ANDEAN TEMPERATE FOREST OWLS: DETECTABILITY, HABITAT
RELATIONSHIPS AND RELIABILITY AS BIODIVERSITY SURROGATES
by
José Tomás Ibarra Eliessetch
B.Sc., Pontificia Universidad Católica de Chile (Agricultural Engineering), 2005 M.Sc.,
Pontificia Universidad Católica 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)
October 2014
© José Tomás 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 20112013, 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 nonlinearly 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, 10781086.
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
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
iv
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 (ψ) and detection
(p) of Strix rufipes and Glaucidium nana in Andean temperate forests, 20112013. 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 (Δ Akaike‟s Information Criterion ≤ 2.0) predicting probability
of occupancy (ψ) 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 Akaike‟s Information Criterion (AIC) for
estimating probability of occurrence (ψ) and detection (p) of two owl species
(a) Strix rufipes and (b) Glaucidium nana in Andean temperate forests. Only the
top model set with Δ 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 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 Akaike‟s Information
Criterion (AIC). Parameter estimates [SE] for covariates present in the top
model set with Δ 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 (ψ) 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 (ψ) 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 Araucanía Region (39°S), 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 (ψ) 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 species……………………………19
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 Araucanía 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
Fig. 4.1. Relationship between probabilities of occurrence (ψ) 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, 20112013…………………………………………………………………..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 Tomás A.
Altamirano, Antonia Barreau, Cristián Bonacic, Alberto Dittborn, Nicolás Gálvez, Jill
Jankowski, Richard Schuster, F. Hernán 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, Tomás A. Altamirano, Antonia Barreau, Cristián
Bustos, Diego Cox, Consuelo Gálvez, Catalina Zumaeta, Fernanda Soffia, Carolina Yáñez,
Ismael Horta, Julián Caviedes, Klaus Kremer, Lina Forero, María Ignacia Ibarra, Mariano
de la Maza, María José Martín, Angélica Pinochet, Gonzalo Vergara, Paloma Corvalán,
Patricio Bahamondes, Santiago Pérez de Castro, Teresa Honorato, Matías Acevedo, María
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 postgraduate scholarship from Comisión Nacional de Investigación Científica y Tecnológica 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 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 Católica de Chile. I thank all of them for their support
and belief in this project.
The Chilean Forest Service (CONAF), Santuario El Cañi (Guías-Cañe: Roberto Sanhueza
y Manuel Venegas), Kawelluco Sanctuary, Francisco Poblete, Ricardo Timmerman,
Mónica Sabugal, Cristina Délano, 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
xii
Robles and Cassandra Storey. My external examiners Ralph J. Gutiérrez, 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 Jofré 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.
“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; Echeverría et al. 2006). However, basic information on owl-habitat
relationships for South American temperate forests is limited (Trejo et al., 2006; but see
Martínez 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 Gutiérrez, 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; Martínez 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 standing trees or
snags, affecting habitat-specialist species that depend on these niche resources (Díaz 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 (Echeverría 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
1.1.2. 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; Lourenço et
al., 2013; Wintle et al., 2005). Therefore, exclusively using the count statistic (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 Gutiérrez, 2007)
and/or landscape-level attributes as covariates (Folliard et al., 2000; Martínez 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; Krüger, 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 owl‟s perception of the environment in order to generate
habitat suitability models (Martínez 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.
1
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 O‟Doherty, 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, (iii) 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 (Díaz
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,
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 forestdwelling 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
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 Araucanía
Region (39º16´S 71ºW) 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 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 (Martínez and Jaksic, 1996). Glaucidium nana is the most widespread and
common owl in Chile (Jiménez 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 (Jiménez and Jaksic, 1989; Martínez 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.
1
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 (ψ) = 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; Martínez 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 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.
1
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.
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 Araucanía Region (39º16´S 71ºW), 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).
1
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 (Lourenço 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; Martínez 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
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).
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
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 360º, 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 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 (ψ) and detection (p) using the program RUnmarked, which allowed the response variables to be functions of covariates (Fiske and
Chandler, 2011). For ψ, 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).
