Vous êtes sur la page 1sur 96

ANDEAN TEMPERATE FOREST OWLS: DETECTABILITY, HABITAT

RELATIONSHIPS AND RELIABILITY AS BIODIVERSITY SURROGATES

by

Jos Toms Ibarra Eliessetch

B.Sc., Pontificia Universidad Catlica de Chile (Agricultural Engineering), 2005


M.Sc., Pontificia Universidad Catlica de Chile (Conservation and Wildlife Management),
2007
M.Sc., University of Kent (Environmental Anthropology), 2010

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF


THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

in

THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES


(Forestry)

THE UNIVERSITY OF BRITISH COLUMBIA


(Vancouver)

December 2014

Jos Toms Ibarra Eliessetch, 2014


Abstract

South American temperate forests are globally exceptional for their high concentration of
endemic species. These forests are among the most endangered ecosystems on Earth
because nearly 70% of them have been lost. Current knowledge of most Neotropical forest
owls is limited. I studied how environmental and habitat conditions might influence the
ecology of sympatric forest owls, and evaluated whether owls can be used as surrogates for
temperate forest biodiversity. Specifically, I examined (i) factors associated with the
detectability, (ii) occurrence rates and habitat-resource utilization across spatial scales, and
(iii) surrogacy reliability of the habitat-specialist rufous-legged owl (Strix rufipes) and the
habitat-generalist austral pygmy-owl (Glaucidium nana) in southern Chile. During 2011-
2013, I conducted 1,145 owl surveys, 505 vegetation surveys and 505 avian point-transects
across 101 sites comprising a range of conditions from degraded habitat to structurally
complex old-growth forest stands. I recorded 292 detections of S. rufipes and 334
detections of G. nana. Detectability for both owls increased with greater moonlight and
decreased with environmental noise, and greater wind speed decreased detectability for G.
nana. Detection of both species was positively correlated with the detection of the other
species. For S. rufipes, occurrence probability ranged from 0.05-1 across sites, and was
positively associated with bamboo density and the variability in diameter at breast height of
trees (multi-aged forests). For G. nana, occurrence ranged from 0.67-0.98, but no habitat
characteristic was related to this species occurrence. Relative to G. nana, S. rufipes had
lower total resource utilization, but achieved similar peak occurrence for resources related
to stand-level forest complexity and forest homogeneity at the landscape scale. I found that
only S. rufipes was a reliable surrogate for all avian biodiversity measures, including
endemism and functional diversity. With increasing occurrence of habitat-specialist owls,
the density of target specialized biodiversity (guilds and communities) increased non-
linearly and peaked at the least degraded sites. This specialist aggregation was driven by
forest-stand structural complexity. Forest management practices that maintain multi-aged
stands with large trees and high bamboo cover will benefit both owl species, and likely will
benefit vulnerable endemic species and specialized avian communities in temperate forests.

ii
Preface

My thesis is written in a manuscript-based format. Chapters 2 through 4 represent


independent chapters that have been or will be submitted in a similar format, except that I
moved the general material on study area, study species and methods, common to the data
sections, to Chapter 1. I took the lead in developing the research framework, conducting the
research, data analysis and manuscript preparation for Chapters 2 to 4.
A version of Chapter 2 has been published: Ibarra, J.T., Martin, K., Altamirano,
T.A., Vargas, F.H., Bonacic, C., 2014. Factors associated with the detectability of owls in
South American temperate forests: implications for nocturnal raptor monitoring. J. Wildl.
Manage. 78, 1078-1086.
A version of Chapter 3 has also been published: Ibarra, J.T., Martin, K., Drever,
M.C., Vergara, G., 2014. Occurrence patterns and niche relationships of sympatric owls in
South American temperate forests: a multi-scale approach. Forest Ecol. Manag. 331, 281-
291.

iii
Table of contents

Abstract ................................................................................................................................. ii
Preface .................................................................................................................................. iii
Table of contents .................................................................................................................. iv
List of tables ......................................................................................................................... vi
List of figures ..................................................................................................................... viii
Acknowledgements ............................................................................................................... x
Dedication ............................................................................................................................ xii
Chapter 1. General introduction and thesis overview ...................................................... 1
1.1. Background ................................................................................................................ 1
1.1.1. Endangered ecosystems and habitat-specialist owls ............................................ 1
1.1.2. Detectability of forest owls .................................................................................. 2
1.1.3. Owl niches and habitat suitability across spatial scales ....................................... 3
1.1.4. Forest owls as reliable biodiversity surrogates .................................................... 4
1.2. Thesis objectives ........................................................................................................ 5
1.3. Study area................................................................................................................... 5
1.4. Study species .............................................................................................................. 6
1.5. General field methods ............................................................................................... 7
1.5.1. Study design and allocation of survey effort ........................................................ 7
1.5.2. Nocturnal raptor surveys ...................................................................................... 7
1.6. Thesis overview .......................................................................................................... 8
Chapter 2. Detectability of owls in South American temperate forests ........................ 10
2.1. Introduction ................................................................................................................. 10
2.2. Methods ........................................................................................................................ 11
2.2.1. Field methods ......................................................................................................... 11
2.2.2. Data analysis ........................................................................................................... 12
2.3. Results ........................................................................................................................... 13
2.3.1. Occurrence and detectability of rufous-legged owls .............................................. 13
2.3.2. Occurrence and detectability of austral pygmy-owls ............................................. 14
2.4. Discussion ..................................................................................................................... 14
2.4.1. Recommendations for owl monitoring ................................................................... 17
2.5. Conclusions .................................................................................................................. 18
Chapter 3. Occurrence patterns of specialist and generalist owls across spatial scales
in South American temperate forests ............................................................................... 26
3.1. Introduction ................................................................................................................. 26
3.2. Methods ........................................................................................................................ 28
3.2.1. Field methods ......................................................................................................... 28

iv
3.2.2. Data analysis ........................................................................................................... 30
3.3. Results ........................................................................................................................... 31
3.3.1. Habitat suitability for owls ..................................................................................... 32
3.3.2. Resource utilization and peak performance by owls .............................................. 33
3.4. Discussion ..................................................................................................................... 33
3.4.1. Habitat suitability across spatial scales .................................................................. 34
3.4.2. Niche width of forest owls ..................................................................................... 36
3.4.3. Recommendations for management ....................................................................... 37
3.5. Conclusions .................................................................................................................. 38
Chapter 4. Reliability of owls as surrogates for biodiversity in South American
temperate forests................................................................................................................. 47
4.1. Introduction ................................................................................................................. 47
4.2. Methods ........................................................................................................................ 49
4.2.1. Field methods ......................................................................................................... 49
4.2.2. Data analysis ........................................................................................................... 51
4.3. Results ........................................................................................................................... 53
4.3.1. Spatial relationships: owls and target biodiversity ................................................. 53
4.3.2. Ecological mechanisms: habitat correlates for owls and target biodiversity.......... 54
4.4. Discussion ..................................................................................................................... 55
4.4.1. Evidence for a reliable surrogacy relationship ....................................................... 55
4.4.2. Untangling ecological mechanisms ........................................................................ 56
Chapter 5. General discussion and conclusions ............................................................... 67
5.1. Thesis summary ........................................................................................................... 67
5.2. Future directions ......................................................................................................... 69
5.2.1. Scaling up owl-habitat relationships: from individuals to landscapes over time ... 69
5.2.2. Understanding surrogacy relationships for functional biodiversity conservation
across spatio-temporal scales............................................................................................ 70
Bibliography ........................................................................................................................ 76

v
List of tables

Table 2.1. Candidate predictors of detectability for forest owls in Andean temperate forests,
2011-2013..21
Table 2.2. Model selection results for estimating probability of occupancy () and detection
(p) of Strix rufipes and Glaucidium nana in Andean temperate forests, 2011-
2013. Site-specific covariates consisted of elevation in meters/1,000 (elev),
and whether the site was 500 m within a protected area or not (Pa). Survey
specific covariates consisted of moonlight (Ml), environmental noise (noise),
whether the other owl species was detected at the unit for the specific survey
(owl), wind speed (wind) and number of days since start of surveys
(day)..23
Table 2.3. Competing models ( Akaikes Information Criterion 2.0) predicting
probability of occupancy () and detection (p) of Strix rufipes and
Glaucidium nana in Andean temperate forests, 2011-2013. The estimated
model coefficients and the lower and upper confidence intervals (LCI and
UCI) are also shown..24
Table 3.1. Stand- and landscape-level resources used to evaluate habitat associations of
forest owls in Andean temperate forests...43
Table 3.2. Model selection statistics based on Akaikes Information Criterion (AIC) for
estimating probability of occurrence () and detection (p) of two owl species
(a) Strix rufipes and (b) Glaucidium nana in Andean temperate forests. Only
the top model set with AIC values < 4 are shown. Parameter estimates are
listed in the order of variable under Model structure column, and beta
estimates in bold font have 95% confidence intervals that do not overlap
0.44
Table 3.3. Mean (SD) values of habitat resources associated with probabilities of
occurrence categorized as low (0 - 0.33), moderate (0.34 - 0.66) and high

vi
(0.67 - 1) for Strix rufipes in Andean temperate forests, based on model-
averaged predictions.46
Table 4.1. Avian species with their geographical and ecological attributes, and stand-level
covariates associated with the density (D) of bird species in Andean
temperate forests, according to model selection statistics based on Akaikes
Information Criterion (AIC). Parameter estimates [SE] for covariates present
in the top model set with AIC values < 2 and with estimates of their 95%
confidence intervals that do not overlap 0, are shown. + and - indicate the
direction of the relation.61
Table 4.2. Ranking of models relating measures of avian diversity and owl probabilities of
occurrence () in Andean temperate forests.64
Table 4.3. Estimated mean [SE] for (i) species richness/site, (ii) density (individuals/ ha x
site) of different diversity measures and (iii) community specialization
index/site, associated with low (0 - 0.33), moderate (0.34 - 0.66) and high
(0.67 - 1) probabilities of occurrence () for Strix rufipes in Andean
temperate forests, based on model-averaged predictions.66

vii
List of figures

Figure 1.1. Distribution of 101 sites (red dots) used to study rufous-legged owls (Strix
rufipes) and austral pygmy-owls (Glaucidium nana) in a mountainous
landscape in Andean temperate forests of the La Araucana Region (39S),
southern Chile. The main map represents the Villarrica watershed and it
shows forests (green), shrublands and grasslands (yellow), alpine areas
(grey), lakes (light blue), wetlands (dark blue), urban areas (black) and snow
and glaciers
(white).9
Figure 2.1. Predicted probabilities of occurrence () and 95% confidence intervals for Strix
rufipes in temperate forests of Chile (2011-2013), in relation to elevation
(meters above sea level) when (a) Pa = 0 (sites located outside of a protected
area) and when (b) Pa = 1 (sites located at least 500 m within a protected
area), using the best model for this owl species19
Figure 2.2. Predicted probabilities of detection (p) and 95% confidence intervals for Strix
rufipes and Glaucidium nana in the temperate forests of Chile (2011-2013),
in relation to moonlight index, number of days since start of surveys and
wind speed (m/s), using the best model for each species. Moonlight index
refers to the amount of light available reduced by the proportion of sky
obscured by clouds. Ml = (1 - cloud cover) moon phase/100...20
Fig. 3.1. (a) Specialized species (continuous line) have smaller niche widths than
generalized species (dashed line) across resource gradients. Specialists can
reach either a (b) higher level of performance (i.e. traditional model of
relative niche width between specialists and generalists) or (c) similar level
of performance (i.e. alternative model of relative niche width) than
generalists, under a subset of resources that are relatively stable.39
Fig. 3.2. Averaged predictions of occurrence for (a) Strix rufipes and (b) Glaucidium nana
in the Villarrica watershed of the La Araucana Region, Chile. Red depicts

viii
areas of higher habitat suitability whereas yellow represents areas of lower
habitat suitability or probability of occurrence. Alpine areas (grey), large
lakes (blue) and boundaries of public protected areas (green), are
shown40
Fig. 3.3. Predicted probabilities of occurrence of Strix rufipes and Glaucidium nana in
Andean temperate forests (2011-2013), in relation to environmental
resources. The curves are representative of models developed for each owl
using only the corresponding resource..41
Fig. 3.4. (a) Total resource utilization (i.e. average area under the curve SE) and (b)
predicted peak performance SE, for Strix rufipes (grey bars) and
Glaucidium nana (white bars) in relation to habitat niche resources42
Fig. 4.1. Relationship between probabilities of occurrence () of habitat-specialist owls
Strix rufipes and (i) species richness/site, (ii) density (individuals/ha x site)
and (iii) community specialization index/site, in Andean temperate forests,
2011-2013..60

ix
Acknowledgements

I would have not been able to finish my dissertation project without the friendly
assistance, cooperation and encouragement of many people. I would like to thank my
supervisor Kathy Martin for her continuous guidance and advice during all the phases of
my PhD. Without her support this project would never have happened. Mark C. Drever and
Peter Marshall were excellent mentors. Many of the ideas behind this research arose from
discussions with my aforementioned committee members, and also with Toms A.
Altamirano, Antonia Barreau, Cristin Bonacic, Alberto Dittborn, Nicols Glvez, Jill
Jankowski, Richard Schuster, F. Hernn Vargas, Alejandra Vermehren and Gonzalo
Vergara. All their advice, experience and encouragement were more helpful than they can
possibly imagine.
I am especially grateful to the following people who helped me with my field
project in southern Chile: Alejandra Vermehren, Toms A. Altamirano, Antonia Barreau,
Cristin Bustos, Diego Cox, Consuelo Glvez, Catalina Zumaeta, Fernanda Soffia, Carolina
Yez, Ismael Horta, Julin Caviedes, Klaus Kremer, Lina Forero, Mara Ignacia Ibarra,
Mariano de la Maza, Mara Jos Martn, Anglica Pinochet, Ricardo Porfiri, Gonzalo
Vergara, Paloma Corvaln, Patricio Bahamondes, Santiago Prez de Castro, Teresa
Honorato, Matas Acevedo, Mara Cecilia Rivera, Isabel Mujica and Francisco Reygadas.
All of them were enthusiastic, congenial and extremely helpful while conducting never-
ending owl nocturnal surveys, vegetation plots and early morning bird counts.
Funding for this project was provided by The Peregrine Fund, Rufford Small Grants
for Nature Conservation, Chilean Ministry of the Environment (FPA 9-I-009-12),
Cleveland Metroparks Zoo, Cleveland Zoological Society and Environment Canada. I
received a post-graduate scholarship from Comisin Nacional de Investigacin Cientfica y
Tecnolgica de Chile (CONICYT) and the following research awards from UBC: NSERC
CREATE Program in Biodiversity Research, Faculty of Forestry Dissertation
Completion Fellowship, Donald S. McPhee Fellowship Award, Mary and David Macaree
Fellowship Award, UBC Graduate Student International Research Mobility Award, and the

x
Werner and Hildegard Hesse Fellowship in Ornithology. Institutional support throughout
this project was provided by the Centre of Local Development (CEDEL-Campus Villarrica;
especially by Gonzalo Valdivieso and Antonio Hargreaves) and Fauna Australis Wildlife
Laboratory, both from the Pontificia Universidad Catlica de Chile. I thank all of them for
their support and belief in this project.
The Chilean Forest Service (CONAF), Santuario El Cai (Guas-Cae: Roberto
Sanhueza y Manuel Venegas), Kawelluco Sanctuary, Francisco Poblete, Ricardo
Timmerman, Mnica Sabugal, Cristina Dlano, Jerry Laker (Kodkod: Lugar de
Encuentros) and many other communities and landowners selflessly allowed me to work in
their lands. I am indebted to the Rayen Lelfun indigenous community that hosted my wife
and I at Menetue for a few months and taught us a lot about birds and forests.
My thesis also benefited from discussions with members of the Martin Lab Group:
Amanda Adams, Alice Boyle, Kristina Cockle, Amanda Edworthy, Jennifer Greenwood,
Andrew Huang, Amy Koch, Elizabeth Macdonald, Michaela Martin, Andrea Norris, Hugo
Robles and Cassandra Storey. My external examiners Ralph J. Gutirrez, Diane Srivastava
and John S. Richardson, and exam chair Dolph Schluter provided helpful feedback on the
exam copy.
My family (including parents, siblings, nieces, nephews and my in-laws Barreau
Daly) have always been willing to help. Liza Jofr kindly helped us taking care of Nahuel
during critical final stages of thesis writing. I am indebted to all of them for their
unconditional support. The tribe was our Vancouverite family while living in Canada. I
treasure the many happy hours we spent together in British Columbia over these years.
Two people have been essential in bringing this project to fruition: Antonia and
Nahuel, who fulfil my life with joy, love and companionship.

xi
Dedication

I see a likeness between the old animist forest, where one could not be sure whether an Owl's
call came from a bird or an Omah, and the evolutionary forest, with its unclear distinctions
between tree and fungus, flower and fir cone. The tree-fungus relationship is as mysterious in its
origins and implications as the Owl-Omah one. Both belong to a world that goes deeper than
appearances, where a buried interconnectedness of phenomena renders behaviour ambiguous,
where one cannot walk a straight line (Wallace 1983)."

Para Sandra, Yuyo, Antonia y Nahuel.


Para el bosque templado de Los Andes del sur y la gente sencilla.

xii
Chapter 1. General introduction and thesis overview

1.1. Background
1.1.1. Endangered ecosystems and habitat-specialist owls
South America hosts the southernmost temperate forests in the world (Armesto et al.,
1998). These ecosystems are recognized as a biodiversity hotspot because of their high
concentration of endemic species, and are subject to conservation concern due to high rates
of anthropogenic degradation and fragmentation (Myers et al., 2000). Here, forest owls
inhabit an area that has been degraded and fragmented by intensive land use practices,
mainly logging, land clearance for agriculture and replacement of native stands by exotic
tree plantations (Armesto et al. 1998; Echeverra et al. 2006). However, basic information
on owl-habitat relationships for South American temperate forests is limited (Trejo et al.,
2006; but see Martnez and Jaksic, 1996; Ibarra et al. 2012).
The potential effects of forest degradation (i.e. a decrease in the quality of forest
attributes at the stand-level) and fragmentation (i.e. a reduction in the amount of forest
accompanied by changes in the configuration of remaining patches) on the relationships
between owls and their habitats are a major concern for raptor conservation (Newton, 2007,
1979; Wiens, 1994). If forest-dwelling owls select specific habitat types to increase their
fitness (Cody, 1985), habitat degradation and fragmentation may affect their ecological
niche (i.e. volume in the environmental space that permits occurrence and positive
population growth; Hinam and St. Clair, 2008; Hutchinson, 1957; Seamans and Gutirrez,
2007). In a niche context, the effects of such processes may be especially likely for habitat-
specialist species that depend strictly on forests for nesting or other life requirements
(Greenwald et al., 2005; Martnez and Jaksic, 1996; Trejo et al., 2006). In contrast,
degraded and fragmented habitats may provide a diversity of nesting and feeding resources
and thus can have positive effects for habitat-generalist species (Fahrig, 2003; Grossman et
al., 2008; Martin et al., 2004).
Intensive land use practices have degraded stand-level distribution and density of
structural attributes of South American temperate forests such as the understorey
vegetation, availability of large canopy trees, volume of coarse woody debris and dead

1
standing trees or snags, affecting habitat-specialist species that depend on these niche
resources (Daz et al., 2005; Ibarra et al., 2012, 2010; Reid et al., 2004; Rozzi et al., 1996).
At the landscape scale, temperate forest landscapes exist as a patchwork of small retention
patches from previously continuous habitat, creating a growing proportion of fragmented
forests across spatially heterogeneous mosaics of habitats (Echeverra et al., 2006; Vergara
and Armesto, 2009; Willson et al., 1994).
To understand how land use practices can influence forest owl ecology and their
broader implications to the rest of the biological community in South American temperate
forests, study designs must allow thorough assessments of how sympatric owls are spatially
associated with both (i) suitable forest habitats and available resources, and (ii) other co-
occurring elements of forest biodiversity (Groce and Morrison, 2010; Sergio et al., 2008).
However, the precision of the estimates resulting from these assessments will depend on the
survey methods used to both detect and analyse presence-absence data for forest owls
(Andersen, 2007). Site-occupancy models offer a reliable approach to study forest owls and
provide the basis for stronger inferences by removing the need to rely on assumptions of
perfect detectability (MacKenzie et al., 2006).

1.1.2. Detectability of forest owls


Studies based on counts at specific sampling points have been the most common
method for assessing occurrence, potential distribution and habitat use of owls (Andersen,
2007). The design of these studies often include the assumption that owl detectability is
perfect or not variable (Olson et al., 2005). However, this assumption is not valid because
nocturnal owls are elusive and occur at low densities, and thus surveys are likely to result in
imperfect detection or false absences of individuals (MacKenzie et al., 2006).
Imperfect detection occurs when a researcher fails to detect individuals (or a species)
that are actually present but not detected owing to environmental conditions (e.g. adverse
weather, noise, temperature or wind), random chance (e.g. temporary absence of
individuals from the location) or are non-responsive due to behavioural factors, such as
breeding condition or the presence of a larger dominant owl species (Kissling et al., 2010;
Loureno et al., 2013; Wintle et al., 2005). Therefore, exclusively using the count statistic

2
(i.e. the number of individuals seen or heard) will result in an underestimation of the
proportion of sites where the species of interest occurs (MacKenzie et al., 2002).
Occupancy models correct occurrence data for false absences in the detection process
by assessing the history of detections across multiple surveys of the same sites (MacKenzie
et al., 2005). They can also include temporal, abiotic and biological covariates related to
detectability (Andersen, 2007; Kissling et al., 2010; Sberze et al., 2010). Occupancy models
can simultaneously incorporate stand-level (Greger and Hall, 2009; Seamans and Gutirrez,
2007) and/or landscape-level attributes as covariates (Folliard et al., 2000; Martnez et al.,
2003; Sergio and Newton, 2003) for owl occurrence probabilities. These models provide a
sound approach to identify the key covariates to which owls respond and with which to
develop habitat suitability models that can be used to predict occurrence as function of
available niche resources (MacKenzie et al., 2006).

1.1.3. Owl niches and habitat suitability across spatial scales


Habitat suitability models presuppose that the observed occurrence of an owl at a
site reflects its ecological niche requirements (Hirzel and Le Lay, 2008). While forest owls
are wide ranging and select habitat at multiple scales to meet their requirements (Flesch and
Steidl, 2010; Krger, 2002; Sergio et al., 2004a), few studies have linked forest degradation
and fragmentation to habitat suitability and niche relationships of sympatric owls across
spatial scales (Hirzel and Le Lay, 2008). Such an examination may be useful to identify: (i)
significant scales concerning an owls perception of the environment in order to generate
habitat suitability models (Martnez et al., 2003; Sergio et al., 2003), and (ii) how
vulnerable are owls when their habitats are rapidly degraded and fragmented such as is
occurring in South America forests. Understanding these linkages can facilitate effective
recommendations for forest owl conservation (Bierregaard, 1998; Hinam and St. Clair,
2008; Lamberson et al., 1992). Furthermore, the implementation of plans for owl habitat
management can ultimately deliver broader biodiversity benefits when owls serve as
surrogates for areas of high biodiversity.

