british columbia birds - wordpress.combritish columbia birds is published annually ....
TRANSCRIPT
Journal of the British Columbia Field OrnithologistsVolume 23 • 2013
British Columbia
Birds
Sapsucker Distribution and Density - Gyug et al.
Volume 23, 2013 ISSN 1183-3521
British Columbia BirdsJournal of the British Columbia Field Ornithologists
PO Box 45507, Westside RPO, Vancouver, B.C., V6S 2N5
Editor: Arthur M. Martell
251-3399 Crown Isle Drive, Courtenay, BC V9N 9X7
e-mail: [email protected]
Immediate Past Editor: John B. Sprague, Salt Spring Island, B.C.
Production Editor: Neil K. Dawe, Parksville, B.C.
Editorial Board: Neil F. Bourne, Nanaimo, B.C. R.A. (Andy) Buhler, Vernon, B.C.
Rob Butler, New Westminster, B.C. Mark Phinney, Dawson Creek, B.C.
Mary J. Taitt, Delta, B.C.
B.C. Field Ornithologists Directors 2012–2013:
George Clulow (President) Larry Cowan (Vice-President)
Mary Taitt (Recording Secretary) Mike Fung (Treasurer)
Jude Grass (Past President) Art Martell
June Ryder Wayne Weber
Membership dues:
Individual memberships or library subscriptions: $30.00/year;
Junior memberships (age under 18): $20.00/year;
U.S. and International memberships or subscriptions $35.00 US/year.
Memberships are for the calendar year.
To join or for further information, contact Larry Cowan, Membership Co-ordinator.
B.C.Field Ornithologists, PO Box 45507, Westside RPO, Vancouver, B.C., V6S 2N5
British Columbia Field Ornithologists is a non-profit society; membership is open to anybody interested in
the study and enjoyment of wild birds in B.C. Our objectives include fostering co-operation between amateur
and professional ornithologists, promoting co-operative bird surveys and research projects, and supporting
conservation organizations in their efforts to preserve birds and their habitats.
British Columbia Birds is published annually. Members/subscribers also receive a quarterly newsletter, B.C. Birding.
Papers and notes published in British Columbia Birds have been reviewed by a member of the editorial board,
a qualified outside reviewer, and the editor. However, views expressed in any paper, note, or book review are
those of the author(s) and do not necessarily reflect views of the editor, editorial board, or BCFO board.
Papers and notes in this volume and previous volumes can be viewed online at www.bcfo.ca.
Unless copyright restrictions are indicated, any paper, note or review (or excerpts from them) may be repro-
duced in another publication provided that both the author(s) and British Columbia Birds are credited fully.
Front cover: Barrow’s Goldeneye (Bucephala islandica) at Stanley Park, Vancouver, B.C., 13 December 2009. Based on recent
surveys, their numbers have shown a decrease between years in waters off the Stanley Park seawall (see page 41). Photograph
by Robyn Worcester.
British Columbia Birds
1
Volume 23, 2013
Contents
Bird distribution and climate change in British Columbia ..................................................................... 2
FRED L. BUNNELL, ARNOLD MOY, MICHAEL I. PRESTON, RALPH W. WELLS
Effects of fire on bird abundance in Okanagan Mountain Provincial Park, British Columbia .......... 16
LES W. GYUG
One size does not fit all: differential responses of waterfowl species to impacts of climate change in
central British Columbia ........................................................................................................................ 27
FRED L. BUNNELL, RALPH W. WELLS, BRUCE HARRISON, AND ANDRE BREAULT
Bird observations by Dr. J.E.H. Kelso in the West Kootenay area of British Columbia, 1913–1932 .......... 39
BILL MERILEES
Changes in the abundance of wintering waterbirds along the shoreline of Stanley Park, Vancouver,
British Columbia, between 2001/2002 and 2010/2011. ......................................................................... 41
ROBYN WORCESTER
Acknowledgements & editor’s comments .................................................................... 44
Photo essay ................................................................................................ Inside back cover
LAURE W. NEISH
British Columbia Birds
2
Volume 23, 2013
Introduction
Many species respond to environmental changes, in-
cluding climate, by shifting their geographic ranges (War-
ren et al. 2001, Walther et al. 2002, Root et al. 2003, Bunnell
et al. 2008). Climate envelopes are composites of prevail-
ing meteorological conditions within an area. A growing
literature uses climate envelopes of current ranges to pre-
dict how geographic ranges of birds will change in re-
sponse to projected changes in climate (e.g. Doswald et
al. 2009, Marini et al. 2009, Willis et al. 2009, Jiménez-
Valverde et al. 2011). Conservation efforts attempt to an-
ticipate changes in distribution (e.g. Bunnell et al. 2011b),
but there has been relatively little empirical testing of such
projections. Green et al. (2008) used retrodiction to evalu-
ate model accuracy.
We examined changes in 10 climate variables meas-
ured within the geographic ranges of 32 bird species in
British Columbia between two decades: the 1960s and the
1990s. That permitted estimates of empirical shifts of range
in response to documented changes in climate and pro-
vided a test of the accuracy of climate in predicting range
shifts. Our objectives were to: 1) describe the degree to
which species’ ranges shifted between the 1960s and
1990s, 2) compare climate variables, especially mean spring
temperature, measured within species’ ranges in the 1960s
and 1990s, 3) test predictions of the general model of avian
response to climate proposed by Bunnell et al. (2005, 2008)
and 4) illustrate limitations of climate in predicting changes
in the geographic distribution of species.
Data and methods
Bird distribution
Data on bird distributions were obtained from the
Biodiversity Centre for Wildlife Studies (BCWS), Victoria,
British Columbia. British Columbia spans about 12° latitude
(48° 30´ to 60° N) and 20° longitude (120° to 140° W). Each 1°
of latitude and 2° of longitude represents a 1:250,000 NTS
(National Topographic Survey) cell that can be further di-
vided into 16 1:50,000 NTS cells (0.25° latitude by 0.5° longi-
tude); there are 1,171 such cells in British Columbia, of which
16 are primarily ocean. Sample units used in our analyses
were the 1:50,000 NTS cells sampled during the breeding
seasons of both decades (1960s and 1990s) for a given spe-
cies. Numbers of cells meeting that criterion ranged from 236
for Lewis’s Woodpecker (Melanerpes lewis) to 441 for Pa-
cific Loon (Gavia pacifica) and Brown-headed Cowbird
(Molothrus ater) (Appendix I).
The digital database includes data from journal and gov-
ernment publications, theses and consultants’ reports, but
many were reported opportunistically by volunteer natural-
Bird distribution and climate change in British Columbia
Fred L. Bunnell1, 3, Arnold Moy1, Michael I. Preston2, Ralph W. Wells1
1 Faculty of Forestry, University of British Columbia, 2424 Main Mall, Vancouver, B.C. V6T 1Z4; email: [email protected] Stantec, 2042 Mills Road Unit 11, Sidney BC V8L 5X43 Corresponding author
Abstract: We evaluated predictions of birds’ response to climate for 32 bird species in British Columbia between the
1960s and 1990s. Of the 32 species tested, 20 showed expansion north when tested between 51° to 60° N, but expansion
was significant for only seven. Four species remained south of 51° N in both decades. Ten spring and summer climate
variables were evaluated; mean spring temperature was most informative. Temperature variables were highly correlated
with other climate variables. Natural history attributes of species had major impacts on both range expansion and ability to
project future ranges. Attributes having strong influences on range expansion included migratory behaviour, food type,
breeding habitat and latitudinal distribution of the range prior to projection. As predicted, waterbirds showed the greatest
tendency to expand ranges. Winter climate variables appear to have influenced those species showing southward shifts in
range, apparently through shorter duration of ice cover. Estimated climate of the 1960s distribution was an inconsistent
predictor of bird distribution during the 1990s.
Keywords: bird distribution, British Columbia, climate envelopes
Birds and climate change - Bunnell et al.
British Columbia Birds
3
Volume 23, 2013
ists. BCWS assesses the accuracy of all records submitted
and discards suspect observations. We reduced potential
influences of variable opportunity and effort on reporting
rates among cells and between decades by restricting the
data base to a single record for any location on any date.
That is, one record was considered sufficient to confirm the
presence of a species in a given area on a given date even
when multiple records existed. This restriction avoided bias
and pseudo-replication resulting from intensive or long-term
study in any particular area.
Tests of spatial occupancy (range) examined changes
between the two decades in numbers and distribution of
occupied 1:50,000 NTS cells (e.g. Figure 1). All sampled cells
are shown in Figure 1; only cells sampled in both decades
for a particular species were included in analyses. Our pri-
mary interest was northward expansion, but statistical tests
were two-tailed to account for possible southward shifts.
Shifts in distribution were evaluated at 0.25° latitudinal in-
crements using the non-parametric Kolmogorov-Smirnov test
to accommodate a variety of possible distributions. For each
species, we constrained testing of distribution to the north-
ernmost observation of either decade, and used counts of
occupied cells in each increment of 0.25° latitude for each
decade. The northernmost latitude at which species were
recorded as present occurred during the 1990s for all species
except Red-throated Loon (Gavia stellata). An example of a
significant northward shift is shown for American Wigeon
(Anas americana) in Figure 1; two species showed a south-
ward shift.
Bird distribution and climate variables
ClimateWNA software (http://www.genetics.forestry.ubc.ca/
cfcg/ClimateWNA/ClimateWNA.html) was used to estimate
values for 10 climate variables. ClimateWNA extracts and
downscales historical monthly and seasonal data for user-
identified locations using PRISM (parameter-elevation re-
gressions on independent slopes modeling; Daly et al. 2002).
We extracted climate for the lowest elevation of each occu-
pied NTS cell.
Ten climate variables were examined: mean spring tem-
perature, minimum spring temperature, maximum spring tem-
perature, average summer temperature, minimum summer tem-
perature, maximum summer temperature, total spring precipi-
tation, total summer precipitation, degree days >5 °C and a
summer heat-moisture index (SHM). Spring was defined as
March through May and summer as June through August.
The SHM index is generated by ClimateWNA using the equa-
tion: SHM = MWMT / (MSP/1000), where MWMT is the
Mean Warmest Month Temperature (°C) and MSP is Mean
Summer Precipitation (mm). SHM is a derived variable, used
as a proxy for direct measures of humidity or evaporation
and transpiration that often are unavailable (Tuhkanen 1980).
We evaluated climate variables within NTS cells occu-
pied and unoccupied during the 1960s and 1990s for 32 spe-
cies. For any species, only cells sampled in both decades
were included in the analyses, so total cells sampled were
the same for each decade for that species. Cells analyzed for
a given species differed as a function of the distribution of
that species. In four species, all or most of the population
occurred south of 51° N in both decades. We evaluated range
expansion over two latitudinal ranges for all species: 48° 30´
to 60° N and 51° to 60° N. For a few restricted species, expan-
sion over the latitudes 48° 30´ to their northernmost occu-
pied latitude was tested.
We examined which climate variables exhibited the great-
est statistical differences between occupied and unoccu-
pied cells and correlations among climate variables (using
decadal means extracted from ClimateWNA for each of the
1155 terrestrial cells). Based on those analyses (see Results),
we emphasized responses to mean spring temperature (MST).
We tested expectations derived from the assumption that
MST was a dominating influence on range expansion. Sta-
Figure 1. National
Topographic Survey
cells of British Colum-
bia occupied by the
American Wigeon: a)
in the 1960s, b) in the
1990s. Gray cells
were occupied; hollow
cells were sampled
but found no Ameri-
can Wigeon; areas
without cells were not
sampled.
a) b)
Birds and climate change - Bunnell et al.
British Columbia Birds
4
Volume 23, 2013
tistical analyses of differences in MST and other climate vari-
ables were restricted to simple, paired t-tests (unequal vari-
ance) of values in occupied and unoccupied cells and be-
tween decades.
Southward movement of two waterbird species prompted
us to include winter variables (average annual snowfall and
mean winter temperature during December through Febru-
ary) to index ice conditions on lakes or wetlands.
Tests of predictions
The general model of avian response to climate of Bunnell
et al. (2005, 2008) offered a priori predictions of probable
responses based on migratory pattern, food habits, habitat
and body size. Five classes of migratory pattern were recog-
nized: resident, partial (e.g. winter at sea and move inland to
breed), short distance (<1000 km), long distance (1000 to
4500 km) and very long distance (>4500 km). Body size was
expected to influence primarily reproductive measures. The
four broad factors are nested within each other, so a sample
of only 32 species limits testing. Three of the predictions
offered by Bunnell et al. (2005, 2008) could be addressed:
• Resident, partial and short distance migrants
should show greater range expansion northward than
long and very long distance migrants. The former three
migratory classes have greater familiarity with regional
climate so can respond more readily.