19
To obtain the best model for each owl species, I first fit models using each covariate
singly to predict ψ or p. I also fit a model with ψ constant across sites and p constant across
surveys (i.e. null models). I ranked models using an information-theoretic approach (Akaike‟s
Information Criterion [AIC]; Burnham and Anderson, 2002). After I fit the single-covariate
models, I assessed more complex models containing different combinations of the bestsupported 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 model‟s 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 20112012 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: ψ = 0.69-0.99, Pa = 0: ψ = 0.27-0.93). However, protected
area status explained little variation given the 95% confidence intervals of the 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.
20
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 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; Murúa, 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 westernscreech 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 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
21
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 Muñoz-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, Lourenço 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 (Martínez,
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).
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 [ψ] ~ 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.
22
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.
Figure 2.1. Predicted probabilities of occurrence (ψ) 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.
23
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
Temporal
Abiotic
Type of variable (code)
Description
Reason for consideration
Days (day)
Number of days since start of
surveys
Time (time)
Number of minutes after 21
hours
Year (year)
Nesting season 2011-2012 or
2012-2013
Temperature (temp) *
ºC
Owl calling behaviour
may change throughout
the nesting season a, b, c, d, j
Owl calling behaviour
may change during the
night a, c, d, j
Owl calling behaviour
may change between
years d, k
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
reduced for that obscured by
clouds. Ml = [(1-cloud) ×
(moon phase/100)].
Moon phase refers to %
Owl and prey behaviour a,
b, d, e, j, l, n
where full moon = 100%
Environmental noise (noise)
Sound carry c
0 = quiet
1 = substantial (dogs barking,
and/or river and stream
noise)
Covariate
Biotic
Type of variable (code)
Description
Other owl species detected
0 = none
(owl)
1 = other owl detected
Reason for
consideration
Owl behaviour f, g, i, k
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 Lourenço et al. (2013), j Morrell et
al. (1991), k Olson et al. (2005), l Penteriani et al. (2010), m Wintle et al. (2005), n O‟Donnell (2004).
* Pairs of strongly inter-correlated (Pearson's r > 0.7) covariates.
20
Table 2.2. Model selection results for estimating probability of occupancy (ψ) and detection
(p) of Strix rufipes and Glaucidium nana in Andean temperate forests, 2011-2013. Sitespecific 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
Strix rufipes
Glaucidium nana
Model
Ka
ΔAIC b
wi c
ψ(elev + Pa), p(Ml + noise + owl)
9
0
0.58
ψ(elev), p(Ml + noise + owl)
8
1.43
0.28
8
3.02
0.13
9
0
0.43
ψ(.), p(Ml + noise + day2 + owl)
8
0.88
0.28
ψ(.), p(Ml + noise + day + owl)
8
2.69
0.11
ψ(Pa), p(Ml + noise + owl)
ψ(.), p(wind + Ml + noise + day2 + owl)
a
Number of parameters estimated.
b
ΔAIC is the difference in AIC values between each model and the lowest AIC model.
c
AIC model weight.
26
Table 2.3. Competing models (Δ Akaike‟s Information Criterion ≤ 2.0) predicting
probability of occupancy (ψ) 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
ψ
Intercept
-1.012
-2.182
0.159
Elevation
Protected area
Intercept
Moonlight
Noise
Other owl
Intercept
Elevation
Intercept
Moonlight
Noise
Other owl
2.441
1.818
-0.441
0.006
-0.508
0.513
-1.276
3.285
-0.449
0.006
-0.503
0.514
0.190
-0.498
-0.744
0.0009
-0.837
0.178
-2.379
1.251
-0.752
0.0009
-0.832
0.179
4.692
4.134
-0.139
0.010
-0.180
0.848
-0.172
5.319
-0.147
0.010
-0.175
0.849
Intercept
Intercept
Wind
Moonlight
Noise
1.415
-1.04
-0.46
0.011
-0.579
0.772
-1.350
-0.873
0.007
-0.885
2.058
-0.730
-0.046
0.015
-0.273
Days2
0.00005
-0.00001
0.00011
Other owl
0.749
0.426
1.07
ψ
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
Strix rufipes
Model 1
p
Model 2
ψ
p
Glaucidium nana
Model 1
ψ
p
Glaucidium nana
Model 2
27
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 (Martínez et al., 2003; Sergio et al., 2003), and
(b) the level of sensitivity of species in habitats subject to rapid degradation and
fragmentation.