3
1.1.4. Forest owls as reliable biodiversity surrogates
The complexity of environmental problems, like forest degradation and
fragmentation, has led to the development of surrogates to track changes in biodiversity
(Lindenmayer and Likens, 2011). Those species which by virtue of their presence or
abundance predict high biodiversity or presence of a threatened community are considered
surrogates of those conditions (Caro and ODoherty, 1999; Lindenmayer and Likens,
2011). Forest owls can serve as biodiversity surrogates because they (i) frequently represent
the top of food chains thus their presence may have effects that cascade down through the
ecosystem, (ii) require large areas that cover populations of other less area-demanding
species, (iii) provide indications of subtle changes within ecosystems, like habitat
degradation and fragmentation processes, that may affect biodiversity structure or function,
(iv) commonly select habitats with high structural or topographic complexity and (v) may
indirectly provide essential resources, such as carrion or safe breeding sites, for other
species (see Sergio et al., 2008 and references therein).
Although studies conducted in the northern hemisphere consistently show a positive
association between forest owls and high avian species richness (e.g. Burgas et al., 2014;
Sergio et al., 2006, 2004; but see Ozaki et al., 2006), less attention has been paid to
understanding whether these avian predators correlate with other measures of biodiversity,
including functional diversity and endemism. Functional diversity -defined as the variety of
functional (morphological, behavioural, physiological, phenological) traits present in a
community- plays an important role linking species richness to ecosystem functioning
(Daz and Cabido, 2001). Functional diversity should be more informative than species
richness in conservation strategies that prioritize ecosystem function or stability (Cadotte et
al., 2011). Furthermore, endemic species commonly have small populations and few areas
assigned to conservation, thus they are especially vulnerable to habitat degradation and
fragmentation (Gaston, 1998). Therefore, there is a need to broaden the examination on the
value of forest owls as surrogates from species richness to other important ecological and
conservation phenomena such as endemism and functional diversity.
Studies from multiple localities are needed to verify the generality of being able to
use owl occurrence as an efficient surrogate for biodiversity (Sergio et al., 2008). Further,

4
addressing the mechanisms underlying positive spatial correlations between surrogates and
the target biodiversity for which they are considered to be proxies is as important as
identifying reliable surrogates (Lindenmayer and Likens, 2011).

1.2. Thesis objectives

The overall goal of my PhD is to compare the ecology of two sympatric forest-
dwelling owls, and to evaluate whether they can be used as reliable surrogates for
biological diversity in South American temperate forests. In my PhD research, I used
methods typical for studying forest owls to conduct local ecological research with global
significance. My objectives were to identify the factors associated with the detectability of
owls (Chapter 2), to assess their occurrence rates and niche relationships across spatial
scales (Chapter 3) and to explore whether occurrence rates of owls are spatially correlated
with avian taxonomic diversity, endemism and functional diversity in temperate forests
(Chapter 4). My thesis has broad implications for the ecology and conservation of forest
owls and associated biodiversity. In Chapter 5, I discuss the major theoretical and practical
implications of my research. I identify critical gaps in knowledge and provide
recommendations for future research on owls and on the surrogate concept, in order to
enhance the empirical selection and applicability of biodiversity surrogates in South
American temperate forests and elsewhere.

1.3. Study area

I conducted my study within the Araucarias Biosphere Reserve (UNESCO, 2010),


specifically 2,585 km2 within the Villarrica watershed in the Andean zone of the La
Araucana Region (3916S 71W) of southern Chile (Fig. 1.1). I chose this watershed
because accessibility is relatively good and its landscapes are representative of Andean
temperate forests. The climate is temperate with a short dry season (<4 months) and a mean
annual precipitation of 1,945 mm (Di Castri and Hajek, 1976). Elevation ranges from 200
to above 2,800 m, with forests distributed from 200 to 1,500 m of altitude. Forests are

5
dominated by deciduous Nothofagus species (southern beeches) at lower altitudes and
mixed deciduous with coniferous Araucaria araucana (monkey puzzle tree) at higher
elevations (Gajardo, 1993). This area is characterized by steep, rugged topography with
valley floors used for agriculture, interspersed with small to medium-sized villages and
towns. Most public protected areas at high elevations (>700 m) are forested, whereas
lowlands (<700 m) are dominated by agriculture and human settlement. However, several
private protected areas have been established during recent decades in the lowland areas.

1.4. Study species

I studied two sympatric owls that hypothetically differ in site-occurrence patterns


and sensitivity to forest degradation and fragmentation (Trejo et al., 2006): rufous-legged
owls (Strix rufipes) and austral pygmy-owls (Glaucidium nana). Both species occur
throughout South American temperate forests (spanning 35-55 S latitude). In the Andean
portion of temperate forests, these two owls are the most common of the five species of
resident owls occurring; therefore, I anticipated I could obtain a sufficient number of
detections to model detectability and occurrence of these two species. Strix rufipes is one of
the least known owls in South America and its populations are suspected to be declining
populations due to the loss of native forests (Martnez and Jaksic, 1996). Glaucidium nana
is the most widespread and common owl in Chile (Jimnez and Jaksic, 1989). Previous
research suggests that S. rufipes is more of a habitat specialist and G. nana is more of a
habitat generalist (Ibarra et al., 2012). Strix rufipes is associated with multi-stratified forest
stands > 100 years old, whereas G. nana is associated with a broad range of environments
including forests, forest-steppe ecotones, shrublands and occasionally urban parks (Jimnez
and Jaksic, 1989; Martnez and Jaksic, 1996; Trejo et al., 2006). However, whether these
two owl species actually differ in either occurrence rates or levels of habitat-specialization
has not been tested. Their efficacy as biodiversity surrogates has also never been assessed.

6
1.5. General field methods

1.5.1. Study design and allocation of survey effort


I defined a site (i.e. sampling unit) as the area within a 500-m detection radius of
the sampling point, which corresponded to the area within which an owl should have been
able to hear a vocal lure (Sutherland et al., 2010). To determine the number of sites
required to develop robust occurrence models, I simulated different study designs in
GENPRES (Bailey et al., 2007). I defined a standard error SE () = 0.05 as the desired
level of precision for the estimated proportion of sites where an owl occurs. Initial
estimates of the key population parameters (i.e. occurrence and detectability) were
calculated based on published studies of the two owl species (Ibarra et al., 2012; Martnez
and Jaksic, 1996). This assessment generated a minimum number of 86 sites to be
surveyed. To be conservative, I established 95 sites for 2011-2012 and 101 for 2012-2013
(i.e. six new sites in the second year).
The 101 sites were established across an elevation gradient from 221 to 1,361 m
(near the tree-line). This range covered a variety of habitat conditions from degraded and
patchy forests to zones comprising continuous undisturbed forests at higher elevations.
Using ArcGIS 10.1 I identified all the headwaters of smaller basins that were accessible by
rural roads or hiking trails within the Villarrica watershed (N = 19 basins). I randomly
selected 13 of these 19 basins and placed the first site for all basins near the headwater
(within 1 km of tree-line). I systematically established the remaining sites every 1.5 km
descending the drainages (Fig. 1.1).

1.5.2. Nocturnal raptor surveys


I used owl broadcast surveys because they are known to improve detection rates of
the owls I studied (Ibarra et al., 2012; Trejo et al., 2011). I surveyed each site over two
nesting seasons at minimum intervals of 10-days, from mid-October to early February. I
conducted 1,145 nocturnal owl broadcast surveys over two years of sampling. In 2011-
2012, I conducted six surveys at 89 sites, four at four sites and three at two sites ( x = 5.85

7
surveys per site). In 2012-2013, I conducted six surveys at 93 sites, four at seven sites and
three at one site ( x = 5.83 surveys per site). More details are provided in Chapter 2.

1.6. Thesis overview

The overall rationale of my research project was to build layers of knowledge that
are reflected sequentially in the chapters of this dissertation such that results obtained in
one chapter were subsequently used as a baseline to proceed to the next question. In
Chapter 2, I present the temporal and abiotic factors associated with the detectability of the
two sympatric owls. I also assess whether detection of one species was related to the
detection of the other species; thus, the estimated variation in their detection rates allowed
me to draw inferences about their presence across the landscape. Using these results, in
Chapter 3 I address the association between habitat resources and occurrence patterns for
each of these two owls at three spatial scales, and used niche theory to test if habitat-
specialist and generalist owls differ in their total resource utilization and peak performance
(i.e. peak occurrence). This examination allowed me to identify key niche resources
associated with owls and thus enabled me to provide reliable recommendations for owl
conservation in South American temperate forests.
In Chapter 4, I used results from Chapters 2 and 3 to further compare the
effectiveness of these two sympatric owls as surrogates for taxonomic avian diversity,
endemism and functional diversity. In addition, I examined whether the surrogate
candidates and target biodiversity have similar habitat correlates and responses to
anthropogenic forest degradation.
My thesis has broad implications for the ecology, monitoring and conservation of
both forest owls and avian forest diversity. This work also raised several new questions that
would be valuable areas for additional research. In Chapter 5, I provide a thesis summary
and suggest scaling-up studies on owl ecology. Further, I provide recommendations for
improving the empirical identification and conservation reliability of biodiversity
surrogates.

8
Figure 1.1. Distribution of 101 sites (red dots) used to study rufous-legged owls (Strix
rufipes) and austral pygmy-owls (Glaucidium nana) in a mountainous landscape in Andean
temperate forests of the La Araucana Region (3916S 71W), southern Chile. The main
map represents the Villarrica watershed and it shows forests (green), shrublands and
grasslands (yellow), alpine areas (grey), lakes (light blue), wetlands (dark blue), urban areas
(black), and snow and glaciers (white).

9
Chapter 2. Detectability of owls in South American temperate forests1

2.1. Introduction

Compared with other avian groups, owls are difficult to study and are typically not
covered by land-bird monitoring programs because of their low densities, elusive behaviour
and nocturnal habits (Fuller and Mosher, 1987). As a result, inferences about the spatial and
temporal variation in owl occurrence could be misleading if researchers do not account for
incomplete detectability or false absences (MacKenzie et al., 2006; Wintle et al., 2005).
Detectability of owls may be affected by several temporal, abiotic and biotic factors
(Andersen, 2007). For example, intraseasonal breeding phenology and social status, which
are commonly correlated with prey availability, can affect calling rates of owls (Hardy and
Morrison, 2000; Kissling et al., 2010; Morrell et al., 1991). Unfavourable detection
conditions such as wind speed, environmental noise and cloud cover can influence the
ability of researchers to detect owls (Andersen, 2007; Fisher et al., 2004), and lunar cycles
appear to influence communication and activity patterns of owls and their prey (Clarke,
1983; Penteriani et al., 2010). Furthermore, the calling rates of owls may be affected by the
risk of being detected by an intraguild predator (Loureno et al., 2013), or by the presence
of a dominant owl in the area (Olson et al., 2005). Thus the number of sites occupied by an
owl species of interest and their detection probabilities can be underestimated if
environmental or social factors are not considered.
Few studies have investigated habitat use and abundance of owls in the temperate
forests of South America and none have examined occurrence and detectability (Ibarra et
al., 2012; Martnez and Jaksic, 1996). Two species, the rufous-legged owls (Strix rufipes)
and austral pygmy-owls (Glaucidium nana), inhabit an eco-region that is among the most

1
A version of this chapter has been published. Ibarra, J.T., Martin, K., Altamirano, T.A., Vargas, F.H.,
Bonacic, C., 2014. Factors associated with the detectability of owls in South American temperate forests:
implications for nocturnal raptor monitoring. J. Wildl. Manage. 78, 1078-1086.

10
threatened on earth because nearly 70% of forest cover has been lost because of large-scale
farming and plantation forestry (Lara, 1996; Myers et al., 2000). In Chile, the great
majority of remaining forests inhabited by these owls are located in high-elevation
protected areas, whereas forests in lowland areas have varying levels of degradation and
fragmentation (Armesto et al., 1998).
I examined factors associated with the probability of detecting S. rufipes and G. nana
in southern Chile to improve monitoring protocols for these raptors. I estimated owl
detectability as a function of survey-specific temporal, abiotic and biotic conditions
(MacKenzie et al., 2006). Quantifying sources of variation in detection rates can provide
more reliable estimates for addressing research questions and may improve monitoring
programs for owls in the region (e.g. Andersen, 2007; Manning, 2011).

2.2. Methods

2.2.1. Field methods

Nocturnal raptor surveys


I conducted surveys along rural roads and trails. I broadcasted calls of both species
beginning approximately 15 minutes after sunset until 03:45 hours. I used a portable
amplifier (Mipro MA-101C, Mipro, Chiayi, Taiwan; 27-watt) for broadcasting owl calls
with a volume adjusted to 100 db at 1 m in front of the speaker measured using a digital
sound-level meter (Extech 407730, Extech Instruments, Nashua, NH; Fuller and Mosher,
1987). Each survey started with a one-minute passive listening period, followed by
playback of calls of both species played in a random sequence. For each species, I
broadcasted vocalizations for 30 seconds while rotating the amplifier 360, then listened for
one minute so that I broadcasted calls for each species twice and followed each time with
one minute of listening (Kissling et al., 2010). At the end of each survey, I took two
minutes to record time, temperature (C), relative humidity (%) and wind speed (m/s) at a
height of 2 m using a hand-held weather monitor (Kestrel 4,200, Kestrelmeters,
Birmingham, MI). I measured cloud cover using okta units (i.e. eighths of sky covered by

11
clouds) and assessed the presence (1) or absence (0) of considerable environmental noise
(e.g. stream or river sound, barking dogs).
I obtained the moon phase (%, where full moon = 100%) for each night surveyed
(http:// kwathabeng.co.za/travel/moon/moon-phase-calendar.html?country=Chile). Because
the amount of ambient light was affected negatively by the presence of clouds, I quantified
moonlight (Ml) as the proportion of illumination relative to the maximum possible at full
moon, reduced by that obscured by clouds and computed as Ml = (1 - cloud cover) (moon
phase/100) (Kissling et al., 2010). I repeated surveys of each site at intervals of
approximately 10 days, and broadcasted owl calls always from the same location at the
centre of the site.

2.2.2. Data analysis


I used a multi-season occupancy framework for open populations using detection
histories of the owls during the study period (MacKenzie et al., 2003). I modeled the data
for each owl species independently (i.e. single-species occupancy models; MacKenzie et
al., 2006). I estimated probabilities of occurrence () and detection (p) using the program
R-Unmarked, which allowed the response variables to be functions of covariates (Fiske and
Chandler, 2011). For , I considered two covariates across the altitudinal gradient: mean
elevation of the site (meters above sea level/1,000), and Pa, a binary covariate indicating
the site was located within 500 m of a protected area (1) or not (0). To identify potential
covariates that may be associated with detectability, I used covariates that were correlated
with owl detection in other studies (Clark and Anderson, 1997; Crozier et al., 2006, 2005;
Hardy and Morrison, 2000; Morrell et al., 1991; Wintle et al., 2005). I modeled the
probability of detection (p) assessing nine temporal, abiotic and biotic covariates (Table
2.1). I also included quadratic terms for number of days since the start of surveys and
moonlight because the influence of these covariates on calling behaviour of owls might not
be linear throughout the breeding season (Ganey, 1990; Kissling et al., 2010). I considered
that pairs of collinear variables (r > 0.7) were estimates of a single underlying factor;
therefore, I did not use collinear variables in the same model. I retained only the covariate
that was expected to be more influential to owl detectability in the analysis (Table 2.1).

12
To obtain the best model for each owl species, I first fit models using each covariate
singly to predict or p. I also fit a model with constant across sites and p constant across
surveys (i.e. null models). I ranked models using an information-theoretic approach
(Akaikes Information Criterion [AIC]; Burnham and Anderson, 2002). After I fit the
single-covariate models, I assessed more complex models containing different
combinations of the best-supported covariates, on the basis of model weights and the
precision of the estimated coefficients (from the single-covariate model). From this base
model, I added extra covariates and evaluated each models weight following every
addition. I continued to add covariates until all supported covariates not in the base model
had been considered. I considered models within 2 AIC units of the top model as the
competitive set of best-supported models. I computed model weights (wi), reflecting the
relative weight of evidence for model i, and considered the best model to be that with the
highest weight and lowest AIC value (Burnham and Anderson, 2002).

2.3. Results

I obtained 292 detections (148 for 2011-2012 and 144 for 2012-2013) of S. rufipes
and 334 (173 for 2011-2012 and 161 for 2012-2013) of G. nana. From 493 surveys where
at least 1 owl was detected during both seasons, 133 (27%) were co-detections (i.e. both
species recorded during a survey), 159 (32%) were S. rufipes alone and 201 (41%) were G.
nana alone. Strix rufipes were detected at 59 (62%) of 95 sites in 2011-2012 and 56 (55%)
of 101 sites in 2012-2013. Glaucidium nana were detected at 68 (72%) of 95 sites in 2011-
2012 and 78 (77%) of 101 sites in 2012-2013.

2.3.1. Occurrence and detectability of rufous-legged owls


I assessed 24 models for S. rufipes. Probabilities of occurrence for S. rufipes were
positively associated with elevation and with sites located within 500 m of a protected area
(Pa; Table 2.2, Fig. 2.2). Probability of occurrence varied among sites located either inside
or outside protected areas (Pa = 1: = 0.69-0.99, Pa = 0: = 0.27-0.93). However,
protected area status explained little variation given the 95% confidence intervals of the

13
coefficient included 0 (Table 2.3). The best approximating models indicated that the
probability of detecting a S. rufipes increased with moonlight (i.e. brighter nights with
waxing moon and little cloud) and was negatively associated with environmental noise
(Table 2.3, Fig. 2.3). Furthermore, the detectability of S. rufipes increased when a G. nana
was detected at the same site during the same survey (Table 2.3). The detection probability
ranged from 0.39-0.52 when a G. nana was not detected; it increased to 0.52-0.65 when the
latter owl was detected.

2.3.2. Occurrence and detectability of austral pygmy-owls


I assessed 20 models for G. nana. Probabilities of occurrence were not associated
with elevation and did not vary among sites located either inside or outside protected areas.
In contrast, detectability increased with moonlight, decreased with both environmental
noise and wind speed (Tables 2.2 and 2.3, Fig. 2.3), and increased throughout the season
from a minimum detectability (p = 0.36 0.04) at the beginning of the sampling season to a
peak (p = 0.47 0.08) during the surveys from 23 January-7 February (Fig. 2.3). Further,
probability of detecting a G. nana increased when a S. rufipes was detected at the same site
during the same survey (Table 2.3). The probability of detecting a G. nana ranged from
0.17-0.40 when a S. rufipes was not detected; it increased to 0.34-0.62 when the latter owl
was detected.

2.4. Discussion

I identified sources of variation associated with detection probabilities for the two
most common owls in Andean temperate forests, and the patterns were similar between
species (e.g. moonlight intensity increased detectability of both S. rufipes and G. nana, and
the detection of both species was positively correlated with the detection of the other
species). Although these two owl species have different broad habitat associations (Ibarra et
al., 2012; Trejo et al., 2006), similarity in both nocturnal prey base and tree-cavities used
for nesting (Beaudoin and Ojeda, 2011; Figueroa et al., 2006; Ibarra et al., 2014a) may be
potential causal mechanisms explaining similar patterns of calling activity, responses to

14
covariates and resultant probabilities of detection. Further, the fact that environmental noise
decreased detectability of both owls suggests that this factor may affect the range of vocal
broadcasts, the capacity of researchers to detect responding owls, the rates of owls calling
or all of these (Hardy and Morrison, 2000; Morrell et al., 1991).
I used moonlight rather than moon phase per se to depict nocturnal illumination
because it corrects ambient light estimates, derived from moon phase, and adjusts for the
reducing effect of clouds on light intensity (Kissling et al., 2010). I found that owl calling
rates were positively associated with clear nights as reported by Morrell et al., (1991) for
great horned owls (Bubo virginianus) when cloud cover was less than 50%, but unlike saw-
whet owls (Aegolius acadicus) whose calling rates increased when cloud cover was >50%
(Clark and Anderson, 1997). Likewise, some studies have reported either moon phase or
moonlight were positively correlated with owl calling (Clark and Anderson, 1997; Clarke,
1983; Kissling et al., 2010; Morrell et al., 1991; Penteriani et al., 2010), but other studies
have not (Ganey, 1990; Hardy and Morrison, 2000). The fact that brighter moonlight was
positively correlated with detection rates for the study species suggests a general preference
to be active during more illuminated nights. The efficiency of owl hunting may increase as
moonlight waxes to full moon cycle, because predators need less time to capture prey
(Clarke, 1983). However, prey may reduce their activity in full moonlight as an anti-
predatory response (Ylonen and Brown, 2007). Little is known about nocturnal activity
periods of owl prey in South American temperate forests. However, as most small mammal
prey (e.g. Dromiciops gliroides, Irenomys tarsalis, Abrothrix olivaceus, Abrothrix
longipilis, Oligoryzomys longicaudatus) of the two owls I studied are chiefly nocturnal
(Franco et al., 2011; Mura, 1995), I expect the amount of moonlight available during
night-time to be a primary driver of owl and prey activity patterns in temperate forests.
Environmental noise reduced detectability of both owls, and wind speed decreased
austral pygmy-owl detection rates. Similar effects of both covariates were found for
western-screech owls (Megascops kennicottii) and A. acadicus in southeastern Alaska, with
reductions on detection rates of nearly two-thirds under considerable noise and also under
moderate winds (< 3 km per hour; Kissling et al., 2010). Because I systematically
established sites 1.5 km apart, several were located near streams, rivers and human

15
habitation (where frequently dogs barked during our nocturnal surveys); the considerable
noise produced by these factors may have reduced detectability for the two owls.
The peak period for detecting G. nana was the end of the survey season in February
when, according to the breeding phenology for the species, chicks had already fledged
(Ibarra et al., 2014a). Owls call more when they are territorial (e.g. they are searching for
suitable sites for reproduction or have established pair bonds) and seldom vocalize when
eggs are in the nest (e.g. long-eared, boreal and saw-whet owls; Clark and Anderson, 1997).
For example, Morrel et al., (1991) reported that B. virginianus were more likely to respond
earlier in the breeding season than later as a function of the chronology of the breeding
activity. One explanation for my result is that adult G. nana frequently emit territorial calls
after fledging to stimulate juveniles to disperse from their natal sites (Norambuena and
Muoz-Pedreros, 2012).
The best supported models for detectability of both species indicated that calling
rates of each species was positively correlated with the other species although the effect
was stronger for G. nana. Previous studies have inferred that detection probabilities of
spotted owls (Strix occidentalis) were lower at sites where the more aggressive barred owls
(Strix varia) are undergoing expansion into spotted owl habitat (Bailey et al., 2009; Crozier
et al., 2006; Olson et al., 2005). Furthermore, Loureno et al., (2013) suggested that the
detectability of tawny owls (Strix aluco) decreased at sites where eagle owls (Bubo bubo),
their predators, were present. In contrast, the few studies that have reported higher calling
rates in response to the calls of another owl species have been associated with mobbing
behaviour or inter-specific territoriality (Boal and Bibles, 2001; Crozier et al., 2005; Ganey,
1990). My results did not support the hypotheses that either G. nana constrain the calling
rate of S. rufipes (Martnez, 2005) or S. rufipes negatively influence calling by G. nana
because of predation risk. However, I explored only the association of interspecific calls on
the probabilities of detecting the other owl species, not the spatial patterns of species co-
occurrence. The latter may have been influenced by factors other than antagonistic
behaviour or intraguild predation, such as common environmental (e.g. habitat) choices
(Brambilla et al., 2010).

