured the
amount and composition of vegetation cover to test for herbicide effects on vegetation
among application intensities. Using linear mixed models, I compared three measures of
air temperature (mean daily minimum, mean, and maximum) and their associated
coefficients of variation (CVs). Additionally, I used linear mixed models to confirm
differences in total vegetation and broadleaved vegetation cover. Although mean total
vegetation cover generally decreased, it did not significantly differ among herbicide
treatments; in contrast, mean broadleaved vegetation cover was significantly reduced in
the moderate and intensive treatments. Herbicide treatment was a significant predictor of
maximum and mean temperatures, but minimum temperatures did not differ with
herbicide treatment. Although there was an effect of herbicide treatment on air
temperature, corrected pairwise comparisons indicated no significant differences among
treatments. I note that though my power to detect statistical differences among treatments
was limited, these differences were quite small (< 0.5°C) and confidence intervals were
generally relatively narrow (< 1.5°C), suggesting that temperature did not differ among
herbicide treatments in biologically meaningful ways. Furthermore, I did not detect an
effect of herbicide treatment on temperature variability. Estimated differences in
temperature variability among treatments were small (< 0.5%) and confidence intervals
covered a relatively broad range of values, indicating that I did not have enough
statistical power to detect effects. I found no uniform pattern in the direction (positive or
negative) of the effect of herbicide treatment on temperature or CVs among treatment
intensities.
Daily air temperatures can strongly influence the reproductive output of earlyseral songbirds if temperatures exceed physiological tolerances of offspring, decreasing
physiological performance (e.g., excessively high and low temperatures can alter
metabolic rates) and survival. Moreover, the abundance and composition of early-seral
forest vegetation can influence songbird reproductive output through changes in nest
predator communities, food resources, or both. Thus, I further hypothesized that intensive
forest management practices could combine with intraseasonal air temperatures to impact
reproductive output in an insectivorous cavity-nesting songbird, the House Wren
(Trogolodytes aedon) (Chapter 3). If this were the case, then House Wren nest survival,
the number of offspring produced, and the quality of those offspring would decline with
greater management intensity and increasing air temperatures. To test these predictions, I
monitored 283 nests on 24 intensively managed early-seral forest stands (8.6 - 18.9 ha) in
the northern Oregon Coast Range, USA representing a gradient in intensive forest
management (i.e., no-spray control, light, moderate and intensive herbicide application).
I used data from Chapter 2 to test for combined effects of air temperature and herbicidedriven vegetation changes on House Wren reproductive output. Using linear mixed
models within a model selection framework, I did not find support for combined effects
of temperature and herbicide treatment on nest survival. After accounting for the effects
of herbicide-driven vegetation changes on reproductive output, air temperature effects
were negligible. My results suggest that post-harvest vegetation management likely does
not influence the number of young produced nor their quality (as indicated by body
condition) in intensively managed early-seral forests, but may influence nest survival.
However, nest survival did not decline along a gradient of herbicide intensity as I
expected. Instead, mean nest survival was greatest in the control and most intensively
managed stands (failure was greatest in the light treatment); however, these effects were
so variable as to not be statistically significant.
My results suggest that post-harvest vegetation management in intensively
managed forests may be linked to minor changes in microclimatic air temperatures, but
that there is high variation in these effects and effects are likely small. Therefore, there
appears to be limited potential for vegetation control strategies through herbicides to
buffer expected climate change effects on organisms in early-seral forest. My results also
suggest that the potential for combined effects of herbicide application and air
temperature on early-seral cavity-nesting songbirds is limited. Under current local
climate patterns, air temperature appears to exert negligible effects on House Wren
reproductive output after accounting for changes in vegetation cover. The effects of
herbicide-driven vegetation changes on early-seral songbirds may be large, though highly
variable, and not an increasing function of herbicide intensity. These effects may be
primarily predator-mediated, as indicated by the large effects of herbicide treatment on
House Wren nest survival but not the number of offspring produced nor their quality. My
finding of limited temperature effects on House Wren reproductive output compared to
the effects of forest management intensity supports predictions that, despite increasing
concerns over the impacts of advancing climate change on animal populations, land cover
change driven by anthropogenic land use will continue to be the primary global change
driver impacting animal populations in the near future.
©Copyright by Kristin N. Jones
November 20, 2015
All Rights Reserved
Combined Effects of Intensive Forest Management and Microclimate on Reproduction in
a Cavity-Nesting Songbird
by
Kristin N. Jones
A THESIS
Submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Master of Science
Presented November 20, 2015
Commencement June 2016
Master of Science thesis of Kristin N. Jones presented on November 20, 2015
APPROVED:
Co-Major Professor, representing Forest Ecosystems and Society
Co-Major Professor, representing Forest Ecosystems and Society
Head of the Department of Forest Ecosystems and Society
Dean of the Graduate School
I understand that my thesis will become part of the permanent collection of Oregon State
University libraries. My signature below authorizes the release of my thesis to any reader
upon request.
Kristin N. Jones, Author
ACKNOWLEDGEMENTS
This project would not have been possible without the use of land provided by
Weyerhauser Company, Hancock Natural Resource Group, Plum Creek Timber
Company, and the Oregon Department of Forestry; collaboration with Jake Verschuyl
and the National Council of Air and Stream Improvement, Inc.; and funding provided by
grants from the United States Department of Agriculture, Agriculture Food and Research
Initiative grant (AFRI-2009-04457), the National Council for Air and Stream
Improvement, the Nobel Fund, Giustina Land and Timber, the Oregon Forest Industry
Council, and the Oregon State University College of Forestry Fish and Wildlife Habitat
in Managed Forests Research Program.
Thanks to Daniel Ardia of Franklin & Marshall College for lending equipment
and protocols and providing guidance on temperature-related measures; and to T.
Manning, J. Bailey, K. Zummo, T. Barron, C. Adlam, H. Beyl, B. Cooney, G. Cummins,
S. Doorly, E. Eve, D. Ferraro, B. Hardt, E. Keyes, D. Millican, L. Natola, E. White, D.
Backlund, T. Stokely, C. Fitzmorris, N. Volpe, K. Wilson, S. Jordan, C. Loucks, E.
Pokrivka, S. Campbell, R. Hepner, N. Garlick, A. Muniz, and J. Powell, who were
absolutely instrumental to data collection. Thanks also to Forest Industry Biologists A.J.
Kroll and Jack Giovanni for providing supporting data, and to cooperating forest industry
biologists and managers J. Johnson, S. Keniston, M. Rochelle, T. Tompkins, J. Bakke, J.
DeRoss, R. Frazzini, A. Heimgartner, T. McBride, J. Thiemens, A. Weathers, E. Finnell,
T. Mortensen, M. Taylor, J. Travers, D. Irons, J. Light for providing further project
support.
I am deeply grateful to my advisors, James Rivers and Matthew Betts, for their
boundless and creative insight into the research process, and for constantly pushing me to
be a better scientist. I am also sincerely thankful to my committee members, Kerry
Grimm and Joan Hagar for their support and research input. Thanks also to the Betts lab
group, especially Thomas Stokely, for providing guidance and insight into research and
analysis. I would like to give an especially massive thank you to Ariel Muldoon for her
instrumental assistance with programming and analysis and seemingly endless patience.
Thanks also to Lisa Ganio for analysis assistance and always being happily willing to
discuss research, statistics, and professional development. I would also like to thank John
Bliss for his wisdom and guidance, as well as Lisa Shipley and Keith Blatner for their
instrumental roles in my placement in the master’s program in the College of Forestry at
Oregon State University. Finally, thanks especially to my parents, Korina Layne-Jones
and Christopher Jones, for keeping me “sane” throughout my master’s program and for
unfailingly providing support and encouragement throughout all of life’s adventures.
TABLE OF CONTENTS
Page
CHAPTER 1: GENERAL INTRODUCTION ................................................................... 1
CHAPTER 2: INTENSIVE FOREST MANAGEMENT EXERTS WEAK AND
VARIABLE INFLUENCES ON MICROCLIMATE IN EARLY-SERAL FOREST ....... 9
2.1 Abstract ..................................................................................................................... 9
2.2 Introduction ............................................................................................................. 10
2.3 Materials and methods ............................................................................................ 15
2.3.1 Study area......................................................................................................... 15
2.3.2 Experimental design and treatments ................................................................ 16
2.3.3 Sampling design ............................................................................................... 17
2.3.3.1 Air temperature ......................................................................................... 18
2.3.3.2 Vegetation abundance and composition ................................................... 18
2.3.4 Data filtering and summary ............................................................................. 20
2.3.5 Statistical analysis ............................................................................................ 20
2.4 Results ..................................................................................................................... 22
2.4.1 Vegetation ........................................................................................................ 22
2.4.2 Mean air temperature ....................................................................................... 23
2.4.3 Variability in mean air temperature ................................................................. 24
2.5 Discussion ............................................................................................................... 25
2.6 Management implications ....................................................................................... 31
CHAPTER 3: INTENSIVE FOREST MANAGEMENT HAS A GREATER IMPACT
THAN CLIMATE ON REPRODUCTIVE OUTPUT IN A CAVITY-NESTING
SONGBIRD ...................................................................................................................... 41
3.1 Abstract ................................................................................................................... 41
3.2 Introduction ............................................................................................................. 42
3.3 Materials and methods ............................................................................................ 48
3.3.1 Study area and species ..................................................................................... 48
3.3.2 Experimental design......................................................................................... 50
3.3.3 Sampling design ............................................................................................... 51
TABLE OF CONTENTS (Continued)
3.3.3.1 Environmental measurements ................................................................... 51
3.3.3.2 Nest survival and productivity .................................................................. 54
3.3.4 Statistical analysis ............................................................................................ 55
3.3.4.1 Vegetation ................................................................................................. 56
3.3.4.2 Reproductive output .................................................................................. 56
3.4 Results ..................................................................................................................... 59
3.4.1 Effects of herbicide treatments on temperature and vegetation ....................... 59
3.4.2 Influence of herbicide treatment and temperature on nest survival ................. 60
3.4.3 Influence of herbicide treatment and temperature on the number of young
produced .................................................................................................................... 62
3.4.4 Influence of herbicide treatment and temperature on nestling body
condition ................................................................................................................... 63
3.5 Discussion ............................................................................................................... 64
CHAPTER 4: CONCLUSIONS ....................................................................................... 86
LITERATURE CITED ..................................................................................................... 88
APPENDICES ................................................................................................................ 100
Appendix A: Herbicide Prescriptions ......................................................................... 101
LIST OF FIGURES
Figure
Page
Figure 2.1. Location of the eight study blocks used to examine the influence of intensive
forest management on early-seral forest biodiversity in the Oregon Coast Range. Blocks
used in this study to assess the influence if intensive forest management on microclimatic
air temperature, May-August 2014, are indicated by orange squares............................... 35
Figure 2.2. Representative examples of stands spanning the gradient in herbicide intensity
for (A) Control, (B) Light, (C) Moderate, and (D) Intensive treatments in the Oregon
Coast Range, US, 2014 ..................................................................................................... 36
Figure 2.3. Ratio of mean percent cover among control, light, moderate, and intensive
herbicide treatments, with Bonferroni-adjusted 95% confidence intervals. The dashed line
at 1 represents no statistical difference. The left panel (A) depicts differences in
broadleaved cover. The center panel (B) depicts differences in conifer cover. The right
panel (C) depicts differences in total vegetation cover ..................................................... 37
Figure 2.4. Average daily air temperature fluctuations among herbicide treatments at each
experimental block in the Oregon Coast Range, US. Lines represent the means of all
sample points (8) for May-August, 2014 .......................................................................... 38
Figure 2.5. Mean air temperature differences among control, light, moderate, and
intensive herbicide treatments in the Oregon Coast Range, May-August 2014, with
Bonferroni-adjusted 95% confidence intervals. The left panel (A) depicts differences in
mean air temperature. The right panel (B) depicts differences in air temperature
variability (CVs). The dashed line at zero represents no statistical difference. Estimates
on the right side of the dashed line correspond to decreases in air temperature with
increasing herbicide treatment intensity, while estimate on the left side correspond to
increases in air temperature .............................................................................................. 39
Figure 3.1. (A) A House Wren (Troglodytes aedon) nestling approximately 8 days after
hatching being prepared for measurement of body condition, and (B) a completed clutch
of House Wren eggs during incubation inside a nest box ................................................. 76
LIST OF FIGURES (Continued)
Figure 3.2. General location map of the eight study blocks used to examine the influence
of intensive forest management on early-seral forest biodiversity in the Oregon Coast
Range. Blocks used in this study to assess the influence if intensive forest management
on House Wren reproductive output May-August 2014, are indicated by orange squares
........................................................................................................................................... 77
Figure 3.3. Mean daily maximum temperature (Tmax) across all study sites in the Oregon
Coast Range during the House Wren breeding season, May-August 2014. Measurements
were taken every 15 min, averaged across a 24-hour period, and then averaged over all
nest boxes (n=192) on 24 study sites in the Oregon Coast Range. The gray band shows
the 95% confidence interval.............................................................................................. 78
Figure 3.4. Representative examples of stands spanning the gradient in herbicide
intensity, including ranging from (A) Control (no-spray), (B) Light, (C) Moderate, and
(D) Intensive treatments in the Oregon Coast Range, US, 2014 ...................................... 79
Figure 3.5. Plots depicting the effect of herbicide treatment and Tmax and on House Wren
nest survival during the breeding season in the Oregon Coast Range, 2014. (A)
Differences in odds ratio estimates of nest survival between the effect of herbicide
treatments relative to control stands; odds ratios were averaged over all models in the
candidate set. The dashed horizontal line represents odds ratios of the control for
comparison with herbicide treatments; 95% CIs that overlap one indicate lack of
significant differences relative to the control treatment. Numbers above confidence
intervals indicate the number of nests monitored per treatment; there were 73 nests in the
control treatment. (B) Boxplots depicting Tmax values from nests that either failed (pink)
or fledged offspring (blue) across all four treatments. Vertical bars within boxes
represent medians, boxes are interquartile ranges, whiskers are 1.5xinterquartile range
and dots are outlying data ................................................................................................. 81
LIST OF FIGURES (Continued)
Figure 3.6. Number of House Wren young produced per successful summarized by (A)
herbicide treatment and (B) Tmax during the breeding season in the Oregon Coast Range,
2014. (A) Boxplots depict the number of young produced by treatment. Bars are medians,
yellow diamonds are means, boxes are interquartile ranges, and whiskers are
1.5xinterquartile range. Numbers above whiskers indicate the number of successful nests
per treatment. (B) Relationship between Tmax and the number of fledglings produced per
successful nest. The blue line is a fitted linear regression line with 95% confidence
intervals ............................................................................................................................. 83
Figure 3.7. House Wren day 8/9 body condition summarized by (A) herbicide treatment
and (B) Tmax during the breeding season in the Oregon Coast Range, 2014. (A) Boxplots
depict nestling body mass by treatment. Bars are medians, yellow diamonds are means,
boxes are interquartile ranges, and whiskers are 1.5xinterquartile range. Numbers above
whiskers indicate the number nests in which nestlings were measured per treatment. (B)
Relationship between Tmax and the number of fledglings produced per successful nest.
The blue line is a fitted linear regression line with 95% confidence intervals ................. 85
LIST OF TABLES
Table
Page
Table 2.1 Means (± 1 SD) for vegetation and stand-location attributes for herbicide
treatments in the Oregon Coast Range, 2014. Vegetation was measured June-early
August 2014 during the height of the growing season. Vegetation values are percent (%)
cover and were averaged over three sample points, then averaged over treatment units.
Elevation and Aspect measured at each temperature sampling point and then averaged
over treatment units. All values are stand-level means..................................................... 34
Table 3.1. A priori candidate models describing herbicide treatment and air temperature
effects on House Wren nest survival, the number of young produced, per nest, and
nestling body condition in the Oregon Coast Range, US, 2014 ....................................... 70
Table 3.2. Means (± 1 SD) for vegetation and stand location attributes for stands
subjected to different herbicide treatments in the Oregon Coast Range, 2014. Vegetation
(percent cover) was measured June-August 2014 during the height of the growing season,
and was averaged over three sample points at each nest box. Measurements from each
box were then averaged then averaged over all herbicide treatments units. Temperature
(Tmax), elevation and aspect were measured at each nest box and then an averaged
average was calculated for all boxes within stands subjected to each herbicide treatment
over treatment units. All values are stand-level means..................................................... 72
Table 3.3. Model selection results from a priori candidate models describing the effects
of herbicide treatment and Tmax on House Wren nest survival, the number of young
produced per nest, and nestling condition in the Oregon Coast Range, 2014. Models are
ranked in ascending order by Akaike’s Information Criterion adjusted for small sample
sizes (AICc). The number of parameters (k), difference between the best model and all
other models (ΔAICc), the relative likelihood of a model, AICc weights (wi), and
evidence ratio (ER) are given for each model .................................................................. 73
Table 3.4. Results of models testing for effects of herbicide treatment and Tmax on House
Wren nest survival, the number of young produced per nest, and nestling body condition
in the Oregon Coast Range, 2014. Model coefficients (β) and 95% confidence intervals
are given for each model. For nest survival models, odds ratios are also given. All model
coefficients are model-averaged estimates ....................................................................... 74
Combined Effects of Intensive Forest Management and Microclimate on Reproduction in
a Cavity-Nesting Songbird
1
CHAPTER 1: GENERAL INTRODUCTION
Plant and animal populations are being markedly altered by climate change
effects (Root et al., 2003; Parmesan, 2006; Cahill et al., 2012). Increasing temperatures
and weather variability associated with climate change and can have direct and indirect
effects on organisms. Indirect effects include earlier timing of spring events and poleward
shifts in species distributions (Parmesan, 2006). Direct effects of climate change on
organisms can result from precipitation and temperatures that exceed physiological
tolerances (e.g., hyperthermia and water stress), potentially altering metabolic rates,
decreasing physiological performance, and resulting in increased mortality rates
(Helmuth et al., 2005; Buckley et al., 2012; Cahill et al., 2012). By the end of the 21st
century, climate change is expected to increase global surface temperatures by more than
1.5-2 °C (IPCC, 2013, see IPCC, 2013 for full range of emissions scenarios); therefore,
predicting how wildlife populations will respond to ongoing climatic shifts is critical to
guiding successful conservation decisions (Dawson et al., 2011).
Predictions of animal population responses to future climate change are often
based on incomplete information (Sears et al., 2011; Potter et al., 2013). Some of the
most widely used measures of climate change effects on populations are shifts in species
ranges and phenology; however, these factors do not address the physiological and
demographic mechanisms that directly determine the likelihood of species’ persistence on
the landscape (Pearson and Dawson, 2003; Cooke et al., 2013). Evaluating a species’
conservation status requires information not only on species’ response to environmental
pressures along spatial and temporal continuums, but also, and more importantly, on the
2
actual and potential capacity of individuals to respond to environmental pressures and
how those responses impact population demographic rates (e.g., birth rates) (Vié et al.,
2009; Bellard et al., 2012). Moreover, predictions of climate change impacts on animal
populations rely largely on large-scale climate envelope models that do not account for
the local climatic conditions to which organisms most closely respond—microclimates
(Potter et al., 2013)—which can profoundly differ from surrounding macroclimates
(Geiger et al. 2003). Failure to consider microclimates can lead to misleading predictions
of how animal populations will respond to climate change (Gillingham et al., 2012;
Varner and Dearing, 2014) by over- or under-estimating the availability of thermally
suitable habitat across landscapes (Ashcroft, 2010; Suggitt et al., 2011).
Microclimates are driven by several local terrain elements, including elevation,
aspect, and vegetation cover (Dobrowski, 2011). Differences in the abundance and
structure of vegetation cover produce different biophysical effects (e.g., albedo and
surface roughness) on climate, changing the relative amount of incident solar radiation
that is used by the land surface to heat the immediate environment (sensible heat) and
dissipated to broader atmospheric scales (latent heat) (Bonan 2008a). Microclimate can
also be influenced by vegetative composition (von Arx et al., 2012; Zhao and Jackson,
2014), which can alter biophysical characteristics such as canopy conductance (the ease
with which plants can transpire water) and evaporative fractions (the fraction of available
radiation used to evaporate water) to enhance or diminish heat dissipation. Therefore, it is
likely that climate change effects on animal populations will be mediated by the
microclimatic effects of local habitat characteristics.