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 landuse 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 (Díaz
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
(Echeverría et al., 2006).
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.
28
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 raptorhabitat 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 (Martínez and Jaksic, 1996).
Glaucidium nana are the most widespread and common owls in Chile (Jiménez 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 habitatspecialists because of their affiliation with multi-stratified forest stands > 100 year 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 (Jiménez
and Jaksic, 1989; Martínez 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.
29
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; Martínez 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-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 (Pearson‟s 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 (Martínez, 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.
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 (ψ) was defined as the probability that at least one individual owl
occurred at a site.
To evaluate ψ, I assessed collinearity to reduce the number of covariates presented in
Table 3.1. With strongly correlated covariates (Pearson‟s 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 ψ modelling: four at the stand- level (SD
30
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 Akaike‟s 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 ψ model for each owl, I first fit models using
each covariate singly to predict ψ and also fitted a model with ψ 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 Δ AIC ≤ 2 were considered the models best supported by the
data. I addressed model selection uncertainty by averaging models with Δ 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 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 Moran‟s 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, nondetection = 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
ψ 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 ψ value for each covariate and compared the responses between owls using Student's ttests. Data were normally distributed based on
Kolmorogov-Smirnov test (P < 0.05).
31
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 ranged from
0.62 (59 sites out of 95 total sites) in 2011-2012, to 0.55 (56 out of 101 total sites) in 20122013. 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 ψ (mean ± standard error) ranged between 0.05 ± 0.04 and
1.00 ± 0.00 across sites. The models with highest support (Δ AIC ≤ 2) for S. rufipes contained
two to four covariates for ψ (Table 3.2a). Model selection results indicated that ψ 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 ψ (Table 3.2a); however, beta coefficient for this covariate also overlapped with zero
(Table 3.2a). The averaged predictions of ψ 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 ψ for S. rufipes
in Andean forests based on averaged model predictions (Table 3.3).
For G. nana, predicted ψ ranged between 0.67 ± 0.18 and 0.98 ± 0.04 across sites. The
models with highest support (Δ AIC ≤ 2) for G. nana contained two or three covariates for ψ,
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 0.09
and 0.07 units higher than models without this term (compare Table 3.2b). Model selection
results indicated that ψ 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 ψ 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 ψ 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 ψ 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 ψ for S. rufipes over the range of all
covariate values. Averaged results of peak performance (i.e. peak ψ) for all 16 covariates
32
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
ψ 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 (Martínez 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 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 multiaged 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;
Martínez 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).
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
33
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; Martínez 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 (Lõhmus, 2003; Seamans and Gutiérrez, 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 (Echeverría et al., 2006; Smith-Ramírez, 2004). The long-term survival of S.
rufipes is jeopardized in these areas of its range (Martínez, 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 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 toppredators (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.,
34
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 microhabitat 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 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; Martínez 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 habitatspecialists. 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 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 (Díaz 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
35
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.
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.