16
I found that the occurrence rates of the forest specialist S. rufipes were positively
associated with elevation. In the study area and in Chile generally, some of the last
remaining continuous and structurally complex forests (e.g. stands maintaining a multi-
storied vertical structure dominated by old shade-tolerant large trees with emergent
pioneers) were restricted to high elevations in the Andes. At lower elevations, forests were
mostly degraded and patchy (Armesto et al., 1998). The gradient of decreasing forest
disturbance and increasing forest cover and complexity with higher elevation may have
partially explained my results on the distribution and occurrence patterns of S. rufipes.

2.4.1. Recommendations for owl monitoring


Developing efficient wildlife monitoring protocols is critical in regions subject to
rapid habitat change such as South American temperate forests. For future owl monitoring
programs in this eco-region, I recommend broadcast surveys with a multi-species design.
This approach has the advantage of being economically efficient as well as increasing
detection rates of each species. To obtain reliable estimates of occupancy (i.e. standard
error SE [] ~ 0.05) and allow modelling detection probabilities of owls in temperate
forests, I recommend 3-4 surveys per season at a minimum number of 86 sampling units
(MacKenzie and Royle, 2005). I also recommend that survey designers avoid sampling
noisy areas (e.g. human habitation with barking dogs, near streams and rivers) and conduct
surveys under favorable weather conditions (e.g. low wind speeds < 5 km per hour,
relatively cloudless sky, no precipitation). In addition, observers should conduct surveys
across several moon phases, but record the moon phase for each survey (easily obtained
from moon phase calendars available online). With data on cloud cover and moon phase,
researchers will be able to calculate moonlight to depict illumination available for nocturnal
owls and use this variable to model detectability. These recommendations could be
implemented in other areas of temperate forests where surveys for more than one species of
owls are desirable.

17
2.5. Conclusions

In summary, detectability for both owls increased with greater moonlight and
decreased with environmental noise, and for pygmy-owls greater wind speed decreased
detectability. The probability of detecting pygmy-owls increased non-linearly with number
of days since the start of surveys and peaked during the latest surveys of the season (23 Jan-
7 Feb). Detection of both species was positively correlated with the detection of the other
species. I have suggested that similarity in both nocturnal prey base and tree-cavities used
for nesting may be potential causal mechanisms explaining similar patterns of calling
activity, responses to covariates and detectability. Future studies should view these
correlations as hypotheses that may be tested in further experimental studies exploring the
factors influencing calling behaviour of forest owls. Identifying the causal mechanisms
responsible for the positive association between calling rates of these owls also warrants
further research. In Chapter 3, I incorporate the estimated variation in detectability for these
owl species and include specific stand- and landscape-level covariates to identify
environmental resources related to occurrence patterns and niche relationships of these
sympatric owls.

18
Figure 2.1. Predicted probabilities of occurrence () and 95% confidence intervals for Strix
rufipes in Andean temperate forests (2011-2013), in relation to elevation (meters above sea
level) when (a) Pa = 0 (sites located outside of a protected area) and when (b) Pa = 1 (sites
located at least 500 m within a protected area), using the best model for this owl species.

19
Figure 2.2. Predicted probabilities of detection (p) and 95% confidence intervals for Strix
rufipes and Glaucidium nana in Andean temperate forests (2011-2013), in relation to
moonlight index, number of days since start of surveys and wind speed (m/s), using the best
model for each species. Moonlight index refers to the amount of light available reduced by
the proportion of sky obscured by clouds. Ml = (1 - cloud cover) moon phase/100.

20
Table 2.1. Candidate predictors of detectability for forest owls in Andean temperate forests,
2011-2013.
Covariate Type of variable (code) Description Reason for consideration

Temporal Days (day) Number of days since start of Owl calling behaviour

surveys may change throughout

the nesting season a, b, c, d, j

Time (time) Number of minutes after 21 Owl calling behaviour

hours may change during the

night a, c, d, j

Year (year) Nesting season 2011-2012 or Owl calling behaviour

2012-2013 may change between

years d, k

Abiotic Temperature (temp) * C Owl behaviour b, d, h, j, m

Wind (wind) m/s Owl behaviour, visibility,

sound carry b, c, h, j

Relative humidity (Hu) * % Owl behaviour n

Moonlight (Ml) Amount of light available Owl and prey behaviour a,


b, d, e, j, l, n
reduced for that obscured by

clouds. Ml = [(1-cloud)

(moon phase/100)].

Moon phase refers to %

where full moon = 100%

Environmental noise (noise) 0 = quiet Sound carry c

1 = substantial (dogs barking,

and/or river and stream

noise)

21
Covariate Type of variable (code) Description Reason for consideration

Biotic Other owl species detected 0 = none Owl behaviour f, g, i, k

(owl) 1 = other owl detected


a
Ganey (1990), b Hardy and Morrison (2000), c Kissling et al. (2010), d Clark and Anderson (1997), e Clarke
(1983), f Crozier et al. (2005), g Crozier et al. (2006), h Fisher et al. (2004), i Loureno et al. (2013), j Morrell
et al. (1991), k Olson et al. (2005), l Penteriani et al. (2010), m Wintle et al. (2005), n ODonnell (2004).
* Pairs of strongly inter-correlated (Pearson's r > 0.7) covariates.

22
Table 2.2. Model selection results for estimating probability of occupancy () and detection
(p) of Strix rufipes and Glaucidium nana in Andean temperate forests, 2011-2013. Site-
specific covariates consisted of elevation in meters/1,000 (elev), and whether the site was
500 m within a protected area or not (Pa). Survey specific covariates consisted of
moonlight (Ml), environmental noise (noise), whether the other owl species was detected at
the unit for the specific survey (owl), wind speed (wind) and number of days since start of
surveys (day).
Species Model Ka AIC b wi c

Strix rufipes (elev + Pa), p(Ml + noise + owl) 9 0 0.58

(elev), p(Ml + noise + owl) 8 1.43 0.28

(Pa), p(Ml + noise + owl) 8 3.02 0.13

Glaucidium nana (.), p(wind + Ml + noise + day2 + owl) 9 0 0.43

(.), p(Ml + noise + day2 + owl) 8 0.88 0.28

(.), p(Ml + noise + day + owl) 8 2.69 0.11


a
Number of parameters estimated.
b
AIC is the difference in AIC values between each model and the lowest AIC model.
c
AIC model weight.

23
Table 2.3. Competing models ( Akaikes Information Criterion 2.0) predicting
probability of occupancy () and detection (p) of Strix rufipes and Glaucidium nana in
Andean emperate forests, 2011-2013. The estimated model coefficients and the lower and
upper confidence intervals (LCI and UCI) are also shown.
Species Response Variables Coefficients LCI UCI

Strix rufipes

Model 1 Intercept -1.012 -2.182 0.159

Elevation 2.441 0.190 4.692

Protected area 1.818 -0.498 4.134

p Intercept -0.441 -0.744 -0.139

Moonlight 0.006 0.0009 0.010

Noise -0.508 -0.837 -0.180

Other owl 0.513 0.178 0.848

Model 2 Intercept -1.276 -2.379 -0.172

Elevation 3.285 1.251 5.319

p Intercept -0.449 -0.752 -0.147

Moonlight 0.006 0.0009 0.010

Noise -0.503 -0.832 -0.175

Other owl 0.514 0.179 0.849

Glaucidium nana

Model 1 Intercept 1.415 0.772 2.058

p Intercept -1.04 -1.350 -0.730

Wind -0.46 -0.873 -0.046

Moonlight 0.011 0.007 0.015

Noise -0.579 -0.885 -0.273

Days2 0.00005 -0.00001 0.00011

24
Species Response Variables Coefficients LCI UCI

Glaucidium nana Other owl 0.749 0.426 1.07

Model 2 Intercept 1.338 0.733 1.943

p Intercept -1.082 -1.374 -0.790

Moonlight 0.112 0.007 0.016

Noise -0.604 -0.909 -0.299

Days2 0.00005 -0.00002 0.0001

Other owl 0.870 0.540 1.190

25
Chapter 3. Occurrence patterns of specialist and generalist owls across
spatial scales in South American temperate forests2

3.1. Introduction

Niche theory has a long history in ecology and it is helpful for assessing the
condition of ecological communities (Clavel et al., 2011; Hirzel and Le Lay, 2008). In a
niche context, specialist species have a narrower width in resource use than generalists (i.e.
generalists utilize a greater variety of resources, Fig. 3.1a). Nevertheless, specialists can
reach either a higher or similar level of peak performance (e.g. occurrence, density) than
generalists under a subset of relatively stable resources (Fig. 3.1b and c; Devictor et al.,
2010; Peers et al., 2012). Narrower niches render specialists more prone to be negatively
affected by habitat degradation and fragmentation, than generalists (Clavel et al., 2011).
Therefore, identifying habitat attributes where specialist species have higher peak
performance is essential for the development of management guidelines that conserve a
diversity of species within a community.
Owls act as apex predators within forest communities, and the implementation of
plans for their conservation may deliver enhanced biodiversity benefits (Sergio et al.,
2006). To meet their niche requirements, forest owls usually require different habitat
patches for breeding and foraging, and thus they select habitat resources from the stand- to
the landscape-level (Flesch and Steidl, 2010). Therefore, multi-scale approaches can be
useful to identify: (a) relevant scales concerning individual perception of the environment
so as to generate habitat suitability models (Martnez et al., 2003; Sergio et al., 2003), and
(b) the level of sensitivity of species in habitats subject to rapid degradation and
fragmentation.

2
A version of this chapter has been published. Ibarra, J.T., Martin, K., Drever, M.C., Vergara, G. 2014,
Occurrence patterns and niche relationships of sympatric owls in South American temperate forests: a multi-
scale approach. Forest Ecol. Manag. 331: 281-291.

26
South America hosts the southernmost temperate forests in the world (Armesto et
al., 1998). These ecosystems are recognized as a biodiversity hotspot because of their
high concentration of endemic species, and are subject to conservation concern due to high
rates of anthropogenic degradation and fragmentation (Myers et al., 2000). Here, intensive
land-use practices have degraded stand-level availability of structural attributes such as the
volume of coarse woody debris, large decaying trees and understory vegetation, and thus
wildlife populations depending on these niche resources have been negatively affected
(Daz et al., 2005; Reid et al., 2004). At the landscape scale, southern temperate ecosystems
have been reduced and fragmented, converting continuous forest into a patchwork of
habitat types (Echeverra et al., 2006).
Habitat suitability models offer an operational application of the ecological niche as
they presuppose that the observed occurrence of an owl at a site reflects its ecological
requirements (Hirzel and Le Lay, 2008). However, the relation between niche requirements
and the occurrence patterns of forest owls may sometimes be equivocal as these birds are
elusive and mainly nocturnal, and therefore a non-detection of individuals at a site does not
mean the species is absent. With the exception of Sberze et al., (2010), most studies on
raptor-habitat relations in South America have made the assumption that owl detectability
was perfect. This assumption may underestimate the number of sites where owls achieve
their niche requirements and miss relevant habitat resources (MacKenzie et al., 2003).
One way to compare niches is to develop habitat models of sympatric species
independently and contrast their characteristics (Hirzel and Le Lay, 2008). I studied two
sympatric owls that hypothetically differ in site-occurrence patterns and sensitivity to forest
degradation and fragmentation: rufous-legged owls (Strix rufipes) and austral pygmy-owls
(Glaucidium nana). Both species occur extensively across South American temperate
forests (35-55 S). Strix rufipes are one of the least known owls in South America with
suspected declining populations due to native forest loss (Martnez and Jaksic, 1996).
Glaucidium nana are the most widespread and common owls in Chile (Jimnez and Jaksic,
1989). Previous research suggests that S. rufipes inhabit a more specific range of stand-
level habitat resources than G. nana (Ibarra et al., 2012). Strix rufipes are considered
habitat-specialists because of their affiliation with multi-stratified forest stands > 100 years

27
old, whereas G. nana are considered habitat-generalists as they utilize a range of
environments including forests, forest-steppe ecotones, shrublands and occasionally urban
parks (Jimnez and Jaksic, 1989; Martnez and Jaksic, 1996; Trejo et al., 2006). However,
whether these species actually differ in either occurrence rates or levels of habitat-
specialization has not been tested.
The aims of this study were to (1) examine the association between habitat
resources and occurrence patterns for each of these two sympatric forest owls at three
spatial scales, and (2) test if habitat-specialist and generalist owls differ in their total
resource utilization and peak performance in Andean temperate forests of southern Chile. I
predicted that (1) owl occurrence rates are influenced from local within-stand to landscape
level habitat resources, and (2) S. rufipes have a lower total resource utilization (Fig. 3.1a)
but either a higher (Fig. 3.1b) or similar (Fig. 3.1c) level of peak performance for particular
niche resources, than G. nana. To examine owl occurrence patterns and test our predictions,
I used occupancy models that account for the likelihood that owls occurred at some sites
without detections (i.e. were present but not detected; Chapter 2). My models allowed me
to identify key niche resources to which owls are associated, and thus can provide reliable
recommendations for owl conservation.

3.2. Methods

3.2.1. Field methods


Nocturnal raptor surveys
See Chapters 1 and 2.

Stand- and landscape-level data


Stand-level niche resources (hereafter covariates) included habitat attributes
suggested as important for S. rufipes and G. nana (Ibarra et al., 2012; Martnez and Jaksic,
1996). At every site, I established an L-shaped transect and located five vegetation plots
(22.4 m diameter; 0.04 ha; N = 505 plots). The first plot for each site was located 50 m
away from the centre of the site (where owl calls were broadcasted), at the vertex of the L-

28
shaped transect. The other four plots were established with a distance of 125 m between
each along two 250 m lines directed outwards from the vertex (Affleck et al., 2005). For
each plot I measured: tree density, tree diameter at breast height (DBH), canopy cover,
volume of coarse woody debris, density of bamboo understory and elevation (Table 3.1).
For DBH I calculated the standard deviation (SD) for each plot as it was considered a more
reliable indicator than the average of (a) distribution of tree-age classes in a stand, (b) stand
structural complexity and (c) the diversity of micro-habitats for owls and their prey (Van
Den Meersschaut and Vandekerkhove, 2000). Also, SD of DBH frequently increases with
stand age (McElhinny et al., 2005) and it was correlated (Pearsons r > 0.7) with mean
DBH in the study system. Values of each habitat covariate for the five plots were averaged
and thus a single value was obtained for each site.
I evaluated landscape-level covariates tested in other occurrence studies of forest
raptors (Finn et al., 2002; Henneman and Andersen, 2009). These covariates included:
forest extent, shrubland extent, core habitat, forest-patch shape index and relative habitat
diversity (Table 3.1). Landscape covariates were measured within 180 and 1,206 ha circular
areas around each site. These areas corresponded to the minimum (1.8 km2) and maximum
(12.8 km2) home-ranges reported for S. rufipes (Martnez, 2005). As no information exists
on home-range sizes for G. nana, I used the home range size considered appropriate for S.
rufipes for evaluating habitat associations in an area larger than a nest or roost site for G.
nana, which allowed me to formally compare results between the two species. Spatial
covariates were obtained from a composition of three Landsat (Landsat Enhanced Thematic
Mapper Plus ETM+) scenes: one from January 2012 and two from January 2013. To obtain
a land-use model for the study area, these scenes were corrected in a mixed-thematic
classification process using the program IDRISI Selva (Eastman, 2012), into the following
habitats: forest, shrubland, open area (including water bodies) and snow or glaciers. From
this model, the two circular areas for each site were extracted using ArcGIS 10.1. Finally,
forest patch and landscape metrics were quantified using Fragstat 4.1 (Table 3.1,
McGarigal et al., 2002). Here I use the terms degradation for stand-level and
fragmentation and reduction for landscape-level covariates indicating anthropogenic
alteration of forest attributes.

29
3.2.2. Data analysis

Modelling occurrence probabilities


Presence/absence data were analyzed using multi-season occupancy models
(MacKenzie et al., 2003). I used the program R-Unmarked (Fiske and Chandler, 2011),
which uses maximum-likelihood methods to estimate probabilities of occurrence and
detection. Probability of occurrence () was defined as the probability that at least one
individual owl occurred at a site.
To evaluate , I assessed collinearity to reduce the number of covariates presented
in Table 3.1. With strongly correlated covariates (Pearsons r > 0.7), I retained for analysis
only the one considered to be most biologically meaningful for the study species (Sergio et
al., 2003). In total, nine covariates were used in the final modelling: four at the stand-
level (SD of tree DBH, canopy cover, volume of coarse woody debris and bamboo density),
four at the 180 ha landscape-level (forest extent, shrubland extent, shape index and relative
habitat diversity) and one at the 1,206 ha landscape-level (shape index).
I used multi-model inference and Akaikes Information Criterion (AIC; Burnham
and Anderson 2002) to identify the best model(s) representing arrangements of the
covariates that I defined a priori. All models contained the sources of variation in detection
probabilities previously identified as important for our target species using the same dataset
(see Chapter 2; Table 3.2). To obtain the best model for each owl, I first fit models using
each covariate singly to predict and also fitted a model with constant (i.e. null model)
across sites. More complex models were then built by combining stand- and landscape-
level covariates among the best-supported covariates, on the basis of model weights and the
precision of the estimated beta coefficients. I added covariates until all supported covariates
not in the initial model had been considered. I evaluated 16 models for S. rufipes and 20 for
G. nana. Model weights, referring to the relative weight of evidence for model i, were
computed and the best model was the one that ranked with the highest weight (Burnham
and Anderson, 2002). Models with AIC 2 were considered the models best supported
by the data. I addressed model selection uncertainty by averaging models with AIC 4 in
the final confidence set for each owl, which also accounted for 95% Akaike weight
(Burnham and Anderson, 2002). The model-averaged predictions were used to project the

30
distribution of owls in the study area using the spatial interpolation toolbar Kriging (Oliver
and Webster, 1990), implemented in ArcGIS 10.1.
Residual analyses were used to test for spatial autocorrelation across a set of
distance classes using the Morans Index method (Moran, 1950). I based residuals on
model-averaged estimates for each owl and calculated them as the observed values at site i
(detection = 1, non-detection = 0) minus the predicted probabilities of detecting the species
at least once. I selected distance classes of 3 km (0-3, 3-6, ..., 27-30 km) because this was
twice the distance between nearest sites. Only G. nana showed positively correlated
residuals, thus I calculated an autocovariate (Aut) term for this species following Moore
and Swihart (2005). The inclusion of an autocovariate term resulted in additional six
models (N = 26) for G. nana.

Comparing owl niches


To fully understand differences between specialists and generalists, a broad
spectrum of niche resources should be tested simultaneously (Peers et al., 2012). I
compared total resource utilization and peak performance between S. rufipes and G. nana
by exploring their according to each of the 16 covariates singly (Table 3.1). I considered
each covariate both as a linear and a non-linear quadratic relationship. The response curves
showed the degree of variation in habitat suitability for each covariate. I integrated across
the range of x and y values for all response curves to obtain the total area under each
response curve (i.e. index of total resource utilization; Peers et al., 2012). I divided each
area calculation by the range across the x axis for each covariate to obtain a range from 0 to
1, with values closer to 1 representing a higher utilization of that specific resource (Peers et
al., 2012). I estimated the peak value for each covariate and compared the responses
between owls using Student's t-tests. Data were normally distributed based on
Kolmorogov-Smirnov test (P < 0.05).

3.3. Results
I had 292 and 334 detections to model occurrence patterns of S. rufipes and G.
nana, respectively (Chapter 2). The proportion of sites at which S. rufipes were detected

31
ranged from 0.62 (59 sites out of 95 total sites) in 2011-2012, to 0.55 (56 out of 101 total
sites) in 2012-2013. The proportion of sites at which G. nana were detected ranged from
0.72 (68 out of 95 total sites) in 2011-2012, to 0.77 (78 out of 101 total sites) in 2012-2013
(Chapter 2).

3.3.1. Habitat suitability for owls


For S. rufipes, predicted (mean standard error) ranged between 0.05 0.04 and
1.00 0.00 across sites. The models with highest support ( AIC 2) for S. rufipes
contained two to four covariates for (Table 3.2a). Model selection results indicated that
for S. rufipes was positively associated with the variability (SD) in the DBH distribution of
trees, bamboo density and canopy cover; however, the 95% confidence interval for the beta
coefficient of canopy cover overlapped with zero and thus this covariate was considered
non-informative. Best models also supported a positive association between forest extent at
180 ha and S. rufipes (Table 3.2a); however, beta coefficient for this covariate also
overlapped with zero (Table 3.2a). The averaged predictions of for S. rufipes revealed a
zone of high habitat suitability to the east of the study area, associated with forests located
close to the Andes Range (Fig. 3.2a). Areas of high habitat suitability occurred to the south
and southeast of the study area as well, where a diagonal chain of volcanoes oriented south
to south-east encompasses relatively continuous and old-growth forests between 700 and
1,500 m of elevation. Zones of high suitability were mostly located inside or surrounding
protected areas to the east, south and southeast of the study area (Fig. 3.2a). For easier
implementation in forest management and planning programs, I calculated the values of
covariates associated with predicted low (0-0.33), moderate (0.34-0.66) and high (0.67-1)
values of for S. rufipes in Andean forests based on averaged model predictions (Table
3.3).
For G. nana, predicted ranged between 0.67 0.18 and 0.98 0.04 across sites.
The models with highest support ( AIC 2) for G. nana contained two or three covariates
for , although two of the best models included the autocovariate term (Aut, Table 3.2b).
The spatial auto-covariate term effectively controlled for intra-landscape data dependence;
it improved the AIC weight of two of the best models as those with the auto-covariate were

32
0.09 and 0.07 units higher than models without this term (compare Table 3.2b). Model
selection results indicated that for G. nana was positively associated with the forest-patch
shape index at 1,206 ha and forest cover extent at 180 ha. The third best model had a
negative association between G. nana and shrubland extent at 1,206 ha (Table 3.2b).
However, beta coefficient for all covariates overlapped with zero in all models. The
averaged predictions of for G. nana showed chiefly uniform mid- to high-levels of
habitat suitability across much of the study area, with zones of slightly higher suitability
either inside or close to protected areas (Fig. 3.2b).