• Species foraging in water should show greater
range expansion northward than those foraging on
terrestrial insects. Earlier timing of ‘ice-off’ and greater
drying of more southern wetlands (e.g. Bunnell et al.
2011b) should encourage northward expansion of the
former group. Terrestrial insects also are expected to
be available earlier with warming, but we expected
more rapid response to ‘ice off’. Late winter and early
spring observations of waterbirds inland suggest
frequent monitoring of the timing of ‘ice off’.
• Species breeding in lakes and wetlands should
show greater range expansion northward than those
breeding in upland areas (rationale as for the
preceding prediction).
None of these predictions is tidily discrete. Many partial
migrants both forage in water and breed in aquatic habitats.
A sample size of 32 species does not permit estimation of
dominating factors, nor should dominance be expected given
the natural covariance. The sample does permit extraction of
broad patterns and was sufficient to reject one prediction.
Projected distributions
Mean spring temperature was projected by ECHAM-5
(see Roeckner et al. 2003), under the A2 scenario of the Inter-
governmental Panel on Climate Change (IPCC). ECHAM has
proven to provide good fits to empirical data in northern
regions (Kattsov and Walsh 2000, Wohlfahrt 2010). We found
it accurately predicted the trends in measured water depths
of wetlands in British Columbia (Bunnell et al. 2011a). No
IPCC scenario includes concerted efforts at reducing emis-
sions. The IPCC treats all scenarios as plausible. Climate
variables for an individual NTS cell were downscaled, yield-
ing approximate empirical climate for that cell. Temperature
thresholds for potential occupancy were the lowest mean
value of all occupied cells, whether this occurred in the 1960s
or the 1990s. These values were commonly similar; for exam-
ple, 6.11 and 6.13 °C for Yellow Warbler (Dendroica petechia)
in the 1960s and 1990s, respectively.
Results
Climate data
Among the 10 climate variables, mean spring tempera-
ture (MST) showed the greatest significant difference (paired
t-tests) between means of cells occupied in the 1960s and
1990s, followed by degree days >5 °C. The latter is corre-
lated with minimum, maximum and average spring or summer
temperatures, but most strongly with mean spring tempera-
ture (Table 1). Precipitation variables showed little distinc-
tion between occupied and unoccupied cells.
tav_sp1 tmx_sp tmn_sp tav_sm tmx_sm tmn_sm ppt_sp ppt_sm ddgt5
tav_sp1 1tmx_sp 0.937 1tmn_sp 0.963 0.809 1tav_sm 0.768 0.872 0.622 1tmx_sm 0.534 0.771 0.306 0.894 1tmn_sm 0.826 0.733 0.827 0.829 0.490 1ppt_sp 0.488 0.262 0.620 0.075 -0.218 0.418 1ppt_sm 0.044 -0.118 0.165 -0.210 -0.379 0.065 0.778 1ddgt5 0.923 0.919 0.847 0.924 0.704 0.917 0.354 -0.054 1shm 0.163 0.376 -0.014 0.480 0.647 0.126 -0.554 -0.712 0.3121 tav_sp = mean spring temperature, tmx_sp = maximum spring temperature, tmn_sp = minimum spring temperature, tav_sm = average summer temperature, tmx_sm = maximum summer temperature, tmn_sm = minimum summer temperature, ppt_sp = total spring precipitation, ppt_sm = total summer precipitation, ddgt5 = degree days >5 oC, shm = summer heat-moisture index.
Table 1. Correlation coefficients during the 1990s among the 10 climate variables evaluated. Variables as downscaled
from 1155 1:50,000 NTS cells by Climate WNA. Critical values for n = 1155 are about 0.06 for p < 0.05, and 0.08 for p < 0.01.
Birds and climate change - Bunnell et al.
British Columbia Birds
5
Volume 23, 2013
�
Because the sample size is large (n = 1155 NTS cells),
correlations among climate variables were generally signifi-
cant (Table 1). During the 1990s, correlations among tem-
perature variables were particularly strong and >0.75 for all
but maximum summer temperature, which is unlikely to influ-
ence breeding range occupancy. The strong inter-correla-
tion among variables and their relative ability to discriminate
between occupied and unoccupied cells encouraged us to
rely primarily on mean spring temperature for further testing.
Changes in species’ distribution between the 1960s and 1990s
Focus on northward expansion encouraged us to focus
on cells north of 51° N. The number of cells north of 51° N
newly occupied in the 1990s ranged from five for Band-tailed
Pigeon (Patagioenas fasciata) to 74 for Common Loon
(Gavia immer) (Table 2). The number of newly occupied
cells north of 51° N was not a revealing index of range expan-
sion northward because widely distributed species often
showed new occupancy in cells both north and south of 51°
N. Total cells newly occupied at any latitude ranged from 30
for Band-tailed Pigeon to 132 for Song Sparrow (Melospiza
melodica). Of the 32 species tested, 20 showed expansion
north when tested from 51° to 60° N, but expansion was
significant for only seven (p < 0.05; Table 2).
Table 2 reports tests over 51° to 60° N. When tested over
the entire range of latitude (48° 30´ to 60° N) only four of the
32 species showed different patterns of response; Spotted
Towhee (Pipilo maculatus) p <0.009, Wood Duck (Aix
sponsa) p < 0.067, Least Flycatcher (Empidonax minimus) p
< 0.03 and Red-throated Loon (Gavia stellata) p < 0.011
showed southward movement.
For species largely restricted to areas south of 51° N in
the 1960s, we also tested over the latitudinal range extend-
ing from 48° 30´ N to the northernmost extent of their range
in the 1960s. None of the four species (Band-tailed Pigeon,
Spotted Towhee, Swainson’s Hawk [Buteo swainsoni], Wood
Table 2. Kolmogorov-Smirnov tests of range expansion by 32 bird species in British Columbia between the
1960s and 1990s. N indicates apparent shift northward; indicates no change in distribution (p > 0.40). nd =
the species was not reported north of 51° N in the 1960s.
Birds and climate change - Bunnell et al.
British Columbia Birds
6
Volume 23, 2013
Duck) showed significant (p < 0.05) northward expansion
within the restricted latitudes.
Temperatures of areas occupied in the 1960s and 1990s
Table 3 summarizes downscaled temperature values as
means for occupied and unoccupied NTS cells in the 1960s
and 1990s. We expected 1990s MSTs to be higher than 1960s
MSTs for all cells. Cells unoccupied in the 1990s were con-
sistently warmer in the 1990s than in the 1960s (p < 0.01).
Cells newly occupied in the 1990s were significantly warmer
in the 1990s than the 1960s for all but one species, Northern
Shoveler (Anas clypeata). Cells occupied in the 1960s showed
consistently higher mean MSTs in the 1990s than the 1960s,
significantly so in 26 of 32 species (Table 3). Generally, the
difference in MST between cells occupied and unoccupied
in the 1960s declined with increasing proportion of occupied
cells north of 51° in the 1960s (r2 = 0.61, p < 0.01). Across all
species the difference in estimated MST between cells occu-
pied and unoccupied in the 1960s was a poor predictor of
either the proportion or number of cells newly occupied in
the 1990s (r2 < 0.03).
Specific patterns are expected if MST is a dominant fac-
tor in changes in geographic range. The 1960s MST of cells
newly occupied in the 1990s should be lower than 1960s
MSTs in cells occupied in the 1960s. In 28 of 32 species the
expectation was met, significantly so for 20 (Table 4). That
is, the large majority of species tested entered cells during
the 1990s that in the 1960s were on average cooler than
cells occupied in the 1960s. For two species (Least Fly-
catcher, Lincoln’s Sparrow [Melospiza lincolnii]) the 1960s
MSTs of cells newly occupied in the 1990s were higher
than in cells occupied in the 1960s. These two species had
the highest proportions of occupied cells north of 51° N in
the 1960s, 93% and 73%, respectively. Red-throated Loon
and Red-necked Grebe showed a tendency to move south
between decades and there was no difference in 1960s
MSTs between cells newly occupied in the 1990s and those
occupied in the 1960s.
Newly occupied in Unoccupied inOccupied in 1960s 1990s 1990s
Species 1960s oC 1990s oC 1960s oC 1990s oC 1960s oC 1990s oC
ResidentBand-tailed Pigeon 7.60 8.62 b 7.16 8.20 b 5.30 6.41 b
Fox Sparrow 5.84 6.86 5.04 6.22 b 4.80 6.04 b
Mourning Dove 7.04 8.14 b 5.36 6.57 b 5.09 6.23 b
Red-throated Loon 7.40 8.44 b 7.39 8.42 b 5.42 6.52 b
Song Sparrow 6.20 7.30 b 5.31 6.46 b 3.67 5.03 b
Spotted Towhee 8.10 9.14 b 7.03 8.10 b 5.69 6.77 b
Varied Thrush 5.95 7.06 a 5.69 6.84 b 4.45 5.71 b
Partial migrantsAmerican Wigeon 6.64 7.76 b 5.26 6.44 b 4.59 5.81 b
Brewer’s Blackbird 6.37 7.49 b 5.43 6.64 b 4.70 5.91 b
Common Loon 5.43 6.59 b 4.87 6.09 b 3.89 5.22 b
Gadwall 7.33 8.40 b 6.17 7.30 b 5.20 6.33 b
Horned Grebe 7.03 8.12 b 5.50 6.66 b 4.63 5.84 b
Lesser Scaup 7.47 8.55 b 5.65 6.83 b 4.81 6.00 b
Northern Pintail 6.92 8.02 b 5.51 6.68 b 4.64 5.85 b
Northern Shoveler 6.49 7.62 5.53 6.78 5.17 6.33 b
Pacific Loon 6.56 7.67 a 6.26 7.36 b 4.48 5.73 b
Red-necked Grebe 5.52 6.76 5.51 6.67 b 4.96 6.15 b
Surf Scoter 5.33 6.51 5.12 6.35 b 4.32 5.58 b
Western Grebe 7.16 8.24 b 5.46 6.62 b 5.05 6.18 b
White-winged Scoter 6.66 7.73 b 5.42 6.59 b 4.69 5.90 b
Wood Duck 7.07 8.14 a 5.95 7.09 b 4.75 5.96 b
Short-distance migrantsLincoln’s Sparrow 4.26 5.58 a 4.67 5.89 b 5.03 6.24 b
Lewis’s Woodpecker 7.47 8.58 b 6.09 7.21 b 5.41 6.51 b
Western Meadowlark 6.96 8.05 b 5.52 6.71 b 4.99 6.15 b
Long-distance migrantsBrown-headed Cowbird 6.15 7.29 b 5.37 6.55 b 4.25 5.51 b
Cinnamon Teal 6.86 7.96 a 5.90 7.06 b 4.65 5.87 b
Swainson’s Thrush 5.74 6.90 b 5.01 6.23 b 3.85 5.15 b
Yellow Warbler 6.11 7.24 b 4.88 6.13 b 4.33 5.61 b
Very-long distance migrantsCommon Nighthawk 6.30 7.45 b 5.15 6.35 b 4.23 5.48 b
Least Flycatcher 3.83 5.11 a 4.44 5.78 b 5.36 6.51 b Swainson’s Hawk 7.09 8.24 6.25 7.41 a 4.88 6.09 b
Wilson’s Phalarope 7.00 8.13 5.49 6.70 b 5.01 6.16 ba, b Indicate significant differences between time periods within paired columns (p < 0.05) a (p < 0.01) b
Table 3. Average of mean spring temperature in NTS cells occupied by selected bird species in British Colum-
bia in the 1960s, newly occupied in the 1990s and unoccupied in the 1990s.
Birds and climate change - Bunnell et al.
British Columbia Birds
7
Volume 23, 2013
�
Similarly, if MST is the dominant factor in changes in
geographic range, we expect the 1990s MSTs of newly occu-
pied cells in the 1990s to be at least as high as 1960s MSTs in
cells occupied in the 1960s. The trend was not pronounced
(Table 4). The general trend was to be warmer than 1960s
MSTs of occupied cells (17 of 32 species), but was signifi-
cant for only seven species. For five species there were no
discernable differences in MST, and for 10 species the 1990s
MSTs of newly occupied cells were lower than 1960s MSTs
in cells occupied in the 1960s (never significantly). Cells cooler
than those occupied in the 1960s were frequently occupied
in the 1990s (15 of 32 species).
For cells unoccupied in the 1990s, we expected 1960s
and 1990s MSTs to be lower than 1960s values of MSTs in
occupied cells. That was true for 30 of 32 species for 1960s
MSTs and 27 of 32 species for 1990s MSTs (Table 4). For
some species (shaded in Table 4), cells that apparently met
the mean MST for occupancy were not occupied.