36
Fig. 3.2. Averaged predictions of occurrence for (a) Strix rufipes and (b) Glaucidium nana
in the Villarrica watershed of the La Araucanía 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
models
Description
Tre
Density of all trees with DBH > 12.5 cm
Standard deviation of
tree diameter at breast
height (DBH, cm) a
Dbh
Canopy cover (%) a
Can
Volume of coarse woody
debris (VCWD) a
Cwd
Bamboo understorey
density (NC) a
Und
Elevation (m.a.s.l)
Ele
SD of tree DBH measures the variability in tree size, and
was considered indicative of the diversity of micro-habitats
within a stand for both owl and potential prey (Van Den
Meersschaut and Vandekerkhove, 2000)
Proportion of sky covered by canopy estimated from the
centre of the plot
Calculated based on the length and diameter of each piece
with diameter > 7.5 cm crossing a transect of 22.4 m length
(oriented N-S)
Density of bamboo vegetation up to 3 m high, expressed as
the number of contacts (NC) using the method described in
Díaz et al., (2006), quantified at five points of a transect of
22.4 m length (oriented N-S)
Meters above sea level measured at the center of the plot
Stand-level
Tree density (trees/ha)
Landscape-level
Forest areas (180 haa /
1,206 ha)
Shrubland areas (180 haa
/ 1,206 ha)
Core habitat (180 ha / 1
206 ha)
Forest-patch shape index
(Si) (180 haa / 1,206 haa)
Relative habitat diversity
(180 haa / 1,206 ha)
a
For180 /
For1206
Shr180 /
Shr1206
Cor180 /
Cor1206
Si180 / Si1206
Hd180 /
Hd1206
% Extent of forested area
% Extent of shrubland
Mean size of interior core habitat (≥ 100m from polygon
edge) of all forest patches in plot
Si = 0.25 x p/√A, where p = forest-patch perimeter and A =
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 within a circular plot measured
as Shannon‟s 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
Covariates retained for tests of habitat associations of forest owls after reducing collinearity.
43
Table 3.2. Model selection statistics based on Akaike‟s Information Criterion (AIC) for
estimating probability of occurrence (ψ) and detection (p) of two owl species (a) Strix
rufipes and (b) Glaucidium nana in Andean temperate forests. Only the top model set
with Δ 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.
Species
(a) Strix rufipes
(b) Glaucidium
nana
Δ AIC d
Model structure
Kc
ψ(Dbh + Und + Can), pa
10
1070.16
0.00
0.27
2.59, 0.98, 2.07
ψ(Dbh + Und), pa
9
1070.17
0.02
0.27
3.28, 1.03
ψ(Dbh + Und + For180), pa
10
1070.81
0.65
0.20
2.82, 0.95, 1.55
ψ(Dbh + Und + For180 + Can), pa
11
1071.32
1.17
0.15
2.29, 0.90, 1.25,
1.82
ψ(Dbh + For180), pa
9
1073.17
3.01
0.06
3.07, 2.43
ψ(Si1206 + For180 + Aut), pb
12
1258.64
0.00
0.38
0.72, 0.17, 1.84
ψ(Si1206 + For180), pb
11
1259.16
0.52
0.29
2.74, 0.44
AIC
Wi e
Parameter
estimates
Estimated
95% CI
0.39, 4.79
0.02, 1.94
-0.85, 5.00
1.27, 5.29
0.09, 1.97
0.68, 4.97
-0.02, 1.93
-1.08, 4.17
-0.01, 4.58
-0.10, 1.90
-1.94, 3.94
-1.15, 4.79
1.05, 5.10
-0.15, 5.01
-2.43, 3.87
-2.63, 2.96
-2.00, 3.89
-4.31, 9.78
-2.37, 3.25
ψ(Shr180 + For180 + Aut), pb
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
ψ(Shr180 + For180), pb
11
1260.80
2.16
0.13
-1.84, 0.79
-6.61, 2.94
-2.19, 3.77
a
p(Ml + No + Ow)
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
Δ AIC is the difference in AIC values between each model and the lowest AIC model.
e
AIC model weight.
b
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 (0.67-1) for Strix
rufipes in Andean temperate forests, based on model-averaged predictions.
