ArticlePDF Available

Pande, S.A., J.R. Duncan, and R. Yosef (eds). 2022. Proceedings of the 6th World Owl Conference, 29 November – 2 December 2019, Pune, India. Ela Journal of Forestry and Wildlife 11(1):1066-1243.

Authors:
  • Discover Owls - Education, Research and Conservation

Abstract and Figures

Proceedings of the 6th World Owl Conference held in Pune, India in 2019.
Content may be subject to copyright.
ISSN 2319-4361
(Indexed in Google Scholar)
Volume 11 | Issue 1
January - March 2022
Listed in UGC- CARE
Proceedings
of the
6th World Owl
Conference,
Pune, India,
2019
John Cordon
OENSL
1066 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
This special issue is dedicated to the
Late Anant Gokhale
whose tireless efforts made the
WOC, Pune, India 2019
a great success
| 1067
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
The 6th World Owl Conference started at Pune,
Maharashtra, India on 29th November 2019. During
the 5th World Owl Conference held at Evora, Portugal,
I had requested the WOC managing committee to
give Ela Foundation, our organization, an opportunity
to host the 6th World Owl Conference in India. I was
happy and felt highly honored that the managing
committee unanimously granted my request. I express
my gratitude to Dr. David Johnson and other dignitaries
of WOC managing committee for their support.
Owls are shrouded in superstitious beliefs and are
misunderstood all over the world, and our country is
not an exception. It was therefore very important that
the owl conference was held in India. The congregation
of international experts on owls would attract the much
needed public and media attention towards our owls.
The hosting of an owl conference for the rst time
in Asia, at Pune, India, where I stay, and where we
have been conducting research on owls for over three
decades, would also inspire further research on these
nocturnal and secretive birds that are very difcult to
study.
I accepted the responsibility of Organizing Secretary
of the 6th WOC and I am happy that the one of the
oldest universities in the country, Savitribai Phule Pune
University gladly agreed to be the co-hosts with Ela
Foundation and Maharashtra Forest Department. I must
express my sincere gratitude to Prof. Nitin Karmalkar,
Hon. Vice Chancellor of SPPU and Mr. Nitin Kakodkar,
IFS, Chief Wildlife Warden and PCCF, Maharashtra
Forest Department for their support. Prof. Karmalkar
offered the heritage venue of Sant Dnyaneshwar Hall
with its scintillating chandeliers and adjacent complex
for hosting the conference. We had the main conference
hall, poster presentation hall, exhibition hall and
photographic gallery to showcase the photographs of
owls taken in the natural habitats. A special student
conference was also conducted. Food was served in the
marble hall where owl artefacts, books, and other shop
items were displayed.
Prof. Reuven Yosef from Eilat, Israel and Dr James
Duncan from Manitoba, Canada, unconditionally
offered their expertise as members of the scientic
committee. The tireless efforts of Ela Foundation
volunters made the conference a great success. Excellent
research papers were presented in the 6th WOC, Pune,
India. The conference was attended by owl researchers
from 17 countries.
The WOC, Pune, India was followed by a vist to
Ela Habitat, Pingori, taluka Purandar, Pune, the eld
research station of Ela Foundation which is 61 km from
Pune. Researchers from the international community
visited the 2nd Indian Owl Festival which was hosted
there, back to back with the WOC. This two-day
festival, which is aimed to create public awareness and
education about owls to promote their conservation and
protection, was attended by atleast 15,000 students,
teachers, farmers, housewives, academicians, artists,
and others from various professions. Paintings,
sculptures, book marks, masks, book marks, and
several other artefacts created by students and artists
were displayed. All visitors were offered owl tattoos,
owl mehdi paintings on hands and owl face painting.
The 6th WOC, Pune, India and the 2nd Indian Owl
Festival received generous press and TV coverage.
I look forward to the 7th World Owl Conference
that aims to share and expand research frontiers and
knowledge about owls for their saftey and conservation.
Prof. Dr. Satish A. Pande
MD, DNB, PhD, FMASci., FLS
Organizing Secretary, 6th WOC, Pune, India
Director, Ela Foundation, Pune, India
Editorial
1068 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Savitribai Phule Pune University, Pune, the venue of the 6th World Owl Conference, Pune, India.
The inaugural session of the 6th World Owl Conference, Pune, India, 29th November - 2nd December 2019.
Participants of the 6th World Owl Conference, Pune, India.
6th World Owl Conference, Pune, India : Photo Feature
| 1069
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
The 6th World Owl Conference, Pune, India, address by the Hon. Vice Chancellor, Prof. N. Karmalkar.
Hon. VC with Prof. Dr. Satish Pande, Organizing Secretary, WOC, Pune, India at the photo exhibition.
Cultural program (dance and Mallakhamb) and lighter moments at the 6th World Owl Conference, Pune, India.
Participants of the 6th World Owl Conference, Pune, India.
6th World Owl Conference, Pune, India : Photo Feature
1070 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Patron:
l
Prof. Nitin Karmalkar, Hon. Vice Chancellor, SPPU
Organizing Secretary
Dr. Satish A. Pande,
MB, MD, DNB,PhD, F.M.A.Sci., FLS
Director, Ela Foundation and OENSL,
Pune, India.
Organizers:
Ela Foundaion, Pune
Savitribai Phule Pune University
(Departments of Zoology and Environmental
Sciences, SPPU)
Scientic Committee
l Dr. James Duncan, Canada
l Prof. Reuven Yosef, Israel
l David Johnson, USA
l Prof. Dr. Satish Pande, India
l Prof. Erkki Korpimaki, Finland
Organizing Committee
l Prof. Dr. Satish Pande, Ela Foundation
l Prof. Kalpana Pai, SPPU
l Prof. Suresh Gosavi, SPPU
l Dr. James Duncan, Discover Owls
l Ingrid Kohl, Austria
l Anant Gokhale, Ela Foundation
l Dr. Suruchi Pande, Ela Foundation
l Dr. Narhari Grampurohit, SPPU
l Dr. Kedar Ahire, SPPU
l Dr. Rajeev Dongre, Ela Foundation
Website
l Raghavendra Manavi
Communication
l Anant Gokhale
l Abhiram Rajandekar
Food
l Prashant Deshpande, Ela Foundation
l Sanjay Khatavkar Ela Foundation
Logistics & Tourism
l Swapnil Kumbhojkar - Daffodils Holidays
l Vivek Vishwasrao
l Vaibhav Gandhe
Owl Photo Exhibition
l Pramod Deshpande, Ela Foundation
l Dr. Satish Karmalkar, Ela Foundation
Publications
l Dr. Satish Pande
l Dr. Suruchi Pande
l Kiran Velhankar
l Shekhar Joshi
Venue
l Vivek Vishwasrao
l Dr. Kedar Ahire
l Sanjay Khatavkar
l Sudhir Deshpande
l Kaustubh Mudgal
Banking
l Chandrashekhar Naniwadekar
l Aashish Bakhle
l Anant Gokhale
l Dr. Suruchi Pande
Cultural Programs
l Dr. Nivedita Pande
l Mrunal Khatavkar
l Shirish Patwardhan (Swaroopvardhini)
l Prasad Kulkarni
l Arundhati Vishwasrao
Legal Advisor
l Chandrashekhar Naniwadekar
Liason Committee
l Anant Gokhale
l Prof. Kalpana Pai
l Prof. Suresh Gosavi
l Dr. Narhari Grampurohit
l Dr. Satish Pande
Maharashtra Forest Department
l Nitin Kakodkar, IFS, PCCF-CWLW
l Sunil Limaye, IFS, APCCF-WL
l Satysheel Gujar, IFS, Sahyadri Tiger Reserve
l Dr. Dinseh Tyagi, IFS, Social Forestry
l Vivek Khandekar, IFS
l Ramesh Kumar, IFS
WORLD OWL CONFERENCE 2019 PUNE-INDIA
29, 30 November, 1, 2 December 2019
| 1071
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
l Rajkumar Pawar
l Rahul Lonkar
l Faiyaz Shaikh
l Avishkar Bhujbal
l Pandurang Madane
l Akram Inamdar
l Zaid Attar
l Tausif Attar
l Abhiram Rajandekar
l Vaibhav Gandhe
l Vishal Hingane
l Rajendra Lambate
l Avadhoot Belsare
l Umaji Khomane
l Shamrao Gaikwad
l Rahul Choudhary
l Sachin Zagade
l Mayur Shinde
l Mayuri Shinde
l Suresh Shinde
l Shubhada Bhujbal
l Harshada Bhujbal
l Sana Shaiikh
l Vrushali Mhaske
l Radhika Bhosale
l Prajakta Thopte
l Divya Bhujbal
l Rutuja Shinde
l Pallavi Pawar
l Pratiksha Bhujbal
l Shubhangi Bardade
l Samruddhi Pawar
l Pragati More
l Priyanka Madane
l Mahesh Shinde
l Akshay Roman
l Somesh Pharande
l Adesh Pawar
l Ankit Khoche
l Dr. Omkar Sumant
l Prasanna Shah
l Dr. M.N.Mahajan
l Dr. Nivedita Pande
l Mahesh Bilaskar
l Hrishikesh Sankpal
l Avinash Nagare
l Tejas Kadam
l Sanket Ghate
l Shubham Kadam
l Sanjay Khatavkar
l Shabbir Lokhandwala
l Pramod Deshpande
l Banda Pednekar
l Amol Shinde
l Baba Shinde
l Atmaprakash Sahu
l Sudhanawa Rajurkar
l Manasi Pathak
l Sanjeewani Wad
l Vrushali Gambhir
l Priyamvada Gambhir
l Hema Rao
l Dr. Satish Pande
l Dr. Suruchi Pande
2nd Indian Owl Festival, Ela Habitat, Pingori
(3 - 4 December 2019)
Volunteers
1072 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Opening Remarks: A Brief Overview of the Previous
Five World Owl Conferences
James R. Duncan
Discover Owls, Box 253, Balmoral, Manitoba, Canada R0C 0H0
JDuncan@discoverowls.ca
Date of Publication:
31 March 2022
ISSN 2319-4361
On the morning of 29 November 2019 during the
inaugural session of the World Owl Conference, Pune,
India, I had an opportunity to deliver a brief oral
presentation on the history of the previous ve World
Owl Conferences. The following is a written account of
that presentation:
In March 2018, I had the pleasure, on behalf of
the World Owl Conference Organizing Committee, to
inform Professor Dr. Satish Pande that the committee
had approved India’s proposal for the ELA Foundation
to host the 6th World Owl Conference in beautiful Pune,
India, in 2019.