3
Land cover change—the conversion of natural vegetation to alternate states—
resulting from anthropogenic land use can significantly alter the microclimates to which
organisms are exposed (Saunders et al., 1991; Pielke et al., 2011). It is also currently the
primary driver of global declines in animal populations through habitat loss and
degradation (Pereira et al., 2012; Monastersky, 2014). By decreasing habitat amount and
quality (Fischer and Lindenmayer, 2007), land cover change can indirectly affect animal
populations through many pathways, including altered food availability, predation, and
temperature extremes (Saunders et al., 1991; Bennett and Saunders, 2010). Consequently,
it is broadly expected that changes in land cover and climate change will combine in their
effects on animal populations (Brook et al., 2008). However, like climate change, studies
of land cover change effects on animal populations have largely focused on broad-scale
measures of species change such as richness and abundance (McGarigal and Cushman,
2002), that do not directly influence the demographic rates governing species’ survival.
As anthropogenic land use increases in extent (Lambin and Meyfroidt, 2011) and
intensity (Tilman et al., 2011; Tscharntke et al., 2012) and the effects of climate change
continue to intensify (IPCC, 2013), understanding the combined effects of land cover and
climate change on demographic rates will be essential to conserving wildlife populations
under global change.
Forests are particularly subject to anthropogenic land cover change (Lambin and
Meyfroidt, 2011; Hooke et al., 2012). Although forest cover loss has slowed globally and
forest cover is even increasing in some regions (Hansen et al. 2013, Sloan et al. 2015),
much of the world’s remaining forests have been degraded (Foley et al., 2005; Potapov et
4
al., 2008). Forest degradation is typically thought of in the context of old-growth forest
types; however, early-seral forest degradation is receiving increasing attention (Swanson
et al., 2010, DellaSala et al., 2014; King and Schlossberg, 2014). Compositionally and
structurally diverse early-seral forest is declining in some locations in the northern
hemisphere (Angelstam, 1998; Thomas et al., 2006), often below its historic range of
variability (Spies and Johnson, 2007). This trend is of conservation concern because
early-seral forest is associated with high species diversity and food-web complexity
(Hagar, 2007; Swanson et al., 2010). Furthermore, the population trends of several
vertebrate species have been linked with the availability of structurally and
compositionally diverse early-seral forest (Hunt, 1998; Litvaitis, 1993).
Forest vegetation management alters the abundance, structure, and composition of
early-seral forest vegetation to favor commercially valuable trees (Balandier et al., 2006;
Fortier and Messier, 2006; Wagner et al., 2006). Intensive forest management is a widely
applied forest management system (Fox, 2000; Adams et al., 2005) that often uses
herbicides to suppress the growth of early-seral herbaceous and broadleaf vegetation that
would otherwise compete with trees for resources (e.g., soil nutrients, water, and
sunlight) (Shepard et al., 2004; Wagner et al., 2006). Herbicides can significantly alter
abundance, structure, and composition of early-seral forest stands (Balandier et al., 2006),
potentially resulting in ecological degradation and wildlife population declines (Flueck
and Smith-Flueck, 2006).
Rapid declines in several Pacific Northwest songbird species are attributed to the
increasing scarcity of structurally and compositionally diverse early-seral forest. Early-
5
seral vegetation shifts driven by herbicide use have been demonstrated to decrease
abundance (Betts et al., 2013) and reproductive rates of several early-seral songbirds
(Easton and Martin, 1998; Easton et al., 2002). Herbicides used in intensive forest
management are generally assumed to have no direct effects on songbirds (Tatum, 2004;
Flueck and Smith-Flueck, 2006). However, indirect effects include vegetation-mediated
decreases in food availability quality (Taylor et al., 2006; Hagar 2007) and increases in
predation rates (Easton and Martin, 1998). Furthermore, as land cover can function to
decouple local climatic conditions from large-scale climatic shifts (Dobrowski, 2011;
Varner and Dearing, 2014), herbicide-driven shifts in early-seral vegetation may provide
a complex feedback mechanism that influences how early-seral organisms are impacted
by climate change.
The goal of my research was to examine the effects of forest cover change on
abiotic and biotic processes along a gradient of management intensity. First, I evaluated
the effects of herbicide treatment on microclimatic air temperature along an experimental
gradient in treatment intensity in intensively managed forests in the northern Coast Range
of Oregon, USA (Chapter 2). Second, I also examined the combined effects of herbicide
intensity and air temperature on songbird reproductive output (Chapter 3). I asked the
following research questions:
1. How does air temperature vary along a gradient in herbicide application intensity
in intensively managed forests? (Chapter 2)
I collected microclimatic air temperature data from 20 intensively managed forest
stands throughout the northern Oregon Coast Range. Intensive forest management is a
6
commonly used forest management system in the Pacific Northwest (Adams et al., 2005;
Murphy et al., 2005) and throughout the world (Fox, 2000; Paquette and Messier, 2009),
thus my results may be widely applicable to timber operations globally. My results begin
to fill important knowledge gaps concerning ecological impacts of forest management
practices on a declining habitat type: early-seral forest. This study is a part of one of the
largest experimental studies on intensive forest management practices globally.
Moreover, this is the first study to examine the effects of post-harvest forest vegetation
management on microclimatic air temperature within a gradient of herbicide treatment
intensity. My results suggest that intensive management of early-seral forests exerts
limited influence on the microclimatic air temperatures.
2. How does intensity of herbicide application interact with air temperature to affect
reproductive success in a cavity-nesting songbird? (Chapter 3)
I tested how the reproductive output of a locally abundant cavity-nesting
songbird, the House Wren (Troglodytes aedon), was influenced by the gradient of
herbicide treatment and microclimatic air temperature examined in Chapter 2. I collected
data on nest survival, the number of fledglings produced per nest, and nestling body
condition from 24 intensively managed forest stands throughout the northern Oregon
Coast Range. My results provide insights into current patterns of the relative influence of
global change drivers on animal population demographic rates. Moreover, this is the first
study to examine the combined effects of a gradient in land cover change intensity and
microclimatic air temperature on avian reproduction. My results suggest that
microclimatic air temperature may influence reproductive success for early-seral
7
songbirds, but land cover change will likely continue to be the primary driver of changes
in demographic rates in the near future.
Finally, Chapter 4 summarizes the outcomes of the work described above.
8
INTENSIVE FOREST MANAGEMENT EXERTS WEAK AND VARIABLE
INFLUENCES ON MICROCLIMATE IN EARLY-SERAL FOREST
Kristin N. Jones, James W. Rivers, Matthew G. Betts
9
CHAPTER 2: INTENSIVE FOREST MANAGEMENT EXERTS WEAK AND
VARIABLE INFLUENCES ON MICROCLIMATE IN EARLY-SERAL FOREST
2.1 ABSTRACT
Climate change is expected to have far reaching impacts on global biodiversity.
Negative effects of rising temperatures at large spatial scales have been widely predicted
in plants and animals, but the spatial scale of such predictions is often broad and poorly
matched to the fine scales at which organisms experience the environment. Such
microclimate thermal regimes are often moderated by forest cover; thus, intensive forest
management practices have the potential to either ameliorate or exacerbate climate
change effects on biota. In this study, we examined how air temperature varied across a
gradient of post-harvest vegetation management (via herbicide application) in intensively
managed early-seral forests in the northern Oregon Coast Range. We evaluated standlevel air temperatures in regenerating forest stands subject to no-spray (control), light,
moderate, and intensive herbicide treatments. We examined whether mean daily
temperatures (minimum, mean, and maximum) and their associated coefficients of
variation varied across herbicide treatments. We found that herbicide treatments
significantly influenced mean and maximum air temperatures, but we did not detect an
effect of herbicide treatment on minimum temperature. However, the direction (positive
or negative) and magnitude of these effects were variable and small (< 0.5°C).
Furthermore, we did not detect any differences in mean air temperatures in corrected
pair-wise comparisons across individual herbicide treatment intensities. Additionally,
10
herbicide treatment did not exert a significant effect on within-treatment temperature
variability. Our results suggest that post-harvest vegetation management has limited
impacts on fine-scale air temperatures and is unlikely to either amplify or buffer the
projected effects of climate change in early-seral forests.
2.2 INTRODUCTION
Land cover change resulting from anthropogenic land use is currently the primary
driver of global biodiversity declines (Pereira et al., 2012; Monastersky, 2014). Humans
are estimated to have modified up to 58% of all ice-free land area through land uses such
as urbanization and agriculture (Hooke et al., 2012), and have profoundly altered critical
ecosystem functions such as climate regulation and wildlife habitat (Foley et al., 2005).
At the same time, climate change is modifying ecosystems (Dawson et al., 2011).
Climate change can affect organisms indirectly through shifts in species distributions and
phenology (Parmesan, 2006), and directly through temperatures that exceed physiological
tolerances and decrease physiological performance and fitness (Bellard et al., 2012;
Buckley et al., 2012). Consequently, it is expected that changes in land cover and climate
will combine in their effects on global biodiversity (Brook et al., 2008).
Forests, some of the most diverse ecosystems in the world (Aerts and Honnay,
2011), have been particularly subject to anthropogenic land cover change (Rudel et al.,
2005; Hooke et al., 2012). Through a combination of increasing afforestation and
decreasing deforestation, global forest area is currently increasing (Sloan and Sayer,
11
2015). However, planted forest area is increasing while natural forest area is decreasing
(FAO 2015). Planted forests may represent degraded forest conditions compared to
naturally regenerated forests through an altered ability to support diverse ecological
communities and recover from disturbance (Carnus et al., 2006; Potapov et al., 2008).
Forest degradation is often considered in the context of old-growth forest (Foster et al.,
1996; Spies, 2004; Thomas et al., 2006). However, increasing attention is being paid to
the degradation of early-seral forests (Swanson et al., 2010; DellaSala et al., 2014; King
and Schlossberg, 2014). Compositionally and structurally diverse early-seral forest is
declining in some locations in the northern hemisphere (Angelstam, 1998; Thomas et al.,
2006), often below its historic range of variability (Spies and Johnson, 2007). This trend
is concerning because early-seral forest is associated with high food web complexity and
species diversity (Hagar, 2007; Swanson et al., 2010). Furthermore, the population trends
of several vertebrates have been linked to decreases in the availability of structurally and
compositionally diverse early-seral forest (Litvaitis, 1993; Hunt, 1998).
Forest vegetation management practices can decrease the quality of early-seral
forests as habitat for wildlife (Swanson et al., 2010; Betts et al., 2013). Intensive forest
management, a widely used forest management system (Fox, 2000), often uses herbicides
to suppress the growth of early-seral herbaceous and broadleaved vegetation that compete
with trees for resources (Fortier and Messier, 2006), leading to changes in abundance,
diversity, and composition of early-seral forest (Balandier et al., 2006). The main impact
of herbicides is altering the relative dominance of plant species within the community
(i.e., relative increases in conifer cover and decreases in broadleaved cover) (Balandier et
12
al., 2006), which can result in negative consequences for early-seral organisms (Easton
and Martin 1998; Betts et al., 2013).
Microclimatic shifts are one mechanism by which herbicide-driven changes in the
relative abundance of plant species in early-seral forests may influence habitat quality.
One microclimatic variable important to habitat quality is air temperature (Huey, 1991).
Surface air temperatures and their variability drive many ecological processes, including
nutrient and water cycles and species distributions (Bonan 2008a). Forest cover primarily
influences local air temperature through altering biophysical characteristics such as
surface roughness (vertical irregularities in the canopy) and albedo (the fraction of solar
radiation reflected by a land surface), which strongly influence the relative amount of
incident solar radiation that is used by the land surface to heat the immediate environment
(sensible heat) and dissipated to broader atmospheric scales through evapotranspiration
(latent heat) (Bonan, 2008b; Jackson et al., 2008; Anderson et al., 2010). Previous work
indicates that the relative abundance of plant species can also exert significant effects on
forest microclimatic conditions (von Arx et al., 2012; Zhao and Jackson, 2014). For
example, deciduous tree species have a summer albedo up to 0.1 higher than coniferous
forests and canopy conductance (the ease with which plants can transpire water) and
evaporative fractions (the fraction of available solar radiation used to evaporate water)
approximately twice that of coniferous forests (Anderson et al., 2010). This additional
transpiration and reflectivity of broadleaved vegetation tends to cool air temperatures to a
greater degree than forests dominated by coniferous species. For example, von Arx et al.
(2012) showed that broadleaved and non-conifer pine forests reduced daily maximum
13
temperatures twice as much as coniferous forests. This greater cooling effect of
broadleaved vegetation suggests that forests with greater amounts of broadleaved
vegetation cover may provide superior “micro-refugia” for species under advancing
global climate change (Ashcroft, 2010; Dobrowski, 2011).
To date, relatively few studies (Proe et al., 2001; Devine and Harrington, 2007;
Parker et al., 2012) have examined the effects of herbicides on air temperature.
Additionally, all of these studies have emphasized the average conditions of herbicide
treatments, not the variation within them. Air temperature variation within treatments
may allow temperature-sensitive species to persist in areas of the stand with suitable
microclimates (Sears et al., 2011); conversely, temperature variation may impact earlyseral species that prefer more stable microclimatic conditions (e.g., Checa et al., 2014).
Moreover, no studies have examined air temperature along a gradient of herbicide
treatment intensity to evaluate the extent to which air temperature is moderated by
variation in the relative abundance of broadleaved and coniferous vegetation in earlyseral forest. Furthermore, no studies have examined the effects of forest vegetation
management on air temperature in the framework of intensive forest management, a
management system that is projected to be applied to an increasing proportion of forested
landscapes in coming years to meet the growing demand for wood products (Fox, 2000).
By evaluating the effects of herbicide treatment intensity on air temperature, we can
identify how forest vegetation management practices can be varied to produce thermal
refuges from climate change for species associated with this forest seral stage.
14
In this study, we examined patterns of microclimatic air temperature variation
across a gradient of post-harvest vegetation control in early-seral coniferous forests in the
central Oregon Coast Range. By examining gradients in herbicide treatment intensity, we
allowed for the potential detection of degrees of forest vegetation management that might
minimize trade-offs between increases in the growth of commercially valuable conifers
(via broadleaved vegetation suppression) and climate regulation in intensively managed
forests. The herbicide treatments are part of a broader study that experimentally evaluates
the effects of a gradient of herbicide treatment on early-seral forest biodiversity (Betts et
al., 2013). We evaluated stand-level variation in microclimatic air temperature among
and within herbicide treatments by addressing the following questions: (1) how do standscale patterns of surface air temperature vary with herbicide intensity? and (2) how does
herbicide intensity affect the variability of surface air temperatures in managed forests?
We predicted that herbicide treatments would increase surface air temperature extremes
by shifting stand composition towards predominantly coniferous species, and that these
increases would intensify with increasing herbicide treatment intensity. We also predicted
that herbicide-treated stands would exhibit greater air temperature variability relative to
untreated stands (Baker et al., 2014; Hardwick et al., 2015) by increasing the variability
of day-time high and night-time low temperatures (i.e., the moderating effect of forest
vegetation cover would decrease with less total cover in herbicide-treated stands). We
further predicted that intermediate levels of herbicide treatment would exhibit increased
spatial variability compared to intensive and control treatments because of (1) irregular
shading by residual broadleaved vegetation, especially deciduous hardwood trees that
15
were not subject to herbicides in the light treatment and/or (2) patchy distribution of
herbaceous species (as opposed to bare ground) in the light and moderate treatments,
which can also exert moderating influences on air temperature (Geiger et al. 2003).
2.3 MATERIALS AND METHODS
2.3.1 Study area
We conducted this study at five of the eight experimental blocks that comprise the
broader study (see Figure 2.1) in the northern Oregon Coast Range—Black Rock, Grande
Ronde, Luckiamute, Trask, and Willamina. The maritime climate of the Oregon Coast
Range is characterized by cool, wet winters and mild, dry summers (Franklin and
Dyrness 1988). Mean annual precipitation ranges from 165 to 330 cm with most
precipitation occurring October through March. Mean annual air temperature ranges from
5-19°C (Franklin and Dyrness 1988, Taylor and Hannan 1999) Soils are dominantly
moderately deep to deep well-drained silt loam soils, and are derived from basalts,
sandstone, and siltstones. Topography is characterized by somewhat low, highly
dissected mountains with slopes ranging from 0 to 90 percent (Knezevich 1982, Taylor
and Hannan 1999). All sampled stands are in the western hemlock (Tsuga heterophylla)
zone (Franklin and Dyrness 1973), ranging in elevation from 165 to 765 m. Douglas fir
(Pseudotsuga menziesii) dominates early-seral forest plantations in this region, with
grand fir (Abies grandis), western hemlock (Tsuga heterophylla), and western redcedar
(Thuja plicata) forming minor components. Dominant shrub/woody species include big-
16
leaf maple (Acer macrophyllum), California hazelnut (Corylus cornuta californica),
cascara (Rhamnus purshiana), common snowberry (Symphoricarpos albus), oceanspray
(Holodiscus discolor), red alder (Alnus rubra), and vine maple (Acer circinatum).
Smaller understory broadleaf species include Oregon grape (Mahonia nervosa) salal
(Gaultheria shallon), and Vaccinium spp. The herbaceous community is comprised of
many native and non-native herbaceous plants, with swordfern (Polystichum munitum)
and brackenfern (Pteridium aquilinum) often dominating.
2.3.2 Experimental design and treatments
Our study makes use of a subset of sites that occur within a larger study area that
uses a randomized complete block design in which 32 stands occur in 8 separate blocks,
and each of 4 levels of herbicide treatment were randomly applied to one stand within
each block (Figure 2.1). All blocks are located across a 100 km (N-S) section of the
northern Oregon Coast Range region (Betts et al., 2013). Stands within each block were
sited > 1 km of each other to ensure spatial independence of treatments and no more than
5 km of each other to reduce within-block variation (i.e., slope, elevation, and pre-cut
vegetation composition). From this larger study framework, we constrained our study on
20 forest stands located within 5 of the blocks, with each stand 10.4 - 18.9 ha because of
logistical considerations that prevented us from working on all stands concurrently.
All stands were clear-cut in fall 2009 and replanted in spring 2011 with Douglasfir, the most commonly commercially cultivated species throughout the region, at
approximately 980 trees per hectare. A full suite of herbicides and surfactants typically
17
used in commercial timber harvesting operations was applied to stands between 2009 and
2014 in a manner that created a gradient in management intensity; see Betts et al. (2013)
and Appendix A for a full description of the herbicides used, the timing with which they
were applied, and their concentration. All herbicide spraying occurred in a timeframe that
mimicked the typical timeframe in which vegetation control takes place in commercial
operations
Temperatures were measured around bird nestboxes to assess the effects of
temperature on songbird reproduction in another study (Jones et al., in prep). Following
harvest treatments, 8 cedar nest boxes were mounted on steel garden stakes 1 - 1.5 m
from the ground on each stand (n = 160). Nest boxes were sited with several
considerations in mind; (1) equal distances between boxes (> 50 m separation), (2) even
stand coverage, and (3) sampling logistics (i.e. walking distance between boxes < 30
min). As distance to stand edge was not a factor in nest box siting, but has been shown to
influence air temperature (Baker et al., 2014), we included it as an analysis covariate (see
Statistical Analysis below). Vegetation was not a factor in box siting.
2.3.3 Sampling design
We recorded air temperature during the local songbird breeding season (May August 2014) as part of a related study (Jones et al., in prep), 3 years after trees were
planted. We measured air temperature at nestboxes on each stand to link temperature
effects to the reproductive output of cavity-nesting songbirds because a goal of that study
18
was to evaluate the effects of intensive forest management on air temperatures
experienced directly by organisms (microclimatic air temperatures).