3.3.2. Resource utilization and peak performance by owls


The predicted for our two owl species varied in habitat suitability according to
each environmental covariate (Fig. 3.3). Total resource utilization was lower for S. rufipes
(mean standard error: 0.625 0.080) and higher for G. nana (0.804 0.020; t 30: -27.569,
P < 0.001) in all 16 covariates (Fig. 3.4a), indicating a lower for S. rufipes over the range
of all covariate values. Averaged results of peak performance (i.e. peak ) for all 16
covariates together did not differ between S. rufipes (0.913 0.045) and G. nana (0.894
0.029; t 30: 0.736, P = 0.467; Fig. 3.4b). Considering each covariate separately, S. rufipes
had higher peak for elevation, tree density, SD of DBH, bamboo density, volume of
coarse woody debris, canopy cover, forest extent at 180 and 1,206 ha, and core habitat at
180 and 1 206 ha than G nana. Glaucidium nana had higher peak performance for forest-
patch shape index at 180 and 1,206 ha, shrubland extent at 180 and 1,206 ha, and habitat
diversity at 180 and 1,206 ha than S. rufipes.

3.4. Discussion
Multi-scale approaches, accounting for detection probability, improve
understanding of species habitat suitability and thus perceptions of ecological pressures
under which habitat selection and niche requirements have evolved (Martnez et al., 2003).
The study owls responded to habitat resources at several spatial scales. Strix rufipes
responded more strongly to stand-level whereas G. nana to landscape-level resources;
however, the parameter estimates for the latter owl were imprecise (i.e. 95% CI overlapped

33
zero). Furthermore, the comparison of niche widths suggested that habitat-specialist S.
rufipes had a lower total resource utilization of the 16 resources under consideration while
achieving a similar peak performance than the generalist G. nana. These results on niche
relationships indicate that specialist owls use smaller portions of the potentially available
habitat, and may require specific management considerations in an area subject to rapid
forest degradation (at the stand-level) and fragmentation (at the landscape-level) such as is
happening in South American temperate ecosystems.

3.4.1. Habitat suitability across spatial scales


I identified a set of environmental resources which could drive habitat selection for
S. rufipes. Habitat selection is considered a hierarchical decision-making process occurring
from large to small spatial scales (Hutto, 1985). However, resources at the stand-level were
more influential than landscape-level resources on occurrence rates for S. rufipes in this
study such that habitat selection for this owl species could involve bottom-up choices
(sensu Flesch and Steidl, 2010). Strix rufipes were more likely to occur in structurally
complex multi-aged forest-stands (i.e. higher values of SD of DBH), characterized by the
presence of large trees and relatively high availability of bamboo understorey. My results
support the findings of the two previous studies of S. rufipes habitat use in temperate forests
(Ibarra et al., 2012; Martnez and Jaksic, 1996). Strix rufipes are secondary cavity nesters
that build nests in cavities generated by tree-decay processes or excavated by Magellanic
woodpeckers (Campephilus magellanicus). The reported dimensions of nesting trees (mean
DBH SD = 122.8 36.2 cm) in Andean temperate forests suggest that trees greater than
100 years old are necessary to support suitable-sized cavities for this owl (Beaudoin and
Ojeda, 2011). For its part, the bamboo understorey provides habitat for several endemic
arboreal and scansorial small mammals which constitute the main prey of S. rufipes in
temperate forests (Figueroa et al., 2006). Dense understorey of native bamboos is frequent
under: (a) large canopy gaps generated by natural tree falls, (b) high-elevation (> 900 m
altitude) old-growth stands with relatively open canopies (54 to 81% canopy cover) and (c)
logged forests where the canopy has been opened (Ibarra et al., 2012; Veblen, 1982).

34
The presence of large trees (likely related to breeding and roosting requirements)
and dense bamboo understorey (likely related to food supply), constitute key structural
resources at the stand-level for S. rufipes. However, my results and previous descriptions of
S. rufipes habitat (e.g. multi-storied forest sites > 100 years old, dominant trees with DBH >
28 cm, more than 5 snags/ha and presence of large decaying trees; Martnez and Jaksic,
1996), emphasize the importance of several resources that generate the structural
complexity suitable for these owls. Other Strix species in temperate forests from the
northern hemisphere select similar habitat resources (Lhmus, 2003; Seamans and
Gutirrez, 2007; Singleton et al., 2010), offering similarities between congeneric habitat-
specialist owls in ecologically comparable environments. Across the study area, stand
structural complexity increased from lowland unprotected areas to higher elevations (>700
m) comprising protected areas and their adjacent zones (Ibarra et al., 2012). These changes
generated heterogeneous distributions of resources important for the predicted distribution
of S. rufipes.
Incorporating landscape-level habitat data into the analyses did not substantially
improve our ability to predict S. rufipes occurrence in Andean temperate forests as the
estimated beta coefficients were imprecise. However, the best supported models included
forest cover at 180 ha, and indicated a positive relation with S. rufipes occurrence. I
therefore stress the importance of including this spatial structural resource in further studies
of S. rufipes. The central depression of Chile and coastal range zones are highly deforested
and lack protected areas (Echeverra et al., 2006; Smith-Ramrez, 2004). The long-term
survival of S. rufipes is jeopardized in these areas of its range (Martnez, 2005); thus,
including forest cover measures at the territory scale will be an important first step in
refining landscape-level assessments for habitat use of S. rufipes.
I identified spatial autocorrelation of initial-model residuals for G. nana, revealing
that sites closer together resembled each other more than sites that are further apart.
Although controlling for such spatial dependence allowed improvement of model fit, the
model selection statistics showed moderate uncertainty about the most plausible model for
occurrence of G. nana. As reported for other Glaucidium species (Campioni et al., 2013),
the weak association between resources and G. nana occurrence could indicate a continuum

35
of good habitat conditions across Andean temperate forests. This hypothesis is supported
by the fairly uniform level of habitat suitability for G. nana across much of the Villarrica
watershed. Forest fragment shape irregularity at 1,206 ha, forest extent at 180 ha and
shrubland extent at 180 ha (the latter with a negative association), were the resources
present in the best-supported models. A previous study suggested a positive association
between forest fragment irregularity and G. nana occupancy (Farias and Jaksic, 2011). As
reported for G. gnoma and G. brasilianum, this higher irregularity, as generated by riparian
zones and linear human structures (e.g. roads and fences), may provide these owls with
different resources and hunting perches with extensive views (Campioni et al., 2013;
Piorecky and Prescott, 2006).

3.4.2. Niche width of forest owls


The species distribution models for the study area and analyses of resource
utilization and peak performances for forest owls, indicate that S. rufipes use a narrower
width of environments than G. nana. Similar to an evaluation of habitat-specialization on
other top-predators (e.g. Lynx canadensis and L. rufus), my results of niche width for forest
owls best fit the similar level of peak performance alternative model of niche width (Fig.
3.1c; Peers et al., 2012).
The ability of species to exploit a range of resources and their performance using
each one have usually been approached using the trade-off model that some species are the
jack of all trades (i.e. species that use a greater diversity of resources perform less well on
average) and either the master of none (Caley and Munday, 2003), master of some
(Richards et al., 2006) or master of all (Barkae et al., 2012). For example, Caley and
Munday (2003) reported that specialist coral reef fishes grew faster than generalists in one
or two habitats, but the growth rate of generalists was more consistent between habitats. In
a small-mammal assemblage, the habitat-specialist Ochrotomys nuttali showed stronger
selection of one micro-habitat whereas generalists Tamia striatus and Peromyscus leucopus
were able to exploit a range of micro-habitat types; however, Ochrotomys outperformed the
generalist rodents in the habitat where they were specialized (Dueser and Hallett, 1980).
Studying top-predator felids, Peers et al., (2012) found that the specialist L. canadensis did

36
not have a narrower width for each resource gradient compared to the generalist L. rufus,
but rather had a wider width and higher performance within a subset of resources. My
results were a mixture of these scenarios as I found peak performance estimates for S.
rufipes were slightly higher over a select number of resources associated with stand-level
forest complexity (sensu McElhinny et al., 2005) and forest stability at the landscape scale.
For their part, G. nana had higher estimates of peak performance for resources related to
human-induced forest degradation and landscape fragmentation. Because occurrence of
habitat-specialists is associated with a subset of niche resources, these species usually have
lower occurrence rates across the landscape as there are a smaller number of habitats in
which they perform highly (Devictor et al., 2010; Peers et al., 2012). The projected
distribution of both owls goes in this direction as areas of high habitat suitability for the
habitat-specialist owl were mostly associated with higher elevation forests close to the
Andes Range and located inside or surrounding protected areas.

3.4.3. Recommendations for management


The worldwide decline of habitat-specialist species is a symptom of current global
processes of biotic homogenization (Olden et al., 2004). In Chile, conservation practices
to secure long term survival of the habitat-specialist S. rufipes, that extend beyond
protected areas, are urgently needed for sites where forestry and agricultural activities take
place (Ibarra et al., 2012; Martnez and Jaksic, 1996). My results suggest that these owls
may benefit if management actions are tailored at the stand-level, but that landscape context
also needs to be considered.
Forest management that maintains multi-aged stands with a variety of tree sizes (SD
of DBH = 19.9 9 cm), including large old-growth trees, with relatively high bamboo
understorey cover (34.2 26.6%), will promote high occurrence of S. rufipes. Furthermore,
landscapes that contain forest cover > 63.5% would also promote occurrence by these
habitat-specialists. These desired habitat attributes might be reached by either dispersed or
aggregated retention of large and small trees (the latter for a continuous supply of large
trees over forest generations), together with a dense bamboo understorey maintained by gap
release (Gustafsson et al., 2012). By incorporating these recommendations, forest and

37
wildlife managers will be better able to meet the requirements of habitat-specialist owls and
will likely provide for the generalist G. nana as well, and will also benefit other avian
habitat-specialists of conservation concern in South American temperate forests (Daz et
al., 2005; Reid et al., 2004).

3.5. Conclusions

In this chapter, I combined occupancy models with niche theory to assess


occurrence patterns and compare niche width of two forest owls. Predicted occurrence
probabilities for the habitat-specialist owls varied markedly across sites and were positively
associated with stand-level forest complexity. Habitat-generalist owls showed relatively
high occurrence probabilities across sites and a uniform level of habitat suitability across
the study area. Relative to habitat-generalist owls, the specialists had lower total resource
utilization due to lower occurrence rates over gradients of all covariates, but achieved
similar peak occurrence for resources related with stand-level forest complexity and forest
stability at the landscape scale. Forest management practices that maintain multi-aged
stands with large trees and high bamboo cover will benefit both owls. In Chapter 4, I use
the predicted occurrence rates across sites for both owls to assess whether these owl species
can reliably be used as surrogates for taxonomic diversity, endemism and functional
diversity in temperate forests.

38
Fig. 3.1. (a) Specialized species (continuous line) have smaller niche widths than
generalized species (dashed line) across resource gradients. Specialists can reach either a
(b) higher level of performance (i.e. traditional model of relative niche width between
specialists and generalists) or (c) similar level of performance (i.e. alternative model of
relative niche width) than generalists, under a subset of resources that are relatively stable.

39
Fig. 3.2. Averaged predictions of occurrence for (a) Strix rufipes and (b) Glaucidium nana
in the Villarrica watershed of the La Araucana Region, Chile. Red depicts areas of higher
habitat suitability whereas yellow represents areas of lower habitat suitability or probability
of occurrence. Alpine areas (grey), large lakes (blue) and boundaries of public protected
areas (green), are shown.

40
Fig. 3.3. Predicted probabilities of occurrence of Strix rufipes and Glaucidium nana in
Andean temperate forests (2011-2013), in relation to environmental resources. The curves
are representative of models developed for each owl using only the corresponding resource.

41
Fig. 3.4. (a) Total resource utilization (i.e. average area under the curve SE) and (b)
predicted peak performance SE, for Strix rufipes (grey bars) and Glaucidium nana (white
bars) in relation to habitat niche resources.

42
Table 3.1. Stand- and landscape-level resources used to evaluate habitat associations of
forest owls in Andean temperate forests.
Resource Abbreviation for Description
models
Stand-level
Tree density (trees/ha) Tre Density of all trees with DBH > 12.5 cm
Standard deviation of Dbh SD of tree DBH measures the variability in tree size, and
tree diameter at breast was considered indicative of the diversity of micro-habitats
height (DBH, cm) a within a stand for both owl and potential prey (Van Den
Meersschaut and Vandekerkhove, 2000)
Canopy cover (%) a Can Proportion of sky covered by canopy estimated from the
centre of the plot
Volume of coarse woody Cwd Calculated based on the length and diameter of each piece
debris (VCWD) a with diameter > 7.5 cm crossing a transect of 22.4 m length
(oriented N-S)
Bamboo understorey Und Density of bamboo vegetation up to 3 m high, expressed as
density (NC) a the number of contacts (NC) using the method described in
Daz et al., (2006), quantified at five points of a transect of
22.4 m length (oriented N-S)
Elevation (m.a.s.l) Ele Meters above sea level measured at the center of the plot

Landscape-level
Forest areas (180 haa / For180 / % Extent of forested area
1,206 ha) For1206
Shrubland areas (180 haa Shr180 / % Extent of shrubland
/ 1,206 ha) Shr1206
Core habitat (180 ha / 1 Cor180 / Mean size of interior core habitat ( 100m from polygon
206 ha) Cor1206 edge) of all forest patches in plot
Forest-patch shape index Si180 / Si1206 Si = 0.25 x p/A, where p = forest-patch perimeter and A =
(Si) (180 haa / 1,206 haa) forest-patch area. Si is an estimator of forest-patch shape
irregularity and edge effects, describing the extent to which
patches depart from a geometrically simple compact
configuration of the same area (for raster maps, square: Si
= 1)
Relative habitat diversity Hd180 / Relative habitat diversity within a circular plot measured
(180 haa / 1,206 ha) Hd1206 as Shannons diversity index, which equals zero when
there is only one patch and increases as the # of patch
types or the proportional distribution of patch types
increases
a
Covariates retained for tests of habitat associations of forest owls after reducing collinearity.

43
Table 3.2. Model selection statistics based on Akaikes Information Criterion (AIC) for
estimating probability of occurrence () and detection (p) of two owl species (a) Strix
rufipes and (b) Glaucidium nana in Andean temperate forests. Only the top model set with
AIC values < 4 are shown. Parameter estimates are listed in the order of variable under
Model structure column, and beta estimates in bold font have 95% confidence intervals that
do not overlap 0.
Parameter Estimated
Species Model structure Kc AIC AIC d Wi e
estimates 95% CI
a
(a) Strix rufipes (Dbh + Und + Can), p 10 1070.16 0.00 0.27 2.59, 0.98, 2.07 0.39, 4.79
0.02, 1.94
-0.85, 5.00
a
(Dbh + Und), p 9 1070.17 0.02 0.27 3.28, 1.03 1.27, 5.29
0.09, 1.97
a
(Dbh + Und + For180), p 10 1070.81 0.65 0.20 2.82, 0.95, 1.55 0.68, 4.97
-0.02, 1.93
-1.08, 4.17
a
(Dbh + Und + For180 + Can), p 11 1071.32 1.17 0.15 2.29, 0.90, 1.25, -0.01, 4.58
1.82 -0.10, 1.90
-1.94, 3.94
-1.15, 4.79
(Dbh + For180), pa 9 1073.17 3.01 0.06 3.07, 2.43 1.05, 5.10
-0.15, 5.01

(b) Glaucidium (Si1206 + For180 + Aut), pb 12 1258.64 0.00 0.38 0.72, 0.17, 1.84 -2.43, 3.87
nana -2.63, 2.96
-2.00, 3.89

(Si1206 + For180), pb 11 1259.16 0.52 0.29 2.74, 0.44 -4.31, 9.78


-2.37, 3.25
b
(Shr180 + For180 + Aut), p 12 1259.92 1.28 0.20 -0.57, 0.51, 1.94 -5.81, 4.67
-2.63, 3.64
-0.41, 4.30
b
(Shr180 + For180), p 11 1260.80 2.16 0.13 -1.84, 0.79 -6.61, 2.94
-2.19, 3.77

44
a
p(Ml + No + Ow)
b
p(Wi + Ml + No + Da2 + Ow). Important detection covariates were identified in Chapter 2 using the same
data set, and consisted of moonlight (Ml), environmental noise (No), whether the other owl species was
detected at the unit for the specific survey (Ow), wind speed (Wi) and number of days since start of surveys
(Da).
c
Number of parameters estimated.
d
AIC is the difference in AIC values between each model and the lowest AIC model.
e
AIC model weight.

45
Table 3.3. Mean (SD) values of habitat resources associated with probabilities of
occurrence categorized as low (0-0.33), moderate (0.34-0.66) and high (0.67-1) for Strix
rufipes in Andean temperate forests, based on model-averaged predictions.
Predicted probability of occurrence
Habitat resource 0-0.33 0.34-0.66 0.67-1
Stand-level
Tree density (#/ha) 225.4 (223.2) 435.7 (300.0) 487.4 (244.7)
SD of diameter at breast height (cm) 5.4 (2.4) 11.2 (2.0) 19.9 (9.0)
a
Bamboo understorey density (NC) 0.1 (0.1) 0.2 (0.3) 2.8 (2.7)
Volume of coarse woody debris (m3) 0 (0.0) 0.2 (0.7) 0.4 (0.4)
Canopy cover (%) 29.9 (22.4) 50.3 (18.8) 65.5 (17.8)
Landscape-level
Forest extent 180 ha (%) 26.4 (20) 49.9 (23.1) 63.5 (23.5)
Forest extent 1,206 ha (%) 35.3 (17.4) 53.2 (19.5) 66.5 (20.0)
Forest shape index 180 ha 1.2 (0.2) 1.3 (0.3) 1.4 (0.3)
Forest shape index 1,206 ha 1.4 (0.1) 1.4 (0.1) 1.6 (0.5)
Shrubland extent 180 ha (%) 27.2 (15.7) 22.9 (14.4) 17.2 (13.4)
Shrubland extent 1,206 ha (%) 24.8 (8.8) 22.2 (9.7) 16.8 (10.4)
Habitat diversity 180 ha (Shannon index) 0.9 (0.4) 1 (0.4) 0.7 (0.4)
Habitat diversity 1,206 ha (Shannon index) 1.1 (0.2) 1.1 (0.2) 0.8 (0.4)
Core habitat 180 ha (# of ha) 5.4 (10.0) 19.1 (26.5) 40.8 (41.6)
Core habitat 1,206 ha (# of ha) 107.9 (105.3) 223.4 (192.2) 389 (271.3)
a
For easier implementation in forest management, the values of this resource in percentage of coverage
approximate 1.6 3% (low), 3.4 5.4% (moderate) and 34.2 26.6% (high probability of occurrence).

46
Chapter 4. Reliability of owls as surrogates for biodiversity in South
American temperate forests

4.1. Introduction

To prevent the further loss of species from eco-regions that are subject to increasing
rates of anthropogenic degradation, conservation biologists often invoke the use of
biodiversity surrogates to identify sites in need of protection (Caro, 2010). Surrogates are
organisms with parameters (e.g. occurrence) that can be used as proxy measures of
biodiversity status and trends (Lindenmayer and Likens, 2011); these species can be used to
monitor the effects of management on other species (Caro, 2010). Criteria on which
surrogates are being selected for testing should be specified explicitly and the surrogate
candidate should meet as many of the criteria as possible (Caro and ODoherty, 1999). For
a species to be a reliable surrogate key criteria should be (i) represent either ecological or
conservation important phenomena, (ii) sufficiently sensitive to indicate anthropogenic
habitat degradation, (iii) able to provide an estimate of the status of target biodiversity
across wide environmental gradients, (iv) easy and cost-effective to survey and (v)
distributed over a broad geographical area (Caro, 2010; Noss, 1990).
A suite of studies conducted in the northern hemisphere has shown that avian top
predators possess useful characteristics as biodiversity surrogates. Top predators represent
the apex of food chains with effects that can cascade through the ecosystem, require large
areas that encompass the ranges or populations of less area-demanding species, often select
areas with high structural complexity and provide early indications of habitat degradation
when they decline (see Sergio et al., 2008 for details). However, not all members of the
top predator guild will be reliable surrogates because those that are habitat-generalists
may use biologically degraded habitats (Ozaki et al., 2006; Rodrguez-Estrella et al., 2008).
The spatial correlation of single predator occurrences and high taxonomic diversity
has supported the use of predators as biodiversity surrogates in forest ecosystems (Burgas
et al., 2014; Sergio et al., 2006). However, conservation focused on taxonomic diversity

47
(e.g. species richness) alone will not necessarily maintain stable levels of ecosystem
functioning (Daz and Cabido, 2001; Naeem et al., 2012). Conservation approaches based
on functional diversity, as a complement to taxonomic diversity, include the value and
range of functional traits (e.g. phenological, behavioural, physiological or morphological)
present in a community, and therefore link diversity with ecosystem stability and processes
(Daz et al., 2007; Julliard et al., 2006; Trivellone et al., 2014). For example, higher
densities of habitat-specialist species in a community engenders higher complementarity in
resource utilization, potentially increasing ecosystem productivity (Tilman et al., 2001) and
indicating higher ecosystem stability (Julliard et al., 2006).
Taxonomic diversity may also not be spatially correlated to biodiversity hotspots
where exceptional concentrations of endemic species coexist (Kerr, 1997). Endemic species
commonly have small populations and few areas assigned to conservation; thus, they are
intrinsically vulnerable to extinction (Gaston, 1998). Functional diversity and endemism are
under even greater threat from anthropogenic habitat degradation than overall species
richness (Flynn et al., 2009; Myers et al., 2000); therefore, identifying the potentially
contrasting role of specialist and generalist predators as surrogates for functional diversity
and endemism would contribute significantly to conservation biology.
Few studies have attempted to elucidate the ecological mechanisms behind a spatial
correlation between top predators and enhanced biodiversity (Lindenmayer and Likens,
2011). Some forest structural attributes increase stand-level complexity (sensu McElhinny
et al., 2005), providing the necessary conditions for wildlife in general, of which predators
are a component, and hence result in a positive association between the surrogate candidate
and target biodiversity (Drever et al., 2008; Lindenmayer et al., 2014a). A further limitation
of previous studies arises when researchers assume that the species of interest, either the
top predator or target biodiversity, are absent from a site when they were not detected. The
probability of detecting both avian top predators and their prey depends on many factors,
including wind speed, temperature, and date and time of survey (Ibarra et al., 2014b). Thus,
the failure to detect individuals at a site does not mean they were absent unless detection
probability is perfect (MacKenzie et al., 2003). The assumption of perfect detectability can

48
result in misleading inferences about the consistency of cross-taxon congruence and,
therefore, of biodiversity surrogacy relationships.
To test whether top predators can be employed reliably to achieve ecosystem-level
conservation targets in South American temperate forests, I (i) compare the reliability of
Strix rufipes (habitat-specialists) and Glaucidium nana (habitat-generalists) as surrogates
for taxonomic diversity, endemism and functional diversity, and (ii) examine whether and
which surrogate candidate and target biodiversity measures have similar habitat correlates
and responses to anthropogenic habitat degradation. I predict (i) the habitat-specialist owl
will outperform the generalist as a surrogate for all target biodiversity measures and (ii)
forest stand-level complexity will be positively correlated with both the occurrence of
reliable surrogate owls and target biodiversity. To test my predictions I used models that
adjust for detectability of both the surrogate candidates and the target biodiversity
measures, providing the basis for stronger inferences by removing the need to rely on
assumptions of perfect detectability.