The tendency to abandon cells by the 1990s that experi-
enced greater snowfall and lower temperatures was general
and significant for all partial migrants feeding in water (e.g.
145.5 versus 97.4 mm of snow, p < 0.05; Table 5). Only Red-
throated Loon showed significant southward movement.
Cells occupied by Red-throated Loon in the 1960s and aban-
doned by the 1990s were those with significantly more snow-
fall and colder temperatures in the 1990s (p < 0.01). Red-
Select cells New cells MSTs of cells
with formerly > MSTs unoccupied in 1990s
Species cooler MSTsa of 1960s cellsb relative to 1960s cellsc
Resident
Band-tailed Pigeon � � * � ** / � **Fox Sparrow � � � / �Mourning Dove � ** � � ** / � *Red-throated Loon ↔ � ** � ** / � **Song Sparrow � ** � � ** / � **Spotted Towhee � ** ↔ � ** / � **Varied Thrush � � * � ** / �Partial migrantsAmerican Wigeon � ** � � ** / � *Brewer’s Blackbird � * � � ** / �Common Loon � �* � ** / �Gadwall � ** ↔ � ** / � **Horned Grebe � ** � � ** / � **Lesser Scaup � ** � � ** / � **Northern Pintail � ** � � ** / � **Northern Shoveler � � � * / �Red-necked Grebe ↔ � * � / �Surf Scoter � � �* / �Pacific Loon � � � ** / �Western Grebe � ** � � ** / � **White-winged Scoter � ** � � ** / � *Wood Duck � ** ↔ � ** / � **Short-distance migrantsLincoln’s Sparrow � � ** � / � **Lewis’s Woodpecker � ** � � ** / � **Western Meadowlark � ** � � ** / � *Long-distance migrantsBrown-headed Cowbird � * � � ** / � *Cinnamon Teal � * � � ** / � **Swainson’s Thrush � * � � ** / �Yellow Warbler �** ↔ � ** / �Very-long distance migrantsCommon Nighthawk � ** ↔ � ** / � *Least Flycatcher � � ** � ** / � **Swainson’s Hawk � � � ** / �Wilson’s Phalarope � * � � ** / �
a Compares 1960s MST values of cells newly occupied in 1990s with 1960s MSTs of cells occupied in 1960s. � indicates lower 1960s
MSTs in newly occupied cells; � indicates higher values in the 1960s.b Compares 1990s MSTs of newly occupied cells to 1960s MSTs of cells occupied in the 1960s. � indicates higher MSTs in 1990s; � indicates lower MSTs in 1990s.
c Compares 1960s (first arrow) and 1990s (second arrow) MSTs of unoccupied cells in 1990s with 1960s MSTs of cells occupied in
1960s. � indicates higher MSTs in 1990s; � indicates lower MSTs in 1990s.
Table 4. Apparent role of MST in creating changes in geographical distribution. indicates difference
in mean MST of < 0.05 °C. Average values of MST tested are from Table 2. * = p < 0.05; ** = p < 0.01.
Shaded cells indicate species that, on average, did not enter cells in the 1990s that were warm enough
to host them based on the 1960s range.
Birds and climate change - Bunnell et al.
British Columbia Birds
8
Volume 23, 2013
necked Grebe (Podiceps grisegena) showed a non-signifi-
cant tendency to move south. In the 1960s, the grebe occu-
pied cells that were significantly colder than other partial
migrants (p < 0.01). Cells occupied in the 1960s and aban-
doned by the 1990s were those with greater snowfall (134.0
mm versus 107.2 mm, p < 0.10; Table 5), but warmer mean
winter temperatures (not significant).
Testing a priori predictions
Bunnell et al. (2005, 2008) predicted greater likelihood of
range expansion among resident species, partial migrants
and short-distance migrants simply because those species
are more intimately exposed to changing temperatures on
the breeding range. The prediction was wrong; generally,
partial and longer distance migrants showed a greater ten-
dency to range expansion (19 of 22) than did resident or
short-distance migrants (two of 10) (Table 2).
The general model invoked food habits as a predictor of
reproductive success because success may be influenced
by asynchrony of arrival dates and food availability. We
also expected that piscivores and species foraging on aquatic
invertebrates would frequently show range expansions north-
ward, because arrival so soon after the ice leaves implies
frequent monitoring of inland waterbodies. Among the par-
tial migrants, 12 of the 14 species winter at sea, then move to
inland lakes and wetlands to breed. Of the 13 species forag-
ing on fish or aquatic invertebrates, 11 expanded their range
northward. The two species that did not (Red-necked Grebe,
Red-throated Loon) tended to shift their ranges south. Red-
throated Loon was classified as resident because it breeds
on lakes close to the ocean and may forage at sea in both
winter and summer. We evaluated whether the shift south-
ward was in response to climate variables that determine ice
cover on lakes and wetlands.
Both snowfall and winter temperature contribute to the
timing of ‘ice off’. In cells occupied in the 1960s, significant
declines (p < 0.001) in snowfall were evident between the
1960s and 1990s for all species foraging in water. There is
strong evidence that cells abandoned by the 1990s were
those with greater snowfall (Table 5). There is no evidence
that the effect was stronger in the two species moving south
over the period. The pattern differs for mean winter tempera-
ture. Across ‘other species’, the general warming is evident
between the 1960s and 1990s for cells occupied in the 1960s
(p < 0.001). Temperatures of cells occupied in the 1960s and
abandoned by the 1990s were insignificantly warmer than
cells in which occupancy persisted (1990s values for ‘other
species’ of -1.12 versus -1.03 °C; Table 5). For species forag-
ing in water, 1990s snowfall was a better predictor of cell
occupancy than was mean winter temperature. Generally,
winter climate variables of the 1960s were inadequate to pre-
dict future occupancy of species foraging in water.
The sample of 32 bird species was not large enough to
evaluate the broad classes nested within each other, but
differences among primary breeding habitats were apparent.
Aquatic food habits and aquatic breeding are conflated, but
reveal a strong influence of aquatic versus upland habitats.
The 16 species breeding on lakes and wetlands showed much
greater likelihood of range expansion (13 showed a tendency
to move north, seven significantly so, p < 0.05) than did the
16 upland breeding species (four showed a tendency to move
north, none significantly) (Table 2). Among species breed-
ing primarily in shorter vegetation (natural grasslands,
shrublands, agricultural areas) two of five expanded their
ranges northward, five of 11 forest-dwelling species expanded
northward and 13 of 16 species nesting on lakes or wetlands
expanded their ranges northward.
Projecting further range expansion
We illustrate four patterns of response to MST illustrated
by groups of species (Figure 2) and the potentially mislead-
ing consequences of projecting simple climate envelopes
(Figures 3 and 4). Horned Grebe (Podiceps auritus) shows
the pattern typical of partial migrants that frequently showed
range expansion northward (Figure 2a). Cells show a gradual
increase in MST from unoccupied in the 1990s to occupied
in the 1990s to occupied in the 1960s, with 1990s tempera-
tures consistently higher. Cells occupied in the 1990s showed
lower MSTs than cells occupied in the 1960s. Cells not occu-
pied in the 1990s showed markedly lower temperatures than
Occupied in Occupied inOccupied in 1960s 1960s & not in 1990s 1960s & 1990s
Species 1960s 1990s 1960s 1990s 1960s 1990s
Snowfall (mm)Other species a 142.7 114.8 181.0 145.5 113.5 97.4RNGR 155.0 123.8 170.7 134.0 129.4 107.2RTLO 145.2 106.5 183.7 129.5 77.8 66.3Winter Temperature ( oC)Other species a -1.84 -1.06 -2.58 -1.12 -1.54 -1.03RNGR -4.77 -3.81 -4.22 -3.26 -5.67 -4.70RTLO 2.67 3.35 1.92 2.72 3.99 4.46a Other partial migrants are Anas Americana, Gavia immer, Anas strepera, Podiceps auritus, Aythya affinis,
Anas acuta, Anas clypeata, Gavia pacifica, Melanitta perspicillata, Aechmophorus occidentalis, and Aix sponsa.
Table 5. Snowfall and mean winter temperature for two periods (1960s and 1990s) in
cells occupied by partial migrants foraging in aquatic habitats during breeding.
Birds and climate change - Bunnell et al.
British Columbia Birds
9
Volume 23, 2013
cells occupied in the 1960s. Other examples of this response
are American Wigeon, Common Loon, Gadwall (Anas
strepera) and Northern Pintail (Anas acuta).
Lewis’s Woodpecker illustrates the pattern for a species
whose northward expansion is somehow constrained (Fig-
ure 2b). The central 50% of the range of MST values in occu-
pied cells is compressed and there is very little difference in
1990s MSTs of cells occupied in the 1960s and the 1990s.
Species showing a similar pattern include Band-tailed Pi-
geon, Spotted Towhee, Swainson’s Hawk and, to a lesser
extent, Western Meadowlark (Sturnella neglecta). None of
these species expanded their ranges northward.
Two species shifted their distribution somewhat south-
ward (Red-throated Loon, Red-necked Grebe) or strongly
augmented their representation south of 51° N (Northern
Shoveler, Yellow Warbler). MSTs for cells occupied in the
1990s differ little from those occupied in the 1960s, but the
range in the 1990s is greater than in the 1960s (Figure 2c).
Like other groups, estimated MST is an inadequate predic-
tor of the area occupied in the 1990s for these species. The
fourth group illustrated includes species for which there was
a marked increase in temperature of cells occupied in the
1990s above that of cells occupied in the 1960s. Least Fly-
catcher is illustrated (Figure 2d); only one other species
showed this pattern (Lincoln’s Sparrow). The pattern is simi-
lar to Figure 2c, but appears to result for different reasons.
These two species showed the largest proportion of occu-
pied cells north of 51° N in the 1960s where warming is greater.
Not illustrated in Figure 2 are species such as Fox
Sparrow (Passerella iliaca) and Yellow Warbler, for which
MSTs of cells occupied in the 1990s spanned 10 °C or
more and were significantly wider than envelopes of
ranges in the 1960s.
Figure 3 illustrates potential forms of projection for
Lewis’s Woodpecker, a constrained species. All projections
employ the ECHAM-5 global circulation model under the A2
scenario. Between the 1960s and 1990s, the spatial extent of
Lewis’s Woodpecker range extended northward, but not sig-
nificantly so (Table 2). The effort-corrected relative density
within that range increased significantly northward (Bunnell
et al. 2008). Mean spring temperature of the 36 cells occu-
pied in the 1960’s was 7.47 °C. That temperature does not
predict occupied range well in the 1960s (Figure 3a) or the
1990s (Figure 3b). Instead, the woodpecker’s range approxi-
�� Horned Grebe b) Lewis’s Woodpecker
c) Red-throated Loon d) Least Flycatcher
�� �� �� � � � � � � � �� �� ��
�����������
�����
��������� ��
�������
����������� ��
�������
����������������� ��������
�� �� �� � � � � � � � �� �� ��
�����������
�����
��������� ��
�����
����������� ��
�������
����������������� ��������
�� �� �� � � � � � � � �� �� ��
�����������
������
��������� ��
������
����������� ��
�������
����������������� ��������
�� �� �� � � � � � � � �� �� ��
�����������
������
��������� ��
�����
����������� ��
�����
����������������� ��������
Figure 2. Box charts
of mean spring tem-
peratures (MST) for
occupied and unoc-
cupied cells for se-
lected bird species in
British Columbia. The
box encompasses
the range in which
50% of the observa-
tions fell; edges of the
box are at the first and
third quartiles. The
circle within the box
represents the mean;
the vertical line the
median. X = outliers
(values outside 1.5
times the inter-quar-
tile range). For each
pair of bars, the upper
bar is the 1960s MST
and the lower bar the
1990s MST.
Birds and climate change - Bunnell et al.
British Columbia Birds
10
Volume 23, 2013
mates the distribution of ponderosa pine (Pinus ponderosa) and
western larch (Larix occidentalis) in the province (Figure 3c).
We illustrate Yellow Warbler as an example of species al-
ready well-distributed through the province in the 1960s (Fig-
ure 4). Between the 1960s and 1990s, the spatial extent of the
species’ range had extended northward (p < 0.065; Table 2), and
effort-corrected relative density had shifted significantly north-
ward (Bunnell et al. 2008). Mean MST of the 97 cells occupied
in the 1960’s was 6.11 °C. In the 1960s, Yellow Warbler fre-
quently occupied cells below the mean MST of occupied cells
(Figure 4a). Yellow Warbler currently arrives on the northern
part of its range in British Columbia in May and June. Minimum
May and June temperatures of cells occupied during the 1960s
or 1990s did no better at predicting range occupancy in the
1990s than did the 1960s mean MST (the May minimum for the
1990s is illustrated in Figure 4b).