This series of international owl conferences, started
in Winnipeg, Manitoba, Canada in 1987, brings together
scientists and others from around the world to share
their owl expertise, knowledge, and wisdom to further
the study, appreciation and conservation of owls.
The location of the international owl conferences
has varied, as has the attendance which has ranged
from 130 to 193 delegates from a variety of countries
as follows:
1987 – 150 delegates – 10 countries – 52 papers –
Manitoba Wildlife Branch, Winnipeg, Manitoba,
Canada
1997 – 180 delegates – 15 countries – 91 papers –
Manitoba Wildlife Branch, Winnipeg, Manitoba,
Canada
2000 – 130 delegates – 10 countries – 51 papers –
Australasian Raptor Association, Canberra, Australia
2007 – 193 delegates – 32 countries – 86 papers
– Worldwide Owl Conference 2007, Groningen,
Netherlands
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
Keywords:
World Owl Conferences; Owl Research and Conservation; India; Canada; Australia; Netherlands; Portugal.
| 1073
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
2017 – 155 delegates – 30 countries – 89 papers –
Universidade de Evora, Evora, Portugal
It is noteworthy that each conference has produced
a proceedings containing peer-reviewed published
papers that serve as a continuing legacy and source
of knowledge to further the conservation of owls
around the world. These proceedings include:
Nero, R.W., Clark, R.J., Knapton, R.J. and R.H.
Hamre (Eds.). 1987. Biology and Conservation of
Northern Forest Owls: Symposium proceedings. 3-7
February 1987. Winnipeg, Manitoba. USDA Forest
Service Gen. Tech. Report RM 142. 309 p.
Duncan, J.R., Johnson, D.H. and T.H. Nicholls (Eds.).
1997. Biology and Conservation of Owls of the
Northern Hemisphere: 2nd International symposium,
5-9 February 1997. Winnipeg, Manitoba. USDA
Forest Service Gen. Tech. Rep. NC-190. 635 p.
Newton, I., Kavanagh, R., Olsen, J. and I. Taylor
(Eds.). Ecology and Conservation of Owls.
Symposium proceedings from Owls 2000: The
biology, conservation and cultural signicance of
owls. 19-23 January 2000, Manning Clark Centre,
Australian National University, Canberra, Australia.
Csiro Publishing, Collingwood, Australia. 363 p.
Johnson, D.J., Van Nieuwenhuyse, D. and J.R.
Duncan (Eds.). 2009. Owls – Ambassadors for the
Protection of Nature in their Changing Landscapes.
Proceedings of the fourth World Owl Conference.
31 October – 4 November 2007, Groningen, The
Netherlands. Ardea 97(4): 395-649.
Roque, I.M.F., Duncan, J.R., Johnson, D.H. and Van
Nieuwenhuyse, D. (Eds.). 2021. Proceedings of the
2017 World Owl Conference. Evora, Portugal. Airo
29.
Many of us here are assisting with the 2019 World
Owl Conference and will be helping to prepare the
follow-up proceedings. This rich history of international
collaboration and support from many sponsors has
culminated in bringing us to this historic point in
time and will guarantee to make the 2019 World Owl
Conference a success.
Marie Jaworski
Saw-whet Owl
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1074 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Small mammal prey and owl predator population dynamics over
a 20-year period (1991-2010) in Manitoba, Canada and adjacent
Minnesota, USA, and an evaluation of a citizen science project.
James R. Duncan
Discover Owls, Box 253, Balmoral, Manitoba, Canada R0C 0H0
JDuncan@discoverowls.ca
Keywords: Small mammal monitoring; owl monitoring; predator prey dynamics; citizen science; Manitoba; Canada
[Keynote oral presentation submitted as a research article for the Proceedings for the 2019 WOC]
Citation:
Duncan James R., Small mammal prey
and owl predator population dynamics over
a 20-year period (1991-2010) in Manitoba,
Canada and adjacent Minnesota, USA, and an
evaluation of a citizen science project.
Ela Journal of Forestry and Wildlife
Vol.11 (1): 1074-1090
Date of Publication:
31 March 2022
ISSN 2319-4361
Abstract
From 1986 to 2010 annual small mammal
relative abundance indices were obtained by a snap-
trapping program in southeastern Manitoba, Canada
and adjacent Minnesota, USA. From 1991 to 2015
owl populations were surveyed extensively across
Manitoba by volunteers in a citizen science program.
Thus, both small mammal prey and owl predator
populations were sampled concurrently for 20 years
(1991-2010) enabling an exploration of predator-prey
dynamics in North America’s boreal forest and adjacent
ecosystems. Pooled relative abundance indices for
11 owl species varied signicantly with those for 9
small mammal species (R2 = 0.514, p = 0.0004) and
generally uctuated synchronously over time. Further
analysis of owl and small mammal prey species indices
revealed that only the two most common species of
owls detected, the Great Horned Owl Bubo virginianus
and the Northern Saw-whet Owl Aegolius acadicus,
varied signicantly with pooled small mammal prey
abundance. Conversely only microtine rodents (voles,
subfamily Arvicolinae), but not shrews (Family
Soricidae) or mice (suborder Myomorpha), varied
signicantly over time with changes in the relative
abundance indices for pooled owl species, Great Horned
Owl, Northern Saw-whet Owl, and the Great Gray Owl
Strix nebulosa. Akaike Information Criterion analysis
corrected for small samples (AICc) selected Meadow
Vole Microtus pennsylvanicus and pooled small
mammal indices as frequent model predictors of owl
species abundance indices and had best approximating
models consistent with signicant linear regression
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1075
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
results for only three of eight owl species. These
analyses indicate that the Manitoba citizen science owl
monitoring program provided valuable time series data
for at least three commonly and/or regularly detected
owl species in relation to prey population changes
over time. This conclusion can be used to encourage
continued participation in this citizen science project.
Introduction
Long-term wildlife monitoring programs are
costly and time consuming but help dene ecological
processes, assess the conservation status of species, and
inform management action planning (Takats et al. 2001;
M
anley et al. 2004; Holthausen et al. 2005; Witmer 2005;
Duncan et al. 2009). Such programs are typically either
intensive (i.e., conducted by scientists at small spatial
scales) or extensive (i.e., conducted by volunteer citizens
over large areas) but rarely, if ever, both (Bonney et al.
2009; Proença et al. 2017). Such divergent approaches
to monitoring have relative strengths and weaknesses
regarding sample sizes and geographic relevance for
extrapolation over areas and through time (Manley et
al. 2004; Quinn and Keough 2002; Proença et al. 2017).
This study examined two independent, concurrent, and
sympatric monitoring programs (one intensive and one
extensive) in Manitoba, Canada and adjacent Minnesota,
USA (Fig. 1) to determine if they informed each other in
an ecologically meaningful manner. Of specic interest
was assessing if and how extensively collected citizen
science owl population abundance indices related to
intensively collected small mammal prey population
indices over a 20-year period.
Methods – Owl Survey
Owl species scientic names followed Handbook of
the Birds of the World and BirdLife International (2019).
The Manitoba Nocturnal Owl Survey methods were
described in detail in Duncan (2021) and are therefore
only briey summarized herein. Survey participants
were provided resources to learn the territorial calls
of the 12 Manitoba owl species, instructions, and
data sheets (https://www.researchgate.net/project/
Manitoba-Nocturnal-Owl-Survey). Their ability to
identify owl species was not tested. Survey routes (Fig.
2) were assigned in a non-random manner as access
to roads in early spring was limited due to thick snow
cover and/or spring ooding. Volunteers surveyed at
least one survey/route per year between mid-March to
mid-April starting at least 30 min after sunset. At each
survey stop individual owls detected (heard or seen) and
the owl’s estimated distance and direction from the stop
were recorded. Some individual owls could be heard
from multiple stops. Therefore, surveyors recorded if
an owl detected was also detected at a previous stop or
stops to prevent the same owl from being counted more
than once. Additional information recorded at each stop
included time, an odometer reading, noise interference
and the number of passing cars.
The estimated number of individual owls detected
per route were pooled annually and used to calculate
total annual owl species indices (number of individuals
of a given species detected/km surveyed/year) to
standardize variable annual survey effort. An overall
annual owl abundance index was calculated by pooling
all individuals of all species detections/km surveyed/
year. The owl survey expanded over time and the survey
protocol changed as follows (see also Duncan 2021).
• 1991 – 1999: Survey stops were 0.8 km apart
and took a minimum of 3 min and 40 s to
complete: 1 min of listening, 20 s playback
of Boreal Owl Aegolius funereus or Northern
Saw-whet Owl Aegolius acadicus territorial
calls, 1 min of listening, 20 s playback of Great
Gray Owl Strix nebulosa or Eastern Screech
Owl Megascops asio territorial calls, 1 min
of listening. No standard survey route length
(number of stops) was prescribed. Boreal Owl
and Great Gray Owl playback was used in
Boreal Forest habitats, whereas Northern Saw-
whet Owl and Eastern Screech Owl playback
was used in other habitats.
• 2000 – 2015: A standardized Canadian volunteer
owl survey protocol was developed in 1999
and implemented in 2000 (Takats et al. 2001)
where a survey route consisted of 10 survey
stops spaced 1.6 km apart. Surveyors listened
for owls passively (no playback used) for 2 min
at each stop.