2.3.3.1 Air temperature
To measure ambient temperature, we placed iButton temperature loggers (n =
160) on the underside of nestboxes erected on study sites at a density of 8 boxes/stand.
Each iButton was programmed to record temperature every 15 min and was affixed to the
underside surface of each nestbox with a zip tie so that it hung freely 5 cm below the box.
We covered iButtons with a section of white 10 cm diameter PVC tube with ventilation
holes so as to prevent direct exposure of the iButton to solar radiation and moisture, allow
for airflow, and minimize heat accumulation. Due to logistical constraints, we used two
iButton models that varied slightly in their accuracy (± 1.0°C accuracy, n = 71 boxes: ±
0.5°C accuracy, n = 89 boxes). The two models were distributed irregularly among
stands; thus, we note that estimated differences in air temperature between stands are
within approximately 1.0°C of their true value. All iButtons were validated against an
independent digital thermometer (Omega HH609R, Omega Engineering, Stamford,
Conneticut, USA) prior to placement on stands; any iButtons that deviated by ≥ 0.5°C
during his validation procedure were not used.
2.3.3.2 Vegetation abundance and composition
We measured vegetation cover in each treatment to quantify changes in
vegetation abundance and composition among the different herbicide treatments. At each
19
temperature sampling point, we measured vegetation at three distinct sampling points (3
m radius each): one was centered directly on the temperature sampling point, and the
other two were located 25 m distant from the temperature sampling point. The azimuth of
the initial off-nestbox point was chosen randomly, with the second point located 180°
away. At each vegetation sampling point, we visually estimated percent cover to genus
(in the case of Rubus, Ribes, and forbs) or species in each of three distinct vegetative
strata; herbaceous (0 - 0.5 m), shrub (0.5 - 2.0 m), and canopy (> 2.0 m) layers. We
aggregated some closely-related species that are functionally similar (i.e., Rubus and
Ribes), as well as all herbaceous forbs, grasses, and ferns. For analysis, we summed the
amount of cover for all plant species and genera in each of the three vegetation sampling
points, then calculated an cover value, grouping plant types together to estimate the total
amount of (1) broadleaved cover (as defined by Ellis and Betts, 2011), (2) conifer cover,
and (3) vegetative cover. At each sampling point, we also measured aspect, which we
transformed to “southwestness” (i.e., cos(aspect degrees - 225); Franklin et al., 2000),
elevation, and distance to edge, as these variables have been shown to strongly influence
microclimate (Dobrowski, 2011). Aspect was transformed because it is not an
incremental metric (e.g., 0° is more similar to 350° [both north-facing] than 90° [eastfacing]). Conversely, southwestness is an index ranging from 1 (signifying southwest) to
-1 (signifying northeast) that can indicate the environmental aridity with more arid
environments producing greater values (Huang et al., 2012).
20
2.3.4 Data filtering and summary
Temperature data were initially inspected and assessed for errors, indicated by
extreme atypical values (> 50°C or < -10°C) caused by instrument malfunction or when
logging stations were found to be disturbed (i.e., nestbox damaged by wildlife). For
example, data were considered erroneous when temperatures logged by the same iButton
within an hour were greater or less than 10°C of the record in question or temperatures
recorded on the same stand at other nestboxes were 20°C greater or less than the recorded
temperature in question. This led us to remove ≤ 5% of temperature values.
We used iButtons to identify the minimum (Tmin), mean (Tmean), and maximum
(Tmax) daily temperature at each nest box for each day during a 24-hour period starting
at midnight. For each of these three measures, we then computed means at each nestbox
during each 24-hour period. We used these data to calculate a stand-level mean and
coefficient of variation (CV; Tmincv, Tmeancv, Tmaxcv) for all 6 air temperature response
variables.
2.3.5 Statistical analysis
All models were fit using the lme function of the nlme package (Pinheiro et al.
2015) in ‘R’ v3.2.0 (R Development CoreTeam 2015). First, we used mixed model
analysis to test whether forest vegetation abundance and composition differed
significantly among herbicide treatments. We constructed models for three response
variables of forest vegetation (broadleaved, conifer, and total vegetation cover) that
included fixed effects for treatment (4 levels: control, light, moderate, intensive) and one
21
random effect (block) to reflect our randomized complete block design. All vegetation
responses were log transformed prior to analysis to correct for heterogeneity of variance
among herbicide treatments. We back-transformed values so that values can be
interpreted as the mean multiplicative increase or decrease in cover between treatments.
Therefore, a treatment contrast of 1 indicates that cover is equal between treatments.
Second, we used a randomized block ANCOVA with one random effect (block)
to test whether mean daily air temperature and within-treatment variability differed
significantly among herbicide treatments. We constructed models for 6 air temperature
response variables (Tmin, Tmean, Tmax, and their associated CVs) that included fixed
effects for treatment. Covariates included were stand-level averages for elevation,
southwestness, and the distance to stand edge, taken over all nestboxes in a stand. We
included elevation as a covariate because stands within a block could vary > 100 m due to
the highly heterogeneous terrain of the Oregon Coast Range. Southwestness and distance
to stand edge were included as covariates to account for any systematic variation in
temperatures due to nestbox siting as, along with elevation, both factors can influence
surface air temperatures (Geiger et al. 2003). No covariates were correlated (r < 0.25 in
all cases). Diagnostic tests revealed minor violations of assumptions of normality and
homogeneous variance among treatments, but these were not enough to warrant data
transformation (Zurr et al. 2009, Ramsey and Schafer 2013).
We also conducted 5 pairwise comparisons using Bonferroni-corrected 95%
confidence intervals and p-values in the ‘estimable’ function of the gmodels package in R
(Warnes 2013). We compared each of the 3 treatments where herbicide was applied to
22
the control (3 comparisons) to evaluate how increasing management intensity influenced
air temperature. In addition, we compared the moderate herbicide treatment to the light
and intensive treatments (2 comparisons) because the moderate treatment this treatment
that is closest to current operational practices (Betts et al. 2013). Thus, these comparisons
(5 total) allowed us to compare how alternative herbicide application may lead to changes
in air temperature. We report means and associated confidence intervals; effects were
considered significant at P < 0.05 for all tests.
2.4 RESULTS
2.4.1 Vegetation
Herbicide treatments had a strong influence on vegetation composition (Figure
2.2) with broadleaved vegetation cover generally decreasing with increasing herbicide
treatment intensity (F3, 11 = 18.68, P < 0.001; Figure 2.3a) while conifer cover increased
(F 3, 11 = 9.19, P = 0.003; Figure 2.3b). However, after correcting for multiple
comparisons, we only detected significant differences in broadleaved cover between the
control and the moderate (t 11 = 4.88, P < 0.001, 𝛽̂ = 1.88 [0.68, 3.08]) and control and
intensive treatments (t 11 = 5.04, P < 0.001 , 𝛽̂ = 1.93 [0.74, 3.12]) and the light and the
moderate treatment (t 11 = 4.89, P < 0.001, 𝛽̂ = 1.87 [0.68, 3.05]) (Figure 2.3a).
Furthermore, we only detected a difference in conifer cover between the control and
intensive treatments (t 11 = -4.69, P < 0.001, 𝛽̂ = -0.95 [-1.58, -0.32]). We did not detect
any statistically significant differences in total vegetation cover among all herbicide
23
treatments (F3, 11 = 2.070 P = 0.162; Figure 2.3c). Lastly, vegetation cover within
vegetative strata (e.g., canopy cover) was similar among herbicide treatments (Table 2.1).
2.4.2 Mean air temperature
Air temperature varied little among blocks (Figure 2.4), and was statistically
indistinguishable among treatments for Tmin, Tmean, and Tmax (Figure 2.5a). Herbicide
treatments exerted a significant effect on most temperature variables, but the direction of
this effect was inconsistent. We did not detect an effect of treatment on Tmin (F3, 9 =
1.039, P = 0.421) but we did find that treatment influenced on Tmean (F3, 9 = 7.655, P =
0.008). However, this effect was driven by the significance of the intercept (control
treatment) in Tmean models (𝛽̂ = 18.6, SE = 1.0, t9 = 18.6, P < 0.001). Pairwise
comparisons revealed no statistically significant differences in Tmean among herbicide
treatments (Figure 2.5a). All estimated differences in mean Tmean were ≤ 1.1°C, with the
majority < 0.5°C (Figure 2.5a), and the direction of these differences (positive or
negative) was inconsistent (Figure 2.5a). We also detected an effect of treatment on
Tmax (F3, 9 = 5.34, P = 0.022). However, this effect was driven by statistically
significant Tmax differences between the control and the light treatment (𝛽̂ = 1.1, SE =
0.4, t 9 = 2.67, P = 0.026) and the control and the moderate treatment (𝛽̂ = 1.0, SE = 0.4, t
9
= 2.34, P = 0.044). Mean Tmax monotonically increased with increasing treatment
intensity (Figure 2.5a).
Elevation was a significant predictor of Tmean (F1, 9 = 30.97, P < 0.001); Tmean
decreased with increasing elevation (𝛽̂ = -0.3°C/100 m, SE = 0.06, t 9 = -4.819 , P <
24
0.001). Elevation was also a significant predictor of Tmax (F1, 9 = 114.68, P < 0.001);
Tmax decreased with increasing elevation (𝛽̂ = -0.7°C/100 m, SE = 0.08 , t 9 = -8.523 , P
= 0.003). In addition, distance to stand edge was a significant predictor of Tmax (F1, 9 =
6.95, P = 0.027); for every 10 m from the stand edge, Tmax increased by 0.4 (± 0.15 SE)
°C.
2.4.3 Variability in mean air temperature
Temperature variability did not statistically differ among herbicide treatments
(Figure 2.5b). We did not detect a significant difference in Tmincv among treatments (all
CIs overlapped zero). Nevertheless, compared to the control, mean Tmincv was lower in
the moderate and intensive treatments, but more variable in the light treatment. Among
treatments, mean Tmincv was greater in the light and moderate treatments compared to
the intensive treatment (Figure 2.5b). We also did not detect a significant difference in
Tmeancv among treatments (all CIs overlapped zero; Figure 2.5b). Nevertheless, mean
Tmeancv was lower in all herbicide-treated stands compared to the control treatment
(Figure 2.5b). Among herbicide treatments, mean Tmeancv was greater in the light
compared to the moderate treatment (𝛽̂ = 0.06% [-1.87, 2.0]); conversely, mean Tmeancv
was greater in the intensive treatment compared to the moderate treatment (𝛽̂ = -0.09% [2.09, 1.91]). Finally, we did not detect a significant difference in Tmaxcv among
treatments (all CIs overlapped zero; Figure 2.5b). Nevertheless, mean Tmaxcv was lower
in all herbicide-treated stands compared to the control (Figure 2.5b). Among herbicide
treatments, mean Tmaxcv was greater in the light treatment compared to the moderate
25
treatment and in the moderate treatment compared to the intensive treatment (Figure
2.5b). Estimated differences among treatments for mean Tmincv, Tmeancv, and Tmaxcv
were ≤ 1.5%, with the majority < 0.5% (Figure 2.5b).
2.5 DISCUSSION
Our results suggest that herbicide treatment has little to no influence on
microclimatic air temperatures in early-seral forests. We found that herbicide treatment
exerted a statistically significant, though minor, effect on Tmean and Tmax, but not on
Tmin. However, corrected pairwise comparisons indicated no significant differences in
mean air temperatures among herbicide treatment intensities. Furthermore, air
temperature variability (Tmincv, Tmeancv, and Tmaxcv) was not significantly influenced
by herbicide treatment. Given the significant differences in broadleaved cover among
treatments, that we detected no more than limited differences among treatments is
surprising and runs counter to our initial predictions. Previous studies have found
broadleaved vegetation to significantly cool air temperature relative to coniferous
vegetation (von Arx et al., 2012; Zhao and Jackson, 2014) via increased evaporative
fractions and higher albedo. Thus, we expected decreases in broadleaved vegetation from
herbicides to result in increased air temperatures. Indeed, our results suggest that Tmax
and Tmean were influenced to a degree by herbicide treatment; however, temperature
differences among treatments were small, highly variable, and exhibited no uniform
direction (positive or negative). For example, Tmax was greater in all herbicide
26
treatments compared to the control, but also was greater in the light and lower in the
intensive treatment compared to the moderate treatment. Therefore, herbicide treatment
seems to exert minimal influence on stand-level microclimatic air temperatures in earlyseral intensively managed forests in the Oregon Coast Range.
One possible explanation for decreased broadleaved cover not influencing air
temperature is that temperature in early-seral stands may depend more on total vegetation
cover than relative composition. Indeed, the largest moderating influences on air
temperature of forest vegetation are generally associated with changes in measures of
total vegetation abundance such as leaf area index (Aussenac, 2000). The lack of
difference in total cover among herbicide treatments was likely driven by increases in
exotic ruderals, such as species in the genera Cirsium, Epilobium, and Aster (Jones et al.,
unpublished data). This group, which comprised the majority of non-woody vegetation in
the understory of our herbicide-treated stands (Figure 2.2), substantially increased with
increasing herbicide treatment (Table 2.1), though this trend was highly variable.
Furthermore, total vegetation cover in early-seral forest stands is dependent on sitespecific factors, including site topography, hydrology, and seed bank composition
(Lautenschlager and Sullivan, 2002); sites with more favorable growing conditions will
regenerate vegetation more quickly. Thus, total vegetation cover may begin to recover
fairly quickly on highly productive sites such as those in the Oregon Coast Range
(Waring and Franklin, 1979). Indeed, as herbicides only alter the relative abundance of
target vegetation and do not necessarily kill competitive and weedy species (Miller and
Miller, 2004), such plants can rebound within 2-5 years following treatment
27
(Lautenschlager and Sullivan, 2002; Miller and Miller, 2004). Thus, increased growth of
rapidly establishing exotic ruderals on our stands may have played a compensatory role
in thermal regulation on sites with decreased broadleaved cover.
An additional explanation for our inability to detect an effect of decreased
broadleaved cover on air temperature could be that the differences in vegetation structure
and composition we observed among treatments were not large enough to produce a
detectable cooling effect. We only detected significant decreases in broadleaved cover in
3 of 5 comparisons among treatments, and these differences were highly variable such
that the range of possible values included estimates not much different from a 1:1 ratio.
(i.e., a treatment contrast of one). Previous work (von Arx et al., 2012; Zhao and Jackson,
2014) demonstrating variation in local (< 4 ha) air temperature as a function of
broadleaved cover gives little indication of the exact amount of broadleaved cover
required to produce detectable cooling effects. However, both of these studies compared
differences in air temperature between land areas (Zhao and Jackson, 2014) and stands
(von Arx et al., 2012) principally dominated by broadleaved vegetation or land cover
principally dominated by other vegetation types (e.g., coniferous). From this we may be
able to conclude that larger differences in broadleaved cover are required to produce
detectable air temperature differences in early-seral forests.
We did not detect significant effects of herbicide treatment on air temperature
variability (Tmincv, Tmeancv, and Tmaxcv). This may not be surprising as temperature
variability is most strongly associated with vegetation structure rather than composition
(Baker et al., 2014; Hardwick et al., 2015), and we did not detect any differences in the
28
total cover among treatments (Figure 2.3c). However, there were some broad patterns in
air temperature variability among treatments. Contrary to expectations, air temperature
variability generally decreased in herbicide-treated stands compared to control stands.
The greater air temperature variability in control treatments is likely caused by a greater
spatial variability in regenerating vegetation compared to the herbicide treatments
(Heithecker and Halpern, 2006; Ma et al., 2010; Xu et al., 1997). Specifically, a greater
number of broadleaved hardwood trees (e.g., Acer macrophyllum and Alnus Rubra) left
after harvest have grown into the canopy (Figure 2.2) in the control treatments, providing
patchy shading that may increase the spatial variability of air temperatures. However,
consistent with our predictions, air temperatures in the light treatment were generally
more variable than the moderate treatment and air temperatures in the moderate treatment
were generally more variable than the intensive treatment. This pattern is also likely
driven by relatively higher spatial variability of regenerating vegetation in the less
intensive herbicide treatments. Not only does the light treatment have a greater amount of
residual broadleaved hardwood trees to provide patchy shading (Table 2.1), but a more
variable distribution of herbaceous species and bare ground in the understory (Jones et
al., unpublished data). The distribution and amount of understory cover also influences
the surface energy fluxes, where air temperatures are generally more tightly coupled to
atmospheric conditions (and thus more variable) over bare ground compared to ground
covered with vegetation (Geiger et al. 2003). These results suggest that less intensive
herbicide treatments may have a greater capacity to provide microhabitats for
29
temperature-sensitive species to persist in areas of the early-seral forest stands with
suitable microclimates (Sears et al., 2011).
Despite the overall patterns in air temperature variability we observed, none of
these effects were significant. The small magnitude of the estimated differences in
temperature variability among treatments, as well as their wide confidence intervals,
indicates that, although statistical power to detect an effect may have been low (Steidl et
al., 1997), herbicide treatments likely minimally influenced temperature variability. This
may be because of the rapid mixing of air masses that occurs within forest clearings
(Geiger et al. 2003; Bonan 2008a) and the relatively small differences in canopy cover
among herbicide treatments, which is a principal driver of air temperature variability in
forests (Chen et al., 1999). It is also possible that the spatial scale of our sampling
influenced our inability to detect treatment effects on air temperature. The distance
between sample points (> 50 m) may have been too large to capture differences in air
temperature variability driven by canopy cover. It is possible that smaller scales of
sampling (e.g., 1 - 5 m) may allow detection of air temperature variability associated with
finer-scale differences in canopy cover.
Lastly, potential differences in mean air temperatures and their variability among
treatments may also be diminished by the matrix in which they occur (Dobrowski, 2011).
Many factors influence air temperature within forest clearings, such as amount of solar
radiation received, vegetation height, and cold air pooling (a climatic process that can
occur in topographic depressions, resulting in a topographically-confined, stagnant air
layer that is cooler than the air aloft; Geiger et al. 2003, Whiteman et al. 2001). This
30
matrix can reduce the effect of microclimatic differences from changes in surface
roughness and albedo driven by decreased vegetation cover. One critical factor is the size
of the clearing relative to the height of the surrounding trees (Geiger et al. 2003). As
clearing size increases, incident solar radiation (the amount of solar radiation energy
received by a land surface during a given time) increases and temperatures increase.
Simultaneously, however, the clearing becomes less protected from the wind and the
mixing of air masses increases, which carries away excess heat (Geiger et al. 2003,
Bonan 2008a). The increase in solar radiation dominates until the clearing size reaches a
threshold (i.e., a distance: height ratio of 1.8; Geiger et al. 2003); mixing dominates after
this point and temperature excesses of clearings compared to forests decrease. As all
treated stands were generally the same size, there was some size variation (10.4 - 18.9
ha), and the age of the forest adjacent to our stands varied in age. Therefore, it is possible
that the variation in the size of the clearing relative to the height of the surrounding stands
made it impossible for us to detect differences in mean air temperature or air temperature
variability or that the distance:height ratio across all/most stands exceeded the threshold
at which the processes influencing air temperature in forest clearings dominate. As this
variable was beyond the scope of this study, we did not measure it. However, for the
reasons stated above, it may be an important consideration for future studies of air
temperature in intensively managed forests.
Our finding of generally weak and inconsistent effects of herbicide treatment on
air temperature in early-seral forests compliments the few studies to examine the effects
of post-harvest vegetation management on microclimatic air temperatures. In an
31
experimental study examining the effects of vegetation control practices on microclimate
in post-clearcut Sitka spruce (Picea sitchensis) stands in the United Kingdom, Proe et al.
(2001) found that herbicide-driven vegetation changes only influenced near surface air
temperatures when combined with other post-harvest treatments (fertilizer and whole-tree
harvesting). Similarly, in a study that independently varied woody and herbaceous
vegetation control in pine-dominated shelterwood stands in central Ontario, Parker et al.
(2012) found minimal vegetation-mediated influences of herbicide on air temperature.