Discussion
Climate variables
We focused on mean spring temperature because it
was most discriminatory and climate variables generally
were highly inter-correlated. The relative discrimination
between occupied and unoccupied cells by the 10 climate
variables was as expected from natural history features.
Spring temperature is most likely to determine forage avail-
ability through timing of insect emergence for insectivores
or ice melt for waterbirds. Heavy spring precipitation can
increase mortality of young birds, but only after the com-
mitment to breeding. The strong correlations among cli-
mate variables (Table 1) imply little insight is gained from
combining them in projections.
Figure 3. Cells deemed favourable and cells occupied by the Lewis’s Woodpecker in British Columbia. a) 1960s: cells
occupied (black) and unoccupied cells meeting the average MST of cells occupied in the 1960s (gray). b) 1990s: cells
occupied (black) and unoccupied cells meeting the average MST of occupied cells in the 1960s (gray). c) 1990s: cells
occupied relative to the boundaries of the Ponderosa Pine and Interior Douglas-fir Biogeoclimatic zones sensu Meidinger
and Pojar 1991(gray).
Figure 4. Cells deemed
favourable and cells oc-
cupied by the Yellow
Warbler in British Co-
lumbia. a) 1960s: cells
occupied (black) and
unoccupied cells meet-
ing the average MST of
cells occupied in the
1960s (gray). b) 1990s:
cells occupied (black)
and unoccupied cells
meeting the minimum
mean May temperature
of cells occupied in the
1990s (gray).
�� �� ��
a) b)
Birds and climate change - Bunnell et al.
British Columbia Birds
11
Volume 23, 2013
Changes in distribution
Potential confounding was reduced by restricting obser-
vations for a cell to a simple binary state: occupied or unoc-
cupied. By restricting analyses to cells sampled in both the
1960s and 1990s, much of the potential confounding due to
greater sampling effort in the north during the 1990s was
eliminated. We cannot, however, correct for the likelihood
that more effort in a given cell is more likely to document a
particular species in that cell. Nor can we distinguish be-
tween increases in bird density increasing the likelihood of
detection from range expansion. Land use practices, particu-
larly forestry and agriculture, created disproportionately more
early-seral and grassland habitat in northern regions in the
1990s as compared to the 1960s. Few of the species analyzed
exploit such habitat, and those that do (e.g. Common
Nighthawk [Chordeiles minor], Mourning Dove [Zenaida
macroura]) showed no significant expansion northward.
The migratory groups least likely to show northward
range extension were resident species and short-distance
migrants; only two of 10 species showed northward expan-
sion; none significantly so (Table 2). Four of these species
occupy ranges largely constrained to southern portions of
the province (Band-tailed Pigeon, Spotted Towhee, Lewis’s
Woodpecker, Western Meadowlark). Southern British Co-
lumbia represents the northern extent of their range (the
meadowlark extends farther north in Alberta), but there is
little commonality in habitat. Whatever the constraint, it does
not appear to be directly influenced by climate. For some of
these species, increased availability of supplemental food in
the south, as from bird feeders and ornamental fruiting trees,
may be slowing potential northward range expansion. Of the
resident and short-distance migrants we analyzed, seven of
10 are known to regularly visit feeders and ornamental fruit-
ing trees, compared to two of 22 species in the other three
migratory groups.
Partial migrants do not necessarily migrate short dis-
tances. Some winter along the coast of southern British Co-
lumbia and then move well north into Alaska and Canadian
Territories. At least some partial migrants using aquatic habi-
tats are reported inland during late winter and early spring
before many lakes and wetlands have thawed, so must make
periodic forays inland, apparently assessing conditions there.
Among partial migrants and other species typically migrat-
ing over 1000 km, 19 of 22 species showed northward shifts
in distribution. One of those not extending northward,
Swainson’s Hawk, was largely constrained south of 51° N in
British Columbia. For several of these species, the range
already extended north of 60° N in the 1960s (e.g. Brown-
headed Cowbird, Common Loon, Lesser Scaup [Aythya
affinis], White-winged Scoter [Melanitta fusca]).
We did not anticipate the southward shift in distribution
of Red-throated Loon (significant) or Red-necked Grebe (non-
significant). The two waterbirds showing a tendency to shift
their range southward between the two decades may have
been evading factors that delay ‘ice off’ – deeper snowfalls,
colder temperatures or both. The significant southward shift
in range by Red-throated Loon would have yielded earlier
‘ice off’ and a longer ice-free period. The non-significant
shift by Red-necked Grebe may have yielded the same, pro-
vided reduced snow depth permitted radiation onto lake ice
earlier. Other waterbirds, however, expanded their range north-
ward and achieved the same end (Table 5).
Our inability to effectively account for differences in ef-
fort between decades is unlikely to be a major cause for our
failure to detect strong shifts in distribution. Working with a
far more amenable data set, 92 tree species in more than
43,000 inventory plots across 31 states, Zhu et al. (2012:1042)
found “…no consistent evidence that population spread is
greatest in areas where climate has changed most”. Their
results showed highly variable responses: about 59% of the
tree species showed range contraction at both northern and
southern boundaries, about 21% showed a northward shift
and 16% a southern shift. Only 4% showed expansion at
both range limits. There are compelling reasons why we ex-
pect trees of a forest to migrate more like a herd of cats than
a herd of buffalo (Bunnell and Kremsater 2012). We expect
some of those reasons to be equally well expressed among
birds. Too many things happen at once during climate change
for a unidirectional response to be likely.
Temperatures between decades
Because occupied cells entered analysis only if they were
surveyed in both the 1960s and 1990s, the same cells are
compared in both decades for any species or group of spe-
cies. Across all cells, average MST increased 1.25 °C be-
tween the 1960s and 1990s. Findings of Table 3 affirm the
general warming trend; MSTs of cells occupied in the 1960s
were significantly warmer in the 1990s than in the 1960s in 26
of 32 tests. Non-significant differences occurred among spe-
cies whose range in British Columbia already spanned con-
siderable distance north in the 1960s, so had greater area to
select across, thus yielding higher variability. For example,
Fox Sparrow (Passerella iliaca) extended north of 58°, North-
ern Shoveler north of 56°, Red-necked Grebe to 59°, Wil-
son’s Phalarope (Phalaropus tricolor) to 57° and Surf Scoter
(Melanitta perspicillata) to 60°. The exception was
Swainson’s Hawk, whose known range was south of 51° N
in the 1960s. Findings suggest that in the 1960s, some spe-
cies already had selected cells within their range with higher
MSTs. For only one species, Northern Shoveler, were cells
newly occupied in the 1990s not significantly warmer in the
1990s than in the 1960s. Of the 24 cells newly occupied by
the Northern Shoveler in the 1990s, 13 were south of 51° N
and 11 were north of 51° N.
The tendency to occupy cells in the 1990s with MSTs at
least as high as those of cells occupied in the 1960s was
general but not strongly expressed (Table 4), indicating that
mean temperatures of occupied cells in the 1960s did not
consistently reflect the mean temperature for future cell oc-
cupancy. For four species (Least Flycatcher, Lincoln’s Spar-
Birds and climate change - Bunnell et al.
British Columbia Birds
12
Volume 23, 2013
row, Red-necked Grebe, Red-throated Loon; Tables 3 and 4),
mean 1960s MSTs of cells newly occupied in the 1990s were
as warm, or warmer, than MSTs of cells occupied in the 1960s.
These cells may have been occupied in the 1970s or 1980s.
All four species showed expansion of their range in the south
that would encompass cells already warm in the 1960s. Least
Flycatcher showed a strong tendency to expand its range
northward (Table 2) but an equally strong expansion of its
range in the south; it was reported from two cells south of
51° N in the 1960s, but occupied 25 cells south of 51° N in the
1990s. A southward shift in range was significant only for
Red-throated Loon when tested over 48° 30´ to 60° N (Table
2). Generally, the selection of newly occupied cells was not
strongly driven by MST or, presumably, by any of the other
highly correlated variables of Table 1.
The only apparent commonality in the natural history of
the five species that did not enter cells warm enough to host
them based on the 1960s range (Table 4) is that ranges of all
of them extend well north of 60° N and three often seek
subalpine or montane habitat (Fox Sparrow, Red-necked
Grebe, Lincoln’s Sparrow). These five species did not oc-
cupy many of the cells warm enough to host them, implying
other factors were acting. One factor could be that climate
variables were derived for the lowest elevation of each cell,
exposing the difficulties in projecting range occupancy in
rugged terrain.
Combined, the results indicate that: 1) during the 1990s
some species did not occupy cells warm enough to host
them based on the 1960s’ range, 2) other species occupied
cells cooler than the 1960s’ range, 3) the reliability of the
temperature envelope decreased with the latitudinal extent
of the range in the 1960s, indicating that degree of range
expansion was a function of existing range (differences in
MST between cells occupied and unoccupied in the 1960s
declined significantly with increasing proportion of occu-
pied cells north of 51°) and 4) differences in MST between
cells occupied and unoccupied in the 1960s was a poor
predictor of range expansion (r2 < 0.03). In short, knowl-
edge of climate of the 1960s’ range was insufficient to pre-
dict future distribution.
There is an important caveat. Where long-term weather
stations are sparsely distributed and terrain is rugged,
downscaling climate variables accurately becomes increas-
ingly error prone. The problem has been exposed in northern
British Columbia (Flanagan et al. 2005, Mbogga et al. 2010,
Wotton et al. 2010), and likely holds in any region with few
long-term weather stations and rugged terrain. Unexpected
patterns appeared more frequently where the northern por-
tion of the province that has experienced more rapid warm-
ing comprised a larger portion of cells used in analyses.
Tests of predictions
Contrary to predictions of the general model, partial and
longer distance migrants showed a greater tendency to range
expansion (19 of 22), than did resident or short-distance
migrants (two of 10). Combining resident and short-dis-
tance migrants is appropriate. Short-distance migrants were
defined as species migrating no more than 1000 km in the
1960s. By the 1990s, some of these species showed an in-
creasing tendency to overwinter within British Columbia
(Bunnell et al. 2008). Apparently, proximity to local climate
did not encourage range expansion northward; the Red-
throated Loon (a resident) shifted its range southward. A
higher portion of resident and short-distance migrants ap-
pears constrained by features other than climate than was
true of other migratory classes (four of the five species
largely constrained to southern portions of the province in
the 1990s are resident or short-distance migrants). It is pos-
sible that long-distance migrants are more likely to show
range expansion northward because they have traits (e.g.
dietary requirements) that are particularly sensitive to tem-
perature—hence they migrate.
Predictions based on foraging habits were upheld: 11
of 13 species foraging on fish or aquatic invertebrates ex-
panded their range, while only six of 13 species foraging on
terrestrial insects expanded theirs. Lack of selection for
warmer cells during spring was most strongly expressed
among partial migrants, particularly those breeding in
wetlands and lakes (Table 3). That occurred despite the
fact that six of seven statistically significant expansions of
ranges northward were among these same partial migrants
(Table 2). The response may represent ashkui, the Innu
name for sites of open water in river and lake systems within
frozen spring landscapes (e.g. Baillie et al. 2004). Red-
throated Loon shifted its range southward. It is largely resi-
dent inland, so could assess localized conditions, such as
ashkui, relatively easily.
There are at least two competing hypotheses for why
waterbirds show a greater tendency for range expansion: 1)
inherently amenable to ‘wandering’ thus expansion or 2) more
scattered distribution permitting more opportunities for ex-
pansion. We expect that both are acting. About 70% of the
province is forested and only 7% is covered by lakes and
wetlands. The greater response of aquatic foragers may re-
flect newly available, formerly limited, habitat provided by
earlier ice-off and longer ice-free periods, whereas forests
are accessible year round. Table 5 indicates that among
aquatic foragers, portions of range occupied in the 1960s,
but not in the 1990s, generally experienced greater snowfall
and lower temperatures. The difference between forest and
aquatic foragers illustrates the inconsistent applicability of
simple temperature envelopes across species. The point is
affirmed by differences across breeding habitats. Fifteen of
18 species breeding in lakes and wetlands expanded their
range while only four of 14 breeding on land expanded theirs.
The simple predictions of the general model were rejected
in one instance and affirmed in two, but presence of domi-
nating influences could not be exposed with a sample of
only 32 species. Results indicate that species and groups of
species respond differently to changing climate.
Birds and climate change - Bunnell et al.