4.2. Methods
4.2.1. Field methods

Nocturnal raptor surveys


See Chapters 1 and 2.

Bird surveys and measures of functional diversity


To test the congruence of owl-target biodiversity relationships, I compared
probability of occurrence of owl species with three measures of avian biodiversity (i)
taxonomic diversity, (ii) endemism and (iii) functional diversity. I chose birds because they
are often used in biodiversity monitoring and are known to be affected, either
taxonomically or functionally, by habitat degradation in temperate forests and elsewhere
(Daz et al., 2005; Flynn et al., 2009). Five multispecies point-transect surveys
systematically separated by 125 m from adjacent point-transects were established along
narrow trails within each of the 101 sites. Each site was surveyed once, from November to

49
January, over the two seasons. Eighty-one sites (80.2%) were surveyed in 2011-2012 and
20 sites (19.8%) were surveyed in 2012-2013. Therefore, I conducted 505 point-transects of
50 m radius, recording all species heard or seen for a period of 6 min. The distances to all
birds detected were recorded and grouped into two distance intervals (0-25 and 26-50 m)
for analysis. Surveys were conducted from dawn to 10:30 h. At each point-transect, I
recorded date, time, temperature (C) and wind speed (m/s) using a hand-held weather
monitor (Kestrel 4200, Kestrel-meters, Birmingham, MI). I assigned habitat type within 50
m of each point-transect as follows: old growth, mid-successional or early successional
forest, mixed shrubland, exotic forestry plantation or openfield.
I used both discrete and continuous measures of functional diversity (Petchey and
Gaston, 2006). I followed Daz et al., (2005) to categorize each species into habitat-use
guilds based on its primary use of temperate forest structure for nesting and/or feeding at
the stand-level. These guilds were: large-tree users, understorey users, vertical-profile
generalists and shrub users. Further, I followed Altamirano et al., (2012) to classify species
into cavity-nesting and non cavity-nesting species, as the former guild is known to be
sensitive to logging (Drever et al., 2008). For continuous functional traits, I followed
Julliard et al., (2006) to quantify the degree of habitat-specialization for a species (SSI) as
the coefficient of variation (standard deviation/average) of its estimated densities across the
six habitat types described above.

Vegetation and habitat measures


I used previous studies conducted in southern temperate forests to choose habitat
attributes (hereafter covariates) associated with stand-level complexity that could drive a
spatial correlation between owls and avian diversity (Daz et al., 2005; Ibarra et al., 2012;
Martnez and Jaksic, 1996; Reid et al., 2004). Around each point count-transect, I
established a vegetation plot (22.4 m diameter) with the point count-transect located at the
centre of the plot (N = 505 plots; Table 3.1). As forest raptors commonly require different
habitat patches for breeding and foraging, for owl analyses I added spatial covariates that
were correlated with occurrence of raptors in other studies (see Chapter 3 for details).

50
4.2.2. Data analysis

Modelling owl occurrence and bird density


Presence/absence data for owls were analyzed using a multi-season occupancy
framework (MacKenzie et al., 2003), and bird counts using distance sampling in a
multinomial-Poisson mixture model framework (MPMM; Royle et al., 2004). I used the
program R-Unmarked (Fiske and Chandler, 2011), which uses maximum-likelihood
methods to estimate probabilities of occurrence () and detection (p) for owls (Chapter 3),
and density (D) and detection (p) for other bird species.
To model D for each bird species (other than owls), I assessed collinearity and
reduced the number of covariates (Table 3.1; same protocol as for owls). I did not use
collinear (r > 0.7) covariates in the same model. I used four covariates (SD of tree DBH,
canopy cover, volume of coarse woody debris and bamboo density) to model D. I first used
AIC to identify whether the half-normal or the hazard- rate was the most suitable distance
function for each species (Royle et al., 2004). The half-normal function always received
stronger support; thus, I used it in all further analyses. To estimate detectability (p), I used
four covariates potentially associated with p (covariates hypothesized to affect the scale
parameter of the detection function): date (number of days since start of surveys in
October), time of survey (minutes since 05:00 h), wind speed (m/s) and temperature (C).
For each species, I used the stepwise covariate selection procedure (without parameterizing
D) described above, and then ranked each model by AIC to select top-ranked models for
further D modelling. To obtain the best models for D (covariates affect the Poisson mean)
for each bird species, I used a stepwise covariate selection procedure linked to R-Unmarked
to create a candidate set of models based on model weights (wi) and the precision of the
estimated coefficients, using an information-theoretic approach (Akaikes Information
Criterion [AIC]; Burnham and Anderson, 2002). Models within 2 AIC units of the top
model were considered as the competitive set of best-supported models (Burnham and
Anderson, 2002). I evaluated a range from 16 to 20 D models for each bird species. Using
the Akaike weights as a weighting factor, I averaged models with AIC 4 in the final
confidence set for each bird (Burnham and Anderson, 2002). Averaged models were used

51
to predict bird D for each point-transect. Values of D for the five point-transects conducted
per site were averaged to obtain one value of D/site for each species.

Testing owl surrogacy


I used generalized linear models (GLMs) to relate the averaged values of habitat-
specialist and generalist owls (derived from Chapter 3) to avian diversity measures across
sites. I also tested second order polynomial models as the association between surrogates
and target biodiversity may follow curvilinear relationships (Lindenmayer et al., 2014a). I
first explored bird richness as a response variable, without accounting for detectability, to
examine whether the most commonly reported correlation between my two test species and
taxonomic diversity (Burgas et al., 2014; Jenkins et al., 2012; Sergio et al., 2006) was also
observed in my system. I then related owl occurrence () to (i) D of endemic species, (ii) D
of avian habitat-use guilds, (iii) D of cavity-nesting species and (iv) community
specialization index (CSI) across sites, where CSI for a site k was



= ,

with N as the total number of bird species in the analysis, SSI as the species specialization
index for species i, and D as the estimated density of species i (Devictor et al. 2008). To
assess the strength of evidence for each tested model, we calculated the value of AIC for
small sample sizes (AICc) and model weight (wi). The latter was used to compare pairs of
models by calculating evidence ratios (Burnham and Anderson 2002). All bird data were
log10 (x + 1) transformed before statistical analyses to improve normality and variance
homogeneity. Species with too few observations to use multinomial-Poisson mixture
models were excluded from the analysis.

52
4.3. Results

I recorded 292 detections of S. rufipes and 334 detections of G. nana across 1,145
owl surveys at 101 sites over two years (Chapter 2). Probabilities of occurrence (; mean
standard error) across sites ranged from 0.05 0.04 to 1.00 0.00 for S. rufipes, and from
0.67 0.18 to 0.98 0.04 for G. nana (Chapter 3).
I recorded 48 species of birds (other than owls). Among these, 21 allowed
estimation of their density (D) using a multinomial-Poisson mixture model (Table 4.1).
Nine (42.9%) species inhabit areas restricted to the southern portion of South America and
eight (38.1%) were endemic to South American temperate forests. Seven (33.3%) species
were large-tree users, six (28.6%) vertical profile generalists, four (19%) understorey users
and four (19%) shrub users. Eleven species (52.38%) were cavity-nesters. The highest
degrees of habitat-specialization were for austral parakeets (Enicognathus ferrugineus; SSI
= 2.68) and Magellanic woodpeckers (Campephilus magellanicus; SSI = 1.96), and the
lowest for austral thrush (Turdus falcklandii; SSI = 0.03) and fired-eyed diucon (Xolmis
pyrope; SSI = 0.09; Table 4.1).

4.3.1. Spatial relationships: owls and target biodiversity


All biodiversity measures had stronger associations with the specialist owl S. rufipes
than with the generalist owl G. nana. Models using model-averaged of S. rufipes as an
independent variable for species richness, D of endemic species and measures of functional
diversity, had stronger support from the data than those using of G. nana, according to
AICc values and model weights (wi; Table 4.2; Fig. 4.1). For overall species richness, the
evidence ratio for the linear model S. rufipes relative to a non-linear model was 2.93,
indicating these two models had similar support, such that model fit did not improve with
the addition of a second degree quadratic function of S. rufipes (Table 4.2). Similarly, for
the D of shrub users I found that the evidence ratio for the non-linear model S. rufipes + S.
rufipes2 relative to a linear model was 1.95, indicating these two models had similar
support. The relationship between D of shrub users and of S. rufipes was negative. There
was strong evidence of non-linearity (second degree quadratic function) for the correlation

53
between of S. rufipes and the D of (i) endemic species, (ii) cavity-nesting species, (iii)
large-tree users, (iv) understorey users and (v) the community specialization index (Table
4.2). For all these response variables, I found peak values when of S. rufipes
approximated a value of 1 (Fig. 4.1). For easier interpretation, I calculated the estimated
values of target avian biodiversity that spatially co-occurred with predicted low (0-0.33),
moderate (0.34-0.66) and high (0.67-1) of S. rufipes in Andean forests based on averaged
model predictions (Table 4.3).

4.3.2. Ecological mechanisms: habitat correlates for owls and target biodiversity
Model selection results indicated that for S. rufipes responded positively to both
the variability (SD) in the DBH distribution of trees (logit-scale estimates: intercept [SE] =
-3.29 [0.93], beta coefficient = 2.59 [1.12]) and bamboo understorey density (beta
coefficient = 0.98 [0.49]; Chapter 3). For G. nana, the spatial auto-covariate term (beta
coefficient [SE] = 1.84 [1.04]) controlled for intra-landscape data dependence as it
improved the AIC weight of best models (0.09 and 0.07 units higher than models without
the Aut term). The best models for G. nana indicated that responded positively to both
forest shape index at 1,206 ha (intercept [SE] = -1.09 [2.22], beta coefficient = 0.72 [1.61])
and forest cover extent at 180 ha (beta coefficient = 0.16 [1.43]), and negatively to
shrubland cover at 180 ha (intercept [SE] = -0.02 [1.47], beta coefficient = -0.57 [2.67];
Chapter 3). However, beta coefficients for all covariates were non-informative for G. nana
as they overlapped zero, and thus were not good predictors for their occurrence (see
Chapter 3 for details).
Model selection results for all other birds indicated that canopy cover, followed by
the variability (SD) in the DBH of trees were the most frequent forest stand-level
components associated with density of birds, with twelve species responding negatively and
three positively to canopy cover. Nine species responded positively to SD in DBH of trees,
three species responded positively to bamboo density, and two species responded
negatively and one responded positively to volume of coarse woody debris (Table 4.1).

54
4.4. Discussion

4.4.1. Evidence for a reliable surrogacy relationship


I found that owls were reliable predictors of local avian endemism, species richness

and functional biodiversity in South American temperate forests, consistent with the

hypothesis that avian top predators act as reliable biodiversity surrogates at the forest stand-

level (Burgas et al., 2014; Sergio et al., 2006). However, my empirical assessment stresses

that the degree of habitat-specialization of owls will likely affect their reliability as

biodiversity surrogates (Cabeza et al., 2007). This is the first study that has accounted for

detectability of both surrogates and target biodiversity. My results indicate that habitat-

specialist S. rufipes always had a stronger relationship with target biodiversity than did the

habitat-generalist G. nana. The broad range of predicted occurrence probabilities () across

our sites for specialist owls suggests that they may act as surrogates across a wide range of

sites in temperate forests: from highly degraded habitat to structurally complex old-growth

forests, with low and high values for both owl occurrence () and avian diversity,

respectively. In contrast, the relatively high for generalist owls across sites makes them

less reliable surrogates. This result for the generalist owl is similar with patterns suggested

by Ozaki et al. (2006) for Accipiter gentilis, who found that this raptor was not an efficient

biodiversity surrogate because it often used anthropogenically degraded habitat that were

poor in taxonomic diversity. The latter study, however, did not account for the detectability

of both the surrogate candidate and target biodiversity, nor tested for other dimensions of

target biodiversity such as endemism or functional diversity.

55
My results suggest that the habitat-specialist S. rufipes could be used as surrogate of
avian endemism. Biodiversity in South American temperate forests evolved in isolation
from other similar forests within the continent since at least the late Tertiary (Axelrod et al.,
1991). This isolation explains the particularly high rates of endemism (e.g. 41% for forest
bird species; Vuilleumier, 1985). I included a relatively high proportion (38.1%) of
endemics, all of which have been reported as habitat-specialists in previous studies (Daz et
al., 2005).
The use of functional diversity rests on the assumption that the density of functional
traits will probably provide insight into ecosystem stability and processes beyond that given
by taxonomic diversity (Devictor et al., 2010; Petchey and Gaston, 2006). Focusing on a
diversity of functional traits rather than on species richness facilitates the synthesis between
community ecology and ecosystem ecology (e.g. monitoring from functional traits through
niche relationships to communities can generate links to an ecosystem-based view), and
allows conservation biologists to make predictive statements of community assembly that
may help policy makers make informed conservation decisions (McGill et al., 2006). I
chose discrete and continuous habitat-specialization traits for all target avian species, which
can inform us about niche relationships and ecosystem stability (Julliard et al., 2006). For
example, in a niche context, species fall on different places along a specialist-generalist
continuum, with specialists being favoured under relatively stable conditions and
generalists under unstable conditions and degraded habitats (Devictor et al., 2010). I found
that measures of avian functional diversity that were associated with habitat-specialization
increased non-linearly and peaked at sites with highest occurrence of habitat-specialist
owls. Julliard et al., (2006) also found that specialist species tend to disproportionately
aggregate at sites that are expected to be more stable. My results suggest that such a pattern
of specialist aggregation is driven by stand structural complexity in temperate forests.

4.4.2. Untangling ecological mechanisms


Understanding the mechanisms underpinning surrogacy relationships is a key
component for improving the use of the surrogate species in conservation efforts
(Lindenmayer and Likens, 2011). In Andean temperate forests, relatively low densities of

56
canopy cover, broad tree-size (DBH) distribution and high density of bamboo understorey
may drive a positive correlation between owls and avian diversity. When compared to early
and mid-successional forests, structurally complex older forest stands tend to have less
dense canopy cover because of a mosaic of canopy gaps produced by tree-falls in mid-
elevations (500-900 m altitude), and naturally open canopies (54-81% canopy cover) and a
lack of shade-tolerant trees in high-elevations (> 900 m altitude; Ibarra et al., 2012; Veblen
et al., 1980). Complex Andean forest-stands are also characterized by their broad tree-size
distribution, with relatively high frequency of large old-living and dead trees, combined
with a dense clumpy bamboo understorey in mid-elevations and a homogeneously
distributed bamboo understorey in high-elevation stands (Veblen et al., 1980).
Stand structural complexity in Andean temperate forests relates to the habitat
requirements of both the surrogate owl, S. rufipes, and target biodiversity species. For
example, nesting S. rufipes occupy cavities available in large trees (mean DBH = 122.8
36.2 cm) that are > 100 years old (Beaudoin and Ojeda, 2011). Similarly, species in the
large tree-user guild of birds also rely on large trees for both nesting and foraging (Daz et
al., 2005). Most cavity nesters (N = 28 species) in southern temperate forests nest in
cavities created by natural tree-decay processes (75% of cavity-nests), with large dead trees
(DBH > 57.26 cm) as the most common (58%) nest substrate (Altamirano, 2014). Cavity-
excavating species C. magellanicus and stripped woodpeckers (Veniliornis lignarius) feed
disproportionately more on larger and more decayed trees than on smaller diameter trees
(Altamirano, 2014; Ojeda et al., 2007). Other cavity excavators such as Chilean flickers
(Colaptes pitius) and white-throated treerunners (Pygarrhichas albogularis) show a strong
preference for large snags or large dead branches for nesting (Altamirano, 2014). These
four excavators produce cavities that are subsequently used by several secondary cavity
nesters, including owls, parakeets, swallows, rayaditos, ducks and small mammals
(Altamirano et al., 2012; Ibarra et al., 2014a). I found that S. rufipes was more likely to
occur in forest stands with a relatively dense bamboo understorey, similar to results from
previous studies on these owls (Chapter 3; Ibarra et al., 2012; Ibarra et al., 2014c; Martnez
and Jaksic, 1996). Native bamboo provides habitat for arboreal and scansorial rodents and
marsupials, which constitute the main prey of S. rufipes in temperate forests (Figueroa et

57
al., 2006). Bamboo understorey has also been identified as providing critical protective
cover and feeding habitat for ground-gleaner birds such as black-throated huet-huets
(Pteroptochos tarnii) and Chucao tapaculos (Scelorchilus rubecula), and leaf-gleaners
Magellanic tapaculos (Scytalopus magellanicus) and Des Murs`s wire-tails
(Sylviorthorhynchus desmursii); all species with poor flying abilities (Reid et al., 2004). In
this study, bamboo understorey was a good predictor of the density of P. tarnii and S.
rubecula. Further, the four understorey users were positively correlated with the variability
(SD) in DBH. Because P. tarnii, S. rubecula and S. magellanicus nest in cavities available
in trees with DBH ranging from 61.3 to 193.8 cm (Altamirano, 2014), my results suggest
that understorey users, as well as S. rufipes, require stands that combine a relatively dense
bamboo understorey with large old-living trees and snags (Chapter 3); therefore, the co-
occurrence of these species results from similar habitat requirements.
Because specialist predators are more sensitive to habitat degradation, meeting their
needs is also expected to provide the requirements of generalist species (Landres et al.,
1988). This may be the case in the study system as habitat-specialist owls were positively
associated to species included in the vertical-profile generalist guild. These species use the
entire vertical profile of forests (canopy, sub-canopy and understory vegetation) for most of
their activities (Daz et al., 2005). Interestingly, I found a negative spatial correlation
between forest-specialist owls and species included in the shrub user guild, which
comprises species that exploit degraded areas but occasionally use forests; therefore, S.
rufipes can be potentially considered as anti-surrogates (sensu Lindenmayer et al., 2014b)
for species using degraded or open stands in temperate forests.
Reliable surrogates should not be selected exclusively on the basis of whether they
are specialists or generalists; they should exhibit a confirmed association to habitat
attributes of interest (Landres et al., 1988). I focused on stand-level forest attributes as
drivers of surrogacy relationships for the following reasons: (i) although there are marked
floristic compositional changes across southern temperate forests from 35 to 55 S
(Gajardo, 1993), bird species inhabiting these ecosystems show relatively broad patterns of
altitudinal and latitudinal distribution; thus, most species occur across the temperate forest
range (Vuilleumier, 1985), (ii) the degradation of structural attributes of temperate forests

58
rather than the lack of certain plant species is affecting habitat quality and long-term
suitability for several species inhabiting this biodiversity hotspot (Daz et al., 2005; Reid et
al., 2004), and (iii) although a surrogacy relationship between owls and biodiversity can
exist at broader landscape-levels (Sergio et al., 2004a), I tested whether this relationship
existed at the stand-level where logging takes place. Further research is warranted on
broader-scale surrogacy relationships in South American temperate forests, but the
consistency of the cross-taxon congruence will need reliable spatial datasets on biodiversity
distribution (Margules and Pressey, 2000), which are commonly not available for areas
subject to increasing levels of anthropogenic disturbance (Rodrigues and Brooks, 2007).

59
Fig. 4.1. Relationship between probabilities of occurrence () of habitat-specialist owls

Strix rufipes and (i) species richness/site, (ii) density (individuals/ha x site) of species in

different habitat-use guilds, and (iii) community specialization index/site, in Andean

temperate forests, 2011-2013.