Woody vegetation control produced the only difference in air temperature among forest
stands but this effect only occurred in 1 out of 4 growing seasons and was relatively small
(0.5°C). Taken together with the results of this study, it appears that post-harvest
vegetation control in intensively managed temperate conifer forests produces negligible
differences in the thermal environment during early-seral stages.
2.6 MANAGEMENT IMPLICATIONS
Intensive forest management practices such as the use of herbicides to control
competing vegetation during stand establishment are widely used to increase forest
production (Wagner et al., 2006). Globally, the areal extent of intensively managed
forests is estimated to be 135 million ha (Fox, 2000), and is projected to increase in the
coming decades to meet growing demands for wood products (Fox, 2000). Although
herbicide application has become a widely adopted post-harvest vegetation management
strategy in silvicultural operations, the ecological costs of herbicide use in forests are not
32
well-understood (Shepard et al., 2004) and can result in decreased habitat quality for
forest organisms (Swanson et al., 2010). One of the ways in which forest cover change
can impact habitat quality is through altered climatic regimes (Lehtinen et al., 2003). Our
results indicate that changes in the relative composition of broadleaved and coniferous
vegetation cover driven by herbicide application do not cause detectable changes in air
temperature in intensively managed early-seral forests. Comparisons among herbicide
treatments indicated no significant differences in Tmin, Tmean, Tmax, nor their
variability. This suggests that management practices in early-seral forest habitats may not
alter air temperatures in a way that strongly impacts early-seral organisms. However, no
data exist on this topic so future studies should focus on testing the link between
intensive management practices, temperature, and measures of organism performance
(e.g., fitness). Conversely, our findings also indicate there is little potential for postharvest vegetation management to ameliorate expected climate change impacts (IPCC,
2013). However, the general pattern for less intensive herbicide treatments (i.e., our light
and moderate herbicide treatments) to exhibit greater air temperature variability than
more intensive herbicide treatments, may indicate that less intensive herbicide
applications have a greater capacity to provide microhabitats that support the persistence
of temperature-sensitive species in early-seral forest stands (Sears et al., 2011). However,
as our study only spans one growing season and early-seral forests are highly dynamic
ecosystems (Swanson et al., 2010), studies spanning multiple years are required to build
upon our results. Nevertheless, our study provides strong evidence that post-harvest
vegetation management has limited impacts on fine-scale air temperatures and is unlikely
33
to either amplify or buffer the projected effects of climate change on biodiversity in
early-seral forests.
34
Table 2.1 Means (± 1 SD) for vegetation and stand-location attributes for herbicide treatments in the Oregon Coast Range, 2014.
Vegetation was measured June-early August 2014 during the height of the growing season. Vegetation values are percent (%)
cover and were averaged over three sample points, then averaged over treatment units. Elevation and Aspect measured at each
temperature sampling point and then averaged over treatment units. All values are stand-level means.
Treatment
Control
Light
Moderate
Intensive
Broadleaved (%)
59.4 (32.6)
59 (32.4)
13.5 (16.4)
9 (7.1)
Conifer (%)
5.6 (4.5)
7.8 (6.7)
10.2 (5.1)
15 (9.4)
Non-woody vegetation (%)
49 (8)
45.8 (46)
81.8 (50)
85.6 (45.5)
Canopy (%)
3.6 (6.4)
4.2 (7.6)
0.4 (0.8)
0.9 (2.2)
Understory
141.1 (28.4)
136 (50.4)
113.2 (48)
114.9 (37)
Elevation
480.7 (181.4)
452.8 (138.8)
417.1 (134.8)
526.4 (159)
Aspect
198.9 (92.7)
188.5 (101.1)
155.7 (116.6)
138.9 (81.4)
35
Figure 2.1. Location of the eight study blocks used to examine the influence of intensive
forest management on early-seral forest biodiversity in the Oregon Coast Range. Blocks
used in this study to assess the influence if intensive forest management on microclimatic
air temperature, May-August 2014, are indicated by orange squares.
36
Figure 2.2. Representative examples of stands spanning the gradient in herbicide
intensity for (A) Control, (B) Light, (C) Moderate, and (D) Intensive treatments in the
Oregon Coast Range, US, 2014.
37
Figure 2.3. Ratio of mean percent cover among control, light, moderate, and intensive herbicide treatments, with Bonferroniadjusted 95% confidence intervals. The dashed line at 1 represents no statistical difference. The left panel (A) depicts differences
in broadleaved cover. The center panel (B) depicts differences in conifer cover. The right panel (C) depicts differences in total
vegetation cover.
38
Figure 2.4. Average daily air temperature fluctuations among herbicide treatments at each experimental block in the Oregon
Coast Range, US. Lines represent the means of all sample points (8) for May-August, 2014.
39
Figure 2.5. Mean air temperature differences among control, light, moderate, and intensive
herbicide treatments in the Oregon Coast Range, May-August 2014, with Bonferroniadjusted 95% confidence intervals. The left panel (A) depicts differences in mean air
temperature. The right panel (B) depicts differences in air temperature variability (CVs).
The dashed line at zero represents no statistical difference. Estimates on the right side of the
dashed line correspond to decreases in air temperature with increasing herbicide treatment
intensity, while estimate on the left side correspond to increases in air temperature.
40
INTENSIVE FOREST MANAGEMENT HAS A GREATER IMPACT THAN
CLIMATE ON REPRODUCTIVE OUTPUT IN A CAVITY-NESTING SONGBIRD
Kristin N. Jones, Matthew G. Betts, James W. Rivers
41
CHAPTER 3: INTENSIVE FOREST MANAGEMENT HAS A GREATER
IMPACT THAN CLIMATE ON REPRODUCTIVE OUTPUT IN A CAVITYNESTING SONGBIRD
3.1 ABSTRACT
Shifts in avian species distributions and phenology that are driven by global
climate change are well-documented, yet we lack a strong understanding of the
mechanisms behind these shifts. Increasing air temperatures, the strongest evidence of
ongoing global climate change, can combine with anthropogenic land cover change to
influence the demographic rates of animal populations. Therefore, identifying the
individual and combined effects of climate and land management practices is essential to
understand how populations will respond to global change. In a large-scale study that
experimentally manipulated intensive forest management practices, we examined effects
of air temperature and degree of vegetation management (via herbicides) on reproductive
output in a cavity-nesting songbird, the House Wren (Trogolodytes aedon). We tested the
relative effects of maximum air temperature and herbicide intensity on nest survival and
the variation in the quantity and quality of offspring produced along a gradient of
management intensity. We did not detect statistical differences in any measure of
reproductive output across herbicide treatments, though our power to detect herbicide
effects was low. Nest survival decreased from the control in the light treatment, but
counter to our predictions, appeared to rebound with further management intensity (i.e.,
in the moderate and intensively managed treatments). Our results also suggest that the
42
effects of maximum air temperature on reproductive output in early-seral forests are
negligible; we detected no link between nest temperature and measures of nest survival,
number of young produced, and offspring quality temperature at the nest, nor did we find
evidence for combined effects of temperature and herbicides on reproductive output.
Overall, neither climate nor land cover change exerted statistically significant effects on
reproductive output in house wrens. Herbicide intensity appeared to have a greater effect
on reproductive output compared to temperature. We hypothesize that land use change
will continue to be the primary global change driver affecting this species in the near
future.
3.2 INTRODUCTION
Habitat loss and degradation resulting from anthropogenic land use are the
principal causes of global declines in animal populations (Vié et al., 2009; Pereira et al.,
2012; Monastersky, 2014). Land cover change driven by anthropogenic land use can
negatively impact demographic rates in some species (Selwood et al., 2014) as a result of
altered habitat availability and quality (Pereira et al., 2012; Selwood et al., 2014).
Although the effects of climate change on demographic rates are less clear (Cahill et al.,
2012), an increasing number of studies have shown that climate significantly influences
temporal variation in demographic rates and population size for many species (McCarty,
2001; Gallinat et al., 2015; Williams et al., 2015). Climate change can influence animal
populations differentially, creating both “winners” (i.e., species that benefit from changes
43
in climate) and “losers” (i.e., species whose response to climate change is negative)
(Williams et al., 2015). Thus, it is expected that changes in climate and land cover will
combine (Brook et al., 2008) to produce novel and potentially deleterious environmental
conditions for many species (Hobbs et al., 2006; Mantyka-pringle et al., 2012; Jantz et
al., 2015). Understanding how these pressures combine to affect demographic rates is
essential to conservation planning as both land cover and climate change are expected to
increase in extent and intensity (Lambin and Meyfroidt, 2011; Seto et al., 2011;
Tscharntke et al., 2012; IPCC, 2013).
Increases in air temperature, one of the most prominent components of ongoing
climate change (IPCC, 2013), can influence demographic rates through both direct (e.g.,
physiological performance) and indirect pathways (e.g., changes in critical resources;
Chase et al., 2005; Hansen, 2009; Adamo et al., 2012; Selwood et al., 2014). For
example, elevated temperatures can alter metabolic rates of organisms (Knut SchmidtNielsen 1997), resulting in decreased survival and fitness (Selwood et al., 2014). In
plants, increased drought frequency may decrease fruit set and seedling survival
(Selwood et al., 2014; Williams et al., 2015). In breeding animals, elevated temperatures
can cause heat stress in parents (e.g., inadequate water availability for lactating females)
and drive declines in neonatal survival and fecundity (Selwood et al., 2014). Birds in
particular have reproductive rates that are closely tied to air temperature because of high
metabolic costs incurred by flight and because their offspring originate in eggs (Chase et
al., 2005; Cox et al., 2013a; Becker and Weisberg, 2014). For instance, ambient
temperatures > 36 - 40.5°C can reduce egg viability (DuRant et al., 2013; Wada et al.,
44
2015) and may induce heat stress in developing nestlings (Murphy, 1985; Pipoly et al.,
2013). Heat stress can shift energy allocation to thermoregulation, ultimately reducing
offspring growth (Murphy, 1985; Pipoly et al., 2013; Salaberria et al., 2013) which may
ultimately reduce post-fledgling survival (Schwagmeyer and Mock, 2008). Furthermore,
heat stress may indirectly impact offspring growth via changes in adult foraging (e.g.,
adults may deliver fewer and lower quality food items for the same amount of effort) (du
Plessis et al., 2012; Edwards et al., 2015). The foraging activity of some nest predators
has been shown to increase with increasing temperatures (Cox et al., 2013b; DeGregorio
et al., 2015), resulting in another indirect impact on the survival of nests and developing
young.
Alternatively, slight temperature increases during the nesting period could
enhance avian reproductive rates. Increased temperature may reduce the exposure of eggs
to temperatures below physiological optima (~ 26°C; Conway and Martin, 2000). Eggs
exposed to suboptimal low temperatures may not hatch or produce offspring with
decreased immune function and/or growth rate (Ardia et al., 2010; DuRant et al., 2012),
and thus decreased survival. Songbirds, especially species with female-only incubation,
may be constrained in maintaining their eggs at optimal temperatures (Webb, 1987)
because of tradeoffs between investing resources in energy-intensive incubation or selfmaintenance (i.e., feeding; Reid et al., 2000). Consequently, increases in ambient
temperatures during incubation can increase offspring growth and survival (Pérez et al.,
2008). Furthermore, increased ambient temperatures during the nestling period may
benefit nestling growth and survival by reducing exposure to cold temperatures (~ 11-
45
20°C; Becker and Weisberg, 2014). Warmer temperatures may also increase food
availability for insectivorous birds whose foraging efficiency can be decreased by cooler
temperatures (via reduced food availability and higher thermoregulatory costs), a result
that has been observed in aerial insectivores (Avery and Krebs, 1984; Winkler et al.,
2002).
Avian reproductive rates can also be influenced by habitat loss and degradation
(Lampila et al., 2005; Becker and Weisberg, 2014; DeGregorio et al., 2014). Forested
ecosystems have been particularly subject to loss and degradation resulting from
anthropogenic land cover change (Hooke et al., 2012), with negative consequences for
some populations (Lampila et al., 2005). Although forest cover loss has slowed globally
(Sloan and Sayer, 2015), much of the world’s forests have been and continue to be
degraded (Potapov et al., 2008). In particular, structurally and compositionally diverse
early-seral forest is declining in the northern hemisphere (Angelstam, 1998; Thomas et
al., 2006), a trend that may be linked to the rapid population declines of several bird
species in the Pacific Northwest that require this habitat type during the critical breeding
season (Betts et al., 2010, 2013).
Intensive forest management, a widely used forest management system (Fox,
2000), may reduce the quality of early-seral forest habitat for breeding songbirds by using
herbicides to suppress the growth of early-seral herbaceous and broadleaved vegetation
that compete with merchantable trees for resources (Shepard et al., 2004; Wagner et al.,
2006). Herbicides can significantly change the abundance, structure, and composition of
early-seral forest vegetation (Balandier et al., 2006), in turn altering the relative strength
46
of factors that limit reproductive output (e.g., food, nest predation). Loss of broadleaved
vegetation through intensive management may have especially strong impacts because
this habitat component is positively correlated with higher densities of invertebrates in
leaf litter and in foliage (Willson and Comet, 1996a; Hagar et al., 2007). In particular, the
abundance of lepidopteran larvae, an important food source for developing nestlings
(Arnold et al., 2010), is positively associated with broadleaved vegetation (Hammond and
Miller, 1998). Moreover, nest predation rates may be higher in more homogenous forests
that contain less deciduous vegetation (Easton and Martin, 1998) because of differences
in the distribution of predator communities (Sieving and Willson, 1998) and because
more diverse foliage allows different species of birds to partition nests among specialized
microhabitats, thereby reducing predation risk (Martin, 1993a). Therefore, herbicide use
in early-seral forests may negatively impact the reproductive output of birds through a
reduction in nest survival, as well as offspring growth.
Forest vegetation composition can influence rates of surface cooling through
changes in biophysical factors such as albedo and canopy conductance (von Arx et al.,
2012; Zhao and Jackson, 2014). Thus, the very changes in vegetative composition that
arise from the use of herbicides may also lead to changes in temperature, and both of
these factors have the potential to negatively impact songbird reproductive output
(Sieving and Willson, 1998; Chase et al., 2005). However, a small number of studies
(Easton and Martin, 1998; Easton et al., 2002) have examined the influence of herbicide
use on avian demographic rates, and few have examined the potential for combined
effects of vegetation changes and air temperature due to herbicide use in forested
47
ecosystems (Cox et al., 2013a; Becker and Weisberg, 2014; Flesch et al., 2015).
Nevertheless, effective conservation approaches require more information on how land
management and climate can combine to affect the demographic rates that underlie
species persistence given ongoing changes in the quality of early-seral habitat (Thomas et
al. 2006, Swanson et al. 2010) and rising global temperatures (IPCC, 2013).
In this study, we examined the effects of herbicides and microclimatic air
temperatures in intensively managed coniferous forests on the reproductive output of a
cavity-nesting songbird, the House Wren (Troglodytes aedon). We selected the House
Wren for this study because (1) this species has experienced a strong, significant longterm decline (3% per year; Sauer et al. 2015) in the Pacific Northwest and (2) House
Wren abundance declines with herbicide treatment intensity (Betts et al. 2013), together
suggesting that decreases in the quality of early-seral forest may be linked to long-term
population declines. We tested the hypotheses that reproductive output was influenced by
air temperature, herbicide treatment, or both acting together. We predicted that House
Wren reproductive output, as measured by nest survival and the number and quality (i.e.,
body condition) of young produced, would decrease with increasing herbicide treatment
intensity as a result of reductions in vegetation cover. We also predicted that these
measures would be positively affected by increased maximum daily air temperatures
throughout the breeding season up to a threshold (nestlings: > 30°C; eggs > 38 - 40.5°C)
(Pipoly et al., 2013; Wada et al., 2015), beyond which reproductive output would
decrease. We used mean daily maximum temperature (hereafter Tmax) because: (1) we
were interested in temperature effects on reproductive output at the scale of days
48
throughout the breeding season, (2) maximum temperatures are likely to be correlated
with a range of temperature summaries relevant to birds (i.e., minimum and mean
temperatures) and can be used as an index of such (Cunningham et al., 2013), and (3)
warm daytime temperatures are expected to have the strongest effect on birds through
reductions in activity (e.g. feeding, nestling provisioning) (du Plessis et al., 2012). Taken
in its entirety, our study is the first to evaluate the combined effects of herbicide intensity
and microclimatic air temperature on songbird reproductive output and therefore has the
potential to provide information critical to songbird conservation planning in early-seral
forests.
3.3 MATERIALS AND METHODS
3.3.1 Study area and species
We conducted this investigation within the framework of a larger study that is
examining the tradeoffs between intensive forest management, biodiversity, and
ecosystem function in the northern Oregon Coast Range (see Betts et al. 2013). Mean
annual precipitation ranges from 165 to 330 cm with most precipitation occurring
October through March. Mean annual air temperature ranges from 5 - 19°C (Franklin and
Dyrness 1988, Taylor and Hannan 1999) Soils are dominantly moderately deep to deep
well-drained silt loam soils, and are derived from basalts, sandstone, and siltstones.
Topography is characterized by somewhat low, highly dissected mountains with slopes
ranging from 0 to 90 percent (Knezevich 1982, Taylor and Hannan 1999). All sampled
49
stands are in the western hemlock (Tsuga heterophylla) zone (Franklin and Dyrness
1988), ranging in elevation from 165 to 765 m. Douglas fir (Pseudotsuga menziesii)
dominates early-seral forest plantations in this region, with grand fir (Abies grandis),
western hemlock (Tsuga heterophylla), and western redcedar (Thuja plicata) forming
minor components. Dominant shrub/woody species include big-leaf maple (Acer
macrophyllum), California hazelnut (Corylus cornuta californica), cascara (Rhamnus
purshiana), common snowberry (Symphoricarpos albus), oceanspray (Holodiscus
discolor), red alder (Alnus rubra), and vine maple (Acer circinatum). Smaller understory
broadleaf species include Oregon grape (Mahonia nervosa) salal (Gaultheria shallon),
and Vaccinium spp. The herbaceous community is comprised of many native and nonnative herbaceous plants, with swordfern (Polystichum munitum) and brackenfern
(Pteridium aquilinum) often dominating.
The House Wren is a small (10 - 12 g), insectivorous, long-distance migrant
passerine that is common throughout much of North America (Johnson 2014). It nests in
a wide variety of open wooded habitats in Oregon from mid-April to mid-August,
preferring nest sites < 30 m from woody vegetation. House Wrens are cavity-nesting
songbirds that primarily nest in pre-formed tree cavities and readily use nest boxes
(Johnson 2014). Female House Wrens can be double-brooded and lay clutches of 4 - 8
eggs, of which more than ~ 90% hatch and ~ 70 - 90% successfully fledge from the nest
(Johnson 2014). Females alone incubate the eggs and brood the young (Figure 3.1a;
Figure 3.1b) until some point in the latter third of the nestling stage (Johnson et al.,
1993); however, both adults provision nestlings (Johnson 2014).
50
3.3.2 Experimental design
Our study makes use of a subset of sites that occur within a larger study area that
uses a randomized complete block design in which 32 stands occurred in 8 separate
blocks, and each of 4 levels of herbicide treatments were randomly applied to one stand
within each block (Figure 3.2). All blocks are located across a 100 km (N-S) section of
the northern Oregon Coast Range region (Betts et al., 2013; Figure 3.2). Stands within
each block were sited > 1 km of each other to ensure spatial independence of treatments
and no more than 5 km of each other to reduce within-block variation (i.e., slope,
elevation, and pre-cut vegetation composition). From this larger study framework, we
constrained our study on 24 forest stands located within 6 of the blocks, with each stand
8.6 - 18.9 ha because of logistical considerations that prevented us from working on all
stands concurrently.