British Columbia Birds
13
Volume 23, 2013
Projecting distributions
Among species apparently constrained by features other
than climate, the nature of the constraint is apparent for
Lewis’s Woodpecker. The species is fire-adapted. Important
aspects of breeding habitat include an open canopy, a brushy
understory, dead or downed woody material, available
perches and abundant insects – all of which are encouraged
by fire. The three principal habitats are open ponderosa pine
forest, open riparian woodland dominated by cottonwood
and logged or burned pine (Tobalske 1997). Historically, the
species has been restricted to the more fire-prone ecosys-
tems of the province, specifically the Coastal Douglas-fir,
Interior Douglas-fir zone and Ponderosa Pine zone sensu
Meidinger and Pojar (1991). Its range in British Columbia
approximates that of ponderosa pine and western larch, both
of which are fire-adapated. Habitat in the Coastal Douglas-
fir zone in the southwest of the province is no longer favour-
able for breeding, likely as a consequence of fire suppres-
sion. The species shows no significant spatial expansion of
its range northward (Table 2), but the effort-corrected rela-
tive density has shifted northward within its range (Bunnell
et al. 2008). The Lewis’s Woodpecker appears largely re-
stricted in its northward extension by northern boundaries
of the Ponderosa Pine and Interior Douglas-fir zones (Figure
3c). Ponderosa pine in these zones is predicted to extend
northward (Hamman and Wang 2006) but that is not evident
to date, nor likely unless mountain pine beetle numbers abate
(Bunnell and Kremsater 2012).
Swainson’s Hawk and Western Meadowlark are now
known from northeastern British Columbia and may have
been present but unreported in the 1990s. It is not clear what
constrains Band-tailed Pigeon, Spotted Towhee or Wood
Duck to southern regions of the province.
Projection for species expanding or augmenting south-
ern portions of their range is similarly hindered. Breeding
season climate variables representing temperature were non-
predictive. Nor was there any apparent pattern with precipi-
tation variables; mean spring precipitation was positively
correlated with mean spring temperature (Table 1).
The pattern for Yellow Warbler (Figure 4) was not unique.
Our tests of habitat affinity in the northeast, southeast and
coastal areas of the province found it was a habitat generalist
with statistical associations at the variant level of the
Biogeoclimatic Ecological Classification system (Meidinger
and Pojar 1991) and specific forest types varying across re-
gions (Bunnell 2010). That is borne out more generally (re-
view of Lowther et al. 1999). In any forest type it shows
preference for moist hardwood thickets, particularly those
dominated by Salix sp., but these are occupied in both up-
land and riparian areas. We found little relation between pat-
terns of range occupancy for any of the 10 individual climate
variables tested and additional ones reflecting lower tem-
perature thresholds (e.g. Figure 4). We did not test combina-
tions of climate variables because these were highly inter-
correlated (Table 1).
Conclusions
Most data analyzed were opportunistically collected by
naturalists and illustrate the value of field ornithology. Analy-
ses expose two major challenges to projecting distributions
of species using climate envelopes. The first results from
unreliable downscaling of climate variables where long-term
weather stations are sparse and/or the terrain is rugged. The
second is that natural history features (e.g. migratory pat-
tern, foraging preferences) appear to influence the nature of
response to climate change. Tests of 32 species are insuffi-
cient to reveal all patterns, but are sufficient to indicate that
aggregating species’ responses is likely to be misleading.
We conclude that the climate variable showing the greatest
discrimination between ranges occupied in the 1960s and
1990s (mean spring temperature) was an inconsistent index
of future range occupancy. The strong inter-correlation
among climate variables suggests that combinations of cli-
mate variables are unlikely to be more predictive. There may
be little general ability to accurately predict future ranges
based solely on climate, but it is probable that once these
kinds of analyses become more commonplace and more spe-
cies are treated, clearer patterns will appear. That would be
beneficial, because correctly anticipating range shifts will
aid our efforts at conservation.
Acknowledgements
Comments by A. Farr, L. Kremsater, Mark Phinney, K.
Squires and an anonymous reviewer improved the manu-
script. The British Columbia Forest Sciences Program sup-
ported the work. The effort of the Biodiversity Centre for
Wildlife Studies in Victoria, British Columbia, to collect his-
torical records is greatly appreciated.
Literature cited
Baillie, S.M., C.D. Wilkerson and T.L. Newbury. 2004.
“Ashkui” vernal ice-cover phenomena and their
ecological role in southern Labrador. Canadian Field-
Naturalist 118:267–269.
Bunnell, F.L. 2010. Regional differences in habitat
associations. Technical Report, B.C. Forest Science
Program. http://for.gov.bc.ca/hfd/library/FIA/2010/
FSP_Y103131c.pdf. Accessed 3 March 2012.
Bunnell, F.L., K.A. Squires, M.I. Preston and R.W. Campbell.
2005. Towards a general model of avian response to
climate change. p. 59–70. in Implications of climate
change in BC’s southern interior forests. Workshop,
Revelstoke, B.C., 26–27 May 2005, Columbia Mountains
Birds and climate change - Bunnell et al.
British Columbia Birds
14
Volume 23, 2013
Institute of Applied Ecology, Revelstoke, B.C. http://
cmiae.org/_PDF/ImpofCCinforestsfinal.pdf. Accessed 3
March 2012.
Bunnell, F.L., M.I. Preston and A.C.M. Farr. 2008. Avian
response to climate change in British Columbia –
toward a general model. p. 9–27 in F. Dallmeier, A.
Fenech, D. MacIver and R. Szaro (eds.). Climate
change, biodiversity and sustainability in the
Americas. Smithsonian Institution Scholarly Press,
Washington, DC.
Bunnell, F.L., A. Moy, R.W. Wells and A. Breault. 2011a. Using
measured wetland depths to evaluate climate influences
on wetlands. Canadian Wildlife Service Technical Report,
Pacific & Yukon Region, Delta, B.C. (under review)
Bunnell, F.L., L.L. Kremsater and R.W. Wells. 2011b. Global
weirding in British Columbia – climate change and
terrestrial vertebrates. BC Journal of Ecosystems and
Management 12(2):21–38. http://jem.forrex.org/index.
php/jem/article/view/74/81. Accessed 3 March 2012.
Bunnell, F.L. and L.L. Kremsater. 2012. Migrating like a herd of cats
– climate change and emerging forests in British Columbia.
BC Journal of Ecosystems and Management (in press).
Daly, C., W P. Gibson, G.H. Taylor, G.L. Johnson and P. Pasteris.
2002. A knowledge-based approach to the statistical
mapping of climate. Climate Research 22:99–113.
Doswald, N., S.G. Willis, Y.C. Collingham, D. J. Pain, R.E. Green
and B. Huntley. 2009. Potential impacts of climatic change
on the breeding and non-breeding ranges and migration
distance of European Sylvia warblers. Journal of
Biogeography 36:1194–1208.
Flannigan, M.D., B.D. Amiro, K.A. Logan, B.J. Stocks and
B.M. Wotton. 2005. Forest fires and climate change in
the 21st Century. Mitigation and Adaptation Strategies
for Global Change 11:847–859.
Green, R.E., Y.C. Collingham, S.G. Willis, R.D. Gregory, K.W.
Smith and B. Huntley. 2008. Performance of climate envelope
models in retrodicting recent changes in bird population size
from observed climate change. Biology Letters 4:599–602.
Hamann, A. and T. Wang. 2006. Potential effects of climate
change on ecosystem and tree species distribution in
British Columbia. Ecology 87:2773–2786.
Jiménez-Valverde, A., N. Barve, A. Lira-Noriega, S.P. Maher,
Y. Nakazawa, M. Papes¸ J. Soberón, J. Sukumaran and
A.T. Peterson. 2011. Global Ecology and
Biogeography 20:114–118.
Kattsov,V.M. and J.E.Walsh. 2000. Twentieth-Century trends
of arctic precipitation from observational data and a
climate model simulation. Journal of Climate 13:1362-1370.
Lowther, P.E., C. Celada, N.K. Klein, C.C. Rimmer and D.A. Spector.
1999. Yellow Warbler (Dendroica petechia). The Birds of
North America Online. (A. Poole, ed.). Cornell Lab of
Ornithology, Ithaca, N.Y. http://bna.birds.cornell.edu/bna/
species/454. Accessed 3 March 2012.
Marini, M.A., M. Barbet-Massin, L.E. Lopes and F. Jiguet.
2009. Predicted climate-driven bird distribution changes
and forecasted conservation conflicts in a neotropical
savanna. Conservation Biology 23:1558–1567.
Mbogga, M.S., X. Wang and A. Hamann. 2010. Bioclimate
envelope model predictions for natural resource
management predictions: dealing with uncertainty.
Journal of Applied Ecology 47:731–740.
Meidinger, D.V. and J. Pojar. 1991. Ecosystems of British
Columbia. B.C. Ministry of Forests, Forest Science
Program, Special Report Series 6. http://
www.for.gov.bc.ca/hfd/pubs/Docs/Srs/Srs06.htm.
Accessed 3 March 2012.
Roeckner, E., G. Bäuml, L. Bonaventura, R. Brokopf, M. Esch,
M. Giorgetta, S. Hagemann, I. Kirchner, L. Kornblueh, E.
Manzini, A. Rhodin, U. Schlese, U. Schulzweida and A.
Tompkins. 2003. The atmospheric general circulation
model ECHAM5. Part 1. climateMax-Planck-Institut für
Meteorologie. Report No. 349. http//mpimet.mpg.de/
fileadmin/publikationen/Reports/max_scirep_349.pdf.
Accessed 3 March 2012.
Root, T.L., J.T. Price, K.R. Hall, S.H. Schneider, C. Rosenzweig
and J.A. Pounds. 2003. Fingerprints of global warming
on wild animals and plants. Nature 421:57–60.
Tobalske, B.W. 1997. Lewis’s Woodpecker (Melanerpes lewis).
The Birds of North America Online. (A. Poole, ed.). Cornell
Lab of Ornithology, Ithaca, N.Y. http://birds.cornell.edu/
bna/species/284. Accessed 3 March 2012.
Tuhkanen, S. 1980. Climatic parameters and indices in plant
geography. Acta Phytogeographica Suecica 67:1–105.
Warren, M.S., J.K. Hill, J.A. Thomas, J. Asher, R. Fox, B.
Huntley, D.B. Roy, M.G. Telfer, S. Jeffcoate, P. Harding,
G. Jeffcoate, S.G. Willis, J.N. Greatorex-Davies, D. Moss
and C.D. Thomas. 2001. Rapid responses of British
butterflies to opposing forces of climate and habitat
change. Nature 414:65–69.
Walther, G.R., E. Post, P. Convey, A. Menzel, C. Parmesan,
T.J.C. Beebee, J.M. Fromentin, O. Hoegh-Guldberg and
F. Bairlein. 2002. Ecological responses to recent climate
change. Nature 416:389–395.
Willis, S.G., D.G. Hole, Y.C. Collingham, G. Hilton, C. Rahbek
and B. Huntley. 2009. Assessing the impacts of future
climate change on protected area networks: a method to
simulate individual species’ responses. Environmental
Management 43:836–845.
Wohlfahrt, H. 2010.Validation of climate system models with
mid-Holocene palaeobotanic data for the Northern
extratropics. Reports on Earth System Science. Max-
Plank-Institut für Meteorologie. Hamburg, Germany.
http://mpimet.mpg.de/fileadmin/publikationen/Reports/
WEB_BzE_78.pdf. Accessed 3 March 2012.
Wotton, B.M., C.A. Nock and M.D. Flannigan. 2010. Forest
fire occurrence and climate change in Canada.
International Journal of Wildland Fire 19:253–271.
Zhu, K., C.W. Woodall and J.S. Clark. 2012. Failure to migrate:
lack of tree range expansion in response to climate
change. Global Change Biology 18:1042–1052.
Birds and climate change - Bunnell et al.
British Columbia Birds
15
Volume 23, 2013
Appendix I. Patterns of occupancy of 1:50 000 NTS cells in British Columbia by selected bird species during the 1960s
and 1990s. Total cells for a species is the number of sampled cells at a latitude equal to or south of the northernmost
occupied cell for that species; only cells sampled in both decades are included.