60
Table 4.1. Avian species with their geographical and ecological attributes, and stand-level covariates associated with the density
(D) of bird species in Andean temperate forests, according to model selection statistics based on Akaikes Information Criterion
(AIC). Parameter estimates [SE] for covariates present in the top model set with AIC values < 2 and with estimates of their
95% confidence intervals that do not overlap 0, are shown. + and - indicate the direction of the relation.
Forest stand-structural components

Habitat- Cavity- Species Volume of Bamboo SD of diameter


Canopy
Geographic use nesting specialization coarse woody understorey at breast height
a b c 3
cover (%)
Name distribution guild species index (SSI) debris (m ) density (NC) of trees (cm)
Chilean pigeon SSA VPG N 0.12
(Patagioenas araucana)
Austral parakeet E LTU Y 2.68 + 0.07 [0.03]
(Enicognathus ferrugineus)
Green-backed firecrown SSA VPG N 0.23 - 0.21 [0.11] - 0.12 [0.03]
(Sephanoides sephaniodes)
Striped woodpecker SSA LTU Y 1.21 - 0.64 [0.15] + 0.07 [0.02]
(Veniliornis lignarius)
Chilean flicker SSA LTU Y 0.37 - 0.21 [0.08]
(Colaptes pitius)
Magellanic woodpecker E LTU Y 1.96 - 1.51 [0.35] + 0.19 [0.09]
(Campephilus magellanicus)
Thorn-tailed rayadito E LTU Y 0.41 - 0.23 [0.02]
(Aphrastura spinicauda)
Des Murs`s wire-tail E UU N 0.72 + 0.63 [0.31] + 0.04 [0.01]
(Sylviorthorhynchus desmursii)

61
Forest stand-structural components

Habitat- Cavity- Species Volume of Bamboo SD of diameter


Canopy
Geographic use nesting specialization coarse woody understorey at breast height
a b c 3
cover (%)
Name distribution guild species index (SSI) debris (m ) density (NC) of trees (cm)
White-throated treerunner E LTU Y 0.38 - 0.22 [0.04]
(Pygarrhichas albogularis)
Black-throated huet-huet E UU Y 0.96 - 0.44 [0.06] + 0.13 [0.04] + 0.04 [0.01]
(Pteroptochos tarnii)
Chucao tapaculo E UU Y 0.78 - 0.44 [0.04] + 0.05 [0.01] + 0.02 [0.01]
(Scelorchilus rubecola)
Magellanic tapaculo SSA UU Y 0.57 - 0.25 [0.05] + 0.03 [0.01]
(Scytalopus magellanicus)
White-crested elaenia SA VPG N 0.16 - 0.25 [0.07] - 0.06 [0.01]
(Elaenia albiceps)
Tufted tit-tyrant SA SU N 0.32 + 0.11 [0.04]
(Anairetes parulus)
Fire-eyed diucon SSA SU N 0.11
(Xolmis pyrope)
Chilean swallow SA LTU Y 0.23 - 0.07 [0.03] + 0.02 [0.01]
(Tachycineta meyeni)
Southern house wren PA SU Y 0.24 + 0.12 [0.03]
(Troglodytes musculus)
Austral thrush SSA VPG N 0.03
(Turdus falcklandii)

62
Forest stand-structural components

Habitat- Cavity- Species Volume of Bamboo SD of diameter


Canopy
Geographic use nesting specialization coarse woody understorey at breast height
a b c 3
cover (%)
Name distribution guild species index (SSI) debris (m ) density (NC) of trees (cm)
Patagonian sierra-finch E VPG N 0.60 - 0.34 [0.05]
(Phrygilus patagonicus)
Austral black bird SSA VPG N 0.25 + 0.09 [0.04] + 0.02 [0.01]
(Curaeus curaeus)
Black-chinned siskin SSA SU N 0.09 + 0.06 [0.03]
(Carduelis barbata)
a
E = endemic, SSA = southern South America, SA = wide spread South America, PA = Pan America (Vuilleumier, 1985).
b
SU = shrub user, VPG = vertical profile generalist, LTU = large tree user, UU = understorey user (Daz et al., 2005).
c
Y = yes, N = No. Y considered species relying on tree cavities for more than 10% of their nests (Altamirano, 2014; Altamirano et al., 2012).

63
Table 4.2. Ranking of models relating measures of avian diversity and owl probabilities of
occurrence () in Andean temperate forests.
Model specification Ka AICc AICc b Wi c -2*LLd ER e

Species richness
Strix rufipes 3 -50.10 0.00 0.75 -56.35 2.93
Strix rufipes + Strix rufipes2 4 -47.95 2.15 0.25 -56.36
Glaucidium nana 3 -11.49 38.61 0.00 -17.74
2
Glaucidium nana + Glaucidium nana 4 -9.32 40.77 0.00 -17.74
NULL 2 7.91 58.01 0.00 3.79
Density of endemic species
Strix rufipes + Strix rufipes2 4 -7.15 0.00 1.00 -15.57 518.14
Strix rufipes 3 5.35 12.50 0.00 -0.90
Glaucidium nana 3 55.55 62.70 0.00 49.31
2
Glaucidium nana + Glaucidium nana 4 57.20 64.35 0.00 48.78
NULL 2 126.92 134.07 0.00 122.80
Density of cavity-nesting species
Strix rufipes + Strix rufipes2 4 -104.24 0.00 1.00 -112.66 3610.36
Strix rufipes 3 -87.86 16.38 0.00 -94.11
Glaucidium nana 3 -40.82 63.43 0.00 -47.06
2
Glaucidium nana + Glaucidium nana 4 -38.71 65.54 0.00 -47.12
NULL 2 31.23 135.47 0.00 27.10
Density of large-tree users
Strix rufipes + Strix rufipes2 4 -110.07 0.00 0.99 -118.48 95.99
Strix rufipes 3 -100.94 9.13 0.01 -107.18
Glaucidium nana 3 -50.10 59.96 0.00 -56.35
2
Glaucidium nana + Glaucidium nana 4 -48.57 61.49 0.00 -56.99
NULL 2 20.66 130.73 0.00 16.54
Density of understorey users
Strix rufipes + Strix rufipes2 4 -20.65 0.00 1.00 -29.06 1618.26
Strix rufipes 3 -5.87 14.78 0.00 -12.12
Glaucidium nana 3 46.72 67.37 0.00 40.48
Glaucidium nana + Glaucidium nana2 4 48.51 69.15 0.00 40.09
NULL 2 118.57 139.21 0.00 114.44

64
Model specification Ka AICc AICc b Wi c -2*LLd ER e

Strix rufipes + Strix rufipes2 4 -263.98 0.00 0.94 -272.40 14.47


Strix rufipes 3 -258.64 5.34 0.06 -264.89
Glaucidium nana 3 -230.58 33.40 0.00 -236.83
2
Glaucidium nana + Glaucidium nana 4 -228.58 35.40 0.00 -237.00
NULL 2 -196.34 67.64 0.00 -200.46
Density of shrub users
Strix rufipes + Strix rufipes2 4 -337.99 0.00 0.66 -346.41 1.95
Strix rufipes 3 -336.66 1.33 0.34 -342.91
2
Glaucidium nana + Glaucidium nana 4 -312.78 25.21 0.00 -321.20
Glaucidium nana 3 -300.79 37.21 0.00 -307.03
NULL 2 -284.82 53.17 0.00 -288.94
Community specialization index (CSI)
Strix rufipes + Strix rufipes2 4 -263.76 0 1.00 -272.18 82941.05
Strix rufipes 3 -241.11 22.65 0.00 -247.36
Glaucidium nana 3 -211.39 52.37 0.00 -217.64
2
Glaucidium nana + Glaucidium nana 4 -209.71 54.05 0.00 -218.12
NULL 2 -141.79 121.97 0.00 -145.92
a
Number of parameters estimated.
b
AICc is the difference in AICc values between each model and the lowest AICc model.
c
AICc model weight.
d
-2 * log likelihood.
e
Evidence ratio among two best models.

65
Table 4.3. Estimated mean [SE] for (i) species richness/site, (ii) density (individuals/ ha x site) of different diversity measures
and (iii) community specialization index/site, associated with low (0-0.33), moderate (0.34-0.66) and high (0.67-1) probabilities
of occurrence () for Strix rufipes in Andean temperate forests, based on model-averaged predictions.

Density

Predicted for Strix Species Cavity Large- Understorey Vertical profile Shrub Community
rufipes richness Endemics nesters tree users users generalists users specialization index
Low (0-0.33) 4.77 [1.16] 3.04 [0.81] 10.30 6.48 1.95 [0.53] 15.93 [1.24] 7.46 0.38 [0.05]
[1.22] [0.91] [0.45]

Moderate (0.34-0.66) 5.49 [1.261] 4.08 [1.11] 11.89 7.64 2.66 [0.78] 17.00 [1.14] 7.01 0.44 [0.06]
[1.68] [1.22] [0.60]

High (0.67-1) 7.12 [1.57] 8.06 [3.29] 17.46 10.96 5.42 [2.23] 19.66 [2.18] 6.15 0.67 [0.19]
[4.27] [2.41] [0.57]

66
Chapter 5. General discussion and conclusions

5.1. Thesis summary

In the introduction to his book Population Ecology of Raptors, published over 35


years ago, Newton (1979) points out that Our understanding of any basic ecological
problem depends on the choice of an easy animal to study. No one who was interested
solely in the general principles of population regulation would choose to work on birds of
prey. Nocturnal birds of prey (owls) are difficult to study because they are elusive, chiefly
nocturnal, and have low population densities (Andersen, 2007). Relatively recent research
on statistical methods emphasize the need to incorporate detectability, and associated
sources of variation, in studies of owl occurrence (Kissling et al., 2010; MacKenzie et al.,
2006). This study showed that rufous-legged owls (Strix rufipes) and austral pygmy-owls
(Glaucidium nana) had similar patterns of detectability (e.g. moonlight intensity was
correlated with higher detectability of both owls, and the detection of one species was
positively correlated with the detection of the other species). Similarities in their nocturnal
prey base and tree-cavities used for nesting (Beaudoin and Ojeda, 2011; Figueroa et al.,
2006; Ibarra et al., 2014) may at least partially explain comparable patterns of detectability
for sympatric S. rufipes and G. nana (Chapter 2).
Raptors have been used increasingly as a model group in community ecology for
their value in clarifying niche relationships among sympatric species (e.g. Brambilla et al.,
2010; Jaksic et al., 1992; Jaksic, 1985; Navarro-Lpez et al., 2014; Sergio et al., 2004).
This study showed that peak performance (i.e. occurrence) estimates for S. rufipes were
slightly higher than G. nana over a select number of resources associated with stand-level
forest complexity and forest stability at the landscape scale; however, averaged results of
peak performance for resources assessed did not differ between S. rufipes and G. nana.
Since these results did not conform to the traditional model of relative niche width
differences between specialists and generalists species (Fig. 3.1), we need to re-think
models on how sympatric predators use habitat resources (Peers et al., 2012; Chapter 3). If
generalist predators follow the pattern of a wider tolerance of habitat conditions, that is not

67
hindered by a lower peak performance, they may be favoured during periods of
environmental change, while the more specialized predators should be favoured during
times of stability (Devictor et al., 2010; Peers et al., 2012).
Some raptor species can be used as biodiversity surrogates to identify sites in need
of protection because they either promote or are spatially associated with high biodiversity
levels (Burgas et al., 2014; Jenkins et al., 2012; Sergio et al., 2006, 2005). However, in
previous assessments of surrogacy relationships ecologists frequently assumed equivalency
among different taxa by using taxonomic diversity (i.e. species richness) as the target for
biodiversity (Westgate et al., 2014). I broadened my assessment to include both the density
of endemic species and functional diversity (measured by using avian habitat-use guilds
and the degree of habitat-specialization of the community) in my analyses on surrogacy
relationships between owls and avian biodiversity. My results showed that the habitat-
specialist S. rufipes outperformed the generalist G. nana as surrogates for taxonomic
diversity, endemism and functional diversity in South American temperate forests (Chapter
4). The cross-taxon congruence (sensu Westgate et al., 2014) between S. rufipes and target
diversity measures was high and indicated that avian functional diversity increased non-
linearly and peaked at sites with highest occurrence of the habitat-specialist owl. I
suggested that a pattern of specialist aggregation occurred in structurally complex
temperate forest-stands (sensu Julliard et al., 2006). Finally, I proposed management
actions to promote occurrence rates of S. rufipes (maintenance of multi-aged stands with a
variety of tree sizes, including large old-growth trees, with relatively high bamboo
understorey cover; Chapter 3), which would therefore be linked to enhanced density of
endemic species, specialized communities and, likely, ecosystem stability in Andean
temperate forests (Chapter 4).

68
5.2. Future directions

5.2.1. Scaling up owl-habitat relationships: from individuals to landscapes over time


As one of the first detailed studies on sympatric owls in Andean temperate forests,
my research provided important information about how environmental and ecological
factors are associated with detectability, occurrence patterns and use of habitat resources by
two sympatric owls, across spatial scales. This work also raised several new questions that
would be valuable areas for additional research. Life history of temperate forest birds in
general, and owls in particular, have been poorly studied in south-temperate systems
(Figueroa and Alvarado, 2012; Martin, 2004). Raptors are long-lived species that produce
few fledglings per breeding attempt, with adult survival proposed as the life history trait
that contributes most to fitness and population growth in owls (Lande, 1988; Newton,
1979). Ranking territories on the basis of overall fitness (e.g. offspring survival) has even
been proposed as an efficient tool for conservation planning for forest owls (Peery and
Gutirrez, 2013). Because conservation and wildlife planners make land use decisions to
ensure the long-term viability of species and ecosystems (Burns et al., 2013), a starting
point for future studies on southern temperate forest owls includes assessments of fitness
(e.g. productivity, survival and mortality agents) and demography. Although the owl
occurrence patterns that I studied may be directly associated with overall habitat quality
and fitness (Sergio and Newton, 2003), I recommend that further research should assess
directly the influence of forest degradation and fragmentation on habitat quality because
these anthropogenic processes are known to reduce long-term viability of raptor
populations (Hinam and St. Clair, 2008; Newton, 1998).
In future long-term monitoring programs and studies on owl distribution and
resource use, the rate of change in occurrence patterns across several years may be of more
interest than the proportion of sites where owls occur over a short time frame (one or two
years) in southern temperate landscapes. For these approaches, local colonization and local
extinction provide insights into the mechanisms underlying site occurrence dynamics
(MacKenzie et al., 2003) and population trends of forest owls (Tempel and Gutirrez,

69
2013). Modelling the dynamics of owl distribution across years is vital for informing and
planning effective owl conservation in changing habitats subject to rapid degradation and
fragmentation (Guisan and Thuiller, 2005; Lahoz-Monfort et al., 2014; Lamberson et al.,
1992; Seamans and Gutirrez, 2007). Furthermore, effectively planning for and managing
the impacts of climate change on mountain forest owls will require efficient landscape scale
monitoring and predictions about temporal shifts in their distributions (Glenn et al., 2010;
Kujala et al., 2013; Noon et al., 2012; Peery et al., 2012). In Chile, 90% of national parks
and reserves are located at high Andean locations (> 600 m of altitude; Armesto et al.,
1998). Spatially co-occurring, zones of high habitat suitability for S. rufipes were located
close to the upper elevation forests of the Andes Range (Chapter 3). The integration of
dynamic vegetation and climatic data into niche models for owls may help in assessing the
efficacy of the network of protected areas for capturing owl habitat availability under
current and projected future climate (Carroll, 2010).

5.2.2. Understanding surrogacy relationships for functional biodiversity conservation


across spatio-temporal scales
The global decline in biodiversity caused by anthropogenic degradation and loss of
forests is expected to disrupt important ecological processes and stability in these
ecosystems (Cardinale et al., 2006; Dirzo et al., 2014; Hooper et al., 2012). Direct
measurements of the effects of anthropogenic disturbances on ecological communities are
often complicated, time-consuming and expensive to assess; therefore, the use of
biodiversity surrogates as a shortcut to assess ecological conditions by proxy has been
useful for both environmental science and biodiversity management for decades (Caro,
2010; Lindenmayer and Likens, 2011). However, a recent meta-analysis of cross-taxon
congruence (i.e. reliability of surrogates as proxies of the distribution of unobserved taxa,
and therefore of a wider community of co-occurring species) in species richness and
composition showed that analyses of these comparisons rarely give consistent results
(Westgate et al., 2014). Here, I suggest key areas of work to enhance the empirical
identification and conservation reliability of biodiversity surrogates.

70
Identifying surrogates for functional diversity. Scientists assessing the impacts of forest
degradation and fragmentation often assume that these processes reduce taxonomic
diversity or species richness, resulting in similar losses of ecosystem functioning and
stability (Milder et al., 2008). This assumption, however, has not been validated either
empirically or theoretically (Mayfield et al., 2010; Schwartz et al., 2000). Research on
surrogacy relationships has explored mainly whether the occurrence of surrogates is
indicative of high taxonomic diversity; nevertheless, functional diversity is now known as
an important metric for linking diversity with ecosystem processes and stability, and thus
should be considered when identifying surrogate candidates (Sattler et al., 2014; Trivellone
et al., 2014). This assessment may provide further insights into the effects of anthropogenic
impacts on the services that humans derive from ecosystems (Chapin et al., 2000).

Niche theory for identifying biodiversity surrogates. Niche models have been used in
theoretical and empirical assessments of the condition of ecological communities (Chapters
3 and 4; Clavel et al., 2011; Hirzel and Le Lay, 2008). Ecological specialization is
frequently defined as the restricted niche width of a species (Futuyma and Moreno, 1988).
Species that are specialized and thus sensitive to habitat changes have been recommended
as ideal surrogates (Cabeza et al., 2007; Pearson, 1994) because, relative to habitat-
generalists, they are more prone to be negatively affected by forest degradation and
fragmentation (Clavel et al., 2011). Furthermore, meeting the needs of habitat-specialists is
expected to provide the requirements of generalist species as well, unless the surrogate
species is so highly specialized that it occupies a very narrow niche (Landres et al., 1988).
Few studies have tested directly whether being a specialist is a chief criterion for putative
surrogates (Branton and Richardson, 2011). The commonly used classification of species
into either specialist or generalist results in only two possible discrete specialization
measures. In nature, however, species vary in their degree of ecological specialization such
that community assemblages support a continuum of specializations (Devictor et al., 2010).
A continuous measure of the degree of species specialization (e.g. Chapter 4; Julliard et al.
2006) can be considered as a holistic functional trait because it integrates other life history
traits (e.g. dispersal ability, nest sites, diet or body mass) in a single comprehensive

71
parameter that is useful in community and ecosystem ecology (Devictor et al., 2008;
McGill et al., 2006).

Spatio-temporal scales on surrogacy relationships. The spatial scale of a study will


determine the extent of cross-taxon congruence observed (Eglington et al., 2012; Westgate
et al., 2014). The reliability of potential surrogates has frequently been assessed at local
scales to test either cross-taxon congruency or to identify sites in need of protection within
degraded, fragmented or relatively unaltered ecosystems (Burgas et al., 2014; Maes and
Dyck, 2005; Martikainen et al., 1998; Sergio et al., 2006). In forested ecosystems, these
assessments can provide specific conservation recommendations for managing stand-level
attributes at the scale where tree harvesting takes place (Lindenmayer et al., 2000).
Consistent cross-taxon congruency has been found for woodpeckers, raptors and other
higher-taxa (e.g. butterflies, vascular plants), as surrogate candidates, and target
biodiversity at landscape scales as well (Gaston and Williams, 1993; Mikusiski et al.,
2001; Pearman and Weber, 2007; Roberge et al., 2008; Sergio et al., 2004a; Simil et al.,
2006). Most of these broader-scale studies have used data available from atlases to quantify
distribution patterns of either surrogate candidates or target biodiversity. Surrogate-based
forest planning strategies at broad scales have strong potential for achieving biodiversity
conservation (Rodrigues and Brooks, 2007); however, their implementation critically
depends on large-scale survey datasets on the spatial distribution of biodiversity (Margules
and Pressey, 2000). Such information is still very limited for eco-regions subject to high
levels of degradation and fragmentation, like South American ecosystems, precisely the
areas most in need of conservation planning (Pimm, 2000; Rodrigues and Brooks, 2007).
Therefore, the challenge is to improve spatial databases on surrogacy relationships for
future implementation of conservation planning strategies in South American temperate
ecosystems and elsewhere.
One also needs to consider temporal variation in cross-taxon congruence. A single
species or habitat attribute can show high congruence with target biodiversity in short term
studies. However, because terrestrial ecosystems are characterized by their heterogeneity
and uncertainty (Filotas et al., 2014), any congruency relationship may change over longer

72
time periods (Hess et al., 2006; Thomson et al., 2005; Westgate et al., 2014). For example,
Lindenmayer et al., (2014a) reported that the relationship between the abundance of cavity-
bearing trees (surrogates) and cavity-dwelling marsupials (target biodiversity) remained
positive throughout a 30-year period, but the decline in abundance of cavity-bearing trees
over time weakened the relationship. Long-term research can provide important insights for
both forest ecology and the sustainable management of its biodiversity (Lindenmayer and
Likens, 2009; Magurran et al., 2010); thus, much work is needed to test the strength and
applicability for conservation of the surrogate-target relationship and the temporal
boundaries that may define this relationship.

Ideal surrogate types for South American temperate forests. The selection of a reliable
surrogate to monitor the effects of degradation and fragmentation on forest biodiversity is
critical and its choice will have scientific and conservation policy implications (Heink and
Kowarik, 2010; Noss, 1990). To date, three forest-dwelling bird species have been
proposed as biodiversity surrogates in South American temperate forests: Magellanic
woodpecker (Campephilus magellanicus), chucao tapaculo (Scelorchilus rubecula) and
rufous-legged owl (Strix rufipes), but only the owls have been tested for reliability as
biodiversity surrogates (Chapter 4).
The Magellanic woodpecker (Campephilus magellanicus) has been identified as a
flagship species for the conservation of old-growth temperate forests (Arango et al., 2007),
and proposed as a potential keystone habitat modifier for at least seven other cavity-nesting
bird species, and several mammals and reptiles (Beaudoin and Ojeda, 2011; McBride,
2000; Ojeda and Chazarreta, 2014). However, no empirical tests have been done and C.
magellanicus may not provide an indication of the status of target biodiversity across wide
environmental gradients (a criterion relevant for the selection of a surrogate; Noss, 1990)
because these birds have a strong affiliation to highland and high-latitude old-growth
forests, very restricted habitat-niche width and low densities across much of its range
(Prendergast et al., 1993; Vergara and Schlatter, 2004; J. T. Ibarra, unpublished data). It is
possible that C. magellanicus may not have a disproportionate influence on the tree cavity-
using community to be considered a keystone species (sensu Martin et al., 2004; Paine,

73
1969), as most secondary cavity nesters (N = 28 species) in southern temperate forests use
tree cavities generated by natural tree-decay processes (75% of cavity-nests; Altamirano,
2014). The remaining 25% of nests were in cavities excavated mainly by white-throated
treerunners (Pygarrhichas albogularis; Altamirano et al., 2012), an excavator that may
deserve further investigation as a potential surrogate for the abundance of several small to
mid-sized cavity-nesting species in temperate forests.
The chucao tapaculo (Scelorchilus rubecula), an iconic bird of Andean forests in
Patagonia, has been proposed as a potential surrogate for the group of endemic understorey
birds and mammals in temperate forests. Under the focal species approach (sensu
Lambeck, 1997), Castelln and Sieving (2012) suggest that landscapes designed to meet the
connectivity requirements of S. rubecula would be permeable to movement by several
forest vertebrates. These authors utilize previous behavioural and demographic studies on
S. rubecula to design patch-networks for these understorey birds, and develop algorithms
for scaling up their criteria to design larger-scale connections for other co-occurring
vertebrates. Similar to many other studies, the selection of S. rubecula as a biodiversity
surrogate has been ad hoc and assumptions underlying its selection remain untested
(Andelman and Fagan, 2000). Scelorchilus rubecula may be a poor surrogate for the
presence of other species that have different habitat requirements (e.g. larger home ranges),
dispersal capabilities or sensitivities to forest degradation and fragmentation (Lindenmayer
and Manning, 2002). The potential of C. magellanicus, S. rubecula, P. albogularis or any
other species, group of species (e.g. guild) or habitat attribute (sensu Lindenmayer et al.,
2014a) to serve as a biodiversity surrogate, should be subjected to appropriate criteria for
its selection and meet as many of the criteria as possible (Caro and ODoherty, 1999; Caro,
2010). The selection of empirically-validated surrogates should overcome the criticism that
the use and even the concept of biodiversity surrogates has received (Andelman and Fagan,
2000; Landres et al., 1988).