All stands were clear-cut in fall 2009 and replanted in spring 2011 with Douglasfir, the most commonly commercially cultivated species throughout the region, at
approximately 980 trees per hectare. Application of the 4 levels of herbicide treatment
occurred between 2009 and 2014. A full suite of herbicides and surfactants typically used
in commercial timber harvesting operations was applied to stands between 2009 and 2014
in a manner that created a gradient in management intensity; see Betts et al. (2013) and
Appendix A for a full description of the herbicides used, the timing with which they were
applied, and their concentration. All herbicide spraying occurred in a timeframe that
51
mimicked the typical timeframe in which vegetation control takes place in commercial
operations
3.3.3 Sampling design
3.3.3.1 Environmental measurements
3.3.3.1.1 Air temperature
We placed iButtons temperature loggers (n = 192) on the underside of nestboxes
(see below) erected on study sites at a density of 8 boxes/stand (see below). Each iButton
was programmed to record temperature every 15 min and was affixed to the underside
surface of each nestbox with a zip tie so that it hung freely 5 cm below the box. We
covered iButtons with a section of white 10 cm diameter PVC tube with ventilation holes
so as to prevent direct exposure of the iButton to solar radiation and moisture, allow for
airflow, and minimize heat accumulation. Due to logistical constraints, we used two
iButton models that varied slightly in their accuracy (± 1.0°C accuracy, n = 71 boxes: ±
0.5°C accuracy, n = 89 boxes). The two models were distributed irregularly among
stands; thus, we note that estimated differences in air temperature between stands are
within approximately 1.0°C of their true value. All iButtons were validated against an
independent digital thermometer prior to placement on stands (Omega HH609R, Omega
Engineering, Stamford, Conneticut, USA); any iButtons that deviated by ≥ 0.5°C during
his validation procedure were not used.
52
We recorded air temperature on the outside of nest boxes because: (1) ambient
temperatures recorded by iButtons were simultaneously used as part of another study to
assess the influence of herbicide treatment on microclimatic air temperatures (Jones et al.
in prep), and (2) recording temperatures outside the nest box allowed us to standardize
measurements in a way that was impossible inside nest boxes because of marked
variation in the architecture of individual nests (see McCabe, 1965). Our pilot data
indicated that temperatures inside nest boxes were 4.4°C warmer (SE = 0.8, t 3 = 5.54, P
= 0.012) throughout the breeding season, so our results provide a conservative measure of
the temperatures experienced by eggs and nestlings (e.g., McComb and Noble, 1981).
We used temperature data from iButtons to calculate the mean daily maximum
temperature (hereafter Tmax) during each observation interval (see below) and nestling
cycle starting at midnight; thus, this value represents the average for 96 temperatures
taken during every 24-hour period across the breeding season. Following standard
approaches with small-scale temperature data (e.g., Baker et al., 2014), we removed
erroneous temperature that were caused by instrument malfunction or when logging
stations were found to be disturbed (i.e., nestboxes damaged by wildlife) prior to
calculating Tmax from our calculations. We considered temperature data to be erroneous
when values were > 50°C or < -10°C with no temporal or spatial precedence. For
example, temperatures logged by the same iButton within an hour were greater or less
than 10°C of the record in question or temperatures recorded on the same stand at other
nestboxes were 20°C greater or less than the recorded temperature in question. This led
us to remove ≤ 5% of temperature values.
53
3.3.3.1.2 Vegetation
We measured vegetation cover in each treatment to quantify changes in
vegetation abundance and composition among the different herbicide treatments, as
vegetation cover can influence temperature (see above). At each nest box, we measured
vegetation at three distinct sampling points: one was centered directly on the nest box,
and the other two were located 25 m distant from the nest box; these latter two were
considered to represent vegetation in the core of a wren’s territory. The azimuth of the
initial off-nest box point was chosen randomly, with the second point located 180° away.
At each vegetation sampling point, we visually estimated the amount of cover provided
by each plant species in a 3 m radius in each of three distinct vegetative strata;
herbaceous (0 - 0.5 m), shrub (0.5 - 2.0 m), and canopy (> 2.0 m) layers. We aggregated
some closely-related species that are functionally similar (i.e., Rubus and Ribes), as well
as all herbaceous forbs, grasses, and ferns. Vegetation measured in this way predicts
avian abundance in early-seral conifer forests in our study area (e.g., Ellis and Betts,
2011; Ellis et al., 2012). For analysis, we summed the amount of cover for all plant
species and genera over all three strata in each 3 m radius sampling point, and then
calculated an average cover value, grouping species and genera together to estimate (1)
the total amount of broadleaf cover (as defined by Ellis and Betts 2011) and (2) the total
amount of vegetative cover of all species and genera. We note that summed cover values
>100% can result from this analysis approach.
54
3.3.3.2 Nest survival and productivity
Following harvest treatments, 8 cedar nest boxes were mounted 1 - 1.5 m above
the ground on each stand using t-posts so that the box (n = 192 total boxes). Nest boxes
were sited with several considerations in mind: (1) equal distances between boxes (> 50
m separation), (2) even stand coverage, and (3) sampling logistics (i.e., walking distance
between boxes < 30 min). Existing vegetation was not a factor in box siting.
We monitored boxes every 3 - 4 days throughout the 2014 breeding season
starting from late April to early August. We checked nest boxes to record nest contents
(i.e., the number of eggs and nestlings present during each visit) and to determine each
nest’s fate (success or failure). A nest was considered successful if, upon termination of a
nesting attempt, at least one offspring was thought to have fledged from the nest. We
considered a nest successful if (1) we observed nestlings fledging from the nest, (2) we
observed fledglings close to boxes, or (3) if, upon checking the nest box, no live nestlings
were observed in the nest and the nest cup rim was flattened and/or excrement covered
the walls (Sutherland et al. 2004). We considered a nest to have failed if (1) ≥ 1 eggs
were missing or broken, (2) all nestlings were missing or dead before the expected
earliest date of fledging (Martin and Geupel, 1993), or (3) the nest was abandoned by
parents. A nest was considered to be abandoned if eggs did not hatch in the generally
established incubation period for the species (12 - 13 days) and/or once no female nesting
activity was observed around the nest for at the nest for at least three consecutive nest
observation intervals. In these instances, the nest was recorded as having failed on the
earliest possible end date for the incubation period (Converse et al., 2013). House Wren
55
nestlings may fledge earlier than the general nestling period (17 - 18 days) for this
species (Johnson 2014), so if a nest was found empty and lacking diagnostic signs of a
fledged nest (e.g., presence of pin feather sheaths), the nest was considered to have failed
(Sutherland et al. 2004). Nests with uncertain fates and those who were suspected to have
failed due to human activities were removed from analysis.
To quantify the reproductive output of successful nests, we recorded the number
of offspring produced from each nest and the body condition of those offspring as a
measure of offspring quality. We recorded the number of offspring fledged from a nest as
the number of nestlings present on nestling day 8/9 (where nestling day 0 is the day of
hatch) minus the number of nestlings found dead in the nest after fledging. On nestling
day 8/9, we measured nestling tarsus length and wing chord (± 0.5 mm, dial caliper and
wing chord ruler, respectively), and body mass (± 0.1 g, digital scale) (Sutherland et al.
2004). House Wren offspring do not reach the peak of their growth until later in the
nestling stage (Day 10 - 13) (Zach, 1982), but we were unable to measure nestlings on
our study area at later developmental stages because of the risk of premature fledging
(Jones, personal observation). We averaged across all nestlings in a brood to calculate
mean age, body mass, tarsus length, and wing chord estimates for day 8/9 nestlings for
each nesting attempt. Due to logistical constraints, we measured nestling body condition
on 4 of the 6 study blocks (16 stands).
3.3.4 Statistical analysis
56
3.3.4.1 Vegetation
All models were fit in in ‘R’ v3.2.0 (R Core Development Team 2015). First, we
used mixed model analysis to test whether forest vegetation abundance and composition
differed significantly among herbicide treatments using the lme function of the nlme
package (Pinheiro et al. 2015). We constructed models for three response variables of
forest vegetation (broadleaved, conifer, and total vegetation cover) that included fixed
effects for treatment (4 levels: control, light, moderate, intensive) and one random effect
(block) to reflect our randomized complete block design. To correct for heterogeneity of
variance among herbicide treatments, both broadleaved and total cover were log
transformed prior to analysis. We back-transformed values after analysis; therefore,
values should be interpreted as the mean multiplicative increase or decrease in cover
between treatments. Thus, a treatment contrast that overlaps with 1 indicates there is no
evidence of a statistical difference between two treatments.
3.3.4.2 Reproductive output
All models were fit in in ‘R’ v 3.2.0 (R Core Development Team 2015). Models
for the number of young produced and nestling body condition were fit using the lme
function of the nlme package (Pinheiro et al. 2015); in contrast, nest survival models
were fit using the glmer function of the lme4 package (Bates et al. 2015). We constructed
three sets of linear mixed models representing a priori hypotheses to separately model
the relationship between (1) herbicide treatment treatments and reproductive output
57
measures (i.e., nest survival, number of young produced, and nestling body condition)
and (2) Tmax and reproductive output measures (see Table 3.1). We included models that
did and did not contain a quadratic term for Tmax because the relationship between
reproductive measures and Tmax may not be linear. For example, moderate increases in
Tmax may benefit nestlings, but excessive increases could cause negative consequences.
We included three random effects in all models: (1) study block, (2) treatment stand, and
(3) nestbox. The random effects for block and stand account for potential correlation of
nest fates within blocks and stands, whereas the random effect for each nestbox accounts
for potential correlation of fates between nests occurring in the same nestbox. Elevation
was included as a covariate in all models to control for differences in elevation between
stands and between boxes within stands. Although avian reproductive output often
declines with seasonal advancement (Perrins, 1970; Martin, 1987), the date of nest
initiation was highly correlated (r > 0.8) with Tmax in our study. Recent work has
indicated that model averaged-coefficients based on AIC weights are invalid where there
is multicollinearity among predictor variables (Cade, 2015), so we did not include a
covariate for nest initiation date in our analyses.
We used a second-order Akaike’s Information Criterion, AICc, to quantify the
relative strength of all competing linear regression models for each response variable
(Burnham and Anderson 2002; Anderson 2008). Model parameter estimates were
averaged across the set of candidate models for each response variable (Anderson 2008).
We used the MuMIn package (Barton 2015) to conduct the model selection and
averaging process. Before performing model selection, the global model (i.e., candidate
58
model with the most parameters) for the number of young produced and nestling body
condition response variables was evaluated for compliance with model assumptions (see
below regarding nest survival model checking). Treatment and temperature effects were
considered significant at P ≤ 0.05.
We used the logistic exposure method to estimate daily nest survival rate (Shaffer,
2004), which allows for the modeling of time-dependent covariates (Grant et al., 2005;
Converse et al., 2013). Logistic exposure takes into account the fate of a nest during
observation intervals (between successive nest checks) by using a logistic-exposure link
function that explicitly considers exposure time, as measured by the length of observation
interval (Shaffer 2004). In all logistic exposure models we included a term for mean nest
age (i.e., nest age at the mid-point of each observation interval; see Grant et al. 2005)
because this measure can influence nest survival rates (Thompson, 2007). All continuous
variables were standardized prior to analysis, and subsequently un-standardized to allow
for interpretation of estimates (Hosmer et al. 2013). Currently, no effective methods exist
for testing fit of logistic exposure models (Shaffer and Thompson, 2007; Shaffer,
personal communication); therefore, we used binned residual plots from package ‘arm’
(Gelman et al. 2015) as a qualitative test and found little evidence of model misfit with
regards to normality among model residuals (Gelman 2000).
To model the number of young produced, we fit linear mixed models. For this
measure, we only considered successful nests for assessing the number of young
produced to remove any variation caused by differences nest predation among treatments.
The resulting data subset was a left-censored Poisson distribution of the response, and
59
thus severely underdispersed (dispersion parameter < 0.4, P < 0.001). As the censored
data approximately followed a normal distribution, we treated the data distribution as an
approximation of the normal distribution for analysis (Greene 2005). We evaluated 6
models that tested the relative importance of herbicide treatment intensity and air
temperature effects on the number of young produced (Table 3.1). All models contained a
term for clutch size (i.e., the maximum number of eggs laid during a nest attempt) to
control for any differences in potential brood size.
To model nestling body condition, we fit linear mixed models. Although body
condition indices such as body mass regression residuals are used widely in the literature
(Labocha and Hayes, 2012), recent work suggests that simple measures of body mass
often outperform body condition indices as a measure of energy stores (Schamber et al.,
2009; Labocha and Hayes, 2012). Therefore, we included tarsus length as a size covariate
in all body condition analyses to control for size (Bowers et al., 2014; Paquette et al.,
2014). In addition, we included a covariate for mean nestling age when measured to
control for differences when nestlings were measured for growth.
3.4 RESULTS
3.4.1 Effects of herbicide treatments on temperature and vegetation
During May-August 2014, Tmax on our study sites averaged 24.8 (± 4.8) °C
(range: 17.1 - 32.2°C; Figure 3.3) and varied little among treatments (Table 3.2),
indicating no direct effect of herbicide treatment on air temperature. As expected,
60
herbicide treatments had a strong influence on vegetation composition (Figure 3.4);
broadleaved vegetation cover generally decreased with increasing herbicide treatment
intensity (F3, 11 = 18.68, P < 0.001). After correcting for multiple comparisons, we
detected significant differences in broadleaved cover between the control and the
moderate treatments (t 9 = 4.88, P < 0.001, 𝛽̂ = 1.88 [0.68, 3.08]), the control and
intensive treatments (t 9 = 5.04, P < 0.001, 𝛽̂ = 1.93 [0.74, 3.12]), and the light and the
moderate treatments (t 9 = 4.89, P < 0.001, 𝛽̂ = 1.87 [0.68, 3.05]). We also found that
conifer cover increased with increases in herbicide intensity (F 3, 11 = 9.19, P = 0.003),
yet we only detected a significant difference in conifer cover between the control and
intensive treatments (t 9= -4.69, P < 0.001, 𝛽̂ = -0.95 [-1.58, -0.32]). We also found that
conifer cover increased with increases in herbicide intensity we only detected a
significant difference in conifer cover between the control and intensive treatments ( t 9 =
-4.69, P < 0.001, 𝛽̂ = -0.95 [-1.58, -0.32]). We did not detect any statistically significant
differences in total vegetation cover among treatments (F3, 11 = 2.070 P = 0.162). We
found that vegetation cover within vegetative strata (e.g., canopy cover) was similar
among herbicide treatments, and that non-woody vegetation cover increased with
increasing herbicide treatment primarily because of an increase of exotic ruderals (Table
3.2).
3.4.2 Influence of herbicide treatment and temperature on nest survival
61
The total number of nesting attempts by House Wrens was similar among
treatments (Figure 3.5a). The average number of nests initiated per stand was 11.8 (±
4.4), reflecting a combination of high occupancy of nest boxes and multiple nesting
attempts in individual boxes. Of 282 active nests, 78 (28%) failed, with 95% of these nest
failures classified as predation.
We found some support for an effect of herbicide treatment on nest survival; all
models containing treatment were better supported than the null model (ΔAICc = 2.69,
Evidence Ratio (ER) = 3.83; Table 3.3). However, we did not detect a statistically
significant effect of herbicide treatment on House Wren nest survival (Table 3.4; Figure
3.5a). Nevertheless, compared to the control treatment, herbicide treatment negatively
influenced mean daily survival rate. Treatment most negatively influenced daily survival
rate in the light treatment (Odds Ratio (OR) = 0.35, 95% CI: 0.09, 1.48; Table 3.4), while
its effect was greatly reduced in the intensive treatment (OR = 0.78, 95% CI: 0.29, 2.10),
and intermediate in the moderate treatment (OR = 0.51, 95% CI: 0.16, 1.59; Table 3.4;
Figure 3.5a). Confidence intervals around parameter estimates for herbicide treatment
effects on nest survival were large (Table 3.4), indicating low statistical power to detect
effects.
We also did not detect a significant effect of temperature on nest survival (Table
3.4; Figure 3.5b). Furthermore, we did not find any evidence of quadratic relationship
between temperature and mean daily nest survival rate (𝛽̂ = -0.14, 95% CI: 0.73, 1.04;
Table 3.4). We found some support for combined effects of treatment and temperature on
mean daily survival rate, as the best supported model contained treatment, temperature,
62
and its quadratic term (wi = 0.35; Table 3.3). However, the best supported model
containing treatment and a quadratic effect of temperature was only 1.3x more likely than
the next best model, which contained treatment only (ΔAICc = 0.53, ER = 1.30; Table
3.3). We note that the comparable ΔAICc values of all models included in the model set
(≤ 4.08) in addition to comparable evidence ratios (Table 3.3) strongly indicated that all
hypotheses were equally plausible (Table 3.4).
3.4.3 Influence of herbicide treatment and temperature on the number of young
produced
On average, 4.9 (± 1.6) nestlings were produced per successful nest across all
treatments combined (n = 204 successful nests; Figure 3.6). We did not find support for
an effect of treatment on the mean number of young produced per nest. The bestsupported model was the null model containing elevation and maximum brood size
(AICc wi = 0.62; Table 3.3). All models containing herbicide treatment had AICc weights
< 0.04 and evidence ratios > 16 (Table 3.3). We did not detect a significant effect of
treatment on the mean number of young produced (Table 3.4; Figure 3.6a). We also did
not detect an effect of temperature on the number of young produced per nest (Table 3.4).
Nevertheless, the mean number of young produced generally decreased with increasing
temperature, however, this estimated effect was extremely small and non-significant
(-0.01 nestlings/1°C, 95% CI: -0.07, 0.05; Figure 3.6b). We also did not detect significant
evidence of a quadratic effect of temperature on the number of young produced (95% CI:
-0.01, 0.01). However, comparable ΔAICc values and evidence ratios for temperature
63
(ΔAICc = 2.01, ER = 2.73) and its quadratic term (ΔAICc = 3.81, ER = 6.70) indicated
that temperature is as equally as plausible an explanation as the null model (Table 3.3).
Finally, we did not find support for combined effects of treatment and temperature on the
mean number of young produced (Table 3.4); both models containing treatment and
temperature had ΔAICc values > 7 and evidence ratios > 47 (Table 3.3).
3.4.4 Influence of herbicide treatment and temperature on nestling body condition
On average, nestlings on day 8/9 averaged 8.86 (± 0.76) grams across all
treatments (Figure 3.7a). The null model containing elevation, mean tarsus length, and
mean nestling age was the best-supported model (AICc wi = 0.36; Table 3.3). However,
that the model containing treatment was the second best supported model (ΔAICc = 0.80,
ER = 1.49) suggests some support for an effect of treatment on mean nestling body
condition. We did not detect a significant effect of treatment on nestling body condition
(Table 3.4)
We also did not detect a significant effect of temperature on nestling body
condition (Table 3.4; Figure 3.7b), nor did we did not detect significant evidence of a
quadratic effect of temperature on mean nestling body condition (Table 3.3). Lastly, we
did not find support for combined effects of treatment and temperature on the mean
number of young produced (Table 3.2). However, models containing treatment and
temperature and treatment and a quadratic effect of temperature were only 4.7x and 7.5x
more likely than the null to explain variation in nestling body condition, respectively
(Table 3.3). All hypotheses were equally good at explaining variation in nestling body
64
condition, evidenced by comparable ΔAICc values and ERs of all models in the model set
(Table 3.3).
3.5 DISCUSSION
Our study suggests that in intensively managed early-seral conifer forests, the use
of post-harvest herbicides may influence House Wren nest survival but not the number of
young produced nor their quality. Our results also indicate that the effects of air
temperature on reproductive output in early-seral conifer forests are negligible compared
to herbicide-driven shifts in vegetation. Lastly, our study provides very little empirical
support for a combined effect of herbicide treatment and air temperature on nest survival,
the number of young produced, or offspring quality.