In general, the survey method change did not preclude
the comparison of the long-term data sets herein as
the owl survey indices are relative and not absolute
abundance estimates. However the implications of the
change in owl survey methods for owl detection rates
has been reviewed and discussed in Duncan (2021) and
addressed in the methods and results section below.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1076 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Methods – Small Mammal Survey
Small mammal scientic names follow the Integrated
Taxonomic Information System (http://www.itis.
gov) accessed 1 October 2020. Small mammals were
trapped annually (late September to mid-October 1986–
2010) at two study areas 80 km apart in southeastern
Manitoba and adjacent Minnesota (Fig. 3). The annual
trapping effort at each study area consisted of six trap
lines, in three pairs, with 50 trap stations (spaced 10 m
apart) per line for a total of 300 stations per study area
(Fig. 4). One Museum Special snap-trap (Woodstream
Corp., Lititz, Pennsylvania, USA), baited with peanut
butter, rolled oats and bacon fat, was set at each station
(Fig. 5). The traps were set for three nights and checked
each morning. Trapped mammals were removed
and identied, and traps were reset or re-baited as
required. In each study area a pair of trap lines sampled
a Tamarack Larix laricina dominated forest stand,
while the other two pairs sampled open habitats with
numerous perches used by hunting owls such as open
treed muskeg, burned over areas, or cleared roadsides
and adjacent drainage areas (Duncan 1987, Fig. 6). The
annual trapping effort was quantied as the number of
trap stations available times the number of nights (#
trap nights); occasional missing or broken traps were
not counted, reducing the annual total effort slightly
each year (Fig. 7). Indices (# mammals caught / # trap
nights) estimated the relative abundance of mammal
prey categories (see below) and pooled mammal prey
abundance for each year to standardize variable annual
survey effort. Total annual index values were used for
regression and other analyses.
Trapped small mammals were grouped into four
prey categories for analysis as follows:
Red-backed Voles: Southern Red-backed Vole
Myodes gapperi
Meadow Voles: Meadow Vole Microtus
pennsylvanicus and Northern Bog Lemming
Synaptomys borealis. The latter species was rarely
caught and occupied similar habitats as Meadow Voles
therefore the two species were pooled for this analysis.
Shrews: Masked Shrew Sorex cinereus, American
Pygmy Shrew Sorex hoyi, Arctic Shrew Sorex arcticus,
Northern Water Shrew Sorex palustris, and Northern
Short-tailed Shrew Blarina brevicauda.
Other: Deer Mouse Peromyscus maniculatus,
Woodland Jumping Mouse Napaeozapus insignis.
Methods – Data Analysis
Data analyses were prepared using 2016 Microsoft
Excel Data Analysis Tool Pack and Daniel’s XL
Toolbox version 7.2.13 (http://www.xltoolbox.net) and
Statistix 10 (www.statistix.com, Analytical Software).
More owls were detected in the earlier survey period
(1991-1999), when playback was used, than in the later
period (2000-2010) when no playback was used (Duncan
2021). This ‘trend’ was likely the result of the change
in methods, rather than a change in owl abundance,
and the indices were ‘detrended’ as in Duncan et al.
(2009) and as follows. For each of the two owl survey
periods, the residuals of a linear regression of observed
and expected index values were calculated. Since the
residuals are larger (in absolute terms) for larger indices,
the residuals in the two periods were made comparable
by converting them into residual ratios (dividing each
residual by the mean index for that period). Hereafter
owl indices expressed as residual ratios will be referred
to simply as owl indices.
General linear models were used for modelling owl
– prey relationships and these were contrasted with the
results of simple linear regression analysis (Kassambara
2018; Bevans 2020). A series of general linear models
were t using pooled and select different owl species
as responses, and pooled and different small mammal
indices as predictors. Candidate models were ranked
using Akaike Information Criterion analysis corrected
for small samples (AICc) to identify the relative
inuence of small mammal prey category abundance
indices on owl abundance indices. The models were t
using PROC MIXED in SAS version 9.4 (SAS Institute
Inc.) with an auto-regressive order 1 error correlation
structure. This allowed for temporal correlation among
the residuals of adjacent years.
No models were run for Burrowing Owl Athene
cunicularia, Snowy Owl Bubo scandiacus or Barn Owl
Tyto alba as their indices were zero for 20, 17, and 14
of 20 years, respectively. Due to the small sample size
(20 years), model forms were restricted to maximum of
two additive predictors. A total of 12 Candidate Model
Sets included all one and two factor combinations of
each of the four small mammal prey categories, a null
model and a pooled small mammal index model. Best
approximating models and competing models (within
2 AICc units of the top model) were identied. If null
or intercept only models were best approximating
or highly competitive, this indicated that there were
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1077
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
no sufciently strong linear relationships with small
mammal predictor models.
Results – Owl Surveys
An estimated 6,335 individual owls of 11 species
were detected on a total of 32,549 km of linear point
count surveys over 25 years (1991 to 2015) by ca. 900
surveyors (Duncan 2021). Northern Saw-whet Owls,
Great Horned Owls Bubo virginianus, and Boreal Owls
were detected every year and accounted for 75% of all
detections (Duncan 2021). Other owl species detected
every year included the Great Gray Owl, Long-eared
Owl Asio otus, and Barred Owl Strix varia. The Northern
Hawk Owl Surnia ulula, Eastern Screech Owl, Short-
eared Owl Asio ammeus, Barn Owl, and Snowy Owl
were detected at an order of magnitude lower and were
not detected every year. The Burrowing Owl was the
only Manitoba owl species not detected. Owl species
relative abundance varied annually (Fig. 8).
Results – Mammal Surveys
A total of 6,733 small mammals were trapped during
a total of 44,587 trap nights over 25 years. For analysis,
trapped mammals were tallied into the following four
small mammal categories (see methods): Southern
Red-backed Voles (3,916), Meadow Voles (1,429),
Shrews (1,161) and Other (227). The number of small
mammals trapped varied considerably from year to
year (Fig. 9).
Results – Owl and Mammal Comparisons
The intensive prey and extensive owl data sets
overlapped for 20 years from 1991-2010. The
following tests ensured that the assumptions of simple
linear regression analysis were met. Both the mammal
indices (W = 0.9701, n = 20, P = 0.756) and owl
indices (W = 0.9193, n = 20, P = 0.165) were normally
distributed (Wilkes - Shapiro normal residuals). The
owl – mammal linear regression error terms were not
correlated (Durbin-Watson test P > 0.345) and there
were no signicant autocorrelations (Runs Test) for
either the mammal indices nor for the owl indices and
the data sets were considered independent. The owl and
mammal data sets were time lagged in either direction,
but there were no signicant predator-prey lags.
Pooled owl indices and pooled small mammal indices
showed generally synchronized uctuations over time
(Fig. 10) and their regression was signicant (Fig. 11,
Table 1). There were also signicant linear regressions
for both Great Horned Owls and Northern Saw-whet
Owls indices with pooled small mammal indices (Table
1). Linear regressions were signicant between each of
the Red-backed Vole and Meadow Vole small mammal
prey category indices and pooled owl abundance indices
(Table 1). Lastly, linear regressions were signicant for
the indices of four other owl species – mammal prey
category pairs (Table 1).
A sample of the AICc analysis results for Northern
Saw-whet Owl is presented in Table 2 and Figure 12.
In this case the best approximating model was the
Red-backed Vole index model indicating that it had
the strongest linear relationship with the Northern
Saw-whet Owl index; the pooled small mammal index
was the best competing model (Table 2). A summary
of all the AICc model output for owl species indices
with small mammal prey category indices as model
predictors is presented in Table 3. A comparison of
linear regressions and AICc model results are discussed
below. In decreasing order, Meadow Vole, Other, or All
Mammals were the most frequent best approximating
model predictors followed by Red-backed Vole and
Shrews; one null best approximating model was
selected (Short-eared Owl; Tables 3, 4). The selected
null model indicates that there were no sufciently
strong linear relationships identied between the Short-
eared Owl and any small mammal indices.
Discussion
The signicant predator-prey relationship between
many owl species and small mammals is well known
(Duncan 2003, Konig and Weick 2009). Owl species
detected in this study eat small mammals as prey to
varying degrees depending on body size and behaviour;
some like the Great Gray Owl are vole specialists
(Duncan 1992) while others like the Great Horned
Owl are diet generalists (Duncan 2013). It is generally
recognized that small mammal prey populations
regulate northern owl numbers (Lehikoinen et al, 2011)
by affecting their survival, reproduction and dispersal
(Cheveau et al. 2004, Bowman et al. 2010, Confer et al.
2014). The demonstration of signicant relationships
between owl and mammal indices in this study
documents that extensively collected Manitoba owl
survey data has an ecological connection to intensively
collected small mammal prey data, thereby validated
the ability of citizen scientists to collect meaningful
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1078 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
owl population data, at least for relatively abundant and
regularly detected owl species.
Pooled small mammals and more commonly
trapped small mammal prey species had a signicant
inuence on pooled owl abundance indices over time
based on linear regression analyses (Fig. 11, Table 1).
Furthermore, linear regression demonstrated signicant
relationships between more commonly trapped small
mammal prey species and three of six regularly detected
owl species (Table 1). One of these, the Great Gray
Owl, is a diet specialist that in Manitoba eats primarily
Meadow Voles (Duncan 1992). While Meadow Voles
were not the most frequently trapped small mammal
prey (Fig. 9) their unique inuence on the detection
rates of Great Gray Owls was corroborated by these
results (Table 1).
The AICc analysis frequently selected more
common small mammal prey category (Meadow
Vole, pooled small mammals and Red-backed Vole)
indices as model predictors of owl indices (Table 3, 4).
These included best approximating models for all six
regularly detected owl species, pooled owls and two
less frequently and irregularly detected owl species,
the Northern Hawk Owl and the Eastern Screech Owl
(Table 3). Only two of these (Great Horned Owl and
Northern Saw-whet Owl) and the pooled owl index
were consistent with signicant linear regression
results (Table 1). The AICc results therefore identied
some potentially important predator-prey models or
relationships identied as not signicant by the linear
regression analysis and they tended (for 6 of 9 owl
indices) to pick best approximating model predictors
that matched the lowest, sometimes nearly signicant,
probabilities generated for the corresponding linear
regressions (i.e., Barred Owl – Pooled or All Mammals,
P = 0.071; Tables 1, 3).
Both linear regression and AICc analyses were
valuable approaches to assessing the Manitoba citizen
science owl survey but the AICc identied best
approximating predator – prey models relative only
to other models considered, and not in an absolute
sense. More analyses of these and similar data sets
is encouraged, including the use of simulations and
Bayesian Information Theory (Brewer et al. 2016).