Strix rufipes: a surrogate-validated species. My empirical study suggests that the habitat-
specialist S. rufipes fulfills the criteria to be used as a biodiversity surrogate in southern
temperate forests (details in Chapter 4; Caro, 2010; Noss, 1990): (i) management actions

74
tailored to promote occurrence rates of S. rufipes can be linked to enhanced density of
endemic species and specialized avian communities, (ii) S. rufipes are sensitive to land-use
practices that reduce the availability of large old-living and dead trees, and that remove or
burn the understorey vegetation, similar to several avian species inhabiting temperate
forests (Daz et al., 2005; Ibarra et al., 2012; Ibarra et al., 2014c), (iii) the broad range of
predicted occurrence rates for S. rufipes would make them reliable surrogates across a
broad range of habitat conditions as they can tolerate some habitat disturbance at the stand
and landscape levels, (iv) conducting repeated nocturnal surveys of owls can readily and
cost-effectively be done in temperate forests (S. rufipes has relatively high detection
probabilities and I have made recommendations to improve survey protocols in Chapter 2
and Ibarra et al., 2014b). Repeated surveys may also be established as a volunteer-based
roadside survey program (Takats et al., 2001), and (v) S. rufipes occur extensively across
South American temperate forests and from sea level up to near the tree-line (1,500 m of
elevation; Ibarra et al., 2014b). These broad latitudinal and elevation ranges are shared by
>75% of avian species inhabiting southern temperate forests (Vuilleumier, 1985).
Therefore, this surrogacy relationship may hold across the temperate forest distribution.
This surrogate-validated species can enable wildlife and forest managers to employ S.
rufipes to more reliably monitor biodiversity and, at least partially, address the challenge of
evaluating the impacts of habitat degradation and fragmentation on wildlife communities in
temperate forests of South America.

75
Bibliography
Affleck, D.L.R., Gregoire, T.G., Valentine, H.T., 2005. Design unbiased estimation in line intersect sampling
using segmented transects. Environ. Ecol. Stat. 122, 139-154.
Altamirano, T.A., 2014. Breeding ecology of cavity-nesting birds in the Andean temperate forest of southern
Chile. PhD Dissertation. Department of Ecosystems and the Environment, School of Agriculture and
Forestry Sciences, Pontificia Universidad Catlica de Chile.
Altamirano, T.A., Ibarra, J.T., Hernndez, F., Rojas, I., Laker, J., Bonacic, C., 2012. Hbitos de nidificacin
de las aves del bosque templado andino de Chile. Fondo de Proteccin Ambiental, Ministerio del Medio
Ambiente, Serie Fauna Australis, Pontificia Universidad Catlica de Chile, Santiago.
Altamirano, T.A., Ibarra, J.T., Martin, K., Bonacic, C., 2012. Southern temperate forest cavity-nest web
structure: species richness and the role of tree decay in Patagonia, Chile, in: V North American
Ornithological Conference. Vancouver, B.C. Canada.
Andelman, S.J., Fagan, W.F., 2000. Umbrellas and flagships: efficient conservation surrogates or expensive
mistakes? Proc. Natl. Acad. Sci. 97, 5954-9.
Andersen, D.E., 2007. Survey techniques, in: Bird, D.M., Bildstein, K.L. (Eds.), Raptor Research and
Management Techniques. Hancock House Publishers, Surrey, Canada, pp. 89-100.
Arango, X., Rozzi, R., Massardo, F., Anderson, C.B., Ibarra, J.T., 2007. Descubrimiento e implementacin
del pjaro carpintero gigante (Campephilus magellanicus) como especie carismtica: una aproximacin
biocultural para la conservacin en la Reserva de Biosfera Cabo de Hornos. Magallania. 35, 71-88.
Armesto, J.J., Rozzi, R., Smith-Ramrez, C., Arroyo, M.T.K., 1998. Conservation targets in South American
temperate forests. Science. 282, 1271-1272.
Axelrod, D.I., Kalin, M.T., Raven, P.H., 1991. Historical development of temperate vegetation in the
Americas. Rev. Chil. Hist. Nat. 64, 413-446.
Bailey, L.L., Ines, J.E., Nichols, J.D., MacKenzie, D.I., 2007. Sampling design trade-offs in occupancy
studies with imperfect detection: examples and software. Ecol. Appl. 17, 281-290.
Bailey, L.L., Reid, J.A., Forsman, E.D., Nichols, J.D., 2009. Modeling co-occurrence of northern spotted and
barred owls: accounting for detection probability differences. Biol. Conserv. 142, 2983-2989.
Barkae, E.D., Scharf, I., Abramsky, Z., Ovadia, O., 2012. Jack of all trades, master of all: a positive
association between habitat niche breadth and foraging performance in pit-building antlion larvae. PLoS
One 7, e33506.
Beaudoin, F., Ojeda, V., 2011. Nesting of rufous-legged owls in evergreen Nothofagus forests. J. Raptor Res.
45, 272-274.
Bierregaard, R.O., 1998. Conservation status of birds of prey in the South American tropics. J. Raptor Res.
32, 19-27.
Boal, C.W., Bibles, B.D., 2001. Responsiveness of elf owls to conspecific and great horned owl calls. J. F.
Ornithol. 72, 66-71.
Brambilla, M., Bassi, E., Ceci, C., Rubolini, D., 2010. Environmental factors affecting patterns of distribution
and co-occurrence of two competing raptor species. Ibis. 152, 310-322.
Branton, M., Richardson, J.S., 2011. Assessing the value of the umbrella-species concept for conservation
planning with meta-analysis. Conserv. Biol. 25, 9-20.
Burgas, D., Byholm, P., Parkkima, T., 2014. Raptors as surrogates of biodiversity along a landscape gradient.
J. Appl. Ecol. 51, 786-794.
Burnham, K.P., Anderson, D.R., 2002. Model selection and inference: a practical information-theoretic
approach, 2nd ed. Springer-Verlag, New York, U.S.A.
Burns, E.S., Tth, S.F., Haight, R.G., 2013. A modeling framework for life history-based conservation
planning. Biol. Conserv. 158, 14-25.
Cabeza, M., Arponen, A., Van Teeffelen, A., 2007. Top predators: hot or not? A call for systematic
assessment of biodiversity surrogates. J. Appl. Ecol. 45, 976-980.
Cadotte, M.W., Carscadden, K., Mirotchnick, N., 2011. Beyond species: functional diversity and the
maintenance of ecological processes and services. J. Appl. Ecol. 48, 1079-1087.

76
Caley, M.J., Munday, P.L., 2003. Growth trades off with habitat specialization. Proc. R. Soc. B, Biol. Sci.
270, S175-177.
Campioni, L., Sarasola, J.H., Santilln, M., Reyes, M.M., 2013. Breeding season habitat selection by
ferruginous pygmy owls Glaucidium brasilianum in central Argentina. Bird Study 60, 35-43.
Cardinale, B.J., Srivastava, D.S., Duffy, J.E., Wright, J.P., Downing, A.L., Sankaran, M., Jouseau, C., 2006.
Effects of biodiversity on the functioning of trophic groups and ecosystems. Nature 443, 989-992.
Caro, T., 2010. Conservation by proxy: indicator, umbrella, keystone, flagship, and other surrogate species.
Island Press, Washington, D.C., U.S.A.
Caro, T., ODoherty, G., 1999. On the use of surrogate species in conservation biology. Conserv. Biol. 13,
805-814.
Carroll, C., 2010. Role of climatic niche models in focal-species-based conservation planning: assessing
potential effects of climate change on northern spotted owl in the Pacific Northwest, USA. Biol.
Conserv. 143, 1432-1437.
Castelln, T.D., Sieving, K.E., 2012. Can focal species planning for landscape connectivity meet the needs of
South American temperate rainforest endemics? Nat. Areas J. 32, 316-324.
Chapin, F.S.I., Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek, P.M., Reynolds, H.L., Hooper, D.U.,
Lavorel, S., Sala, O.E., Hobbie, S.E., Mack, M.C., Daz, S., 2000. Consequences of changing
biodiversity. Nature 405, 234-242.
Clark, K.A., Anderson, S.H., 1997. Temporal, climatic and lunar factors affecting owl vocalizations of
western Wyoming. J. Raptor Res. 31, 358-363.
Clarke, J.A., 1983. Moonlights influence on predator/prey interactions between short-eared Owls (Asio
flammeus) and deermice (Peromyscus maniculatus). Behav. Ecol. Sociobiol. 13, 205-209.
Clavel, J., Julliard, R., Devictor, V., 2011. Worldwide decline of specialist species: toward a global functional
homogenization? Front. Ecol. Environ. 9, 222-228.
Cody, M.L., 1985. An introduction to habitat selection in birds, in: Cody, M.L. (Ed.), Habitat Selection in
Birds. Academic Press, New York, U.S.A., pp. 3-56.
Crozier, M.L., Seamans, M.E., Gutirrez, R.J., 2005. The effect of broadcasting great horned owl
vocalizations on spotted owl vocal responsiveness. J. Raptor Res. 39, 111-118.
Crozier, M.L., Seamans, M.E., Gutirrez, R.J., Loschl, P.J., Horn, R.B., Sovern, S.G., Forsman, E.D., 2006.
Does the presence of barred owls suppress the calling behavior of spotted owls? Condor 108, 760-769.
Devictor, V., Clavel, J., Julliard, R., Lavergne, S., Mouillot, D., Thuiller, W., Venail, P., Villger, S.,
Mouquet, N., 2010. Dening and measuring ecological specialization. J. Appl. Ecol. 47, 15-25.
Devictor, V., Julliard, R., Jiguet, F., 2008. Distribution of specialist and generalist species along spatial
gradients of habitat disturbance and fragmentation. Oikos 117, 507-514.
Di Castri, F., Hajek, E.R., 1976. Bioclimatologa de Chile. Direccin De Investigacin, Vice-Rectora
Acadmica, Universidad Catlica De Chile, Santiago, Chile.
Daz, I.A., Armesto, J.J., Reid, S., Sieving, K.E., Willson, M.F., 2005. Linking forest structure and
composition: avian diversity in successional forests of Chilo Island, Chile. Biol. Conserv. 123, 91-101.
Daz, I.A., Armesto, J.J., Willson, M.F., 2006. Mating success of the endemic Des murs wiretail
(Sylviorthorhynchus desmursii, Furnariidae) in fragmented Chilean rainforests. Austral Ecol. 31, 13-21.
Daz, S., Cabido, M., 2001. Vive la diffrence: plant functional diversity matters to ecosystem processes.
Trends Ecol. Evol. 16, 646-655.
Daz, S., Lavorel, S., de Bello, F., Qutier, F., Grigulis, K., Robson, T.M., 2007. Incorporating plant
functional diversity effects in ecosystem service assessments. Proc. Natl. Acad. Sci. U. S. A. 104,
20684-20689.
Dirzo, R., Young, H.S., Galetti, M., Ceballos, G., Isaac, N.J.B., Collen, B., 2014. Defaunation in the
Anthropocene. Science. 345, 401-406.
Drever, M., Aitken, K., Norris, A., Martin, K., 2008. Woodpeckers as reliable indicators of bird richness,
forest health and harvest. Biol. Conserv. 141, 624-634.
Dueser, R., Hallett, J., 1980. Competition and habitat selection in a forest-floor small mammal fauna. Oikos
35, 293-297.
Eastman, J.R., 2012. IDRISI Selva. Clark University, Worcester, U.K.
Echeverra, C., Coomes, D., Salas, J., Rey-Benayas, J., Lara, A., Newton, A., 2006. Rapid deforestation and
fragmentation of Chilean temperate forests. Biol. Conserv. 130, 481-494.

77
Eglington, S.M., Noble, D.G., Fuller, R.J., 2012. A meta-analysis of spatial relationships in species richness
across taxa: birds as indicators of wider biodiversity in temperate regions. J. Nat. Conserv. 20, 301-309.
Fahrig, L., 2003. Effects of habitat fragmentation on biodiversity. Annu. Rev. Ecol. Evol. Syst. 34, 487-515.
Farias, A.A., Jaksic, F.M., 2011. Low functional richness and redundancy of a predator assemblage in native
forest fragments of Chilo Island, Chile. J. Anim. Ecol. 80, 809-817.
Figueroa, R., Corales, S., Martnez, D., Figueroa, R., Gonzlez-Acua, D., 2006. Diet of the rufous-legged
owl (Strix rufipes, Strigiformes) in an Andean Nothofagus-Araucaria forest, southern Chile. Stud.
Neotrop. Fauna Environ. 41, 179-182.
Figueroa, R.A., Alvarado, S.A., 2012. Ecologa y conservacin de los bhos chilenos: avances y desafos de
investigacin. Ornitol. Neotrop. 23, 209-219.
Filotas, E., Parrott, L., Burton, P.J., Chazdon, R.L., Coates, K.D., Coll, L., Haeussler, S., Martin, K.,
Nocentini, S., Puettmann, K.J., Putz, F.E., Simard, S.W., Messier, C., 2014. Viewing forests through the
lens of complex systems science. Ecosphere 5, 1-23.
Finn, S.P., Marzluff, J.M., Varland, D.E., 2002. Effects of landscape and local habitat attributes on northern
goshawk site occupancy in western Washington. For. Sci. 48, 427-436.
Fisher, R.J., Poulin, R.G., Todd, L.D., Brigham, R.M., 2004. Nest stage, wind speed, and air temperature
affect the nest defence behaviours of burrowing owls. Can. J. Zool. 82, 724-730.
Fiske, I.J., Chandler, R.B., 2011. Unmarked: an R package for fitting hierarchical models of wildlife
occurrence and abundance. J. Stat. Softw. 43, 1-23.
Flesch, A.D., Steidl, R.J., 2010. Importance of environmental and spatial gradients on patterns and
consequences of resource selection. Ecol. Appl. 20, 1021-1039.
Flynn, D.F.B., Gogol-Prokurat, M., Nogeire, T., Molinari, N., Richers, B.T., Lin, B.B., Simpson, N.,
Mayfield, M.M., DeClerck, F., 2009. Loss of functional diversity under land use intensification across
multiple taxa. Ecol. Lett. 12, 22-33.
Folliard, L., Reese, K., Diller, L., 2000. Landscape characteristics of northern spotted owl nest sites in
managed forests of northwestern California. J. Raptor Res. 34, 75-84.
Franco, M., Quijano, A., Soto-Gamboa, M., 2011. Communal nesting, activity patterns, and population
characteristics in the near-threatened Monito del monte, Dromiciops gliroides. J. Mammal. 92, 994-
1004.
Fuller, M.R., Mosher, J.A., 1987. Raptor survey techniques, in: Giron Pendleton, B.A., Millsap, B.A., Cline,
K.W., Bird, D.M. (Eds.), Raptor Management Techniques Manual. National Wildlife Federation,
Washington, D.C., U.S.A., pp. 37-65.
Futuyma, D.J., Moreno, G., 1988. The evolution of ecological specialization. Annu. Rev. Ecol. Syst. 19, 207-
233.
Gajardo, R., 1993. La vegetacin natural de Chile: clasificacin y distribucin geogrfica. Editorial
Universitaria, Santiago, Chile.
Ganey, J.L., 1990. Calling behavior of spotted owls in northern Arizona. Condor 92, 485-490.
Gaston, K.J., 1998. Rarity as double jeopardy. Nature 394, 229-230.
Gaston, K.J., Williams, P.H., 1993. Mapping the worlds species-the higher taxon approach. Biodivers. Lett.
1, 2-8.
Glenn, E.M., Anthony, R.G., Forsman, E.D., 2010. Population trends in northern spotted owls: associations
with climate in the Pacific Northwest. Biol. Conserv. 143, 2543-2552.
Greenwald, D.N., Crocker-Bedford, D.C., Broberg, L., Suckling, K.F., Tibbitts, T., 2005. A review of
northern goshawk habitat selection in the home range and implications for forest management in the
western United States. Wildl. Soc. Bull. 33, 120-128.
Greger, P.D., Hall, D.B., 2009. Burrow occupancy patterns of the western burrowing owl in southern Nevada.
West. North Am. Nat. 69, 285-294.
Groce, J.E., Morrison, M.L., 2010. Habitat use by saw-whet owls in the Sierra Nevada. J. Wildl. Manage. 74,
1523-1532.
Grossman, S.R., Hannon, S.J., Snchez-Azofeifa, A., 2008. Responses of great horned owls (Bubo
virginianus), barred owls (Strix varia), and northern saw-whet owls (Aegolius acadicus) to forest cover
and configuration in an agricultural landscape in Alberta, Canada. Can. J. Zool. 86, 1165-1172.
Guisan, A., Thuiller, W., 2005. Predicting species distribution: offering more than simple habitat models.
Ecol. Lett. 8, 993-1009.

78
Gustafsson, L., Baker, S.C., Bauhus, J., Beese, W.J., Brodie, A., Kouki, J., Lindenmayer, D.B., Lhmus, A.,
Martnez Pastur, G., Messier, C., Neyland, M., Palik, B., Sverdrup-Thygeson, A., Volney, W.J.A.,
Wayne, A., Franklin, J.F., 2012. Retention forestry to maintain multifunctional forests: a world
perspective. Bioscience 62, 633-645.
Hardy, P.C., Morrison, M.L., 2000. Factors affecting the detection of elf owls and western screech owls.
Wildl. Soc. Bull. 28, 333-342.
Heink, U., Kowarik, I., 2010. What criteria should be used to select biodiversity indicators? Biodivers.
Conserv. 19, 3769-3797.
Henneman, C., Andersen, D.E., 2009. Occupancy models of nesting-season habitat associations of red-
shouldered hawks in Central Minnesota. J. Wildl. Manage. 73, 1316-1324.
Hess, G.R., Bartel, R. a., Leidner, A.K., Rosenfeld, K.M., Rubino, M.J., Snider, S.B., Ricketts, T.H., 2006.
Effectiveness of biodiversity indicators varies with extent, grain, and region. Biol. Conserv. 132, 448-
457.
Hinam, H.L., St. Clair, C.C., 2008. High levels of habitat loss and fragmentation limit reproductive success by
reducing home range size and provisioning rates of northern saw-whet owls. Biol. Conserv. 141, 524-
535.
Hirzel, A.H., Le Lay, G., 2008. Habitat suitability modelling and niche theory. J. Appl. Ecol. 45, 1372-1381.
Hooper, D.U., Adair, E.C., Cardinale, B.J., Byrnes, J.E.K., Hungate, B.A., Matulich, K.L., Gonzalez, A.,
Duffy, J.E., Gamfeldt, L., OConnor, M.I., 2012. A global synthesis reveals biodiversity loss as a major
driver of ecosystem change. Nature 486, 105-108.
Hutchinson, G.E., 1957. Concluding remarks. Cold Spring Harb. Symp. Quant. Biol. 22, 415-427.
Hutto, R.L., 1985. Habitat selection by nonbreeding, migratory, land birds, in: Cody, M.L. (Ed.), Habitat
Selection in Birds. Academic Press, San Diego, California, U.S.A., pp. 455-476.
Ibarra, J.T., Altamirano, T.A., Glvez, N., Rojas, I., Laker, J., Bonacic, C., 2010. Avifauna de los bosques
templados de Araucaria araucana del sur de Chile. Ecol. Austral 20, 33-45.
Ibarra, J.T., Altamirano, T.A., Martin, K., Vargas, F.H., Bonacic, C., 2014a. Tree-cavity nesting of austral
pygmy-owls (Glaucidium nana) in Andean temperate forests of southern Chile. J. Raptor Res. 48, 82-
85.
Ibarra, J.T., Glvez, N., Gimona, A., Altamirano, T.A., Rojas, I., Hester, A., Laker, J., Bonacic, C., 2012.
Rufous-legged owl (Strix rufipes) and austral pygmy owl (Glaucidium nanum) stand use in a gradient of
disrupted and old growth Andean temperate forests, Chile. Stud. Neotrop. Fauna Environ. 47, 33-40.
Ibarra, J.T., Martin, K., Altamirano, T.A., Vargas, F.H., Bonacic, C., 2014b. Factors associated with the
detectability of owls in South American temperate forests: implications for nocturnal raptor monitoring.
J. Wildl. Manage. 78, 1078-1086.
Ibarra, J.T., Martin, K., Drever, M.C., Vergara, G., 2014c. Occurrence patterns and niche relationships of
sympatric owls in South American temperate forests: a multi-scale approach. For. Ecol. Manage. 331,
281-291.
Jaksic, F.M., 1985. Toward raptor community ecology: behavior bases of assemblage structure. J. Raptor Res.
19, 107-112.
Jaksic, F.M., Jimnez, J.E., Castro, S.A., Feinsinger, P., 1992. Numerical and functional response of predators
to a long-term decline in mammalian prey at a semi-arid Neotropical site. Oecologia 89, 90-101.
Jenkins, J., Simmons, R.E., Curtis, O., Atyeo, M., Raimondo, D., Jenkins, A.R., 2012. The value of the black
harrier Circus maurus as a predictor of biodiversity in the plant-rich Cape Floral Kingdom, South
Africa. Bird Conserv. Int. 23, 66-77.
Jimnez, J.E., Jaksic, F.M., 1989. Biology of the austral pygmy owl. Wilson Bull. 101, 377-389.
Julliard, R., Clavel, J., Devictor, V., Jiguet, F., Couvet, D., 2006. Spatial segregation of specialists and
generalists in bird communities. Ecol. Lett. 9, 1237-44.
Kerr, J., 1997. Species richness, endemism, and the choice of areas for conservation. Conserv. Biol. 11, 1094-
1100.
Kissling, M.L., Lewis, S.B., Pendleton, G., 2010. Factors influencing the detectability of forest owls in
southeastern Alaska. Condor 112, 539-548.
Kroll, A.J., Fleming, T.L., Irwin, L.L., 2010. Site occupancy dynamics of northern spotted owls in the eastern
Cascades, Washington, USA, 1990-2003. J. Wildl. Manage. 74, 1264-1274.