Our finding of a lack of combined effects of air temperature and herbicide-driven
changes in early-seral forest vegetation change differs from a growing number of other
studies (Cox et al., 2013a; Becker and Weisberg, 2014; Flesch et al., 2015). However,
those studies examined the relationship between reproductive output and temperature on
a coarser scale between years whereas our study was focused on changes within the
breeding season. Other investigations have found within-season effects of temperature on
songbird reproduction (Salaberria et al., 2013), but our data indicate that temperatures in
our study area did not reach those that could be deleterious to avian offspring (Pipoly et
al., 2013; Wada et al., 2015). Indeed, Tmax taken across our study sites during the
breeding season never exceed 33°C, which is well within the optimal range for eggs
65
(DuRant et al., 2013). Temperatures above > 30°C have been shown to negatively affect
nestling growth (Murphy, 1985; Pipoly et al., 2013) but not survival. However, that we
did not detect an effect of temperature on House Wren nesting growth may be because
such temperatures were relatively infrequent on our study sites (e.g., 18% [340/1869] of
all observation periods). Thus, even though we had high spatial resolution, full-season
temperature measurements across a managed forest landscape, we were still unable to
detect an effect of Tmax on nest survival. Coupled with the small effects of temperature on
nest survival and body condition and the relatively low variability surrounding these
estimates, our results suggest that any combined effects are limited if they are indeed
present. In contrast, the limited empirical support for combined effects of herbicide
treatment and air temperature for the number of young produced indicates that these two
factors do not combine to influence reproductive output.
Of the factors we assessed, herbicide treatment had the greatest impact on House
Wren reproductive output through reductions in nest survival. This was indicated by our
inability to detect an effect of air temperature on nest survival after considering herbicide
treatment and the relatively large (though variable) effects of herbicide treatment on nest
survival. The difference in House Wren nest survival between control and herbicide
treatments in our study is consistent with the one other study of the effects of herbicidedriven changes in vegetation composition on songbird nest survival. Easton and Martin
(1998) found the nest survival of several species of forest songbirds such as the American
Robin (Turdus migratorius), Dark-eyed Junco (Junco hyema) and Dusky Flycatcher
(Empidonax oberholseri) to decline with herbicide-driven decreases in broadleaved
66
vegetation abundance in managed conifer forests in British Columbia. They found daily
survival probability to be as much as 18% lower between control and herbicide-treated
forest stands. The commercial forest stands we studied were similar to those in Easton
and Martin (1998) in their age, their dominance by conifers, and their location in the
Pacific Northwest, yet our study found nest survival can decline by more than triple that
amount (64% between the control and light treatments) for House Wrens. This strongly
implicates local factors as being important in driving variation in avian nest predation, a
topic that should be the focus of future studies.
House Wren nest survival, although impacted by herbicide treatment, did not
respond to management intensity as we expected. Among herbicide treatments, nest
survival increased substantially; when compared to control stands, daily survival rate was
2.9x lower in the light treatment (64%) than in the intensive treatment (22%). Several
studies have shown a decrease in House Wren nest survival with increasing vegetation
density at the nest (Belles-Isles and Picman, 1986; Finch, 1989; Li and Martin, 1991),
which may provide hiding cover for nest predators (e.g., snakes, small mammals) that are
themselves preyed upon by higher-level predators (e.g., raptors). The idea that greater
vegetation cover leads to reduced nest survival is not supported by our finding that
predation rates were lowest in the control stands. The mechanism behind this result is
unclear, but it may be that a vegetation threshold exists whereby the numbers and/or
activity of nest predators changes (e.g., prey-switching; (Schmidt, 1999). Nest boxes in
our study were used by three other species (Jones et al., unpublished data) and box
occupancy was high on all stands, so perhaps the vegetation on light herbicide stands was
67
enough to provide hiding cover to nest predators, but not so much that it impeded their
ability to search for and locate nests. Regardless of the mechanism responsible for the
variation in nest survival in this study, the relationship between vegetation cover and nest
survival is less clear for cavity-nesting birds than it is for open-cup nesting species
(Martin, 1993b; Thompson, 2007). For example, Li and Martin (1991) examined nest
survival for 14 species of cavity-nesting birds in conifer forest in central Arizona and
found the height of the cavity above the ground and entrance diameter to be the strongest
predictors of nest survival, which was negatively correlated with increasing vegetation
cover. Cockle et al. (2015) also found entrance diameters and nest height to be
significantly more important than nest-site vegetation in predicting the nest survival of 35
species of cavity-nesting birds in northeastern Argentina. Similarly, the nest survival of
Western Bluebirds (Sialia Mexicana) (Kozma and Kroll, 2010) and Lewis’ woodpeckers
(Melanerpes lewis) (Newlon and Saab, 2011) were found to be most strongly related to
factors such as nest-initiation date and clutch size rather than nest-site vegetation
characteristics. The relationship between House Wren nest survival and herbicide
treatment is likely multifaceted; however, the large effects of herbicide treatment on nest
survival we observed indicate that the relationship between post-harvest vegetation
management and nest survival for cavity-nesting songbirds warrants further exploration.
Decreased broadleaved cover may also influence the reproductive output of
breeding songbirds by affecting the amount and quality of food available, as invertebrate
density in leaf litter and on foliage are positively correlated with this vegetation type
(Willson and Comet, 1996b; Hagar, 2007). For example, lepidopteran larvae, an
68
important source of energy and nutrients for songbird nestlings (Arnold et al., 2010), are
associated with increased abundance of herbaceous and deciduous vegetation (Hammond
and Miller, 1998; Miller et al., 2003). Thus, we would have expected food availability to
decrease with increasing herbicide treatment intensity, and that should translate into
decreased nestling body mass and/or the number of young produced (Schwagmeyer and
Mock, 2008). However, that herbicide treatment was as equally likely as all other
hypotheses to explain the number of young produced and their body condition in
combination with its small effects suggests that herbicide treatments likely did not alter
food availability for House Wrens in our study. This may be due to their general foraging
habits and broad nestling diet (Pezzolesi 2000), which may not be as sensitive to
decreased broadleaf vegetation as songbirds that are more strongly associated with
broadleaf vegetation (Betts et al., 2010; Ellis and Betts, 2011).
Our results support the notion that, despite increased concern about climate
change impacts, anthropogenic land use will continue to be the primary global change
driver influencing animal populations in the near future (Pereira et al., 2012). Our study
suggests that current air temperature patterns do not combine with land cover change to
impact the reproductive rates of animal populations in temperate, early-seral coniferous
forests. However, it is plausible that projected increases in climate (IPCC, 2013) may
lead to combined effects of air temperature and forest cover change that reduce
reproductive rates in ways that are apparently not present at the current time. Herbicide
use in early-seral forests may alter the structure and function of early-seral communities
(Flueck and Smith-Flueck, 2006) and predators and prey may respond differently to
69
ongoing climatic shifts. Thus, as anthropogenic land cover change expands and
intensifies (Lambin and Meyfroidt, 2011; Seto et al., 2011; Tscharntke et al., 2012) and
the climate continues to change (IPCC, 2013), recognition is growing that these pressures
must be considered in tandem to more accurately predict how animal populations will
respond to global change. Furthermore, patterns of land use change can influence
populations far more than climate (Pereira et al., 2012), so understanding the relative
influence of each factor is essential to formulating successful conservation strategies
(Brook et al., 2008).
70
Table 3.1. A priori candidate models describing herbicide treatment and air temperature
effects on House Wren nest survival, the number of young produced, per nest, and
nestling body condition in the Oregon Coast Range, US, 2014.
Model name
Parameters
Description
Nest survival
Null*
Elevation + mean nest age
Temperature
Tmax
Documented variation in nest survival with
elevation and nest age
Daily maximum temperature influences nest
survival directly via inducing nestling
metabolic stress and decreased nest survival
Tmax²
Daily maximum temperature influences nest
survival directly via inducing nestling
metabolic stress and decreased nest survival
and may exhibit a quadratic relationship
Management intensity
Herbicide treatment†
Herbicide treatment influences nest survival
via altered vegetative abundance, structure,
and composition and thus decreased food
availability
Temperature +
Management intensity
Tmax + herbicide treatment†
Herbicide treatment influences nest survival
but this effect is compounded daily
maximum temperatures
Tmax² + herbicide treatment†
Herbicide treatment influences nest survival
but this effect is compounded daily
maximum temperatures and may exhibit a
quadratic relationship
Elevation + Maximum brood size
Null model accounts for documented
variation in fledgling brood size with
elevation and the number of nestlings
present in each nest able to survive to
fledging
Tmax
Daily maximum temperature influences
fledgling brood size directly via inducing
nestling metabolic stress and decreased
nestling survival
Tmax²
Daily maximum temperature influences
fledgling brood size directly via inducing
nestling metabolic stress and decreased
nestling survival and may exhibit a
quadratic relationship
Temperature²
Temperature² +
Management intensity
Number of young
produced
Null*
Temperature
Temperature²
71
Management intensity
Herbicide treatment†
Herbicide treatment influences fledgling
brood size via decreased food availability
and thus fewer nestlings survive to fledging
Temperature +
Management intensity
Tmax + herbicide treatment†
Herbicide treatment influences fledging
brood size but this effect is compounded by
daily maximum temperatures
Tmax² + herbicide treatment†
Herbicide treatment influences nest survival
but this effect is compounded daily
maximum temperatures and may exhibit a
quadratic relationship
Elevation + Mean tarsus length +
Mean nestling age
Null model accounts for documented
variation in songbird nestling mass with
elevation with a factor to correct for nestling
size
Tmax
Daily maximum temperature influences
nestling mass directly via inducing nestling
metabolic stress and thus influencing
growth
Temperature²
Tmax²
Daily maximum temperature influences
nestling mass directly via inducing nestling
metabolic stress and thus influencing
growth and may exhibit a quadratic
relationship
Management intensity
Herbicide treatment†
Herbicide treatment influences nestling
mass via decreased food availability and
thus slower nestling growth and lower mass
Temperature +
Management intensity
Tmax + herbicide treatment
Herbicide treatment influences nestling
mass but this effect is compounded by daily
maximum temperatures
Tmax² + herbicide treatment†
Herbicide treatment influences nest survival
but this effect is compounded daily
maximum temperatures and may exhibit a
quadratic relationship
Temperature² +
Management intensity
(Global model)
Nestling body condition
Null*
Temperature
Temperature² +
Management intensity
(Global model)
*All terms in the null model were included in the candidate set
†Herbicide treatment includes four levels of treatment intensity
(Control, Light, Moderate, Intense)
72
Table 3.2. Means (± 1 SD) for vegetation and stand location attributes for stands subjected to different herbicide treatments in
the Oregon Coast Range, 2014. Vegetation (percent cover) was measured June-August 2014 during the height of the growing
season, and was averaged over three sample points at each nest box. Measurements from each box were then averaged then
averaged over all herbicide treatments units. Temperature (Tmax), elevation and aspect were measured at each nest box and then an
averaged average was calculated for all boxes within stands subjected to each herbicide treatment over treatment units. All values
are stand-level means.
Treatment
Control
Light
Moderate
Intensive
Broadleaved (%) Conifer (%)
59.4 (± 32.6)
5.6 (± 4.5)
59 (± 32.4)
7.8 (± 6.7)
13.5 (± 16.4)
10.2 (± 5.1)
9 (±7.1)
15 (± 9.4)
Non-woody (%)
49 (± 38)
45.8 (± 46)
81.8 (± 50)
85.6 (± 45.5)
Canopy (%)
3.6 (± 6.4)
4.2 (± 7.6)
0.4 (± 0.8)
0.9 (± 2.2)
Understory (%)
141.1 (± 28.4)
136 (± 50.4)
113.2 (± 48)
114.9 (± 37)
Temperature (°C) Elevation (m)
24.7 (±1.7)
480.7 (± 181.4)
25.1(±1.5)
452.8 (± 138.8)
25.1(±1.8)
417.1 (± 134.8)
24.3 (±1.4)
526.4 (± 159)
Aspect (degrees)
198.9 (± 92.7)
188.5 (± 101.1)
155.7 (± 116.6)
138.9 (± 81.4)
73
Table 3.3. Model selection results from a priori candidate models describing the effects
of herbicide treatment and Tmax on House Wren nest survival, the number of young
produced per nest, and nestling condition in the Oregon Coast Range, 2014. Models are
ranked in ascending order by Akaike’s Information Criterion adjusted for small sample
sizes (AICc). The number of parameters (k), difference between the best model and all
other models (ΔAICc), the relative likelihood of a model, AICc weights (wi), and
evidence ratio (ER) are given for each model.
Model set
Nest survival
Tmax² + Herbicide treatment
Herbicide treatment
Tmax + Herbicide treatment
Tmax²
Elevation + Mean nest age
Tmax
Number of young produced
Elevation + Brood size
Tmax
Tmax²
Herbicide treatment
Tmax + Herbicide treatment
Tmax² + Herbicide treatment
Nestling body condition
Elevation + Mean tarsus length + Mean nestling age
Herbicide treatment
Tmax
Tmax²
Tmax + Herbicide treatment
Tmax² + Herbicide treatment
k
AICc
ΔAICc
wi
ER
11
9
10
8
6
7
636.39
636.91
638.20
638.56
639.07
640.47
0.00
0.53
1.82
2.17
2.69
4.08
0.35
0.27
0.14
0.12
0.09
0.05
1.00
1.30
2.48
2.97
3.83
7.67
7
8
9
10
11
12
481.31
483.32
485.12
486.93
489.00
490.95
0.00
2.01
3.81
5.62
7.69
9.64
0.62
0.23
0.09
0.04
0.01
0.01
1.00
2.73
6.70
16.84
47.92
124.60
8
11
9
10
12
209.33
210.13
211.74
211.99
212.44
0.00
0.80
2.41
2.66
3.11
0.39
0.26
0.12
0.10
0.08
1.00
1.49
3.34
3.77
4.73
13
213.36
4.03
0.05
7.46
74
Table 3.4. Results of models testing for effects of herbicide treatment and Tmax on House
Wren nest survival, the number of young produced per nest, and nestling body condition
in the Oregon Coast Range, 2014. Model coefficients (β) and 95% confidence intervals
are given for each model. For nest survival models, odds ratios are also given. All model
coefficients are model-averaged estimates.
Model set
β
Odds ratio
L (95% CI)
Intercept
5.40
220.52
87.78
554.02
Temperature
0.00
1.00
0.82
1.21
Temperature (squared)
-0.14
0.87
0.73
1.04
Light treatment
-1.03
0.36
0.09
1.48
Moderate treatment
-0.68
0.51
0.16
1.59
Intensive treatment
-0.25
0.78
0.29
2.10
Elevation
0.00
1.00
0.70
1.43
Mean nest age
0.03
1.03
0.81
1.31
Effect
U (95% CI)
Nest survival
Number of young produced
Intercept
0.48
-0.12
1.08
Temperature
-0.01
-0.07
0.05
Temperature (squared)
0.00
-0.01
0.01
Light treatment
-0.13
-0.56
0.30
Moderate treatment
-0.13
-0.55
0.29
Intensive treatment
0.00
-0.45
0.44
Elevation
0.00
0.00
0.00
Maximum brood size
0.93
0.85
1.00
75
Nestling body condition
Intercept
2.26
-0.16
4.69
Temperature
0.00
-0.02
0.02
Temperature (squared)
0.00
0.00
0.00
Light treatment
0.02
-0.16
0.20
Moderate treatment
0.01
-0.14
0.17
Intensive treatment
0.02
-0.17
0.21
Elevation
0.00
0.00
0.00
Mean tarsus length
2.76
2.63
2.88
Mean nestling age
0.31
0.07
0.55
76
Figure 3.1. (A) A House Wren (Troglodytes aedon) nestling approximately 8 days after hatching being prepared for measurement
of body condition, and (B) a completed clutch of House Wren eggs during incubation inside a nest box.
77
Figure 3.2. General location map of the eight study blocks used to examine the influence
of intensive forest management on early-seral forest biodiversity in the Oregon Coast
Range. Blocks used in this study to assess the influence if intensive forest management
on House Wren reproductive output May-August 2014, are indicated by orange squares.
78
Figure 3.3. Mean daily maximum temperature (Tmax) across all study sites in the Oregon Coast Range during the House Wren
breeding season, May-August 2014. Measurements were taken every 15 min, averaged across a 24-hour period, and then
averaged over all nest boxes (n = 192) on 24 study sites in the Oregon Coast Range. The gray band shows the 95% confidence
interval.
79
Figure 3.4. Representative examples of stands spanning the gradient in herbicide intensity,
including ranging from (A) Control (no-spray), (B) Light, (C) Moderate, and (D) Intensive
treatments in the Oregon Coast Range, US, 2014.
80
Figure 3.5. Plots depicting the effect of herbicide treatment and Tmax and on House Wren nest survival during the
81
breeding season in the Oregon Coast Range, 2014. (A) Differences in odds ratio estimates of nest survival between
the effect of herbicide treatments relative to control stands; odds ratios were averaged over all models in the
candidate set. The dashed horizontal line represents odds ratios of the control for comparison with herbicide
treatments; 95% CIs that overlap one indicate lack of significant differences relative to the control treatment.
Numbers above confidence intervals indicate the number of nests monitored per treatment; there were 73 nests in
the control treatment. (B) Boxplots depicting Tmax values from nests that either failed (pink) or fledged offspring
(blue) across all four treatments. Vertical bars within boxes represent medians, boxes are interquartile ranges,
whiskers are 1.5xinterquartile range and dots are outlying data.
82
83
Figure 3.6. Number of House Wren young produced per successful summarized by (A) herbicide treatment and (B)
Tmax during the breeding season in the Oregon Coast Range, 2014. (A) Boxplots depict the number of young
produced by treatment. Bars are medians, yellow diamonds are means, boxes are interquartile ranges, and whiskers
are 1.5xinterquartile range. Numbers above whiskers indicate the number of successful nests per treatment. (B)
Relationship between Tmax and the number of fledglings produced per successful nest. The blue line is a fitted linear
regression line with 95% confidence intervals.
84
85
Figure 3.7. House Wren day 8/9 body condition summarized by (A) herbicide treatment and (B) Tmax during the
breeding season in the Oregon Coast Range, 2014. (A) Boxplots depict nestling body mass by treatment. Bars are
medians, yellow diamonds are means, boxes are interquartile ranges, and whiskers are 1.5xinterquartile range.
Numbers above whiskers indicate the number nests in which nestlings were measured per treatment. (B)
Relationship between Tmax and nestling mass. The blue line is a fitted linear regression line with 95% confidence
intervals.
86
CHAPTER 4: CONCLUSIONS
My thesis explored the effects of post-harvest vegetation control intensity on
microclimate and avian population demography in intensively managed early-seral
forests. I examined the microclimatic air temperature and its variability along a gradient
of herbicide treatment intensity to test for combined effects of herbicide treatment
intensity and air temperature on reproductive output in the House Wren, an insectivorous
cavity-nesting songbird.
Animal population responses to future climate change are often predicted using
large-scale envelope models that do not account for the local climatic conditions to which
organisms most closely respond (Potter et al., 2013). Predictions from such models can
be incomplete and misleading (Gillingham et al., 2012; Varner and Dearing, 2014);
resulting in misguided wildlife management decisions. Further, they do not give any
information as to how land management decisions may modify the impacts of expected
climatic shifts on organisms. I found that herbicide treatment may influence forest
microclimatic air temperatures and their variability (Chapter 2); however, these effects,
if present, are small and variable. Instead, microtopograhic variables such as elevation
and distance to stand edge were the principal determinants of microclimatic air
temperatures. My results support recent findings that, in addition to differences in
vegetative abundance and structure, differences in forest vegetation composition may
influence microclimatic air temperatures (von Arx et al., 2012; Zhao and Jackson, 2014).
However, the generally small and the variable effects of herbicide-driven vegetation
87
changes on air temperature among treatments in addition to a lack of herbicide treatment
effects on temperature variability indicates little potential for post-harvest vegetation
management to exacerbate or ameliorate ongoing climate change effects (IPCC, 2013) on
early-seral organisms in coniferous forests. My results suggest that microtopographic
elements such as elevation may be the most significant modifiers of future climate
change impacts on early-seral organisms in topographically heterogeneous and forested
environments such as the Oregon Coast Range.