0-0.33
Stand-level
Tree density (#/ha)
SD of diameter at breast height (cm)
Bamboo understorey density (NC) a
Volume of coarse woody debris (m3)
Canopy cover (%)
Landscape-level
Forest extent 180 ha (%)
Forest extent 1,206 ha (%)
Forest shape index 180 ha
Forest shape index 1,206 ha
Shrubland extent 180 ha (%)
Shrubland extent 1,206 ha (%)
Habitat diversity 180 ha (Shannon index)
Habitat diversity 1,206 ha (Shannon
index)
Core habitat 180 ha (# of ha)
Core habitat 1,206 ha (# of ha)
Predicted probability of occurrence
0.34-0.66
0.67-1
225.4 (223.2)
5.4 (2.4)
0.1 (0.1)
0 (0.0)
29.9 (22.4)
435.7 (300.0)
11.2 (2.0)
0.2 (0.3)
0.2 (0.7)
50.3 (18.8)
487.4 (244.7)
19.9 (9.0)
2.8 (2.7)
0.4 (0.4)
65.5 (17.8)
26.4 (20)
35.3 (17.4)
1.2 (0.2)
1.4 (0.1)
27.2 (15.7)
24.8 (8.8)
0.9 (0.4)
1.1 (0.2)
49.9 (23.1)
53.2 (19.5)
1.3 (0.3)
1.4 (0.1)
22.9 (14.4)
22.2 (9.7)
1 (0.4)
1.1 (0.2)
63.5 (23.5)
66.5 (20.0)
1.4 (0.3)
1.6 (0.5)
17.2 (13.4)
16.8 (10.4)
0.7 (0.4)
0.8 (0.4)
5.4 (10.0)
107.9 (105.3)
19.1 (26.5)
223.4 (192.2)
40.8 (41.6)
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).
45
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 O‟Doherty, 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; Rodríguez-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 (e.g.
species richness) alone will not necessarily maintain stable levels of ecosystem functioning
(Díaz 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 (Díaz et al.,
2007; Julliard et al., 2006; Trivellone et al., 2014). For example, higher densities of habitatspecialist 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.
46
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
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 (Díaz 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 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
47
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 Díaz 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 cavitynesting 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 (Díaz et al., 2005; Ibarra et al., 2012;
Martínez 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).
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 (ψ) 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
48
create a candidate set of models based on model weights (wi) and the precision of the estimated
coefficients, using an information-theoretic approach (Akaike‟s 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 Δ AIC ≤ 4 in the final confidence set for each bird (Burnham
and Anderson, 2002). Averaged models were used 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 ψ values of habitatspecialist 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 (ψ) to (i) D of endemic species, (ii) D of avian habitatuse 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. To assess the strength of
evidence for each tested model, I 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.
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 (ψ; 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 habitatspecialization were for austral parakeets (Enicognathus ferrugineus; SSI = 2.68) and
Magellanic woodpeckers (Campephilus magellanicus; SSI = 1.96), and the lowest for austral
49
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 ψ 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 ψ 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 ψ (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 ψ of S. rufipes was negative. There was strong
evidence of non-linearity (second degree quadratic function) for the correlation between ψ 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 ψ 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)
ψ 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 ψ 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 ψ 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 ψ 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).
50
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 (ψ) 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 (ψ) and avian diversity, respectively. In contrast, the relatively high
ψ 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.
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 (Díaz 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.
51
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 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 highelevations (> 900 m altitude; Ibarra et al., 2012; Veblen et al., 1980). Complex Andean foreststands 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 highelevation 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 (Díaz 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; Martínez 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 al., 2006). Bamboo understorey
has also been identified as providing critical protective cover and feeding habitat for groundgleaner 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
52
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
(Díaz 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 rather than the lack of certain
plant species is affecting habitat quality and long-term suitability for several species
inhabiting this biodiversity hotspot (Díaz 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).