79
Krger, O., 2002. Analysis of nest occupancy and nest reproduction in two sympatric raptors: common
buzzard Buteo buteo and goshawk Accipiter gentilis. Ecography. 25, 523-532.
Kujala, H., Moilanen, A., Arajo, M.B., Cabeza, M., 2013. Conservation planning with uncertain climate
change projections. PLoS One 8, e53315.
Lahoz-Monfort, J.J., Guillera-Arroita, G., Wintle, B. a., 2014. Imperfect detection impacts the performance of
species distribution models. Glob. Ecol. Biogeogr. 23, 504-515.
Lambeck, R.J., 1997. Focal species: a multi-species umbrella for nature conservation. Conserv. Biol. 11, 849-
856.
Lamberson, R., McKelvey, R., Noon, B., Voss, C., 1992. A dynamic analysis of northern spotted owl viability
in a fragmented forest landscape. Conserv. Biol. 6, 505-512.
Lande, R., 1988. Demographic models of the northern spotted owl (Strix occidentalis caurina). Oecologia 75,
601-607.
Landres, P.B., Verner, J., Thomas, J.W., 1988. Ecological uses of vertebrate indicator species: a critique.
Conserv. Biol. 2, 316-328.
Lara, A., 1996. La conservacin del bosque nativo en Chile: problemas y desafos, in: Armesto, J.J.,
Villagrn, C., Arroyo, M.T. (Eds.), Ecologa de Los Bosques Nativos de Chile. Editorial Universitaria,
Santiago, Chile, pp. 335-362.
Lindenmayer, D.B., Barton, P.S., Lane, P.W., Westgate, M.J., McBurney, L., Blair, D., Gibbons, P., Likens,
G.E., 2014a. An empirical assessment and comparison of species-based and habitat-based surrogates: a
case study of forest vertebrates and large old trees. PLoS One 9, e89807.
Lindenmayer, D.B., Lane, P.W., Westgate, M.J., Crane, M., Michael, D., Okada, S., Barton, P.S., 2014b. An
empirical assessment of the focal species hypothesis. Conserv. Biol. 28, 1594-1603.
Lindenmayer, D.B., Likens, G.E., 2009. Adaptive monitoring: a new paradigm for long-term research and
monitoring. Trends Ecol. Evol. 24, 482-6.
Lindenmayer, D.B., Likens, G.E., 2011. Direct measurement versus surrogate indicator species for evaluating
environmental change and biodiversity loss. Ecosystems 14, 47-59.
Lindenmayer, D.B., Manning, A.D., 2002. The focal species approach and landscape restoration: a critique.
Conserv. Biol. 16, 338-345.
Lindenmayer, D.B., Margules, C.R., Botkin, D.B., 2000. Indicators of biodiversity for ecologically
sustainable forest management. Conserv. Biol. 14, 941-950.
Lhmus, A., 2003. Do ural owls (Strix uralensis) suffer from the lack of nest sites in managed forests? Biol.
Conserv. 110, 1-9.
Loureno, R., Goytre, F., del Mar Delgado, M., Thornton, M., Rabaa, J.E., Penteriani, V., 2013. Tawny owl
vocal activity is constrained by predation risk. J. Avian Biol. 44, 1-8.
MacKenzie, D.I., Nichols, J.D., Hines, J.E., Knutson, M.G., Franklin, Alan, B., 2003. Estimating site
occupancy, colonization, and local extinction when a species is detected imperfectly. Ecology 84, 2200-
2207.
MacKenzie, D.I., Nichols, J.D., Lachman, G.B., Droege, S., Andrew Royle, J., Langtimm, C.A., 2002.
Estimating site occupancy rates when detection probabilities are less than one. Ecology 83, 2248-2255.
MacKenzie, D.I., Nichols, J.D., Royle, J.A., Pollock, K.H., Bailey, L.L., Hines, J.E., 2006. Occupancy
estimation and modeling: inferring patterns and dynamics of species occurrence. Academic Press,
London, UK.
MacKenzie, D.I., Nichols, J.D., Sutton, N., Kawanishi, K., Bailey, L.L., 2005. Improving inferences in
population studies of rare species that are detected imperfectly. Ecology 86, 1101-1113.
MacKenzie, D.I., Royle, J.A., 2005. Designing occupancy studies: general advice and allocating survey effort.
J. Appl. Ecol. 42, 1105-1114.
Maes, D., Dyck, H. Van, 2005. Habitat quality and biodiversity indicator performances of a threatened
butterfly versus a multispecies group for wet heathlands in Belgium. Biol. Conserv. 123, 177-187.
Magurran, A.E., Baillie, S.R., Buckland, S.T., Dick, J.M., Elston, D.A., Scott, E.M., Smith, R.I., Somerfield,
P.J., Watt, A.D., 2010. Long-term datasets in biodiversity research and monitoring: assessing change in
ecological communities through time. Trends Ecol. Evol. 25, 574-582.
Manning, J.A., 2011. Factors affecting detection probability of burrowing owls in southwest agroecosystem
environments. J. Wildl. Manage. 75, 1558-1567.
Margules, C.R., Pressey, R.L., 2000. Systematic conservation planning. Nature 405, 243-253.

80
Martikainen, P., Kaila, L., Haila, Y., 1998. Threatened beetles in white-backed woodpecker habitats. Conserv.
Biol. 12, 293-301.
Martin, K., Aitken, K., Wiebe, K., 2004. Nest sites and nest webs for cavity-nesting communities in interior
British Columbia, Canada: nest characteristics and niche partitioning. Condor 106, 5-19.
Martin, T.E., 2004. Avian life-history evolution has an eminent past: does it have a bright future? Auk 121,
289-301.
Martin, T.G., Wintle, B.A., Rhodes, J.R., Kuhnert, P.M., Field, S.A., Low-Choy, S.J., Tyre, A.J., Possingham,
H.P., 2005. Zero tolerance ecology: improving ecological inference by modelling the source of zero
observations. Ecol. Lett. 8, 1235-1246.
Martnez, D.R., 2005. El concn (Strix rufipes) y su hbitat en los bosques templados australes, in: Smith-
Ramrez, C., Armesto, J.J., Valdovinos, C. (Eds.), Historia, Biodiversidad Y Ecologa de Los Bosques
Costeros de Chile. Editorial Universitaria, Santiago, Chile, pp. 477-484.
Martnez, D.R., Jaksic, F.M., 1996. Habitat, relative abundance, and diet of rufous-legged owls (Strix rufipes
King) in temperate forest remnants of southern Chile. Ecoscience 3, 259-263.
Martnez, J.A., Serrano, D., Zuberogoita, I., 2003. Predictive models of habitat preferences for the Eurasian
Eagle Owl Bubo bubo: a multiscale approach. Ecography. 26, 21-28.
Mayfield, M.M., Bonser, S.P., Morgan, J.W., Aubin, I., McNamara, S., Vesk, P. a., 2010. What does species
richness tell us about functional trait diversity? Predictions and evidence for responses of species and
functional trait diversity to land-use change. Glob. Ecol. Biogeogr. 423-431.
McBride, P., 2000. Magellanic Woodpecker (Campephilus magellanicus) habitat selection in deciduous
Nothofagus forests of Tierra del Fuego. Msc Thesis, Western Washington University, Bellingham, WA,
U.S.A.
McElhinny, C., Gibbons, P., Brack, C., Bauhus, J., 2005. Forest and woodland stand structural complexity: its
definition and measurement. For. Ecol. Manage. 218, 1-24.
McGarigal, K., Cushman, S.A., Neel, M.C., Ene, E., 2002. FRAGSTATS: spatial pattern analysis program for
categorical maps.
McGill, B.J., Enquist, B.J., Weiher, E., Westoby, M., 2006. Rebuilding community ecology from functional
traits. Trends Ecol. Evol. 21, 178-185.
McKinney, M.L., Lockwood, J.L., 1999. Biotic homogenization: a few winners replacing many losers in the
next mass extinction. Trends Ecol. Evol. 14, 450-453.
Mikusiski, G., Gromadzki, M., Chylarecki, P., 2001. Woodpeckers as indicators of forest bird diversity.
Conserv. Biol. 15, 208-217.
Milder, J.C., Lassoie, J.P., Bedford, B.L., 2008. Conserving biodiversity and ecosystem function through
limited development: an empirical evaluation. Conserv. Biol. 22, 70-79.
Moore, J.E., Swihart, R.K., 2005. Modeling patch occupancy by forest rodents: incorporating detectability
and spatial autocorrelation with hierarchically structured data. J. Wildl. Manage. 69, 933-949.
Moran, P.A.P., 1950. Notes on continuous stochastic phenomena. Biometrika 37, 17-23.
Morrell, T.E., Yahner, R.H., Harkness, W.L., 1991. Factors affecting detection of great horned owls by using
broadcast vocalizations. Wildl. Soc. Bull. 19, 481-488.
Mura, R., 1995. Comunidades de mamferos del bosque templado de Chile, in: Armesto, J.J., Villagrn, C.,
Arroyo, M.T. (Eds.), Ecologa de Los Bosques Nativos de Chile. Editorial Universitaria, Santiago,
Chile, pp. 113-133.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots
for conservation priorities. Nature 403, 853-858.
Naeem, S., Duffy, J.E., Zavaleta, E., 2012. The functions of biological diversity in an age of extinction.
Science. 336, 1401-1406.
Navarro-Lpez, J., Vergara, P., Fargallo, J.A., 2014. Trophic niche width, offspring condition and immunity
in a raptor species. Oecologia 174, 1215-1224.
Newton, A.C., 2007. Forest ecology and conservation: a handbook of techniques. Oxford University Press,
Oxford, U.K.
Newton, I., 1979. Population ecology of raptors. T & AD Poyser, Hertfordshire, U.K.
Newton, I., 1998. Population limitation in birds. Academic Press, San Diego, California, U.S.A.
Noon, B.R., Bailey, L.L., Sisk, T.D., McKelvey, K.S., 2012. Efficient species-level monitoring at the
landscape scale. Conserv. Biol. 26, 432-441.

81
Norambuena, H. V, Muoz-Pedreros, A., 2012. Diurnal activity of the austral pygmy owl (Glaucidium nana)
in southern Chile. Wilson J. Ornithol. 124, 633-635.
Noss, R.F., 1990. Indicators for monitoring biodiversity: a hierarchical approach. Conserv. Biol. 4, 355-364.
ODonnell, R.P., 2004. Effects of environmental conditions on owl responses to broadcast calls. Trans. West.
Sect. Wildl. Soc. 40, 101-106.
Ojeda, V., Chazarreta, L., 2014. Home range and habitat use of Magellanic woodpeckers in an old-growth
forest of Patagonia. Can. J. For. Res. 44, 1265-1273.
Ojeda, V.S., Suarez, M.L., Kitzberger, T., 2007. Crown dieback events as key processes creating cavity
habitat for magellanic woodpeckers. Austral Ecol. 32, 436-445.
Olden, J.D., LeRoy Poff, N., Douglas, M.R., Douglas, M.E., Fausch, K.D., 2004. Ecological and evolutionary
consequences of biotic homogenization. Trends Ecol. Evol. 19, 18-24.
Oliver, M.A., Webster, R., 1990. Kriging: a method of interpolation for geographical information systems.
Int. J. Geogr. Inf. Syst. 4, 313-332.
Olson, G.S., Anthony, R.G., Forsman, E.D., Ackers, S.H., Loschl, P.J., Reid, J.A., Dugger, K.M., Glenn,
E.M., Ripple, W.J., 2005. Modeling of site occupancy dynamics for northern spotted owls, with
emphasis on the effects of barred owls. J. Wildl. Manage. 69, 918-932.
Ozaki, K., Isono, M., Kawahara, T., Iida, S., Kudo, T., Fukuyama, K., 2006. A mechanistic approach to
evaluation of umbrella species as conservation surrogates. Conserv. Biol. 20, 1507-1515.
Paine, R.T., 1969. A note on trophic complexity and community stability. Am. Nat. 103, 91-93.
Pearman, P.B., Weber, D., 2007. Common species determine richness patterns in biodiversity indicator taxa.
Biol. Conserv. 138, 109-119.
Pearson, D.L., 1994. Selecting indicator taxa for the quantitative assessment of biodiversity. Philos. Trans. R.
Soc. London 345, 75-79.
Peers, M.J.L., Thornton, D.H., Murray, D.L., 2012. Reconsidering the specialist-generalist paradigm in niche
breadth dynamics: resource gradient selection by Canada lynx and bobcat. PLoS One 7, 1-10.
Peery, M.Z., Gutirrez, R.J., 2013. Life-history tradeoffs in spotted owls (Strix occidentalis): implications for
assessment of territory quality. Auk 130, 132-140.
Peery, M.Z., Gutirrez, R.J., Kirby, R., LeDee, O.E., LaHaye, W., 2012. Climate change and spotted owls:
potentially contrasting responses in the Southwestern United States. Glob. Chang. Biol. 18, 865-880.
Penteriani, V., Delgado, M.D.M., Campioni, L., Loureno, R., 2010. Moonlight makes owls more chatty.
PLoS One 5, e8696.
Petchey, O.L., Gaston, K.J., 2006. Functional diversity: back to basics and looking forward. Ecol. Lett. 9,
741-58.
Pimm, S.L., 2000. Conservation connections. Trends Ecol. Evol. 15, 262-263.
Pincheria-Ulbrich, J., Rodas-Trejo, J., Almanza, V.P., Rau, J.R., 2008. Estado de conservacin de las aves
rapaces de Chile. Hornero 23, 5-13.
Piorecky, M.D., Prescott, D.R.C., 2006. Multiple spatial scale logistic and autologistic habitat selection
models for northern pygmy owls, along the eastern slopes of Albertas Rocky Mountains. Biol.
Conserv. 129, 360-371.
Prendergast, J.R., Quinn, R.M., Lawton, J.H., 1993. Rare species, the coincidence of diversity hotspots and
conservation strategies. Nature 365, 335-337.
Reid, S., Daz, I., Armesto, J., 2004. Importance of native bamboo for understory birds in Chilean temperate
forests. Auk 121, 515-525.
Richards, C.L., Bossdorf, O., Muth, N.Z., Gurevitch, J., Pigliucci, M., 2006. Jack of all trades, master of
some? On the role of phenotypic plasticity in plant invasions. Ecol. Lett. 9, 981-993.
Roberge, J.-M., Mikusiski, G., Svensson, S., 2008. The white-backed woodpecker: umbrella species for
forest conservation planning? Biodivers. Conserv. 17, 2479-2494.
Rodrigues, A.S.L., Brooks, T.M., 2007. Shortcuts for biodiversity conservation planning: the effectiveness of
surrogates. Annu. Rev. Ecol. Evol. Syst. 38, 713-737.
Rodrguez-Estrella, R., Donzar, J.A., Hiraldo, F., 2008. Raptors as indicators of environmental change in the
scrub habitat of Baja California Sur, Mexico. Conserv. Biol. 12, 921-925.
Royle, J.A., Dawson, D.K., Bates, S., 2004. Modeling abundance effects in distance sampling. Ecology 85,
1591-1597.

82
Rozzi, R., Martnez, D.R., Willson, M.F., Sabag, C., 1996. Avifauna de los bosques templados de
Sudamrica, in: Armesto, J.J., Villagrn, C., Arroyo, M.T. (Eds.), Ecologa de Los Bosques Nativos de
Chile. Editorial Universitaria, Santiago, Chile, pp. 135-152.
Sattler, T., Pezzatti, G.B., Nobis, M.P., Obrist, M.K., Roth, T., Moretti, M., 2014. Selection of multiple
umbrella species for functional and taxonomic diversity to represent urban biodiversity. Conserv. Biol.
28, 414-426.
Sberze, M., Cohn-Haft, M., Ferraz, G., 2010. Old growth and secondary forest site occupancy by nocturnal
birds in a Neotropical landscape. Anim. Conserv. 13, 3-11.
Schwartz, M.W., Brigham, C.A., Hoeksema, J.D., Lyons, K.G., Mills, M.H., Van Mantgem, P.J., 2000.
Linking biodiversity to ecosystem function: implications for conservation ecology. Oecologia 122, 297-
305.
Seamans, M.E., Gutirrez, R.J., 2007. Habitat selection in a changing environment: the relationship between
habitat alteration and spotted owl territory occupancy and breeding dispersal. Condor 109, 566.
Sergio, F., Caro, T., Brown, D., Clucas, B., Hunter, J., Ketchum, J., McHugh, K., Hiraldo, F., 2008. Top
predators as conservation tools: ecological rationale, assumptions, and efficacy. Annu. Rev. Ecol. Evol.
Syst. 39, 1-19.
Sergio, F., Marchesi, L., Pedrini, P., 2004a. Integrating individual habitat choices and regional distribution of
a biodiversity indicator and top predator. J. Biogeogr. 31, 619-628.
Sergio, F., Newton, I., 2003. Occupancy as a measure of territory quality. J. Anim. Ecol. 72, 857-865.
Sergio, F., Newton, I., Marchesi, L., 2005. Top predators and biodiversity. Nature 436, 192.
Sergio, F., Newton, I., Marchesi, L., Pedrini, P., 2006. Ecologically justified charisma: preservation of top
predators delivers biodiversity conservation. J. Appl. Ecol. 43, 1049-1055.
Sergio, F., Pedrini, P., Marchesi, L., 2003. Adaptive selection of foraging and nesting habitat by black kites
(Milvus migrans) and its implications for conservation: a multi-scale approach. Biol. Conserv. 112, 351-
362.
Sergio, F., Rizzolli, F., Marchesi, L., Pedrini, P., 2004b. The importance of interspecific interactions for
breeding-site selection: peregrine falcons seek proximity to raven nests. Ecography. 27, 818-826.
Simil, M., Kouki, J., Mnkknen, M., Sippola, A.-L., Huhta, E., 2006. Co-variation and indicators of species
diversity: can richness of forest-dwelling species be predicted in northern boreal forests? Ecol. Indic. 6,
686-700.
Singleton, P.H., Lehmkuhl, J.F., Gaines, W.L., Graham, S.A., 2010. Barred owl space use and habitat
selection in the eastern Cascades, Washington. J. Wildl. Manage. 74, 285-294.
Smith-Ramrez, C., 2004. The Chilean coastal range: a vanishing center of biodiversity and endemism in
South American temperate rainforests. Biodivers. Conserv. 13, 373-393.
Sutherland, G.D., Smith, J.R., O`Brien, D.T., Waterhouse, F.L., Harestad, A.S., 2010. Validation of modelled
habitat classifications for the northern spotted owl in British Columbia using patterns of historical
occupancy. B.C. Ministry of Forest Range, Forest Science Program, Victoria, B.C. Canada.
Takats, D.L., Francis, C.M., Holroyd, G.L., Duncan, H.J., Mazur, K.M., Cannings, R.J., Harris, W., Holt, D.,
2001. Guidelines for nocturnal owl monitoring in North America. Beaverhill Bird Observatory and Bird
Studies Canada, Edmonton, Alberta, Canada.
Tempel, D.J., Gutirrez, R.J., 2013. Relation between occupancy and abundance for a territorial species, the
California spotted owl. Conserv. Biol. 27, 1087-1095.
Thomson, J., Fleishman, E., Macnally, R., Dobkin, D., 2005. Influence of the temporal resolution of data on
the success of indicator species models of species richness across multiple taxonomic groups. Biol.
Conserv. 124, 503-518.
Tilman, D., Reich, P.B., Knops, J., Wedin, D., Mielke, T., Lehman, C., 2001. Diversity and productivity in a
long-term grassland experiment. Science. 294, 843-845.
Trejo, A., Beaudoin, F., Ojeda, V., 2011. Response of rufous-legged owls to broadcast of conespecific calls in
southern temperate forests of Argentina. J. Raptor Res. 45, 71-74.
Trejo, A., Figueroa, R.A., Alvarado, S.A., 2006. Forest-specialist raptors of the temperate forests of southern
South America: a review. Rev. Bras. Ornitol. 14, 317-330.
Trivellone, V., Schoenenberger, N., Bellosi, B., Jermini, M., de Bello, F., Mitchell, E.A.D., Moretti, M., 2014.
Indicators for taxonomic and functional aspects of biodiversity in the vineyard agroecosystem of
Southern Switzerland. Biol. Conserv. 170, 103-109.

83
UNESCO, 2010. World network of biosphere reserves (WNBR) [WWW Document]. Ecol. Sci. Sustain. Dev.
URL http://www.unesco.org/new/en/natural-sciences/environment/ecological-sciences/biosphere-
reserves/wnbr/
Van Den Meersschaut, D., Vandekerkhove, K., 2000. Development of a stand-scale forest biodiversity index
based on the State Forest Inventory, in: Hansen, M., Burk, T. (Eds.), Integrated Tools for Natural
Resources Inventories in the 21st Century. Dept. of Agriculture, Forest Service, North Central Forest
Experiment Station, St. Paul, Minnesota, U.S.A., pp. 340-350.
Veblen, T.T., 1982. Growth patterns of Chusquea bamboos in the understory of Chilean Nothofagus forests
and their influences in forest dynamics. Bull. Torrey Bot. Club 109, 474-487.
Veblen, T.T., Schlegel, F.M., Escobar, B., 1980. Structure and dynamics of old-growth Nothofagus forests in
the Valdivian Andes, Chile. J. Ecol. 68, 1-31.
Vergara, P., Schlatter, R.P., 2004. Magellanic woodpecker (Campephilus magellanicus) abundance and
foraging in Tierra del Fuego, Chile. J. Ornithol. 145, 343-351.
Vergara, P.M., Armesto, J.J., 2009. Responses of Chilean forest birds to anthropogenic habitat fragmentation
across spatial scales. Landsc. Ecol. 24, 25-38.
Vuilleumier, F., 1985. Forest birds of Patagonia: ecological geography, speciation, endemism, and faunal
history. Ornithol. Monogr. 36, 255-304.
Wallace, D.R., 1983. The Klamath knot: explorations of myth and evolution. Yolly Bolly Press, San
Francisco, California, U.S.A.
Westgate, M.J., Barton, P.S., Lane, P.W., Lindenmayer, D.B., 2014. Global meta-analysis reveals low
consistency of biodiversity congruence relationships. Nat. Commun. 5, 1-8.
Wiens, J.A., 1994. Habitat fragmentation: island v. landscape perspectives on bird conservation. Ibis. 137, 97-
104.
Wiens, J.J., Ackerly, D.D., Allen, A.P., Anacker, B.L., Buckley, L.B., Cornell, H. V, Damschen, E.I.,
Jonathan Davies, T., Grytnes, J.-A., Harrison, S.P., Hawkins, B. A, Holt, R.D., McCain, C.M.,
Stephens, P.R., 2010. Niche conservatism as an emerging principle in ecology and conservation
biology. Ecol. Lett. 13, 1310-24.
Willson, M.F., De Santo, T.L., Sabag, C., Armesto, J.J., 1994. Avian communities of fragmented south-
temperate rainforests in Chile. Conserv. Biol. 8, 508-520.
Wintle, B.A., Kavanagh, R.P., McCarthy, M.A., Burgman, M.A., 2005. Estimating and dealing with
detectability in occupancy surveys for forest owls and arboreal marsupials. J. Wildl. Manage. 69, 905-
917.
Ylonen, H., Brown, J.S., 2007. Fear and the foraging, breeding, and sociality of rodents, in: Wolf, J.O.,
Sherman, P.W. (Eds.), Rodent Societies: An Ecological and Evolutionary Perspective. University of
Chicago Press, Chicago, Illinois, U.S.A., pp. 328-341.

84