The signicant relationship between the dependent
(pooled owl) and the independent (pooled small
mammal) abundance indices is evidence that the
owl data collected extensively by citizen scientists is
ecologically valid. This conclusion held true for the
two most frequently detected owls, the Northern Saw-
whet Owl and Great Horned Owl, but, except for the
Great Gray Owl, not for less commonly or irregularly
detected species (Table 1). The are various reasons
some owl species were detected less frequently. The
Eastern Screech Owl has a limited, largely urban
distribution in Manitoba (Taylor 2003) and the Northern
Hawk Owl is thought to be an irregular and irruptive
migrant and predominantly calls diurnally (Duncan &
Duncan 1998). The Short-eared Owl is relatively rare
in the province in most years, and most likely migrates
quickly through Manitoba to the tundra to breed (Taylor
2003). The Barn Owl is rare and accidental in Manitoba
and the majority of Snowy Owls depart southern
Manitoba for their Arctic breeding range prior to the
survey period (Taylor 2003). Lastly, the endangered
Burrowing Owl was never detected due to its arrival in
Manitoba as a late spring migrant (Taylor 2003) after
the survey period ended. It also has an exceedingly
small range in extreme southwestern Manitoba where
few owl surveys took place. Duncan (2021) concluded
that these ve species require species-specic targeted
survey methods to adequately monitor their populations
in Manitoba. In general, less common species need
other means of monitoring their populations (Manley
et al. 2004; Holthausen et al. 2005).
Recruiting and retaining volunteer citizen scientists
for extensive owl or wildlife monitoring programs
is important (Bonney et al. 2009). Volunteer citizen
scientists were mainly motivated to participate in the
Manitoba owl survey for social reasons, i.e., having fun
with family and friends (Ng et al. 2018). The results of
this current study can be used to reassure and educate
volunteers that they are collecting ecologically relevant
data that is valuable to help manage and conserve owls.
Acknowledgements
Thanks to my wife and colleague, Patricia Duncan,
for her support and assistance with our research on owls.
My former professor Robert W. Nero continues to be an
inspiration for engaging the public in owl research and
conservation. Steve Wilson of Minnesota generously
shared his owl survey protocols that were then adapted
to start the Manitoba owl survey. Llwellyn Armstrong
with Ducks Unlimited Canada assisted with the AICc
analysis. Llwellyn Armstrong, Dean Berezanski,
Patricia Duncan, Riki Kerbrat, and Prachi Mehta
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1079
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
kindly reviewed and greatly improved this manuscript.
Financial and other support came from Bird Studies
Canada’s James Baillie Memorial Fund, the University
of Manitoba’s Alumni Fund, and the Government
of Manitoba. Bird Studies Canada’s Beaverhill Bird
Observatory and Nature Saskatchewan enabled Colin
Murray to kindly prepare the owl survey map. Thanks
to the over 900 hundred volunteer owl surveyors for
participating in the owl survey and to Dr. Satish Pande,
the ELA Foundation and many volunteers for hosting
the 2019 World Owl Conference.
All Mammals RBV MV Shrews Other
Owl Species R2P R2P R2P R2P R2P
All Owls 0.514 0.0004 0.433 0.002 0.633 0.00003 0.072 0.251 0.054 0.323
Great Horned Owl Bubo virginianus* 0.238 0.029 0.115 0.144 0.339 0.007 0.124 0.127 0.021 0.541
Northern Saw-whet Owl Aegolius
acadicus* 0.228 0.033 0.263 0.021 0.229 0.033 0.005 0.761 0.005 0.763
Barred Owl Strix varia* 0.170 0.071 0.104 0.165 0.126 0.125 0.134 0.113 0.028 0.480
Boreal Owl Aegolius funereus* 0.108 0.157 0.113 0.147 0.114 0.146 0.022 0.531 0.057 0.312
Barn Owl Tyto alba 0.081 0.225 0.029 0.470 0.119 0.137 0.073 0.248 0.015 0.607
Northern Hawk Owl Surnia ulula 0.029 0.476 0.006 0.742 0.044 0.373 0.033 0.440 0.000 0.947
Great Gray Owl Strix nebulosa* 0.084 0.216 0.079 0.229 0.207 0.044 0.004 0.788 0.167 0.073
Eastern Screech Owl Megascops asio 0.021 0.544 0.006 0.740 0.004 0.802 0.123 0.129 0.000 0.982
Long-eared Owl Asio otus* 0.055 0.321 0.071 0.257 0.116 0.142 0.015 0.607 0.059 0.304
Short-eared Owl Asio ammeus 0.001 0.880 0.003 0.810 0.054 0.325 0.172 0.069 0.000 0.967
Snowy Owl Bubo scandiacus 0.041 0.392 0.071 0.257 0.033 0.447 0.023 0.526 0.067 0.271
Table 1. Simple linear regression of owl abundance indices as residual ratios and small mammal prey abundance indices
from Manitoba, Canada and adjacent Minnesota, USA (1991-2010). See text for details on abundance indices. Owl
species regularly detected on owl surveys each year are denoted with an asterix (*). P = probability and yellow shading
denotes signicance (P < 0.05). All Mammals = pooled small mammal data; RBV = Southern Red-backed Vole Myodes
gapperi; MV = Meadow Vole Microtus pennsylvanicus and Northern Bog Lemming Synaptomys borealis; Shrews =
Masked Shrew Sorex cinereus, American Pygmy Shrew Sorex hoyi, Arctic Shrew Sorex arcticus, Northern Water Shrew
Sorex palustris, and Northern Short-tailed Shrew Blarina brevicauda; Other = Deer Mouse Peromyscus maniculatus and
Woodland Jumping Mouse Napaeozapus insignis.
Great horned Owl
Ken Stewart
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1080 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Model Predictors k -2 log Likelihood AICc Delta AICc AICc weight
RBV 4 -94.54 -83.87 0.00 0.43
All Mammals 4 -92.75 -82.09 1.79 0.17
RBV + Shrews 5 -94.72 -80.43 3.44 0.08
MV 4 -91.06 -80.40 3.48 0.07
RBV + MV 5 -94.65 -80.37 3.50 0.07
RBV + Other 5 -94.58 -80.29 3.58 0.07
NULL 3 -87.09 -79.59 4.28 0.05
Shrews 4 -87.82 -77.16 6.72 0.01
MV + Other 5 -91.30 -77.02 6.85 0.01
MV + Shrews 5 -91.06 -76.78 7.09 0.01
Other 4 -87.25 -76.58 7.29 0.01
Shrews + Other 5 -87.85 -73.56 10.31 0.00
Table 2. Sample AICc model output for Northern Saw-whet Owl Aegolius acadicus and various small mammal prey
indices as model predictors. The best approximating model (RBV) is highlighted in yellow and the best competing
model (All Mammals) in orange. All Mammals = pooled small mammal data; RBV = Southern Red-backed Vole Myodes
gapperi; MV = Meadow Vole Microtus pennsylvanicus and Northern Bog Lemming Synaptomys borealis; Shrews =
Masked Shrew Sorex cinereus, American Pygmy Shrew Sorex hoyi, Arctic Shrew Sorex arcticus, Northern Water Shrew
Sorex palustris, and Northern Short-tailed Shrew Blarina brevicauda; Other = Deer Mouse Peromyscus maniculatus and
Woodland Jumping Mouse Napaeozapus insignis. k = number of model parameters.
Predictors Appearing in Best Approximating and Top Competing Models
Dependent Variable Best Approximating Model
form 1st Competing Model
form* 2nd Competing
Model form*
All Owl Index Meadow Vole
Short-eared Owl NULL
Northern Saw-whet Owl1Red-backed Vole All Mammals
Northern Hawk Owl Meadow Vole
Long-eared Owl1Other NULL
Great Horned Owl1All Mammals Meadow Vole Shrews
Great Gray Owl1Other
Eastern Screech Owl Shrews NULL
Boreal Owl1Meadow Vole NULL
Barred Owl1** All Mammals Red-backed Vole -
Other
Meadow Vole -
Other
* within 2 AICc units
1 Owl species regularly detected on owl surveys each year
** Barred Owl with three additional successive Competing Models: Other; Red-backed Vole; and Shrews
Table 3. Summary of AICc model output for owl species (see Table 1 for owl scientic names) and various small mammal
prey indices as model predictors. Null indicates that there were no sufciently strong linear relationships identied
between owl and any small mammal indices. All Mammals = pooled small mammal data; Red-backed Vole = Myodes
gapperi; Meadow Vole = Microtus pennsylvanicus and Northern Bog Lemming Synaptomys borealis; Shrews = Masked
Shrew Sorex cinereus, American Pygmy Shrew Sorex hoyi, Arctic Shrew Sorex arcticus, Northern Water Shrew Sorex
palustris, and Northern Short-tailed Shrew Blarina brevicauda; Other = Deer Mouse Peromyscus maniculatus and
Woodland Jumping Mouse Napaeozapus insignis.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1081
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Independent
Variable
Frequency in Best
Approximating AICc Model Frequency in Competing
AICc Model
Frequency in Best
Approximating or
Competing AICc Model
Meadow Vole 3 2 5
Other 2 3 5
All Mammals 2 1 3
NULL 1 3 4
Red-backed Vole 1 2 3
Shrews 1 2 3
Table 4. Frequency of small mammal prey independent model variables in the best approximating or a competing model
in all AICc model sets for owl abundance indices. Null indicates that there were no sufciently strong linear relationships
identied between owl and any small mammal indices. All Mammals = pooled small mammal data; Red-backed Vole =
Myodes gapperi; Meadow Vole = Microtus pennsylvanicus and Northern Bog Lemming Synaptomys borealis; Shrews =
Masked Shrew Sorex cinereus, American Pygmy Shrew Sorex hoyi, Arctic Shrew Sorex arcticus, Northern Water Shrew
Sorex palustris, and Northern Short-tailed Shrew Blarina brevicauda; Other = Deer Mouse Peromyscus maniculatus and
Woodland Jumping Mouse Napaeozapus insignis.