53
Fig. 4.1. Relationship between probabilities of occurrence (ψ) 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 Akaike‟s Information
Criterion (AIC). Parameter estimates [SE] for covariates present in the top model set with Δ 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
Name
Geographic
distributionª
Habitat-use guildᵇ
Cavitynesting
speciesͨ
Species
specialization
index (SSI)
Volume of
coarse
woody
debris (m3)
Canopy
cover (%)
Bamboo
understorey
density (NC)
SD of diameter
at breast height
of trees (cm)
Chilean pigeon
(Patagioenas araucana)
Austral parakeet
(Enicognathus ferrugineus)
Green-backed firecrown
(Sephanoides sephaniodes)
Striped woodpecker
(Veniliornis lignarius)
Chilean flicker
(Colaptes pitius)
Magellanic woodpecker
(Campephilus magellanicus)
Thorn-tailed rayadito
(Aphrastura spinicauda)
Des Murs`s wire-tail
(Sylviorthorhynchus desmursii)
White-throated treerunner
(Pygarrhichas albogularis)
Black-throated huet-huet
(Pteroptochos tarnii)
Chucao tapaculo
(Scelorchilus rubecola)
Magellanic tapaculo
(Scytalopus magellanicus)
White-crested elaenia
(Elaenia albiceps)
Tufted tit-tyrant
(Anairetes parulus)
SSA
VPG
N
0.12
E
LTU
Y
2.68
SSA
VPG
N
0.23
SSA
LTU
Y
1.21
- 0.64 [0.15]
SSA
LTU
Y
0.37
- 0.21 [0.08]
E
LTU
Y
1.96
- 1.51 [0.35]
E
LTU
Y
0.41
- 0.23 [0.02]
E
UU
N
0.72
E
LTU
Y
0.38
- 0.22 [0.04]
E
UU
Y
0.96
- 0.44 [0.06]
+ 0.13 [0.04]
+ 0.04 [0.01]
E
UU
Y
0.78
- 0.44 [0.04]
+ 0.05 [0.01]
+ 0.02 [0.01]
SSA
UU
Y
0.57
- 0.25 [0.05]
+ 0.03 [0.01]
SA
VPG
N
0.16
SA
SU
N
0.32
+ 0.07 [0.03]
- 0.21 [0.11]
- 0.12 [0.03]
+ 0.07 [0.02]
+ 0.19 [0.09]
+ 0.63 [0.31]
- 0.25 [0.07]
+ 0.04 [0.01]
- 0.06 [0.01]
+ 0.11 [0.04]
55
Fire-eyed diucon
SSA
SU
N
0.11
(Xolmis pyrope)
Chilean swallow
SA
LTU
Y
0.23
- 0.07 [0.03]
(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)
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]
(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 (Díaz 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).
+ 0.02 [0.01]
+ 0.02 [0.01]
56
Table 4.2. Ranking of models relating measures of avian diversity and owl probabilities
of occurrence (ψ) in Andean temperate forests.
AICc
Δ AICc b
Wi c
-2*LLd
ER e
3
-50.10
0.00
0.75
-56.35
2.93
Strix rufipes + Strix rufipes
4
-47.95
2.15
0.25
-56.36
Glaucidium nana
3
-11.49
38.61
0.00
-17.74
Glaucidium nana + Glaucidium nana2 4
-9.32
40.77
0.00
-17.74
NULL
2
7.91
58.01
0.00
3.79
Strix rufipes + Strix rufipes2
4
-7.15
0.00
1.00
-15.57
Strix rufipes
3
5.35
12.50
0.00
-0.90
Glaucidium nana
3
55.55
62.70
0.00
49.31
Glaucidium nana + Glaucidium nana2 4
57.20
64.35
0.00
48.78
NULL
2
126.92
134.07
0.00
122.80
Strix rufipes + Strix rufipes2
4
-104.24
0.00
1.00
-112.66
Strix rufipes
3
-87.86
16.38
0.00
-94.11
Glaucidium nana
3
-40.82
63.43
0.00
-47.06
Glaucidium nana + Glaucidium nana2 4
-38.71
65.54
0.00
-47.12
NULL
2
31.23
135.47
0.00
27.10
Strix rufipes + Strix rufipes2
4
-110.07
0.00
0.99
-118.48
Strix rufipes
3
-100.94
9.13
0.01
-107.18
3
-50.10
59.96
0.00
-56.35
Glaucidium nana + Glaucidium nana 4
-48.57
61.49
0.00
-56.99
NULL
2
20.66
130.73
0.00
16.54
Strix rufipes + Strix rufipes2
4
-20.65
0.00
1.00
-29.06
Strix rufipes
3
-5.87
14.78
0.00
-12.12
3
46.72
67.37
0.00
40.48
Glaucidium nana + Glaucidium nana 4
48.51
69.15
0.00
40.09
NULL
Model specification
Strix rufipes + Strix rufipes2
Strix rufipes
Glaucidium nana
Glaucidium nana + Glaucidium nana2
NULL
118.57
AICc
-263.98
-258.64
-230.58
-228.58
-196.34
139.21
Δ AICc b
0.00
5.34
33.40
35.40
67.64
0.00
Wi c
0.94
0.06
0.00
0.00
0.00
114.44
-2*LLd
-272.40
-264.89
-236.83
-237.00
-200.46
Ka
Model specification
Species richness