In Chapter 3, I explored the potential for combined effects of air temperature and
herbicides on reproductive output in a House Wren nestbox population in intensively
managed early-seral coniferous forests. My results suggest that post-harvest vegetation
control appears to influence songbird nest survival, but not the number of young
produced or their quality. However, nest survival did not decline along a gradient of
herbicide intensity as expected. Instead, nest survival rate was lowest in the light
treatment and highest the intensive herbicide treatment; however, these effects, though
large, were not statistically significant. Furthermore, my results suggest that the effects of
air temperature on songbird reproductive output in early-seral forests is negligible
compared to herbicide-driven shifts in vegetation. Lastly, my results suggest that there is
limited evidence for temperature and herbicide-driven vegetation changes to combine in
their effects on songbird reproductive output. My results support the prediction that
anthropogenic land cover change will continue to be the most important global change
driver impacting animal populations in the near-future.
88
As global climate change alters the distribution and abundance of plant species
(Kelly and Goulden, 2008; Pauli et al., 2012) and anthropogenic land cover change
increases in extent and intensity (Hooke et al., 2012; Tscharntke et al., 2012), it is
imperative that future studies model the combined and interacting (De Chazal and
Rounsevell, 2009) effects of climate change and vegetation on population dynamics to
more completely understand how species will respond to global change (Brook et al.,
2008; Oliver and Morecroft, 2014). Thus, mine is the first study to examine the potential
for combined effects of herbicide treatment and microclimatic air temperature on avian
demographic rates in intensively managed early-seral forests, and my study advances our
understanding of the effects of intensive forest management practices on early-seral
forest ecology.
LITERATURE CITED
Adamo, S.A., Baker, J.L., Lovett, M.M.E., Wilson, G., 2012. Climate change and
temperate zone insects: the tyranny of thermodynamics meets the world of limited
resources. Environmental Entomology 41, 1644–52.
Adams, W., Hobbs, S., Johnson, N., 2005. Intensively managed forest plantations in the
Pacific Northwest: Introduction. Journal of Forestry 103, 59–60.
Aerts, R., Honnay, O., 2011. Forest restoration, biodiversity and ecosystem functioning.
BioMed Central Ecology 11, 29.
Anderson, D. R. 2008. Model based inference in the life sciences: a primer on evidence.
Springer Science+Business Media, LLC, New York, New York, USA.
Anderson, R.G., Canadell, J.G., Randerson, J.T., Jackson, R.B., Hungate, B.A.,
Baldocchi, D.D., Ban-Weiss, G.A., Bonan, G.B., Caldeira, K., Cao, L., others, 2010.
Biophysical considerations in forestry for climate protection. Frontiers in Ecology and
the Environment 9, 174–182.
89
Angelstam, P.K., 1998. Maintaining and restoring biodiversity in European boreal forests
by developing natural disturbance regimes. Journal of Vegetation Science 9, 593–602.
Ardia, D.R., Pérez, J.H., Clotfelter, E.D., 2010. Experimental cooling during incubation
leads to reduced innate immunity and body condition in nestling tree swallows.
Proceedings of the Royal Society B: Biological Sciences 277, 1881–1888.
Arnold, K.E., Ramsay, S.L., Henderson, L., Larcombe, S.D., 2010. Seasonal variation in
diet quality: antioxidants, invertebrates and blue tits Cyanistes caeruleus. Biological
Journal of the Linnean Society 99, 708–717.
Ashcroft, M.B., 2010. Identifying refugia from climate change. Journal of Biogeography
37, 1407–1413.
Aussenac, G., 2000. Interactions between forest stands and microclimate:
ecophysiological aspects and consequences for silviculture. Annals of Forest Science
57, 287–301.
Avery, M.I., Krebs, J.R., 1984. Temperature and foraging success of great tits Parus
major hunting for spiders. Ibis 126, 33–38.
Baker, T.P., Jordan, G.J., Steel, E.A., Fountain-Jones, N.M., Wardlaw, T.J., Baker, S.C.,
2014. Microclimate through space and time: Microclimatic variation at the edge of
regeneration forests over daily, yearly and decadal time scales. Forest Ecology and
Management 334, 174–184.
Balandier, P., Collet, C., Miller, J.H., Reynolds, P., Zedaker, S., 2006. Designing forest
vegetation management strategies based on the mechanisms and dynamics of crop tree
competition by neighbouring vegetation. Forestry 79, 3–27.
Barton, K.. 2015. MuMIn: Multi-Model Inference. R package version 1.14.0.
http://CRAN.R-project.org/package=MuMIn.
Bates, D., Maechler, M., Bolker, B.,Walker, S., 2015. Fitting Linear Mixed-Effects
Models Using lme4. Journal of Statistical Software 67, 1-48.
Becker, M.E., Weisberg, P.J., 2014. Synergistic effects of spring temperatures and land
cover on nest survival of urban birds. The Condor 117, 18–30.
Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W., Courchamp, F., 2012. Impacts of
climate change on the future of biodiversity. Ecology Letters 15, 365–377.
Belles-Isles, J.-C., Picman, J., 1986. Nesting losses and nest site preferences in House
Wrens. Condor 88, 483–486.
Bennett, A.F., Saunders, D.A., 2010. Habitat fragmentation and landscape change.
Conservation biology for all 93, 1544–1550.
Betts, M.G., Hagar, J.C., Rivers, J.W., Alexander, J.D., McGarigal, K., McComb, B.C.,
2010. Thresholds in forest bird occurrence as a function of the amount of early-seral
broadleaf forest at landscape scales. Ecological Applications 20, 2116–30.
Betts, M.G., Verschuyl, J., Giovanini, J., Stokely, T., Kroll, A.J., 2013. Initial
experimental effects of intensive forest management on avian abundance. Forest
Ecology and Management 310, 1036–1044.
Bonan, G., 2008a. Ecological climatology: concepts and applications. Cambridge
University Press, Cambridge, UK.
90
Bonan, G.B., 2008b. Forests and climate change: forcings, feedbacks, and the climate
benefits of forests. Science 320, 1444–1449.
Bowers, E.K., Hodges, C.J., Forsman, A.M., Vogel, L.A., Masters, B.S., Johnson, B.G.,
Johnson, L.S., Thompson, C.F., Sakaluk, S.K., 2014. Neonatal body condition,
immune responsiveness, and hematocrit predict longevity in a wild bird population.
Ecology 95, 3027–3034.
Brook, B.W., Sodhi, N.S., Bradshaw, C.J.A., 2008. Synergies among extinction drivers
under global change. Trends in Ecology and Evolution 23, 453–60.
Buckley, L.B., Hurlbert, A.H., Jetz, W., 2012. Broad-scale ecological implications of
ectothermy and endothermy in changing environments. Global Ecology and
Biogeography 21, 873–885.
Burnham, K. P, Anderson, D. R., 2002. Model selection and multimodel inference: a
practical information-theoretic approach. Second edition. Springer-Verlag New York,
Inc. New York, New York, USA.
Cade, B.S., 2015. Model averaging and muddled multimodel inference. Ecology 96,
2370-2382.
Cahill, A.E., Aiello-Lammens, M.E., Fisher-Reid, M.C., Hua, X., Karanewsky, C.J., Ryu,
H.Y., Sbeglia, G.C., Spagnolo, F., Waldron, J.B., Warsi, O., others, 2012. How does
climate change cause extinction? Proceedings of the Royal Society of London B:
Biological Sciences
Carnus, J.-M., Parrotta, J., Brockerhoff, E., Arbez, M., Jactel, H., Kremer, A., Lamb, D.,
OHara, K., Walters, B., 2006. Planted forests and biodiversity. Journal of Forestry 104,
65–77.
Chapin III, F.S., Matson, P.A., Vitousek, P., 2011. Principles of terrestrial ecosystem
ecology. Springer Science and Business Media, LLC. New York, New York, USA.
Chase, M.K., Nur, N., Geupel, G.R., Stouffer, P., 2005. Effects of weather and population
density on reproductive success and population dynamics in a song sparrow
(Melospiza melodia) population: a long-term study. The Auk 122, 571–592.
Checa, M.F., Rodriguez, J., Willmott, K.R., Liger, B., 2014. Microclimate variability
significantly affects the composition, abundance and phenology of butterfly
communities in a highly threatened neotropical dry forest. Florida Entomologist 97, 1–
13.
Chen, J., Saunders, S.C., Crow, T.R., Naiman, R.J., Brosofske, K.D., Mroz, G.D.,
Brookshire, B.L., Franklin, J.F., 1999. Microclimate in forest ecosystem and landscape
ecology. BioScience 49, 288–297.
Cockle, K.L., Bodrati, A., Lammertink, M., Martin, K., 2015. Cavity characteristics, but
not habitat, influence nest survival of cavity-nesting birds along a gradient of human
impact in the subtropical Atlantic Forest. Biological Conservation 184, 193–200.
Converse, S.J., Royle, J.A., Adler, P.H., Urbanek, R.P., Barzen, J.A., 2013. A
hierarchical nest survival model integrating incomplete temporally varying covariates.
Ecology and Evolution 3, 4439–4447.
Conway, C.J., Martin, T.E., 2000. Effects of ambient temperature on avian incubation
behavior. Behavioral Ecology 11, 178–188.
91
Cooke, S.J., Sack, L., Franklin, C.E., Farrell, A.P., Beardall, J., Wikelski, M., Chown,
S.L., 2013. What is conservation physiology? Perspectives on an increasingly
integrated and essential science. Conservation Physiology 1.
Cox, W.A., Thompson, F.R., Reidy, J.L., Faaborg, J., 2013a. Temperature can interact
with landscape factors to affect songbird productivity. Global change biology 19,
1064-1074.
Cox, W.A., Thompson III, F., Reidy, J., 2013b. The effects of temperature on nest
predation by mammals, birds, and snakes. The Auk 130, 784–790.
Cunningham, S.J., Martin, R.O., Hojem, C.L., Hockey, P.A., 2013. Temperatures in
excess of critical thresholds threaten nestling growth and survival in a rapidly-warming
arid savanna: a study of common fiscals. PloS One 8.
Dawson, T.P., Jackson, S.T., House, J.I., Prentice, I.C., Mace, G.M., 2011. Beyond
predictions: biodiversity conservation in a changing climate. Science 332, 53–8.
De Chazal, J., Rounsevell, M.D., 2009. Land-use and climate change within assessments
of biodiversity change: A review. Global Environmental Change 19, 306–315.
DeGregorio, B.A., Weatherhead, P.J., Sperry, J.H., 2014. Power lines, roads, and avian
nest survival: effects on predator identity and predation intensity. Ecology and
Evolution 4, 1589–600.
DeGregorio, B.A., Westervelt, J.D., Weatherhead, P.J., Sperry, J.H., 2015. Indirect effect
of climate change: Shifts in ratsnake behavior alter intensity and timing of avian nest
predation. Ecological Modelling 312, 239–246.
DellaSala, D.A., Bond, M.L., Hanson, C.T., Hutto, R.L., Odion, D.C., 2014. Complex
Early Seral Forests of the Sierra Nevada: What are They and How Can They Be
Managed for Ecological Integrity? Natural Areas Journal 34, 310–324.
Devine, W.D., Harrington, C.A., 2007. Influence of harvest residues and vegetation on
microsite soil and air temperatures in a young conifer plantation. Agricultural and
Forest Meteorology 145, 125–138.
Dobrowski, S.Z., 2011. A climatic basis for microrefugia: the influence of terrain on
climate. Global Change Biology 17, 1022–1035.
Du Plessis, K.L., Martin, R.O., Hockey, P.A., Cunningham, S.J., Ridley, A.R., 2012. The
costs of keeping cool in a warming world: implications of high temperatures for
foraging, thermoregulation and body condition of an arid-zone bird. Global Change
Biology 18, 3063–3070.
DuRant, S.E., Hopkins, W.A., Hawley, D.M., Hepp, G.R., 2012. Incubation temperature
affects multiple measures of immunocompetence in young wood ducks (Aix Sponsa).
Biology letters 8, 108–111.
DuRant, S.E., Hopkins, W.A., Hepp, G.R., Walters, J., 2013. Ecological, evolutionary,
and conservation implications of incubation temperature-dependent phenotypes in
birds. Biological Reviews.
Easton, W.E., Martin, K., 1998. The effect of vegetation management on breeding bird
communities in British Columbia. Ecological Applications 8, 1092–1103.
Easton, W.E., Martin, K., Powell, A., 2002. Effects of thinning and herbicide treatments
on nest-site selection by songbirds in young managed forests. The Auk 119, 685–694.
92
Edwards, E.K., Mitchell, N.J., Ridley, A.R., 2015. The impact of high temperatures on
foraging behaviour and body condition in the Western Australian Magpie Cracticus
tibicen dorsalis. Ostrich 86, 137–144.
Ellis, T.M., Betts, M.G., 2011. Bird abundance and diversity across a hardwood gradient
within early seral plantation forest. Forest Ecology and Management 261, 1372–1381.
Ellis, T.M., Kroll, A.J., Betts, M.G., 2012. Early seral hardwood vegetation increases
adult and fledgling bird abundance in Douglas-fir plantations of the Oregon Coast
Range, USA. Canadian Journal of Forest Research 42, 918–933.
Finch, D.M., 1989. Relationships of surrounding riparian habitat to nest-box use and
reproductive outcome in House Wrens. Condor 91, 848–859.
Fischer, J., Lindenmayer, D.B., 2007. Landscape modification and habitat fragmentation:
a synthesis. Global Ecology and Biogeography 16, 265–280.
Flesch, A.D., Hutto, R.L., van Leeuwen, W.J., Hartfield, K., Jacobs, S., 2015. Spatial,
Temporal, and Density-Dependent Components of Habitat Quality for a Desert Owl.
PloS one 10, e0119986.
Flueck, W.T., Smith-Flueck, J.A.M., 2006. Herbicides and forest biodiversity: an
alternative perspective. Wildlife Society Bulletin 34, 1472–1478.
Foley, J.A., DeFries, R., Asner, G.P., Barford, C., Bonan, G., Carpenter, S.R., Chapin,
F.S., Coe, M.T., Daily, G.C., Gibbs, H.K., others, 2005. Global consequences of land
use. Science 309, 570–574.
Fortier, J., Messier, C., 2006. Are chemical or mechanical treatments more sustainable
for forest vegetation management in the context of the TRIAD? The Forestry
Chronicle 82, 806–818.
Foster, D.R., Orwlg, D., McLachlan, J., 1996. Ecological and conservation insights from
reconstructive studies of temperate old-growth forests. Trends in Ecology and
Evolution 11, 419–424.
Fox, T.R., 2000. Sustained productivity in intensively managed forest plantations. Forest
Ecology and Management 138, 187–202.
Franklin, J., McCullough, P., Gray, C., 2000. Terrain variables used for predictive
mapping of vegetation communities in Southern California, in: Wilson, John P. and
Gallant, John C. (Ed.), Terrain Analysis: Principles and Applications. John Wiley &
Sons, Inc., pp. 331–353.
Franklin, J. F., Dyrness, C. T., 1988. Natural vegetation of Oregon and Washington. Gen.
Tech. Rep. PNW-8. USDA, USFS, Pacific Northwest Forest and Range Experiment
Station.
Friedl, M.A., McIver, D.K., Hodges, J.C., Zhang, X., Muchoney, D., Strahler, A.H.,
Woodcock, C.E., Gopal, S., Schneider, A., Cooper, A., others, 2002. Global land cover
mapping from MODIS: algorithms and early results. Remote Sensing of Environment
83, 287–302.
Gallinat, A.S., Primack, R.B., Wagner, D.L., 2015. Autumn, the neglected season in
climate change research. Trends in Ecology and Evolution 30, 169–176.
Geiger, R., Aron, R.H., Todhunter, P., 2003. The climate near the ground. Sixth edition.
Rowman & Littlefield Publishers, Oxford, UK.
93
Gelman, A, Goegebeur, Y., Tuerlinckx, F., Van Mechelen, I., 2000. Diagnostic checks
for discrete data regression models using posterior predictive simulations. Applied
Statistics 49: 247-268.
Gelman, A., Su, Y., 2015. Arm: data analysis using regression and
multilevel/hierarchicalmodels. R package version 1.8-5. http://CRAN.Rproject.org/package=arm.
Gillingham, P., Huntley, B., Kunin, W., Thomas, C., 2012. The effect of spatial
resolution on projected responses to climate warming. Diversity and Distributions 18,
990–1000.
Grant, T.A., Shaffer, T.L., Madden, E.M., Pietz, P.J., Johnson, D., 2005. Time-specific
variation in passerine nest survival: new insights into old questions. The Auk 122,
661–672.
Greene, W.H., 2005. Censored Data and Truncated Distributions. NYU Working Paper
No. EC-05-08.
Hagar, J.C., 2007. Wildlife species associated with non-coniferous vegetation in Pacific
Northwest conifer forests: A review. Forest Ecology and Management 246, 108–122.
Hagar, J.C., Dugger, K.M., Starkey, E.E., 2007. Arthropod prey of Wilson’s Warblers in
the understory of Douglas-fir forests. The Wilson Journal of Ornithology 119, 533–
546.
Hammond, P.C., Miller, J.C., 1998. Comparison of the biodiversity of Lepidoptera within
three forested ecosystems. Annals of the Entomological Society of America 91, 323–
328.
Hansen, J., Sato, M., Ruedy, R., Lo, K., Lea, D.W., Medina-Elizade, M., 2006. Global
temperature change. Proceedings of the National Academy of Sciences 103, 14288–
14293.
Hansen, M.C., Potapov, P.V., Moore, R., Hancher, M., Turubanova, S., Tyukavina, A.,
Thau, D., Stehman, S., Goetz, S., Loveland, T., others, 2013. High-resolution global
maps of 21st-century forest cover change. Science 342, 850–853.
Hansen, P.J., 2009. Effects of heat stress on mammalian reproduction. Philosophical
Transactions of the Royal Society B: Biological Sciences 364, 3341–3350.
Hardwick, S.R., Toumi, R., Pfeifer, M., Turner, E.C., Nilus, R., Ewers, R.M., 2015. The
relationship between leaf area index and microclimate in tropical forest and oil palm
plantation: Forest disturbance drives changes in microclimate. Agricultural and Forest
Meteorology 201, 187–195.
Heithecker, T.D., Halpern, C.B., 2006. Variation in microclimate associated with
dispersed-retention harvests in coniferous forests of western Washington. Forest
Ecology and Management 226, 60–71.
Helmuth, B., Kingsolver, J.G., Carrington, E., 2005. Biophysics, physiological ecology,
and climate change: does mechanism matter? Annual Review of Physiol. 67, 177–201.
Hobbs, R.J., Arico, S., Aronson, J., Baron, J.S., Bridgewater, P., Cramer, V.A., Epstein,
P.R., Ewel, J.J., Klink, C.A., Lugo, A.E., others, 2006. Novel ecosystems: theoretical
and management aspects of the new ecological world order. Global Ecology and
Biogeography 15, 1–7.
94
Hooke, R.L., Martin-Duque, J.F., Pedraza, J., 2012. Land transformation by humans: a
review. GSA today 22, 4–10.
Hosmer, D. W., Lemeshow, S., 2013. Applied logistic regression. Third edition. Wiley &
Sons Inc., Hoboken, New Jersey, USA.
Huang, C., Asner, G.P., Barger, N.N., 2012. Modeling regional variation in net primary
production of pinyon-juniper ecosystems. Ecological Modelling 227, 82–92.
Huey, R.B., 1991. Physiological consequences of habitat selection. American Naturalist
137, 91–115.
Hunt, P., 1998. Evidence from a landscape population model of the importance of early
successional habitat to the American Redstart. Conservation Biology 12, 1377–1389.
IPCC 2013. The physical science basis. Contribution of Working Group I to the Fifth
Assessment Report of the Intergovernmental Panel on Climate. Cambridge University
Press, Cambridge, United Kingdom and New York, NY, USA.
Jackson, R.B., Randerson, J.T., Canadell, J.G., Anderson, R.G., Avissar, R., Baldocchi,
D.D., Bonan, G.B., Caldeira, K., Diffenbaugh, N.S., Field, C.B., others, 2008.
Protecting climate with forests. Environmental Research Letters 3.