During the 1960s During the 1990s
Guild Total Not Not Newly
Species Ma Fb Bc Cellsd Occupied Occupied Occupied Occupied Occupiedc
Patagioenas fasciata 1 5 1 371 35 336 48 323 30Passerella iliaca 1 3 1 440 25 415 92 348 80Zenaida macroura 1 5 1 407 46 361 64 343 41Podiceps grisegena 1 2 3 354 33 321 42 312 30Melospiza melodia 1 3 1 439 127 312 218 221 132Pipilo maculatus 1 3 1 253 36 217 86 167 58Ixoreus naevius 1 3 1 440 53 387 121 319 96Anas americana 2 5 3 429 66 363 139 290 96Euphagus cyanocephalus 2 3 2 428 60 368 110 318 73Gavia immer 2 2 3 364 78 286 184 180 126Anas strepera 2 5 3 394 30 364 75 319 53Podiceps auritus 2 4 3 424 57 367 122 302 87Aythya affinis 2 4 3 424 24 400 108 316 88Anas acuta 2 4 3 424 66 358 108 316 67Anas clypeata 2 4 3 414 19 395 33 381 24Gavia pacifica 2 2 3 441 41 400 73 368 55Podiceps grisegena 2 2 3 428 29 399 61 367 50Melanitta perspicillata 2 4 3 337 30 307 76 261 61Aechmophorus occidentalis 2 2 3 395 69 326 104 291 62Melanitta fusca 2 4 3 424 75 349 93 331 55Aix sponsa 2 5 3 423 38 385 90 333 62Melospiza lincolnii 3 3 1 440 43 397 91 349 68Melanerpes lewis 3 3 1 236 36 200 58 178 53Sturnella neglecta 3 3 2 414 43 371 68 346 47Molothrus ater 4 3 2 441 88 353 164 277 107Anas cyanoptera 4 4 3 434 36 398 83 351 55Catharus ustulatus 4 3 1 362 105 257 186 176 117Dendroica petechia 4 3 1 345 97 248 182 163 109Chordeiles minor 5 3 2 438 108 330 138 300 72Empidonax minimus 5 3 1 423 27 396 67 356 58Buteo swainsoni 5 1 2 434 8 426 26 408 22Phalaropus tricolor 5 4 3 321 20 301 54 267 31
a Migratory guild: 1= resident, 2 = partial (winter at sea, move inland to breed), 3 = short distance (<1000 km), 4 = long distance (1000-4500 km), 5 = very long distance (>4500 km).
b Forage guild: 1 = predator, 2 = piscivore, 3 = terrestrial invertebrates, 4 = aquatic invertebrates, 5 = herbivores, frugivores and granivores.c Broad breeding habitat: 1 = forest including riparian, 2 = short vegetation (agriculture, grassland, shrubland), 3 = lakes and wetlands.d Number of cells may be greater than in Table 2, because cells of Table 2 are restricted to cells 51-60 oN.
Birds and climate change - Bunnell et al.
British Columbia Birds
16
Volume 23, 2013
Introduction
Single day counts of birds such as the Christmas Bird
Count (CBC) are popular as birding events because they are
group efforts combining the social and scientific aspects of
birding. The strength of the CBC is in its long record (1900
and continuing), and in the large number of counts under-
taken each year (>2000 in 2012, National Audubon Society
2012). There are no similar long-running birding events that
combine the social aspects of birding with the science of
data collection during the breeding season. The summer
Breeding Bird Survey requires surveyors to start half-hour
before dawn and has a strict protocol to standardize results
including only one person doing all the observations during
the survey on any given route. Breeding Bird Atlassing in
given states or provinces do not allow for regular mass par-
ticipation because they are only repeated at long intervals
and often do not combine social activities with birding. One
day breeding bird blitzes of given areas using methods simi-
lar to Christmas Bird Counts, i.e. count all birds detected in a
given area in one day, are therefore popular but have not
been organized on anything but a local scale and do not
contribute data to any formal single large-scale database.
This report summarizes the results of a long-term annual
breeding season bird count in Okanagan Mountain Provin-
cial Park. In 2003 fire burned 99% of the park at varying
levels of intensity. Because of this marked change in local
habitat, analysis of this count provided insight into bird habi-
tat use that goes beyond the local scale.
Okanagan Mountain Provincial Park (hereafter, “the park”)
on the east side of Okanagan Lake was gazetted in 1973 after
concerted efforts by local natural history societies and envi-
ronmental groups, particularly the Okanagan-Similkameen
Parks Society, with support from the Central Okanagan Natu-
ralist’s Club (CONC) of Kelowna, and the South Okanagan
Naturalist’s Club (SONC) of Penticton. In 1989 at the request
of B.C. Parks, CONC started an Annual Pilgrimage hike into
the park in mid-summer as an on-going project for the park’s
support and enhancement (CONC 2001). As many as 45 hik-
ers took part in the hikes from 1989 to 1992. In 1993, these
pilgrimages changed to annual bird counts taking place on
the last weekend in May or the first weekend in June, and
were organized by CONC, SONC, and B.C. Parks (CONC 2001).
Throughout western North America, fire suppression has
Effects of fire on bird abundance in Okanagan Mountain
Provincial Park, British Columbia
Les W. Gyug
Les W. Gyug, 3130 Ensign Way, West Kelowna, B.C. V4T 1T9 Canada; e-mail: [email protected]
Abstract: The Okanagan Mountain Provincial Park bird count is held annually the last weekend of May or the first weekend
of June when parties of observers record all birds detected. In 2003 the Okanagan Mountain fire burned 99% of the park at
varying intensities, providing a unique opportunity to examine long-term changes in bird species abundance affected by
fire. Relative abundance was compared from a period of 11 years before the fire (1993–2003) to a period up to eight years
after the fire (five counts from 2006–2011). In total 165 species have been tallied in the 16 counts. The average number of
species per count was significantly higher after the burn (104.6) than before (96.3). Of 90 species considered common
enough for meaningful statistical analyses, 28 increased in relative abundance after the fire, 11 decreased, and there was
no significant difference for 51 species. Increases were particularly noted among: woodpeckers including Hairy, Black-
backed, American Three-toed and Northern Flicker; some cavity nesters including House Wren, White-breasted Nuthatch,
and Mountain and Western Bluebirds; some insectivores including Olive-sided Flycatcher, Say’s Phoebe and Western
Wood-Pewee; and shrub-occupying birds including Warbling Vireo, Lazuli Bunting, MacGillivray’s Warbler, Song Sparrow
and Lincoln’s Sparrow. Severe declines were noted for forest inhabiting birds including Red-breasted Nuthatch, Golden-
crowned Kinglet and Townsend’s Warbler. For most species the response to fire determined by other studies was confirmed.
Key words: Okanagan Mountain Provincial Park, bird, bird count, burn, fire
Fire and bird abundance - Gyug
First published online — May 2012
British Columbia Birds
17
Volume 23, 2013
become very effective since the 1930’s. In southern interior
B.C., Blackwell and Gray (2003) defined several historical
natural fire regimes. Prior to effective fire suppression,
ponderosa pine forests would tend to be affected by fre-
quent (0–35 year cycle) low severity fires that would kill
understory trees and burn woody debris on the ground, but
leave large moderate to low densities of veteran trees pro-
tected by their thick bark from ground fires. Lodgepole pine
forests at higher elevations would be renewed on a longer
cycle (35–200 years) by higher severity stand-replacement
fires. The new generation of lodgepole pines generally re-
generate from seed as the serotinous cones are opened by
the fire’s heat. In low elevation forests where the frequent
fires have been suppressed, conifer forests would tend to
become older and denser, woody debris will build up on the
forest floor, and the potential for large devastating fires will
increase. The park master plan of 1990 (B.C. Parks 1990) rec-
ognized the build-up of fuels, and the “significant potential
for fire to be devastating to the park”.
Various short-term aspects of fire effects on bird abun-
dance have been studied. Relative abundance in a variety of
recent burns 1–4 years old were examined by Hutto (1995).
Hutto (1995) noted that some studies of birds in burns found
few species changes, but also noted that sample size and fire
intensity were often low in these comparisons. Saab and
Powell (2005) edited a compendium of 10 papers that each
summarized fire effects on avian ecology in different biomes
of North America including Saab et al. (2005) for the Rocky
Mountains, Hannan and Drapeau (2005) for the boreal for-
est, Huff et al. (2005) for the maritime Pacific Northwest.
Each of these biomes shared many bird species with the
park. None of the studies cited in those review papers com-
pared sites before and after wildfire. Studies were limited to
comparing current occupancy of different habitats in post-
hoc evaluations. The only large or comprehensive study to
compare bird relative abundance before a wildfire to the same
place afterwards was Smucker et al. (2005) in northwest
Montana. They compared relative abundance on point counts
surveyed for five years before a mixed-severity fire to the
relative abundance 1–3 years after. No burned sites seem to
have been studied long-term by any study.
The 256 km2 Okanagan Mountain fire of 2003 affected
99% of the park area, burning well beyond the park bounda-
ries and into the City of Kelowna. The fire dramatically al-
tered most of the vegetated habitats in the park, which pro-
vided an unrivalled opportunity to compare the bird popula-
tion of the dense conifer-dominated fire-suppressed habitat
before the fire to the burned forests after the fire. The
Okanagan Mountain Bird Count differed from all other stud-
ies in that there was a long-standing record (11 years) of bird
occurrence and relative abundance prior to the massive and
intense fire of 2003. In this study comparisons were not af-
fected by possible differences in physical setting; the only
difference before and after the fire were the habitat changes
caused by the fire.
Study Area
Okanagan Mountain Provincial Park is a Class A provin-
cial park of 105.8 km2 on the east side of Okanagan Lake with
32 km of lakeshore (Figure 1) (B.C. Parks 1990). It extends
from the lakeshore (342 m elevation) to the top of Okanagan
Mountain (1579 m). The park is dominated by Okanagan
Mountain, with its striking canyons and rugged rock out-
crops at lower elevations. Okanagan Mountain is treed to
the top as it is not high enough to be in the alpine or subalpine
zone. The southern boundary of the park is 20 km north of
downtown Penticton, and the northern boundary of the park
is 13 km south of downtown Kelowna.
The park is almost entirely natural with its wilderness
character persisting (B.C. Parks 1990). There are no develop-
ments in the park except for communications towers at the
top of Okanagan Mountain and a guide-outfitters cabin. The
only public road into the park is Lakeshore Drive extending
4 km into the park along the lakeshore at the north end.
There are three separate blocks of private land holdings along
Okanagan Lake that are not part of the park: the most north-
erly (1.17 km2) is accessible from Lakeshore Road, the others
(0.72 km2, and 0.39 km2) accessible only by boat.
Biogeoclimatic Ecosystem Classification (BEC) zones in
the park (Figure 2) include Ponderosa Pine (PP) at the lowest
elevations, up to 800 m on south slopes, and about 600 m on
north slopes; Interior Douglas-fir xeric (IDFx) at mid eleva-
tions from 800–1150 m on south slopes, and 600–800 m north
Figure 1. Location of Okanagan Mountain Provincial Park,
B.C., and 2003 Okanagan Mountain fire boundaries.
Fire and bird abundance - Gyug
British Columbia Birds
18
Volume 23, 2013
slopes; Interior Douglas-fir dry (IDFd) subzone at mid el-
evations from 1150–1500 m on south slopes, and 800–1350
m on north slopes; and Montane Spruce (MS) zone from
1500 m on south slopes and 1350 m on north slopes to the
top of Okanagan Mountain at 1579 m. Forests in the PP are
dominated by ponderosa pine (Pinus ponderosae), in the
IDFx by a mix of ponderosa pine and Douglas-fir
(Psuedotsuga menziesii), in the IDFd by a mix of Douglas-
fir and lodgepole pine (Pinus contorta), and in the MS by
lodgepole pine.
Prior to the Okanagan Mountain fire of 2003, the upper
elevation forests were mostly closed canopy (>40% crown
closure) while lower elevation forests were mostly of moder-
ate canopy closure (26–40% crown closure) (Figure 3). These
estimates were based on forest cover mapping produced by
the B.C. Forest Service. As the park had very few timber
values and was mostly mapped as Non-Productive Forest,
the polygons tended to be large and hid a lot of variability
from rock outcrop openings within that range of crown clo-
sure. After the 2003 fire, the B.C. Forest Service produced a
highly detailed digital map layer of the fire intensity and
remaining live crown. Less than 1% of the park’s area re-
mained untouched by the fire. High or extreme fire activity,
where almost all mature trees were killed, covered 66% of the
park area and moderate or low fire activity covered 33% (Fig-
ure 2). Extreme and high fire intensities covered 81% of the
upper elevations (MS and IDFd), but only 50% of the lower
elevations (PP and IDFx). After the fire, the majority of stands
at higher elevations had very low crown closure, i.e. all stand-
ing trees had been killed in most stands, and there were only
small remnant live tree patches (Figure 3). The post-fire for-
ests at lower elevations were a mosaic of completely burned
and partially burned patches with many remnant live trees
with a wider spread of crown closures.
Natural regeneration of lodgepole pine at higher eleva-
tions occurred within 2–4 years in most places with many
seedlings 1 m tall by 2011. Regeneration of conifers at lower
elevations appeared to have been much slower. Post-fire
shrub growth of red-stem ceanothus at low and mid eleva-
tions from seed stock dormant in the soil resulted in many
thickets up to 2 m tall by 2011. Burned-over aspen groves
had suckered into dense 2–3 m tall aspen stands by 2011.