Strix rufipes
2
Density of endemic species
518.14
Density of cavity-nesting species
3610.36
Density of large-tree users
Glaucidium nana
2
95.99
Density of understorey users
Glaucidium nana
2
2
Ka
4
3
3
4
2
1618.26
ER e
14.47
57
Density of shrub users
Strix rufipes + Strix rufipes2
4
-337.99
0.00
0.66
-346.41
3
-336.66
1.33
0.34
-342.91
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
Strix rufipes + Strix rufipes2
4
-263.76
0
1.00
-272.18
Strix rufipes
3
-241.11
22.65
0.00
-247.36
3
-211.39
52.37
0.00
-217.64
4
-209.71
54.05
0.00
-218.12
2
-141.79
121.97
0.00
-145.92
Strix rufipes
Glaucidium nana + Glaucidium nana
2
1.95
Community specialization index (CSI)
Glaucidium nana
Glaucidium nana + Glaucidium nana
NULL
2
82941.05
a
Number of parameters estimated.
Δ 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.
b
58
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 (ψ) for Strix rufipes in Andean temperate forests, based on model-averaged predictions.
Density
Predicted ψ for Strix
rufipes
Species
richness
Cavity
nesters
Largetree users
Understorey
users
Vertical profile
generalists
Shrub
users
Community
specialization index
Endemics
66
Low (0-0.33)
4.77 [1.16]
3.04 [0.81]
10.30
[1.22]
6.48
[0.91]
1.95 [0.53]
15.93 [1.24]
7.46
[0.45]
0.38 [0.05]
Moderate (0.34-0.66)
5.49 [1.261]
4.08 [1.11]
11.89
[1.68]
7.64
[1.22]
2.66 [0.78]
17.00 [1.14]
7.01
[0.60]
0.44 [0.06]
High (0.67-1)
7.12 [1.57]
8.06 [3.29]
17.46
[4.27]
10.96
[2.41]
5.42 [2.23]
19.66 [2.18]
6.15
[0.57]
0.67 [0.19]
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-López 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 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.,
71
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).
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 Gutiérrez, 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 Gutiérrez, 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 Gutiérrez,
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).
72
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, timeconsuming 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.
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
73
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; Mikusiński et al., 2001; Pearman and
Weber, 2007; Roberge et al., 2008; Sergio et al., 2004a; Similä 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 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 cavitybearing 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).
74
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,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), Castellón 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 O‟Doherty, 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 habitatspecialist 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
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
(Díaz 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
75
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.
76
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 Católica de Chile.
Altamirano, T.A., Ibarra, J.T., Hernández, F., Rojas, I., Laker, J., Bonacic, C., 2012. Hábitos de nidificación de las
aves del bosque templado andino de Chile. Fondo de Protección Ambiental, Ministerio del Medio
Ambiente, Serie Fauna Australis, Pontificia Universidad Católica 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 implementación del
pájaro carpintero gigante (Campephilus magellanicus) como especie carismática: una aproximación
biocultural para la conservación en la Reserva de Biosfera Cabo de Hornos. Magallania. 35, 71-88.