Figure 1. Location of Manitoba (red outline) in association with major ecosystems in Canada. Source: https://
prosfa.vn/wp-content/uploads/2017/08/515bfe84-d6e7-40c2-b645-44a6bc714c95.jpg accessed 5 October 2020.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1082 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Figure 2. Location (red dots) of
Manitoba Nocturnal Owl Survey
Routes (1991-2015).
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1083
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Figure 3. Location of small mammal survey areas (1986 to 2010) in southeastern Manitoba, Canada (A) and
adjacent Minnesota, USA (B). Figure image source: Google Earth accessed 5 October 2020.
Figure 4. Location of small mammal survey lines in two study areas (1986 to 2010) in southeastern Manitoba,
Canada (A) and adjacent Minnesota, USA (B). Google Earth imagery dates were 1 May 2016 (left) and 9 July 2015
(right) accessed 5 October 2020.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1084 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Figure 5. Examples of small mammals trapped, including Southern Red-backed Vole Myodes gapperi (Left) and
Masked Shrew Sorex cinereus (Right), in southeastern, Manitoba and adjacent Minnesota, USA.
Figure 6. Examples of habitats surveyed for small mammals in southeastern, Manitoba and adjacent Minnesota,
USA. Upper left: Mature Tamarack Larix laricina Forest; Upper right: Treed Muskeg; Lower left: Cut Over and
Burned Areas; Lower right: Drainage Ditch.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1085
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Figure 7. Small mammal survey sample eld data sheets. See text for explanation of prey and habitat category
abbreviations. “# TN’s” = number of total trap nights.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1086 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Figure 8. Owl species survey abundance indices as residual ratios by year in Manitoba,
Canada.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1087
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Figure 9. Small mammal prey category indices by year in southeastern Manitoba, Canada
and adjacent Minnesota, USA. See text for explanation of prey categories.
Figure 10. Pooled owl relative abundance indices as residual ratios and pooled small mammal relative
abundance indices (see text) by year in Manitoba, Canada and adjacent Minnesota, USA.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1088 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Figure 11. Scatter plot and linear regression of pooled owl relative abundance indices as residual ratios and pooled
small mammal relative abundance indices (see text) in Manitoba, Canada and adjacent Minnesota, USA, from
1991 to 2010.
Figure 12. Northern Saw-whet Owl Aegolius acadicus (NSWO) and Red-backed Vole Myodes gapperi (RBV)
relative abundance indices by year in Manitoba, Canada and adjacent Minnesota, USA, including model-predicted
NSWO values and 95% condent limits.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
| 1089
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
References
Bevans, R. 2020. An introduction to the Akaike
information criterion. https://www.scribbr.com/
statistics/akaike-information-criterion/. Accessed 6
June 2020.
Bonney, R., Cooper, C.B., Dickinson, J., Kelling,
S., Phillips, T., Rosenberg, K.V. and Shirk, J. 2009.
Citizen science: a developing tool for expanding
scientic knowledge and scientic literacy.
BioScience 59: 977–984.
Bowman, J., Badzinski, D.S. and Brooks, R.J. 2010.
The numerical response of breeding Northern Saw-
whet Owls Aegolius acadicus suggests nomadism.
Journal of Ornithology 151: 499–506.
Brewer, M.J., Butler, A. and Cooksley, S.L. 2016.
The relative performance of AIC, AICC and BIC in
the presence of unobserved heterogeneity. Methods
in Ecology and Evolution 7: 679–692.
Cheveau, M., Drapeau, P., Imbeau, L. and Bergeron,
Y. 2004. Owl winter irruptions as an indicator of
small mammal population cycles in the boreal forest
of eastern North America. Oikos 107: 190-198.
Confer, J.L., Kanda, L. and Li, I. 2014. Northern
Saw-whet Owl: regional patterns for fall migration
and demographics revealed by banding data. Wilson
Journal of Ornithology 126: 305-320.
Duncan, J.R. 1987. Movement strategies, mortality
and behaviour of radio-marked great gray owls in
southeastern Manitoba and northern Minnesota. Pp.
101-107 in Nero, R.W., Clark, R.J., Knapton, R.J.
and Hamre, R.H. (eds.). Biology and Conservation
of Northern Forest Owls: Symposium proceedings.
February 3-7, 1987. Winnipeg, Manitoba. USDA,
Forest Service, Gen. Tech. Report RM-142. 309p.
Duncan, J.R. 1992. Inuence of prey abundance and
snow cover on great gray owl breeding dispersal.
Ph.D. Thesis. Zoology Department, University of
Manitoba, Winnipeg, Manitoba. 127p.
Duncan, J.R. and Duncan, P.A. 1998. Northern
Hawk Owl (Surnia ulula). In: Poole, A. & Gill, F.
(eds.). The Birds of North America, No. 356. The
Birds of North America, Inc., Philadelphia, PA. 28p.
Duncan, J.R. 2003. Owls of the World. Key Porter
Books, Ltd. Toronto, ON. 310p.
Duncan, J.R., Swengel, S.R. and Swengel, A.B.
2009. Correlations of Northern Saw-whet Owl
(Aegolius acadicus) calling indices from surveys
in southern Wisconsin, USA with owl and small
mammal surveys in Manitoba, Canada, 1986-2006.
Pp. 489-496 in Johnson, D.J., Van Nieuwenhuyse,
D. and Duncan, J.R. (eds.). Owls – Ambassadors
for the Protection of Nature in their Changing
Landscapes. Proceedings of the fourth World
Owl Conference. 31 October – 4 November 2007,
Groningen, The Netherlands. Ardea 97: 395-649.
Duncan, J.R. 2013. The Complete Book of North
American Owls. Thunder Bay Press. San Diego,
CA. 192p.
Duncan, J.R. 2021. An Evaluation of 25 Years of
Volunteer Nocturnal Owl Surveys in Manitoba,
Canada. Pp xx-xx in Roque, I.M.F., Duncan, J.R.,
Johnson, D.H. and Van Nieuwenhuyse, D. (eds).
2017 World Owl Conference Proceedings. Evora,
Portugal. Airo 29:66-82.
Handbook of the Birds of the World and BirdLife
International (2019). Handbook of the Birds of the
World and BirdLife International digital checklist
of the birds of the world. Version 4. Available
at:http://datazone.birdlife.org/userles/le/Species/
Taxonomy/HBW-BirdLife_Checklist_v4_Dec19.
zip. Accessed 1 October 2020.
Holthausen, R., Czaplewski, R.L., DeLorenzo, D.,
Hayward, G., Kessler, W.B., Manley, P., McKelvey,
K.S., Powell, D.S., Ruggiero, L.F., Schwartz, M.K.,
Horne, B.V. and Vojta, C.D. 2005. Strategies for
monitoring terrestrial animals and habitats. General
Technical Report RMRS-GTR-161. Department
of Agriculture, Forest Service. Rocky Mountain
Research Station, Fort Collins. 34p.
Kassambara, A. 2018. Regression Model Validation -
Regression Model Accuracy Metrics: R-square, AIC,
BIC, Cp and more. http://www.sthda.com/english/
articles/38-regression-model-validation/158-
regression-model-accuracy-metrics-r-square-aic-
bic-cp-and-more/. Accessed 1 September 2020.
Konig, C. and Weick, F. 2009. Owls of the World,
2nd ed. Yale University Press. New Haven,
Connecticut. 528p.
Lehikoinen. A., Ranta, E., Pietiäinen, H., Byholm,
P., Saurola, P., Valkama, J., Huitu, O., Henttonen, H.
and Korpimäki, E. 2011. The impact of climate and
cyclic food abundance on the timing of breeding
and brood size in four boreal owl species. Oecologia
165: 349–355.
Manley, P.N., Zielinski, W.J., Schlesinger, M.D. and
Mori, S.R. 2004. Evaluation of a multiple-species
approach to monitoring species at the ecoregional
scale. Ecological Applications. 14: 296-310.
Ng, C.S., Duncan, J.R. and Koper, N. 2018. Who’s
“hooting”? Motivations and scientic attitudes of
Manitoban citizen science owl surveyors. Avian
Conservation and Ecology 13: Art. 9. https://doi.
org/10.5751/ACE-01265-130209.
Proença, V., Martin, L.J., Pereira, H.M., Fernandez,
M., McRae, L., Belnap, J., Böhmg, M., Brummitt,
N., García-Moreno, J., Gregory, R.D., Honrado,
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
1090 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
J.P., Jürgens, N., Opige, M., Schmeller, D.S., Tiago,
P. and van Swaay, C.A.M. 2017. Global biodiversity
monitoring: From data sources to essential biodiversity
variables. Biological Conservation 213: 256–263.
Quinn, G.P. and Keough, M.J. 2002. Experimental
Design and Data Analysis for Biologists. Cambridge
University Press. 537p.
Takats, D.L., Francis, C.M., Holroyd, G.L., Duncan,
J.R., Mazur, K.M., Canings, R.J., Harris, W. and Holt,
D. 2001. Guidelines for Nocturnal Owl Monitoring in
North America. Beaverhill Bird Observatory and Bird
Studies Canada, Edmonton, Alberta. 32p.
Taylor, P. (ed.). 2003. The Birds of Manitoba. Manitoba
Avian Research Committee, Manitoba Naturalists
Society. Winnipeg, Manitoba. 504p.
Witmer, G.W. 2005. Wildlife population monitoring:
some practical considerations. Wildlife Research 32:
259–263.
Proceedings of WOC, Pune, India, 2019 Duncan, J. R.
Long-eared Owl
Leroy Miller
| 1091
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Long-eared Owl Asio otus behaviour, prey provisioning and
diet during the nestling period using a camera trap in 2015 in
Manitoba, Canada.
James R. Duncan1*, Riki Kerbrat2
1 Discover Owls, Box 253, Balmoral, Manitoba, Canada R0C 0H0, 2 Clayton H. Riddell Faculty of
Environment, Earth, and Resources, Wallace Building, Dysart Road, University of Manitoba, Winnipeg,
Manitoba, Canada R3T 2M6
* Corresponding author: JDuncan@discoverowls.ca
Keywords: Long-eared Owl; Asio otus; nesting; prey provisioning; behaviour; diet; camera trap; Manitoba;
Canada. [Research Article for Proceedings for the 2019 World Owl Conference, Pune, India]
Citation:
Duncan James R. and Kerbrat Riki. Long-
eared Owl Asio otus behaviour, prey provisioning
and diet during the nestling period using a camera
trap in 2015 in Manitoba, Canada.