Jantz, S.M., Barker, B., Brooks, T.M., Chini, L.P., Huang, Q., Moore, R.M., Noel, J.,
Hurtt, G.C., 2015. Future habitat loss and extinctions driven by land-use change in
biodiversity hotspots under four scenarios of climate-change mitigation. Conservation
Biology 29, 1122–1131.
Johnson, S.L., 2014. House Wren (Troglodytes aedon), The Birds of North America
Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the Birds
of North America Online:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/380doi:10.2173/bn
a.380
Johnson, L.S., Kermott, L.H., Lein, M.R., 1993. The cost of polygyny in the house wren
Troglodytes aedon. Journal of Animal Ecology 669–682.
Kelly, A.E., Goulden, M.L., 2008. Rapid shifts in plant distribution with recent climate
change. Proceedings of the National Acadademy of Sciences 105, 11823–6.
Knezezich, C. A., 1982. Soil survey of Polk County, Oregon. United States Department
of Agriculture, Soil Conservation Service, Washington DC, USA.
King, D.I., Schlossberg, S., 2014. Synthesis of the conservation value of the earlysuccessional stage in forests of eastern North America. Forest Ecology and
Management 324, 186–195.
Kozma, J.M., Kroll, A.J., 2010. Nest survival of Western Bluebirds using tree cavities in
managed ponderosa pine forests of central Washington. The Condor 112, 87–95.
Labocha, M.K., Hayes, J.P., 2012. Morphometric indices of body condition in birds: a
review. Journal of Ornithology 153, 1–22.
Lambin, E.F., Meyfroidt, P., 2011. Global land use change, economic globalization, and
the looming land scarcity. Proceedings of the National Academy of Sciences 108,
3465–3472.
Lampila, P., Mӧnkkӧnen, M., Desrochers, A., 2005. Demographic responses by birds to
forest fragmentation. Conservation Biology 19, 1537–1546.
95
Lautenschlager, R., Sullivan, T.P., 2002. Effects of herbicide treatments on biotic
components in regenerating northern forests. The Forestry Chronicle 78, 695–731.
Lehtinen, R.M., Ramanamanjato, J.B., Raveloarison, J.G., 2003. Edge effects and
extinction proneness in a herpetofauna from Madagascar. Biodiversity and
Conservation 12, 1357–1370.
Li, P., Martin, T.E., 1991. Nest-site selection and nesting success of cavity-nesting birds
in high elevation forest drainages. The Auk 405–418.
Litvaitis, J.A., 1993. Response of early successional vertebrates to historic changes in
land use. Conservation Biology 7, 866–873.
Ma, S., Concilio, A., Oakley, B., North, M., Chen, J., 2010. Spatial variability in
microclimate in a mixed-conifer forest before and after thinning and burning
treatments. Forest ecology and management 259, 904–915.
Mantyka-pringle, C.S., Martin, T.G., Rhodes, J.R., 2012. Interactions between climate
and habitat loss effects on biodiversity: a systematic review and meta-analysis. Global
Change Biology 18, 1239–1252.
Martin, T.E., 1993a. Nest predation among vegetation layers and habitat types: revising
the dogmas. American Naturalist 897–913.
Martin, T.E., 1993b. Nest predation and nest sites. BioScience 43, 523–532.
Martin, T.E., 1987. Food as a limit on breeding birds: a life-history perspective. Annual
review of ecology and systematics 18, 453–487.
Martin, T.E., Geupel, G.R., 1993. Nest-Monitoring Plots: Methods for Locating Nests
and Monitoring Success Journal of Field Ornithology 507–519.
McCabe, R.A., 1965. Nest construction by house wrens. The Condor 67, 229–234.
McCarty, J.P., 2001. Ecological consequences of recent climate change. Conservation
biology 15, 320–331.
McComb, W.C., Noble, R.E., 1981. Microclimates of nest boxes and natural cavities in
bottomland hardwoods. The Journal of Wildlife management 45, 284–289.
McGarigal, K., Cushman, S.A., 2002. Comparative evaluation of experimental
approaches to the study of habitat fragmentation effects. Ecological applications 12,
335–345.
Miller, J.C., Hammond, P.C., Ross, D.N., 2003. Distribution and functional roles of rare
and uncommon moths (Lepidoptera: Noctuidae: Plusiinae) across a coniferous forest
landscape. Annals of the Entomological Society of America 96, 847–855.
Miller, K.V., Miller, J.H., 2004. Forestry herbicide influences on biodiversity and
wildlife habitat in southern forests. Wildlife Society Bulletin 32, 1049–1060.
Monastersky, R., 2014. Biodiversity: Life-a status report. Nature 516, 159–161.
Murphy, G.E., Sutton, W.R., Hill, D., Chambers, C., Creel, D., Binkley, C., New, D.,
2005. Economics of intensively managed forest plantations in the Pacific Northwest.
Journal of forestry 103, 78–82.
Murphy, M.T., 1985. Nestling Eastern Kingbird growth: effects of initial size and
ambient temperature. Ecology 66, 162–170.
Newlon, K.R., Saab, V.A., 2011. Nest-site selection and nest survival of Lewis’s
Woodpecker in aspen riparian woodlands. The Condor 113, 183–193.
96
O’Connor, R.J. 1984. The growth and development of birds. John Wiley and Sons, New
York, New York, USA.
Oliver, T.H., Morecroft, M.D., 2014. Interactions between climate change and land use
change on biodiversity: attribution problems, risks, and opportunities. Wiley
Interdisciplinary Reviews: Climate Change 5, 317–335.
Paquette, A., Messier, C., 2009. The role of plantations in managing the world’s forests
in the Anthropocene. Frontiers in Ecology and the Environment 8, 27–34.
Paquette, S.R., Pelletier, F., Garant, D., Bélisle, M., 2014. Severe recent decrease of adult
body mass in a declining insectivorous bird population. Proceedings of the Royal
Society of London B: Biological Sciences 281.
Parker, W.C., Pitt, D.G., Morneault, A.E., 2012. Influence of herbaceous and woody
vegetation control on seedling microclimate, leaf gas exchange, water status, and
nutrient relations of Pinus strobus L. seedlings planted in a shelterwood. Forest
Ecology and Management 271, 104–114.
Parmesan, C., 2006. Ecological and evolutionary responses to recent climate change.
Annual Review of Ecology, Evolution, and Systematics 637–669.
Pauli, H., Gottfried, M., Dullinger, S., Abdaladze, O., Akhalkatsi, M., Benito Alonso,
J.L., Coldea, G., Dick, J., Erschbamer, B., Fernández Calzado, R., Ghosn, D., Holten,
J.I., Kanka, R., Kazakis, G., Kollár, J., Larsson, P., Moiseev, P., Moiseev, D., Molau,
U., Molero Mesa, J., Nagy, L., Pelino, G., Puşcaş, M., Rossi, G., Stanisci, A.,
Syverhuset, A.O., Theurillat, J.-P., Tomaselli, M., Unterluggauer, P., Villar, L., Vittoz,
P., Grabherr, G., 2012. Recent plant diversity changes on Europe’s mountain summits.
Science 336, 353–355.
Pearson, R.G., Dawson, T.P., 2003. Predicting the impacts of climate change on the
distribution of species: are bioclimate envelope models useful? Global ecology and
biogeography 12, 361–371.
Pereira, H.M., Navarro, L.M., Martins, I.S., 2012. Global biodiversity change: the bad,
the good, and the unknown.
Pérez, J.H., Ardia, D.R., Chad, E.K., Clotfelter, E.D., 2008. Experimental heating reveals
nest temperature affects nestling condition in tree swallows (Tachycineta bicolor).
Biology Letters 4, 468–471.
Perrins, C., 1970. The timing of birds ‘breeding seasons. Ibis 112, 242–255.
Perry, D.A., Oren, R., Hart, S.C., 2008. Forest ecosystems. Johns Hopkins University
Press, Baltimore, Maryland, USA.
Pielke, R.A., Pitman, A., Niyogi, D., Mahmood, R., McAlpine, C., Hossain, F. and
G.K.K., Nair, U., Betts, R., Fall, S., et al., 2011. Land use/land cover changes and
climate: modeling analysis and
observational evidence. Wiley Interdisciplinary
Reviews: Climate Change 2, 828–850.
Pinheiro J., Bates, D., DebRoy S., Sarkar, D., and R Core Team, 2015. nlme: Linear and
Nonlinear Mixed Effects Models. R package version 3.1-122.
Pipoly, I., Bókony, V., Seress, G., Szabó, K., Liker, A., 2013. Effects of Extreme
Weather on Reproductive Success in a Temperate-Breeding Songbird. PloS one 8.
97
Potapov, P., Yaroshenko, A., Turubanova, S., Dubinin, M., Laestadius, L., Thies, C.,
Aksenov, D., Egorov, A., Yesipova, Y., Glushkov, I., Karpachevskiy, M., Kostikova,
A., Manisha, A., Tsybikova, E., Zhuravleva, I., 2008. Mapping the World’s Intact
Forest Landscapes by Remote Sensing. Ecology and Society 13.
Potter, K.A., Woods, H.A., Pincebourde, S., 2013. Microclimatic challenges in global
change biology. Global change biology 19, 2932–2939.
Proe, M., Griffiths, J., McKay, H., 2001. Effect of whole-tree harvesting on microclimate
during establishment of second rotation forestry. Agricultural and Forest Meteorology
110, 141–154.
Ramsey, F. L., Schafer, D. W., 2013. The statistical sleuth: a course in methods of data
analysis. Third edition. Brooks/Cole, Cengage Learning, Boston, MA, USA.
Reid, J.M., Monaghan, P., Ruxton, G.D., 2000. Resource allocation between reproductive
phases: the importance of thermal conditions in determining the cost of incubation.
Proceedings of the Royal Society of London 267, 37–41.
Root, T.L., Price, J.T., Hall, K.R., Schneider, S.H., Rosenzweig, C., Pounds, J.A., 2003.
Fingerprints of global warming on wild animals and plants. Nature 421, 57–60.
Rudel, T.K., Coomes, O.T., Moran, E., Achard, F., Angelsen, A., Xu, J., Lambin, E.,
2005. Forest transitions: towards a global understanding of land use change. Global
environmental change 15, 23–31.
Salaberria, C., Celis, P., López-Rull, I., Gil, D., 2013. Effects of temperature and nest
heat exposure on nestling growth, dehydration and survival in a Mediterranean holenesting passerine. Ibis 156, 265-275.
Sauer, J. R., Hines, J. E., Fallon, J. E., Pardieck, K. L., Ziolkowski, D. J., Jr., Link, W. A.,
2015. The North American Breeding Bird Survey, Results and Analysis 1966 - 2013.
Version 01.30.2015 USGS Patuxent Wildlife Research Center, Laurel, MD
Saunders, D.A., Hobbs, R.J., Margules, C.R., 1991. Biological consequences of
ecosystem fragmentation: a review. Conservation biology 5, 18–32.
Schamber, J.L., Esler, D., Flint, P.L., 2009. Evaluating the validity of using unverified
indices of body condition. Journal of Avian Biology 40, 49–56.
Schmidt, K.A., 1999. Foraging theory as a conceptual framework for studying nest
predation. Oikos 85, 151–160.
Schmidt-Nielsen, K. 1997. Animal physiology: adaptation and environment. Cambridge
University Press, New York, New York, USA.
Schwagmeyer, P., Mock, D.W., 2008. Parental provisioning and offspring fitness: size
matters. Animal Behaviour 75, 291–298.
Sears, M.W., Raskin, E., Angilletta, M.J., 2011. The world is not flat: defining relevant
thermal landscapes in the context of climate change. Integrative and Comparative
Biology, 1-10.
Selwood, K.E., McGeoch, M.A., Mac Nally, R., 2014. The effects of climate change and
land-use change on demographic rates and population viability. Biological Reviews.
Seto, K.C., Fragkias, M., Güneralp, B., Reilly, M.K., 2011. A meta-analysis of global
urban land expansion. PLoS One 6.
98
Shaffer, T.L., 2004. A unified approach to analyzing nest success. The Auk 121, 526–
540.
Shaffer, T.L., Thompson, F.R., 2007. Making meaningful estimates of nest survival with
model-based methods. Studies in Avian Biology 34, 84–95.
Shepard, J.P., Creighton, J., Duzan, H., 2004. Forestry herbicides in the United States: an
overview. Wildlife Society Bulletin 32, 1020–1027.
Sieving, K.E., Willson, M.F., 1998. Nest predation and avian species diversity in
northwestern forest understory. Ecology 79, 2391–2402.
Sloan, S., Sayer, J.A., 2015. Forest Resources Assessment of 2015 shows positive global
trends but forest loss and degradation persist in poor tropical countries. Forest Ecology
and Management 352, 134–145.
Spies, T.A., 2004. Ecological concepts and diversity of old-growth forests. Journal of
Forestry 102, 14–20.
Spies, T.A., Johnson, K.N., 2007. Projecting Forest Policy and Management Effects
across Ownerships in Coastal Oregon. Ecological Applications 17, 3–4.
Steidl, R.J., Hayes, J.P., Schauber, E., 1997. Statistical power analysis in wildlife
research. The Journal of Wildlife Management 61, 270–279.
Suggitt, A.J., Gillingham, P.K., Hill, J.K., Huntley, B., Kunin, W.E., Roy, D.B., Thomas,
C.D., 2011. Habitat microclimates drive fine-scale variation in extreme temperatures.
Oikos 120, 1–8.
Swanson, M.E., Franklin, J.F., Beschta, R.L., Crisafulli, C.M., DellaSala, D.A., Hutto,
R.L., Lindenmayer, D.B., Swanson, F.J., 2010. The forgotten stage of forest
succession: early-successional ecosystems on forest sites. Frontiers in Ecology and the
Environment 9, 117–125.
Sutherland, W.J., Newton, I., Green, R.E., 2004. Bird ecology and conservation: a
handbook of techniques. Oxford University Press Inc., New York, New York, USA.
Tatum, V.L., 2004. Toxicity, transport, and fate of forest herbicides. Wildlife Society
Bulletin 32, 1042–1048.
Taylor, G.H., Hannan, C., 1999. The climate of Oregon: from rain forest to desert.
Oregon State University Press, Corvallis, USA.
Taylor, R.L., Maxwell, B.D., Boik, R.J., 2006. Indirect effects of herbicides on bird food
resources and beneficial arthropods. Agriculture, ecosystems, and environment 116,
157–164.
Thomas, J.W., Franklin, J.F., Gordon, J., Johnson, K.N., 2006. The Northwest Forest
Plan: origins, components, implementation experience, and suggestions for change.
Conservation Biology 20, 277–287.
Thompson, F.R., 2007. Factors affecting nest predation on forest songbirds in North
America. Ibis 149, 98–109.
Tilman, D., Balzer, C., Hill, J., Befort, B.L., 2011. Global food demand and the
sustainable intensification of agriculture. Proceedings of the National Academy of
Sciences 108, 20260–20264.
99
Tscharntke, T., Clough, Y., Wanger, T.C., Jackson, L., Motzke, I., Perfecto, I.,
Vandermeer, J., Whitbread, A., 2012. Global food security, biodiversity conservation
and the future of agricultural intensification. Biological Conservation 151, 53–59.
Varner, J., Dearing, M.D., 2014. The Importance of Biologically Relevant Microclimates
in Habitat Suitability Assessments. PloS one 9.
Vié, J.-C., Hilton-Taylor, C., Stuart, S.N., 2009. Wildlife in a changing world: an analysis
of the 2008 IUCN Red List of threatened species. IUCN.
Von Arx, G., Dobbertin, M., Rebetez, M., 2012. Spatio-temporal effects of forest canopy
on understory microclimate in a long-term experiment in Switzerland. Agricultural and
Forest Meteorology 166, 144–155.
Wada, H., Kriengwatana, B., Allen, N., Schmidt, K.L., Soma, K.K., MacDougallShackleton, S.A., 2015. Transient and permanent effects of suboptimal incubation
temperatures on growth, metabolic rate, immune function, and adrenocortical
responses in zebra finches. The Journal of Experimental Biology 218, 2847-2855.
Wagner, R.G., Little, K.M., Richardson, B., Mcnabb, K., 2006. The role of vegetation
management for enhancing productivity of the world’s forests. Forestry 79, 57–79.
Waring, R.H., Franklin, J.F., 1979. Evergreen coniferous forests of the Pacific Northwest.
Science 204, 1380–1386.
Webb, D., 1987. Thermal tolerance of avian embryos: a review. Condor 89, 874–898.
Whiteman, C.D., Zhong, S., Shaw, W.J., Hubbe, J.M., Bian, X., Mittelstadt, J., 2001.
Cold pools in the Columbia Basin. Weather and forecasting 16, 432–447.
Williams, C.M., Henry, H.A., Sinclair, B.J., 2015. Cold truths: how winter drives
responses of terrestrial organisms to climate change. Biological Reviews 90, 214–235.
Willson, M.F., Comet, T.A., 1996a. Bird communities of northern forests: patterns of
diversity and abundance. Condor 98, 337–349.
Willson, M.F., Comet, T.A., 1996b. Bird communities of northern forests: ecological
correlates of diversity and abundance in the understory. Condor 98, 350–362.
Winkler, D.W., Dunn, P.O., McCulloch, C.E., 2002. Predicting the effects of climate
change on avian life-history traits. Proceedings of the National Academy of Sciences
99, 13595–13599.
Xu, M., Chen, J., Brookshire, B.L., 1997. Temperature and its variability in oak forests in
the southeastern Missouri Ozarks. Climate Research 8, 209–223.
Zach, R., 1982. Nestling house wrens: weight and feather growth. Canadian Journal of
Zoology 60, 1417–1425.
Zhao, K., Jackson, R.B., 2014. Biophysical forcings of land-use changes from potential
forestry activities in North America. Ecological Monographs 84, 329–353.
Zurr, A. F., Ieno, E. N., Walker, N.J., Saveliev A. A., and G. M. Smith. 2007. Mixed
effects models and extensions with R. Springer Science+Business Media LLC, New
York, New York, USA.
100
APPENDICES
101
APPENDIX A: HERBICIDE PRESCRIPTIONS
Study Treatment
Year post-harvest & Season
Prescription
Chemical and Quantity/Acre
Control
Light
1-10
2 (spring)
Planting (year 2)
Herbaceous
None
2.66 lbs. Velpar (hexazinone)
32 oz 2-4-D (2,4-dichlorophenoxy
acetic acid)
3 (late summer)
Woody vegetation
control
1.125 qt Accord (glyphosate)
20 oz Garlon (triclopyr)
Moderate
1 (late summer)
2 (spring)
Site preparation
Herbaceous control
1.5 oz Escort (metsulfuron methyl)
3 qts Accord (glyphosate)
24 oz Chopper (imazapyr)
3 oz Oust (sulfometuron methyl &
metsulfuron methyl)
24 oz MSO (methylated seed oil)
2.66 lbs. Velpar (hexazinone)
32 oz 2-4-D (2,4-dichlorophenoxy
acetic acid)
3 (late summer)
Woody vegetation
control
1.5 qt Accord (glyphosate)
20 oz Garlon (triclopyr)
4 (late summer)
Intensive
1 (late summer)
2-10 (spring)
Acer macrophyllum
sprout control (as
necessary)
Imazapyr
Acer macrophyllum
sprout control follow-up
(if necessary)
Imazapyr
Site preparation
Herbaceous control
1.5 oz Escort (metsulfuron methyl)
3 qts Accord (glyphosate)
24 oz Chopper (imazapyr)
3 oz Oust (sulfometuron methyl &
metsulfuron methyl)
24 oz MSO (methylated seed oil)
2.66 lbs. Velpar (hexazinone)
32 oz 2-4-D (2,4-dichlorophenoxy
acetic acid)
102
Woody vegetation
control (Annual review
with backpack treatments
as necessary).
1.125 qt Accord (glyphosate)
3-10 (late summer)
20 oz Garlon (triclopyr)
3-10 (late summer)
Acer macrophyllum
sprout control and
follow-up (as necessary)
Imazapyr