Armesto, J.J., Rozzi, R., Smith-Ramírez, 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., Tóth, 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.
77
Caley, M.J., Munday, P.L., 2003. Growth trades off with habitat specialization. Proc. R. Soc. B, Biol. Sci. 270, S175177.
Campioni, L., Sarasola, J.H., Santillán, 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., O‟Doherty, 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, 14321437.
Castellón, 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., Díaz, 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. Moonlight‟s 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 i
Birds. Academic Press, New York, U.S.A., pp. 3-56.
Crozier, M.L., Seamans, M.E., Gutiérrez, 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., Gutiérrez, 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., Villéger, S., Mouquet,
N., 2010. Defining 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. Bioclimatología de Chile. Dirección De Investigación, Vice-Rectoría
Académica, Universidad Católica De Chile, Santiago, Chile.
Díaz, 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 Chiloé Island, Chile. Biol. Conserv. 123, 91-101.
Díaz, 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.
Díaz, S., Cabido, M., 2001. Vive la différence: plant functional diversity matters to ecosystem processes.
Trends Ecol. Evol. 16, 646-655.
Díaz, S., Lavorel, S., de Bello, F., Quétier, 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.
Echeverría, 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.
78
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 Chiloé Island, Chile. J. Anim. Ecol. 80, 809-817.
Figueroa, R., Corales, S., Martínez, D., Figueroa, R., González-Acuña, 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. Ecología y conservación de los búhos chilenos: avances y desafíos de
investigación. 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 vegetación natural de Chile: clasificación y distribución geográfica. 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 world‟s 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., Sánchez-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.
79
Gustafsson, L., Baker, S.C., Bauhus, J., Beese, W.J., Brodie, A., Kouki, J., Lindenmayer, D.B., Löhmus, A.,
Martínez 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 redshouldered 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., O‟Connor, 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., Gálvez, 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., Gálvez, 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., Jiménez, 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.
Jiménez, 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, 10941100.
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.
80
Krüger, 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., Araújo, 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, 849856.
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 conservación del bosque nativo en Chile: problemas y desafíos, in: Armesto, J.J., Villagrán,
C., Arroyo, M.T. (Eds.), Ecología 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.
Lõhmus, A., 2003. Do ural owls (Strix uralensis) suffer from the lack of nest sites in managed forests? Biol.
Conserv. 110, 1-9.
Lourenço, R., Goytre, F., del Mar Delgado, M., Thornton, M., Rabaça, 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.
Martínez, D.R., 2005. El concón (Strix rufipes) y su hábitat en los bosques templados australes, in: SmithRamírez, C., Armesto, J.J., Valdovinos, C. (Eds.), Historia, Biodiversidad Y Ecología de Los Bosques
Costeros de Chile. Editorial Universitaria, Santiago, Chile, pp. 477-484.
Martínez, 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.
Martínez, J.A., Serrano, D., Zuberogoitía, 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.
Mikusiński, 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.
Murúa, R., 1995. Comunidades de mamíferos del bosque templado de Chile, in: Armesto, J.J., Villagrán, C.,
Arroyo, M.T. (Eds.), Ecología 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-López, 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, Muñoz-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.
O‟Donnell, 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., Gutiérrez, R.J., 2013. Life-history tradeoffs in spotted owls (Strix occidentalis): implications for
assessment of territory quality. Auk 130, 132-140.
Peery, M.Z., Gutiérrez, 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., Lourenço, 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 conservación 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 Alberta‟s 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., Díaz, 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., Mikusiński, 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.
Rodríguez-Estrella, R., Donázar, 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., Martínez, D.R., Willson, M.F., Sabag, C., 1996. Avifauna de los bosques templados de Sudamérica,
in: Armesto, J.J., Villagrán, C., Arroyo, M.T. (Eds.), Ecología 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, 297305.
Seamans, M.E., Gutiérrez, 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, 351362.
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.
Similä, M., Kouki, J., Mönkkönen, 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, 686700.
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-Ramírez, 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., Gutiérrez, 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/biospherereserves/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, 97104.
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