Ela Journal of Forestry and Wildlife
Vol.11 (4): 1091-1108
Date of Publication:
31 March 2022
ISSN 2319-4361
Proceedings of WOC, Pune, India, 2019
Abstract
A camera trap was set up at a Long-eared Owl Asio
otus nest in Manitoba, Canada in 2015. This was the
rst time this nocturnal species has been studied in
this manner. An analysis of 128,694 images collected
over 15 d during the nestling period revealed new
information on Long-eared Owl behaviour and
diet. Initially, the male delivered prey to the female
brooding nestlings and the female rarely left the nest.
The female often performed a raised wing display with
erect body plumage when receiving prey from the male
reminiscent, in part, of a male precopulatory display.
Two unsuccessful nestling predation attempts were
recorded. When the oldest nestling was 16-17 d old the
female spent less time on the nest which coincided with
the inferred onset of thermoregulation and observed
ability of nestlings to feed themselves. Prey deliveries
increased up to edging in response to increased
nestling energy requirements and to expedite edging
and maximizing nestling survival. Prey provisioning
at the nest then decreased as successive nestlings
edged and were fed directly outside the view of the
camera. The majority (92.5%) of 106 prey deliveries
were small mammals, especially voles (Cricetidae),
which was consistent with other diet studies derived
from pellet analysis. Only 48.1% of delivered prey were
identied to species. Nestling diet did not signicantly
change with time. Results from this study demonstrate
the potential and limitations of camera traps for future
research on the behaviour and diet of nesting Long-
eared Owls and other nocturnal species.
Duncan and Kerbrat
1092 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Introduction
Studying animal behaviour enhances our
understanding of a species’ biology including habitat
use, interactions with other species and feeding
ecology (Sutherland 1998; Rogers et al. 2005).
This understanding is important to develop species
management and conservation plans to mitigate
increased anthropogenic induced changes to the
environment and climate (Jordan 2005; Berger-Tal et
al. 2011; Caravaggi et al. 2017; Gaglio et al. 2017). Yet
for many species such basic information such as diet
is either non-existent, limited, or biased (Jordan 2005),
especially for nestlings where diet inuences growth,
development, and tness (Robinson et al. 2015).
Directly observing wildlife is time-consuming, can
inuence behaviour, produce observer bias and fatigue,
yield small sample sizes, and may be not be logistically
feasible (Rogers et al. 2005; Reif and Tornberg 2006;
Caravaggi et al. 2017). Indirect methods, such as
the identication of prey remains in nests and pellet
analysis, may also yield limited data and biased results,
especially for raptors (Marti 1974; Marti 1987; Jordan
2005; Rogers et al. 2005; García-Salgado et al. 2015;
Robinson et al. 2015).
To overcome the challenges noted above, researchers
have used simple and affordable non-intrusive camera
traps to continuously document nesting bird diet,
behaviour and predation through high quantity and
quality recorded video or images (Cutler and Swann
1999; Lewis et al. 2004; Cox et al. 2012; García-Salgado
et al. 2015; Caravaggi et al. 2017). Difcult to observe
and rare events or behaviours can be documented while
minimizing disturbance to species during sensitive
times, during inclement weather, and at night (Cutler
and Swann 1999; Lewis et al. 2004; Reif and Tornberg
2006; Caravaggi et al. 2017). Continuously recorded
video and/or image les can be viewed repeatedly by
many reducing observer bias, increase sample sizes
and more accurately estimate prey provisioning rates
and nestling diets (Cutler and Swann 1999; Lewis et
al. 2004; Rogers et al. 2005; Reif and Tornberg 2006;
García-Salgado et al. 2015). Nonetheless cameras may
affect bird behaviour, are subject to mechanical or
battery failure and may bias diet estimates, i.e., blurry
images of prey items, colourless images taken at night,
or images with prey hidden from view (Cutler and
Swann 1999; Lewis et al. 2004; García-Salgado et al.
2015; Caravaggi et al. 2017). Despite these limitations,
nest camera traps remain an effective tool for studying
raptor behaviour and diet (García-Salgado et al. 2015).
The Long-eared Owl is notoriously elusive during
the nesting season and, as with other nocturnal owl
species, unbiased detailed information on its breeding
biology is sparse (Glue 1977). It is found in a variety
of open forested habitat in North America, Eurasia, and
northern Africa between 30° and 65°N latitude (Cramp
1985; Marks et al. 1994). It is a widespread and regular
migratory breeding bird in Manitoba (Holland et al.
2003; Artuso 2019; Duncan 2020). While its diet and
some aspects of its breeding biology have been studied
using pellet analysis, radio-telemetry, and direct
observation it can be sensitive to disturbances making
detailed and unbiased research on its nesting behaviour
challenging (Craig et al. 1988, Marks et al. 1994). This
study is the rst time this species has been studied with
nest camera trap technology to document its breeding
behavior and feeding ecology. Because determining
the limitations of camera trap technology is important,
especially for species active at night, our assessment
examined the degree to which delivered prey could be
identied from black and white images taken at night.
Methods
Study Area
The nest site we monitored was located within a
small forest patch near Balmoral, Manitoba, Canada
(50.2406168°, -97.3027412°) at 254 m asl in a highly
fragmented agriculture-dominated landscape (Fig. 1).
Mean annual temperature ranges from 2-3°C with a
mean 16°C summer temperature and a mean -12.5°C
winter temperature. Mean annual precipitation range
is 450-700 mm (Environment Canada 2020). This area
is within the Lake Manitoba Plain Ecoregion in the
Prairies Ecozone (Wiken et al. 1996) and was originally
characterized as mixed-grass aspen parkland. It is now
a mix of dispersed remnant native and non-native
grasslands and forest patches within a dominant mosaic
of forage and cereal crops.
The nest was in a forested patch dominated by
trembling aspen (Populus tremuloides) mixed with
some burr oak (Quercus macrocarpa) and Manitoba
maple (Acer negundo). Introduced tree species included
scots pine (Pinus sylvestrus), white birch (Betula
papyrifera), jack pine (Pinus banksiana), tamarack
(Larix laricina), balsam r (Abies balsamea), northern-
white cedar (Thuja occidentalis), and white spruce
Proceedings of WOC, Pune, India, 2019 Duncan and Kerbrat
| 1093
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
(Picea glauca). The stick nest used by the owls was
5.5 m above ground in a white spruce and was built
by American crows (Corvus brachyrhynchos) prior to
2015. The nest tree was ca. 50 m north of an occupied
farmhouse (Fig. 1).
Observations of Long-eared Owls at the nest and
in the area helped interpret some behaviours obtained
from camera images and enabled the young to be
aged accurately (Table 1). This included setting up an
observation blind 10 m north of the nest from which to
infrequently observe the nest without ushing the owls.
Data Collection
On 2 June 2015, a Reconyx PC900 HyperFire
Professional High Output Covert IR Trail Camera
(www.reconyx.com) was mounted just above and
1 m from the centre of the nest using a steel angled
bracket attached to the nest tree (Fig. 2) to optimize
the collection of images to document behaviour and
identify prey. Steen et al. (2016) determined that <1 m
missed behaviours and >1 m made prey identication
more difcult. Cameras placed >1 m from nests were
triggered excessively by moving branches, wasted
limited image memory storage space, and depleted
batteries faster (R. Steen pers. comm.). The camera
was focused to 1 m by the manufacturer at purchase
to maintain the warranty. The camera settings used
were: motion sensor = on; sensitivity = high; pictures
per trigger = 3; picture interval = rapid re; quiet
period = no delay; and night mode = fast shutter.
Images included an image number, time (hh:mm:ss),
date (yyyy-mm-dd) and temperature (oC) stamp. Time
was set for the Central Daylight Saving Time Zone.
The 3 images/second capture/motion setting was a
compromise between likelihood of identifying prey to
species level and limits of memory and battery capacity
(R. Steen pers. comm.). Reducing the frequency of
camera memory and battery maintenance also reduced
disturbance to the nesting owls (Caravaggi et al. 2017).
The camera’s SD card and batteries were replaced three
times and 128,694 images were recorded from 08:08
hrs on 2 June 2015 to 15:13 hrs on 18 June 2015 (Table
2). The camera was removed after the last of the young
edged from the nest. Because Long-eared Owls have
an asynchronous hatch (Marks et al. 1994) and are
only moderately sexually size dimorphic (Earhart and
Johnson 1970) we assumed that the largest nestling,
readily identiable in images (Fig. 3), was also the
oldest.
Reviewing Camera Trap Images
All camera trap images collected were reviewed and
behaviours such as prey deliveries were transcribed
into a spreadsheet. The time, date and number of prey
delivery images were recorded to determine their
duration and peak periods. Changes in the number and
type of prey items delivered to the nest as nestlings
aged was assessed by clustering images into three time
periods (see below). The sex of the adult delivering
prey was noted to see if changed as nestlings aged.
While Long-eared Owls exhibit moderate reverse
sexual size and facial disk colour dimorphism (Earhart
and Johnson 1970; Marks et al. 1994; Holt 2016) these
differences or any other differentiating characteristics
(i.e., plumage pattern) were not distinguishable on
images. We presumed that only the female Long-eared
Owl incubated and brooded as only they develop a
brood patch (Marks et al. 1994). Therefore, we noted
the male as delivering prey to the nest when the female
was in the nest. The female was identied as delivering
prey only if she remained in the nest afterwards and if
she subsequently fed or preened the young. Otherwise,
the adult delivering the prey was noted as unknown sex.