Methods
The Okanagan Mountain Park Bird and Critter Count
consisted of a single annual count similar in methods to a
Christmas Bird Count. Parties of observers were assigned
routes to cover in the park and tallied all birds detected by
species. Notes were taken on occurrence of all other animals
as well but that information was not summarized here. For
the first 11 years the count was a single Saturday on the last
weekend in May. After the fire the count became a two-day
Figure 2. 2003 fire severity and Biogeoclimatic Ecosys-
tem Classification (BEC) zones in Okanagan Mountain
Provincial Park, B.C. PP = Ponderosa Pine zone, IDFx =
Interior Douglas-fir xeric subzone, IDFd = Interior Doug-
las-fir dry subzone, MS = Montane Spruce.
Figure 3. Summary of crown closure by area of
Okanagan Mountain Park prior to the 2003 fire and in
the fall of 2003 after the fire from BC Forest Service For-
est Cover mapping.
Fire and bird abundance - Gyug
British Columbia Birds
19
Volume 23, 2013
event either on the last weekend in May or the first weekend
in June. No route was covered more than once in any given
year and the coverage was similar whether the event was
held on one day or over more days. Occasionally a route
could not be covered on count day or count weekend so it
would be covered within 3 days of the count day or week-
end. Only one route is accessible for its entire length by car
on Lakeshore Road at the north end of the park. One route is
along the lakeshore by boat. The remainder of the routes are
hiking or walking trails.
Statistical comparisons based on single day counts of
areas (e.g. Christmas Bird Count) or routes (e.g. Breeding
Bird Survey) are generally only made for many areas or routes
grouped together because of high variance among counts
or routes and high annual variance on any given count or
route. Only with the rather dramatic change in habitat that
occurred in Okanagan Mountain Park was there likely to be
a large enough change in bird abundance by species to be
detectable when comparing a single annual count over time.
Relative abundance for each species was the number
counted divided by the search effort in party-hours. Data
was available from 11 years prior to the 2003 fire (1993–
2003), and for 5 years after the fire (2006–2007, 2009–2011).
After the 2003 fire, the park was closed to the public in 2004
and 2005 so the count did not begin again until 2006. Rela-
tive abundance after the fire in the period from 2006–2011
included any post-fire successional vegetation changes
such as herb, shrub and tree seedling/sapling growth but
did not include the two immediate post-fire years when
there may have been little shrub or tree seedling/sapling
regrowth. No salvage logging of any burned or live timber
took place within the park.
Species considered sufficiently common for meaningful
statistical analyses were those tallied on >60% of the counts
either before or after the fire, and with mean annual count >3
either before or after the fire. Total effort, number of species
counted, number of birds counted, and relative abundance
of each common species was contrasted before and after the
fire using non-parametric Mann-Whitney U tests at the al-
pha = 0.05 level to test for significance.
Changes in the breeding bird populations after the fire
were assumed to be for the most part a result of responses to
habitat change, and not direct mortality to the birds them-
selves from the fire. The fire occurred in late August when
breeding for almost all species would have been finished,
young would have been already fledged, and many would
have already departed on migration. For any birds that re-
mained, we assume that most of them would simply have
flown away from the advancing fire front.
Scientific names of birds mentioned are presented in
Appendix 2.
Results
The number of species counted per year on the Okanagan
Mountain Provincial Park count was significantly higher af-
ter the fire with, on average, 8.3 more species counted per
year (Table 1). Average number of observers decreased by
10 per year after the fire but number of party-hours did not
change significantly. After the fire, available observers were
spread more thinly to continue to attempt consistent cover-
age with an average 4.2 people per party before the fire com-
pared to 2.9 after the fire.
In total, 165 species have been counted in the 16 years of
the count. Ninety of these were common species (Table 2),
and 75 not common enough to compare statistically (Ap-
pendix 1). Twenty-eight of these common species (31%) in-
creased significantly in relative abundance after the fire.
Eleven common species (12%) decreased significantly in rela-
tive abundance after the fire. Fifty-one common species (57%)
did not change significantly in relative abundance.
Only two species (Black-backed Woodpecker and Moun-
tain Bluebird) were never counted prior to the fire but were
relatively common after the fire. Four other species (Western
Bluebird, Gray Catbird, Lincoln’s Sparrow and Lazuli Bunting)
were counted three times or fewer in total on all the pre-fire
counts but were relatively common after the fire.
The biggest increases were for birds that breed success-
Parameter
Before fire
(1993-2003)
After fire
(2006-2011)Statistics1
Mean SD Mean SD U1 U2 Result
Counts (n) 11 5
Observers 41.6 9.5 31.8 7.4 10.5 44.5 0
Effort (party-hours) 54.6 11.9 50.8 11.0 21 34 0
Total Birds Counted 1938 510 1865 357 23 32 0
No. of Species 96.3 8.9 104.6 2.9 49 6 +
1 Mann-Whitney U critical level given samples sizes of 11 and 5 for alpha = 0.05 is ≤9.
0 = no significant difference; + = significantly higher post fire; - = significantly lower post fire.
Table 1. Okanagan Mountain Provincial Park bird count summary statistics, 1993-2011, before and after the
2003 fire. SD = Standard Deviation.
Fire and bird abundance - Gyug
British Columbia Birds
20
Volume 23, 2013
Before Fire
(1993‒2003)
After Fire
(2006‒2011)Statistics1 Other Studies2
Species Mean SD Mean SD U1 U2 Response3 Response3 n
Canada Goose 1.92 1.56 0.45 0.56 47 8 -Mallard 0.57 0.31 0.21 0.13 49 6 -Ring-necked Duck 0.10 0.17 0.04 0.07 35.5 19.5 0Common Merganser 0.18 0.11 0.06 0.12 46 9 -Ruffed Grouse 0.24 0.20 0.08 0.06 49 6 - - 3California Quail 0.18 0.22 0.47 0.32 12 43 0Turkey Vulture 0.13 0.09 0.56 0.28 0 55 +Osprey 0.09 0.05 0.13 0.10 19 36 0Red-tailed Hawk 0.07 0.06 0.20 0.13 7 48 +American Kestrel 0.02 0.03 0.16 0.15 2 53 + + 4American Coot 0.05 0.04 0.19 0.19 21 34 0Spotted Sandpiper 0.49 0.37 0.18 0.15 45 10 0Ring-billed Gull 0.07 0.09 0.04 0.04 31 24 0Mourning Dove 0.47 0.30 0.42 0.11 25 30 0 + 3Vaux's Swift 0.11 0.08 0.03 0.03 45 10 0White-throated Swift 0.28 0.34 0.47 0.27 11 44 0Calliope Hummingbird 1.02 0.34 0.56 0.21 48 7 -Rufous Hummingbird 0.15 0.07 0.09 0.07 39 16 0 0 2Red-naped Sapsucker 0.32 0.19 0.15 0.09 48 7 - 0- 4Downy Woodpecker 0.03 0.02 0.08 0.07 14 41 0 0+ 4Hairy Woodpecker 0.10 0.08 0.58 0.18 0 55 + + 10Black-backed Woodpecker 0.00 0.00 0.08 0.06 0 55 + + 12Northern Flicker 0.51 0.18 0.91 0.34 8 47 + + 8Pileated Woodpecker 0.12 0.08 0.11 0.07 28 27 0 0- 4Am. Three-toed Woodpecker 0.02 0.03 0.08 0.06 6 49 + + 10Olive-sided Flycatcher 0.02 0.02 0.26 0.13 0 55 + + 6Western Wood-Pewee 0.20 0.17 0.63 0.39 6 49 + + 6Hammond's Flycatcher 0.56 0.31 0.72 0.23 17 38 0 m 3Dusky Flycatcher 1.00 0.36 1.93 0.85 8 47 +Pacific-slope Flycatcher 0.11 0.10 0.06 0.06 33 22 0Say's Phoebe 0.01 0.02 0.09 0.06 3 52 +Cassin's Vireo 0.66 0.30 0.48 0.21 38 17 0 0- 3Warbling Vireo 0.51 0.30 1.13 0.45 5 50 + 0- 4Gray Jay 0.09 0.09 0.02 0.04 41 14 0 m 6Steller's Jay 0.06 0.05 0.07 0.04 20 35 0 - 2Clark's Nutcracker 0.30 0.32 0.20 0.07 27 28 0 m 6Black-billed Magpie 0.06 0.04 0.01 0.01 51 4 -American Crow 0.13 0.11 0.11 0.05 28 27 0Common Raven 0.49 0.18 0.41 0.18 33 22 0 0 3Tree Swallow 0.34 0.23 0.39 0.18 24 31 0 + 8Violet-green Swallow 3.38 2.20 1.47 0.51 45 10 0Nor. rough-winged Swallow 0.26 0.13 0.09 0.05 50 5 -Barn Swallow 0.09 0.08 0.13 0.13 21 34 0Black-capped Chickadee 0.31 0.22 0.15 0.08 40 15 0 0 4Mountain Chickadee 0.53 0.22 0.34 0.17 42 13 0 - 4Red-breasted Nuthatch 1.60 0.70 0.72 0.15 55 0 - - 6White-breasted Nuthatch 0.04 0.05 0.17 0.12 6 49 +Pygmy Nuthatch 0.32 0.18 0.30 0.15 30 25 0Rock Wren 0.06 0.05 0.06 0.06 28.5 26.5 0Canyon Wren 0.04 0.04 0.07 0.05 17 38 0House Wren 0.05 0.05 2.15 1.49 0 55 + + 5Pacific Wren 0.25 0.27 0.14 0.07 32 23 0 m+ 6Golden-crowned Kinglet 0.55 0.58 0.02 0.02 55 0 - - 3Ruby-crowned Kinglet 0.33 0.26 0.31 0.15 26 29 0 m- 8Western Bluebird 0.00 0.01 0.21 0.15 0 55 +Mountain Bluebird 0.00 0.00 0.32 0.14 0 55 + + 7Townsend's Solitaire 0.54 0.27 0.40 0.08 32 23 0 m 5Veery 0.09 0.09 0.22 0.11 7 48 +Swainson's Thrush 0.20 0.17 0.37 0.22 13 42 0 0 5Hermit Thrush 0.09 0.08 0.08 0.04 25 30 0 m 8
►
Table 2. Relative abundance (number counted per party-hour) and response to fire-caused changes in habitat
of 90 common bird species in Okanagan Mountain Provincial Park before and after the 2003 fire. Species sorted
by taxonomic order. SD = Standard Deviation.
Fire and bird abundance - Gyug
British Columbia Birds
21
Volume 23, 2013
fully in post-fire habitats, either associated with standing dead
trees, semi-open country, or with dense and abundant shrubs.
Woodpeckers that feed on wood-boring insects increased
significantly in numbers after the fire. Black-blacked Wood-
peckers had never been encountered in the park prior to the
fire but were present after the fire. Relative abundance of
Hairy Woodpeckers and American Three-toed Woodpeck-
ers were four to six times higher after the fire.
Birds found in semi-forested or open habitats were well
up. Both Mountain Bluebirds and Western Bluebirds were
either not observed or rare in the park prior to the fire, but
both were common afterwards. Olive-sided Flycatchers, a
species considered Threatened in Canada because of 40%
declines in numbers in the past 50 years (COSEWIC 2007),
increased ten-fold in abundance. Turkey Vultures, Brew-
er’s Blackbirds, American Kestrels and Say’s Phoebe also
increased. Red-winged Blackbirds appeared to increase be-
cause they were observed using upland burned-over shrub
areas as part of their habitat where they would not have
used the forests prior to the fire. The cat-tail marshes they
breed in appeared to remain relatively unchanged before
and after the fire, but the increase in use of upland habitat
may have allowed higher densities overall in the same habi-
tats. White-breasted Nuthatch increased, even though they
are a species normally inhabiting live ponderosa pine for-
ests as well. The open ponderosa pine forests created by
the fire may be more similar to the more natural open for-
ests in which they likely evolved, rather than the very closed
and dense ponderosa pine forests prior to the fire that may
not be ideal habitat.
Some bird species associated with shrubs increased
greatly in abundance. The most significant increase was
in House Wren numbers, with an average of 2.6 counted
per year before the fire to an average of 101.2 per year
after the fire. Only one Lazuli Bunting had ever been
counted in the entire 11 years of the count prior to the
fire, but it was a regular after the fire with an average of
seven counted per year. Warbling Vireos were particularly
abundant in the suckering aspen groves, doubling in abun-
dance. Other shrub-associate species that increased were
Veery, MacGillivray’s Warbler, Song Sparrow, and Lincoln’s
Sparrow.