Identication of Prey from Camera Trap Images
All prey items delivered to the nest were identied
to the nest taxonomic level possible from images by
consensus of both authors to estimate the contribution
of prey type to the nestlings’ diet. Small mammal
prey were identied using a combination of one or
more visible features including relative tail or hind leg
length, ear and/or eye size, and uniform vs contrasting
pelage shading (Fig. 3). There were few pellets
recovered from in and around the nest after the young
edged. No additional prey species were identied from
pellets to avoid double counting them. Bird prey were
likewise identied by bill, head, body, wing and tail
shape and size and plumage appearance to the extent
possible and by consultation with three bird experts.
Images with bird prey were also posted on iNaturalist
(https://www.inaturalist.org/) to see if yielded crowd-
sourced identication suggestions.
Low lighting, movement resulting in blurring or a
partially blocked eld of view resulted in prey delivery
or handling images (Fig. 3) from which prey were
assigned to one of six possible categories: unidentied,
unidentied mammal, unidentied rodent, unidentied
mouse, unidentied vole, or unidentied passerine. The
proportion of prey items that comprise the nestlings’
Proceedings of WOC, Pune, India, 2019 Duncan and Kerbrat
1094 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
diet was summarized in three equal time segments to
determine if proportion changed as nestlings aged as
follows: week one (June 02-07), week two (June 08-
13), and week three (June 14-18). A chi-square test
= 0.05, Statistical Function, Excel for Microsoft
365 updated April 2020) was used to determine if diet
composition changed over time. The oldest nestling
was 12 d old on 2 June and was 28 d old on 18 June.
The duration of time spent by the female in the nest
was also recorded to assess whether the level of parental
care changed as nestlings aged. The difference in the
date and time associated with the rst and last images
in which the female arrived, stayed, and then left the
nest was recorded as time spent in the nest. A selection
of other behaviours observed in recorded images were
noted and are described herein.
Results
In total, 128,694 camera images were collected
(Table 2). Although the camera was frequently checked,
a battery failure caused the camera to malfunction on 9
June at 01:04 hrs until it was reset on 11 June at 07:29
hrs resulting in an image and data gap when the oldest
nestling was between 19 and 21 d old. The oldest nestling
edged from the nest the following day at 22 d old.
From 2 to 17 June 2015 the average time of sunset
was 21:35:11 hrs (21:29:00 to 21:40:00 hrs) and sunrise
was 05:20:53 hrs (05:19:00 to 05:24:00 hrs) with 7 h and
46 min from sunset to sunrise. The times that the nest
camera switched from colour (day) to black and white
(night) mode varied depending on light levels affected
by cloud cover and vegetation density but averaged
21:03:24 hrs or 32 min before sunset. Likewise, the
times that the camera switched from black and white
to colour mode averaged 05:56:03 hrs or 35 min after
sunrise (Fig. 4).
Prey Deliveries
From the 128,694 camera images collected, 106
prey deliveries were documented. Most (89.6%) prey
deliveries occurred after sunset and before sunrise with
the two largest peaks about 22:00 and 01:00 hrs and a
smaller peak about 03:00 hrs (Fig. 4). The frequency
of prey deliveries generally increased as nestlings aged
until the camera’s batteries failed a few days prior to
when the oldest nestling edged (Fig. 5). From this
point on, the frequency of prey deliveries declined as
subsequent nestlings edged from the nest (Fig. 5) and
edged young received prey away from the nest.
For the 15-d period the camera recorded images the
mean prey delivery rate was 7.1 prey/d (n=106, se =
1.04, range 2-16) for an average of 1.8 prey/nestling/d.
However, as mentioned above, some prey deliveries
occurred away from the nest after the oldest nestling
edged. The mean prey delivery rate was 8.4 prey/d
(n = 76, se = 1.44, range 2-16) for the 9 days all four
nestlings were present for an average of 2.1 prey/
nestling/d.
Of the 106 prey deliveries, 31 were by the male, 20
by the female, and 55 by an unknown adult whose sex
could not be determined. The male initially delivered
most of the prey (Fig. 6) while the brooding female
rarely left the nest (Fig. 7). During this period, the
female accepted and tore up prey to feed the nestlings.
When the oldest nestling was 15 d old the female
began to spend less time at the nest, and this trend
decreased further as the nestlings aged (Fig. 7). At this
time, the arrival of the female became increasingly
indistinguishable from that of the male and more prey
deliveries were noted as being by an “unknown” adult
(Fig. 6). Toward the end of the nestling period, nearly
all prey delivered to the nest was by an unknown
adult as the nestlings no longer required assistance in
consuming prey and duration of time spent at the nest
by the female neared zero (Figs. 6, 7).
Prey Identication and Diet
There was limited consensus or specicity among
the three bird experts we asked to independently
identify bird prey from select images (i.e., Fig. 4);
one suggested all ve bird prey were recently edged
Chipping Sparrows Spizella passerina while two
thought that some of them were either Savannah
Sparrows Passerculus sandwichensis or “unidentied
sparrow” (Passerellidae). The habitat around the owl
nest supported healthy populations of Savannah, Clay-
coloured Spizella pallida and Chipping Sparrows
in spring and summer (J. Duncan unpubl. data). The
same images of bird prey posted on iNaturalist (https://
www.inaturalist.org/) on 9 June 2020 failed to produce
additional identications from website bird expert
moderators over a three-week period. Therefore,
we considered all bird prey to be unidentied small
passerines, likely sparrows.
Of the 106 prey items delivered to the nest, 98 were
small mammals, 6 were sparrows (Passerellidae) or
small passerine birds, and 2 were unidentiable to class
(Table 3). In total, 48.1% of prey items were identied
Proceedings of WOC, Pune, India, 2019 Duncan and Kerbrat
| 1095
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
to one of ve species: Deer Mouse Peromyscus
maniculatus, Meadow Vole Microtus pennsylvanicus,
Arctic Shrew Sorex arcticus, Red-backed Vole Myodes
gapperi and Northern Short-tailed Shrew Blarina
brevicauda (Table 3). The remaining prey were
identied as a chipmunk (Tamius spp.), three jumping
mice (Zapodinae) and one of six other groupings
(Table 3).
Voles (Cricetidae) made up most (65.1%) of the
nestling diet; of the 69 voles, 28 were unidentied, 26
were Meadow Voles, and 15 were Red-backed Voles
(Table 3). Unidentied rodents (Rodentia) made up
11.3% of the nestlings’ diet followed by mice (Muridae)
comprised 10.4% of nestling diet including seven Deer
Mice, three jumping mice, and one unidentied mouse
(Table 3). The remainder of nestling diet was composed
of sparrows (Passerellidae), unidentied mammals,
shrews (Soricidae), and unidentiable prey (Table 3).
Voles consistently made up the greatest proportion
of delivered prey as nestlings aged (Fig. 8, Table 3).
Changes in composition of nestling diet by week did
not change signicantly as nestlings aged (x2 = 3.946, p
= 0.684, Table 3).
Other Notable Behaviours and Events
When the male delivered prey to the female brooding
young his arrival was immediately preceded by the
female erecting her body plumage, especially those
on her back, while raising her wings up such that the
wing tips almost touched over her back, and tilting her
body forward (Fig. 9). This represents a new ritualized
display for this species as the closest similar behaviour
has only been described as a precopulatory male calling
posture (Mikkola 1983). This display has also been
seen repeatedly only in spring when feeding a captive
human-imprinted female Long-eared Owl (J. Duncan
unpublished data).
On several occasions the camera captured nest
sanitation behaviour by the female where she inspected
the nest and used her bill to pick up and either eat or
dispose uneaten prey or pellets over the side of the nest
(Fig. 10). This was consistent with observations made
by others (Murphy 1992; Scott 1997). Images of the
female removing saclike nestling feces, a behaviour
reported by Murphy (1992), were not recorded.
We considered an adult Cooper’s Hawk Accipiter
cooperii landing near the nest on 4 June 2015 to be a
predation attempt, as this species is known to kill and eat
Long-eared Owls (Bloom 1994). Immediately after the
hawk’s arrival the female owl exhibited an exaggerated
inverted wing display with erected body feathers while
swaying left to right after which the Cooper’s Hawk
dispersed (Fig. 11) after 1 m and 8 s.
Another predation attempt involving a Short-tailed
Weasel Mustela erminea was documented whereby
the weasel appeared to attempt to take a nestling at
06:27 hrs on 12 June 2015 (Fig. 12). This mammal
regularly climbs trees to the nests of even large birds
for prey (King 1989). This event was concurrently
and independently photographed by Skip Shand (Fig.
12) which documented that both the male and female
participated in the successful defence of the nestlings.
This incident appears to be the rst documented
predation attempt by a weasel on a Long-eared Owl
nest. A week earlier, at 05:08 hrs on 5 June 2015, the
brooding female peered intently at some stimulus
below the nest, but the object of her attention was not
visible (Fig. 12).
Discussion
This effort is the rst camera trap study of Long-
eared Owl nesting behaviour and analysis of temporal
changes in prey provisioning rates and nestling diet
composition. Prey deliveries for this nocturnal species
(Marks et al. 1994) were expected to largely occur
at night (Fig. 4). The peak prey delivery times we
observed (Fig. 4) were generally consistent with, but
started an hour earlier than, those observed by Craig et
al. (1988) where nesting Long-eared Owls were most
active between 22:00 to 05:00 hrs. The Long-eared
Owls we observed perhaps had to start hunting earlier
as they had less time (7 h and 45 min) to do so between
sunset and sunrise than those studied in Idaho (9 h and
45 min; Craig et al. 1988).
Craig et al. (1988) also reported lulls in prey
deliveries from 20:00 to 22:00 hrs and from 05:00 to
06:00 hrs whereas the rst lull we detected was an
hour later at 23:00 hrs (Fig. 4). The occasional day-
time prey deliveries we recorded (Fig. 4) were likely
opportunistic, energy efcient prey captures close to
the nest. The prey deliveries just prior to sunset (Fig.
4) were consistent with Bayldon’s (1978) observation
that Long-eared Owls begin hunting before sunset,
especially during brood-rearing.
The observed increase in prey provisioning as
nestlings aged (Fig. 5) correlated with an inferred
increase in their energy needs (Steen et al. 2012) and
Proceedings of WOC, Pune, India, 2019 Duncan and Kerbrat
1096 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
represents a means by which adults increase nestling
edging success and survival (Ricklefs 1968). The
provisioning rate we obtained (8.4 prey/d) was higher
than that reported for two nests in Utah, USA, (3.5-
4 prey/d) by Delong (1982) but the number of young
present in these nests was not given. We nor Delong
(1982) measured prey availability in our respective
study areas, which likely inuence prey delivery rates.
As expected, the subsequent progressive decline in
prey provisioning (Fig. 5) coincided with the successive
edging of nestlings as adults increasingly fed young
owls outside of the nest and the camera detection range;
edged Long-eared Owls are fed by the female until the
young are 6.5-8 weeks old and 2-3 weeks afterwards
by the male (Marks et al. 1994). The observed daily
uctuations in prey delivery rates (Fig. 5) may relate
to weather as other have reported that rain and wind
reduced Long-eared Owl hunting success to only one
capture for every six or more attempts (Cramp 1985).
Most raptors have asymmetric parental roles in which
the female incubates, broods and feeds the nestlings,
while the male hunts for prey and is usually assisted by the
female for the latter part of the nestling period (Sonerud
et al. 2014). This study found similar results as the male
was the main provider of prey during the beginning of
the nestling stage (Fig. 6). As the nestlings aged toward
the latter part of the nestling period the female also
delivered prey and ultimately spent little time at the nest
(Figs. 6, 7), however, the extent to which she captured
prey was unknown. One of many hypotheses for the
evolution of reversed size dimorphism in owls is that the
larger female is better able to defend eggs or nestlings
from predators (Mueller 1986). The female was observed
playing a primary and immediate role in defending the
nestlings on at least two separate occasions, therefore it
is assumed that she also remained close to the young in
the latter part of the nestling and post-edgling periods
and had fewer opportunities than the male to forage for
prey. This conjecture is supported by Ulmschneider’s
(1990) observation that radio-marked males made 2.5
times more food deliveries than did similarly marked
and more protective female Long-eared Owls during the
post-edgling period.
The decrease in the duration of time on the nest and
increase in prey delivery by the female was similar
to, but more delayed than, that inferred by Craig et
al. (1988) from the movements of two radio-marked
breeding female Long-eared Owls. This decrease was
inuenced by two non-mutually exclusive factors, the
ontogeny of thermoregulation and self-feeding ability
of the nestlings. While away from the nest, the female
presumably perched and hunted opportunistically
nearby. Female absence from the nest varied with the
age of the nestlings: when the oldest nestling was 13-14
d old, the female was away from the nest <10% of the
time; when it was 15-16 d old, she was away for 26-
29% of the time; and when it was 17-18 d old she was
away for 76-90% of the time (Fig. 7). Barn Owl Tyto
alba nestlings can maintain their own body temperature
when 15-20 d old and the absence of the female on
the nest before this date has adverse consequences
for the young (Durant et al. 2004). Similarly, nestling
Eastern Screech Owls Megascops asio achieved
thermoregulation when 14-16 d old (Lohrer 1985).
Older and larger nestlings able to thermoregulate can
transfer heat to younger and smaller nestlings (Durant
et al. 2004). Therefore, we infer that Long-eared Owl
nestlings can thermoregulate when they are 15-18 d
old based on the female’s behaviour (Fig. 7) which
is much later than what Singer (1988) observed in
Idaho (11–12 d old) based on behavioural observations
of young. Singer (1988) did not observe brood or
group homeothermy at an earlier age to individual
homeothermy and considered that their results were
inuenced by relatively mild ambient temperatures.
As the female spent less time on the nest, the
behaviours that distinguished her from the male
(i.e., remaining on the nest and feeding young) also
decreased, thereby increasing the number of prey
deliveries denoted as “unknown adult” (Fig. 6).
Capturing and marking a breeding adult could mitigate
this limitation (Craig et al, 1988; Ulmschneider 1990).
Identifying delivered prey to class (mammal or
bird) from camera trap images was successful but more
specic identications were often challenging (Fig. 3,
Table 3). Challenges included harsh shadows or low
lighting, movement resulting in blurred images, or the
prey item was partially or wholly blocked by an owl
(Fig. 3). In many cases, prey were identied from one
or more of a series of images taken after prey delivery,
such as feeding events. In addition, almost all deliveries
occurred at night while the camera was in black and
white mode, which prevented the use of colour in prey
identication.
The diet of the Long-eared Owl nestlings in this
study was dominated by small mammals, especially
Proceedings of WOC, Pune, India, 2019 Duncan and Kerbrat
| 1097
Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
voles, while birds made up a relatively small portion,
and parallels other studies in North America, Eurasia
and Africa (Marti 1976; Klippel and Parmalee 1982;
Marks 1984; Marks et al. 1994; Birrer 2009; Mori
and Bertolino 2015). Voles may have been the most
common or available prey in our study area, as raptors
often select prey relative to their availability (Boshoff
et al. 1994). The composition of nestling Long-eared
Owl diet did not signicantly change with age (Table 3)
suggesting that the adults captured and delivered prey
opportunistically and that relative prey availability did
not change over the same period. This interpretation is
consistent with other research which state that Long-
eared Owls feed opportunistically on available prey
rather than being a specialized vole predator (Marks
1984; Craig et al. 1988; Marks et al. 1994). In contrast,
one study (Ulmschneider 1990) noted a gradual increase
in pocket mice (Perognathus spp.) in Long-eared Owl
nestling diets but also concluded that this increase was
related to an increase in this prey type’s availability.
Birds may contribute to a limited portion of nestling
diet as they may be less active at night or parents may
avoid feeding them to nestlings as they may take longer
for nestlings to handle and eat (Sonerud et al. 2014).
Bird prey may also be more dangerous for nestlings
to eat. Duncan and Nero (1998) documented that a
recently edged Eastern Screech-owl choked on and
died while eating an intact red-breasted nuthatch Sitta
canadensis; the wedged bill of the nuthatch pierced the
upper pallet and skull of the owlet.
Camera traps are an innovative way to study animal
behaviour (Steen 2009; Trolliet et al. 2014; Caravaggi
et al. 2017). Their use, and other new sources of
images and data, such as internet forums and social
media (Lourenço 2019), are yielding new data on
wild species. The camera trap as used in this study
was useful in obtaining a better understanding of key
aspects of the breeding ecology of the Long-eared
Owls such as provisioning rates, diet, nest sanitation,
predation attempts, and one new ritualized display
behaviour. These results would be logistically difcult
or impossible to obtain by direct observation or pellet
analysis for this nocturnal owl species. The female
stayed away from the nest for nearly an hour after the
initial camera set-up during a time when she rarely
left the nest. However, we concluded that overall, the
camera trap was non-intrusive and all four nestlings
edged from the nest at approximately 22 days old.
The camera trap was an efcient way to capture
images for subsequent study. Images can be reviewed
by multiple researchers or experts, enhancing the
accuracy of the results, and archived for later access.
We recommend that the same methods be used at other
Long-eared Owl nest sites to determine the generality
of our ndings (Johnson 2002). Camera trap research
should use standard research and reporting methods
(Meek et al. 2014; Robinson and Prostor 2017). Camera
traps are subject to malfunctions, such as the battery
failure we experienced, creating data gaps. Additional
limitations included difculties in the identication of
prey to species. However, we conclude that the benets
of using camara traps outweighed their limitations and
that they compliment more traditional methods to study
animal behaviour. Camera traps provide detailed results
which positively impact conservation management and
research efforts.
Acknowledgements
Thanks to Patricia Duncan for help setting up the
camera trap and monitoring the nest and Daryll Hedman
for making the camera support arm. The Lady Gray’l
Fund, The Winnipeg Foundation and Manitoba Nature
provided funding and administrative support for this
study. This research was also supported by donations
from school children, schools, and other organizations
to Discover Owls (www.DiscoverOwls.ca). Skip Shand
kindly let us use his pictures. Thanks to Christian Artuso,
Peter Taylor and Spencer Sealy for examining bird
prey images. Thanks to Nicola Koper, Rick Baydack,
Dean Berezanski, Brooks Duncan and Patricia Duncan
for insightful discussions and/or improvements to the
manuscript. Thanks also to the organizing committee,
volunteers, and the ELA Foundation that made the 2019
World Owl Conference possible.
Proceedings of WOC, Pune, India, 2019 Duncan and Kerbrat
1098 | Ela Journal of Forestry and Wildlife | www.elafoundation.org | www.mahaforest.nic.in | Vol. 11 | Issue 1 | January - March 2022
Table 1. Chronology of observed and estimated* events at a Long-eared Owl Asio otus nest in Manitoba,
Canada in summer 2015.
Date Event or Observation
29 March Long-eared Owls detected in adjacent areas during nocturnal owl surveys
05 April Male Long-eared Owl courtship call heard 130 m W of nest site
22-28 April Estimated range of dates eggs that hatched were laid
09 May Six eggs observed in nest 5.5 m above ground in a 11 m high White Spruce Picea glauca tree
14-16 May Observation blind erected 10 m N of nest site over a 3-d period
17 May Spring snow blizzard with an overnight low of -4 oC, 60 kph NNE winds and 12 cm snow
cover likely froze two of six eggs positioned on north side of the stick nest
21 May Estimated rst egg hatch
25 May Three eggs and three nestlings observed
27 May Estimated date fourth egg hatches
30 May Two eggs and four nestlings observed
02 June Two eggs and four nestlings observed and camera mounted
04 June Cooper’s Hawk Accipiter cooperii predation attempt at 10:51 hrs
11-12 June Oldest nestling edged at estimated 22 d old
12 June Short-tailed Weasel Mustela erminea predation attempt at 06:27 hrs
14 June Oldest nestling (edged 11-12 June) observed 22 m ESE of nest tree
15 June Third oldest nestling edged at estimated 23 d old to branch 1.3 m above nest
17-18 June Last two of four nestlings edged, the youngest at estimated 21-22 d old
* Estimated based on observations and 2 d egg laying intervals and 28 d incubation (Marks et al. 1994).
Table 2. Summary of camera trap recorded image les at a Long-eared Owl Asio otus nest in Manitoba,
Canada in summer 2015.
Date and Time
Checked
Date/Time Range of Recorded mages Number of Days Number of Images