Within the 2006–2011 post-fire period, there were ob-
vious increases for only three species. House Wren rela-
tive abundance increased from 0.92 per party hour in 2006
to 2.92 in 2011; MacGillivray’s Warbler increased from 0.51
to 1.26; and Spotted Towhee increased from 0.47 to 1.49.
The relative abundance of Spotted Towhee was higher in
Fire and bird abundance - Gyug
◄ Table 2 Before Fire
(1993‒2003)
After Fire
(2006‒2011)Statistics1 Other Studies2
Species Mean SD Mean SD U1 U2 Response3 Response3 n
European Starling 0.25 0.16 0.41 0.13 12 43 0Orange-crowned Warbler 0.21 0.14 0.41 0.37 24 31 0 0 2Nashville Warbler 0.89 0.42 0.55 0.30 41 14 0Yellow Warbler 0.13 0.08 0.12 0.10 30 25 0Yellow-rumped Warbler 1.58 0.70 1.95 0.87 17 38 0 m 8Townsend's Warbler 1.03 0.56 0.22 0.07 55 0 - 0 4Northern Waterthrush 0.03 0.03 0.08 0.05 11 44 0MacGillivray's Warbler 0.30 0.18 0.87 0.36 2 53 + m 2Wilson's Warbler 0.24 0.19 0.21 0.10 27 28 0Western Tanager 0.75 0.40 0.71 0.19 24 31 0 m 8Spotted Towhee 0.58 0.25 1.30 0.86 9 46 +Chipping Sparrow 1.61 0.80 1.82 0.69 22 33 0 m 8Vesper Sparrow 0.05 0.04 0.07 0.10 29.5 25.5 0Song Sparrow 0.09 0.08 0.72 0.54 0 55 + - 2Lincoln's Sparrow 0.00 0.01 0.15 0.11 0 55 + m 3White-crowned Sparrow 0.04 0.11 0.26 0.17 5 50 + + 3Dark-eyed Junco 1.75 1.03 1.65 0.62 26 29 0 m+ 8Lazuli Bunting 0.00 0.00 0.14 0.10 0 55 + 0+ 2Red-winged Blackbird 0.11 0.09 0.40 0.30 5 50 +Brewer's Blackbird 0.03 0.04 0.23 0.05 0 55 +Brown-headed Cowbird 0.88 0.31 0.90 0.15 20 35 0 0 2Bullock's Oriole 0.09 0.11 0.16 0.10 9 46 +Cassin's Finch 0.24 0.18 0.12 0.06 40 15 0 + 4House Finch 0.39 0.31 0.31 0.15 29 26 0Red Crossbill 0.75 0.93 0.05 0.04 42 13 0 0- 5Pine Siskin 1.13 0.80 0.56 0.37 40.5 14.5 0 m 7American Goldfinch 0.07 0.09 0.16 0.09 15.5 39.5 0Evening Grosbeak 0.93 1.17 0.33 0.28 35 20 0 m- 21 Mann-Whitney U critical level given samples sizes of 11 and 5 for alpha ꞊ 0.05 is ≤9. 2 Responses from 16 other studies as summarized in Saab et al. (2005), Hannon and Drapeau (2005), Huff et al. (2005) and Smucker et al. (2005). Only
species with more than one study cited. The number of studies (n) that gave a result for each species if given.3 Responses to fire: 0 ꞊ no significant difference; + ꞊ significantly higher post fire; - ꞊ significantly lower post fire, m ꞊ mixed results. Where there were
several results from previous studies, the dominant response was given first, followed by subdominant trend.
British Columbia Birds
22
Volume 23, 2013
2009-2011 than in 2006-2007 and higher than in any year
prior to the fire.The response of Spotted Towhee to post-
fire shrub succession may have been slightly slower than
other species, and was only beginning to be shown six
years post-fire.
Declines were significant for 11 species after the fire. The
largest declines were for those species that inhabit mature
closed forests. Abundance of Ruffed Grouse, Red-breasted
Nuthatch, and Townsend’s Warbler were down to 25–67%
of pre-fire levels. Golden-crowned Kinglets declined to 3%
of their pre-fire level. Red-naped Sapsuckers also declined
by 45% as most of the aspen groves they tended to inhabit
were burned severely. Several other uncommon forest spe-
cies including Boreal Chickadee, Brown Creeper and Varied
Thrush appeared to decline as well (see Appendix 1). Sum-
ming up the totals for all counts, only one Boreal Chickadee
was counted after the fire in five years compared to 17 before
in 11 years. No Brown Creepers or Varied Thrushes were
counted after the fire compared to 18 and 24 before the fire.
Brown Creepers have been observed occasionally in the park
after the fire (Pers. comm., Deirdre and Jim Turnbull,
Naramata), so it would appear that Brown Creepers 3–8 years
after the fire were still so uncommon that they had yet to be
detected on count days.
Significant declines were noted for some water birds
including Canada Goose, Mallard and Common Mergan-
ser. Most of the Canada Geese counted were flying over-
head, rather than in breeding, foraging or resting habitat,
as there is almost no Canada Goose breeding or foraging
habitat within the park. The decline in Canada Goose rela-
tive abundance may represent the results of an egg-ad-
dling program to reduce numbers in the Okanagan Valley
which would result in fewer geese commuting to and from
foraging habitat in Penticton and Kelowna. The shoreline
in general is too rocky and plunges directly to depth with
virtually no shoreline marshes suitable for waterfowl. Rea-
sons why Mallard or Common Merganser numbers may
have declined are unknown.
No significant changes were noted in some of the sig-
nature birds of the park such as Canyon Wren and White-
throated Swift that are associated with cliff and talus rather
than with conifer forests. Yellow-rumped Warbler, which
one might ordinarily associate with conifer forests,
showed no decline at all and appeared just as abundant in
forests of standing dead trees as in the dense pre-fire
conifers. Similarly, many other common species that we
might ordinarily associate with forests such as Western
Tanager, Mountain Chickadee, Nashville Warbler, Pacific
Wren, Hermit Thrush and Swainson’s Thrush showed no
significant changes in relative abundance. None of the
common finches including Cassin’s Finch, House Finch,
Red Crossbill, Pine Siskin, American Goldfinch and
Evening Grosbeak showed any significant change in rela-
tive abundance before and after the fire.
Discussion
Within the period from 3–8 years after the Okanagan
Mountain fire, no common bird species were eliminated from
Okanagan Mountain Park because of the habitat changes
caused by the 2003 fire even with what would appear to any
observer to be major habitat changes. The majority of bird
species showed no great changes in numbers. Since habitat
was created for additional species without the loss of any
common species, overall bird diversity increased and the
number of species counted per year was higher after the fire.
Overall about twice as many bird species increased after the
fire than decreased, similar to the results found after fire in
northwestern Montana (Smucker et al. 2005).
The relative abundance of most common bird species
was assumed to be relatively static in the 11 years prior to
the fire because there were no major habitat changes. It will
be difficult to show any effect of fire on species such as
seed-eating finches whose local abundance may fluctuate
markedly on an annual basis. However, the annual abun-
dance of most other species tended to be relatively stable in
this study prior to the fire.
Relative abundance of some species such as the House
Wren, MacGillivray’s Warbler and Spotted Towhee were not
static and were changing at a comparatively quick pace after
the fire. As the habitats change with natural post-fire suc-
cession, the abundance of many other bird species will also
be expected to change. Future analyses of the count data
will have to take increases and decreases with habitat suc-
cession into account in regression and trend analyses, rather
than the relatively simple comparisons of before and after
presented here.
Hutto (1995) identified bird species relatively restricted
to early post-fire conditions as Olive-sided Flycatcher, Ameri-
can Three-toed Woodpecker, Black-backed Woodpecker,
Clark’s Nutcracker and Mountain Bluebird. Clark’s Nutcracker
was just as common in the park before and after the fire so
would not appear to be an early post-fire specialist. Each of
the other species increased significantly in the park after the
fire, or in the case of Black-backed Woodpecker and Moun-
tain Bluebird, were found only after the fire.
Saab et al.(2005) summarized fire response of 66 species
from eight other studies in the Rocky Mountains; Hannon
and Drapeau (2005) summarized response of 69 species from
five other studies in the boreal forest; Huff et al. (2005) sum-
marized the response of 26 species from three other studies
in Pacific maritime forests. In general, the results matched
those of this study (Table 2) as discussed below. However
these cross-study comparisons must be interpreted with
caution because when species show different responses to
fire in different studies, this is more likely to be a response to
different fire severity or time since fire in different studies
(Smucker et al. 2005).
Fire and bird abundance - Gyug
British Columbia Birds
23
Volume 23, 2013
Of the 14 park species with significant increases in this
study, and for which >1 study was cited in Table 2, 11 were in
general agreement, i.e. most studies indicated the species in-
creased after fire. For the three species for which there was no
agreement, the fault appeared to be in the length of the stud-
ies. Short-term studies of <4 years after fire tended to indicate
only a few species significantly increase after fire (e.g. Hutto
1995). However, longer term studies such as this one indicate
a much wider variety of species increase after fire, in particular
as shrub regrowth creates new habitat. For example, species
such as Warbling Vireo, Dusky Flycatcher, and MacGillivray’s
Warbler can be very abundant in dense 2–3 m tall deciduous
shrubs in regenerating clearcuts (Gyug 2000) and these all
exhibited increases after the fire in the park. However, if a
post-fire study does not last long enough for the shrubs to
reach 2–3 m tall, then no effect will likely be shown.
Of the five park species with significant decreases in this
study, and for which there was >1 study (Table 2), three were
in general agreement with none of those species showing in-
creases after fire. Only for Townsend’s Warbler was no de-
crease shown in any of those studies after fire. In the park
Townsend’s Warbler relative abundance decreased to 21% of
the pre-fire abundance. There was no doubt about the sensi-
tivity to fire as the few places with Townsend’s Warblers were
almost always remnant forest patches. Red-naped Sapsuckers
showed no response to fire in two of three studies, and a
negative response in one. The response in this study was
negative as most aspen patches were burned over but any
unburned aspen patches did tend to retain sapsuckers.
Of the 27 park species for which no fire effect was shown
in this study, and for which >1 study was cited by any of
those three papers (Table 2), 22 were in general agreement,
with either no response shown, mixed responses, or differ-
ent responses in different studies, i.e. particular site effects
may be stronger than the any response to fire for those spe-
cies. Of the five species that did not match, the effects of fire
tended to be positive for Mourning Dove, Downy Wood-
pecker, Tree Swallow and Cassin’s Finch, and the effects
tended to be negative for Mountain Chickadee, but no ef-
fects could be shown on these species in this study.
For some species 2003 may have coincided with another
event causing long-term decreases or increases in popula-
tion abundance that might confound the results of this study.
However, with the exception of Canada Goose, the possibil-
ity of any such single event significantly affecting abun-
dance to a greater degree than the major effects of the fire on
habitat were likely to be small. Examination of an 18-year
period will tend to account for both inter-annual fluctua-
tions that add statistical noise to the analysis and for long
and steady declines or increases in abundance. On the longer
term trend and regression analysis will be required for many
species to determine how post-fire successional patterns
affect abundance. However, over the 18-year term of this
study, the strongest effects on abundance before and after
2003 were likely to be those of the fire alone.
The forests of the park are going to be continually chang-
ing for many years in the course of natural succession as
young conifers and aspens grow into forests again and par-
ticularly as these overtop the current shrub growth that is
very dense in many places. If continued, this annual count
will provide an excellent long-term record of the changes in
bird populations in response to those habitat changes.
Whether the park will ever again get to the rather unnatural
state where older conifer forests dominate the entire area will
depend on any habitat management that might go on within
the park and, of course, when fires may strike again. If some
fires happen within the park relatively soon, they may not
sweep over the whole park as happened in 2003 because
fuels in general have been reduced. The results may be a
mosaic of open and closed forests at lower elevations more
resembling the forests prior to widespread fire suppression
and with bird communities to match.
Acknowledgements
The author would like to thank all who have participated
in the Okanagan Mountain Park count over the years for
their efforts and contributions, whether on behalf of the Cen-
tral Okanagan Naturalists’ Club, the South Okanagan Natu-
ralists’ Club, or on their own behalf. In particular, many stal-
wart volunteers deserve thanks for keeping the Okanagan
Mountain Park count going and successful and for giving
freely of their time and efforts. The following people are de-
serving of particular mention for providing support and or-
ganization in the early years: Don Gough as regional man-
ager of B.C. Parks; Eileen Dillabough, Brenda Thomson,
Gwynneth Wilson, Judy Latta, Denise Brownlie and Eileen
Chappell in the central Okanagan; and Eva Durance and
Laurie Rockwell in the south Okanagan. The author thanks
Mary Taitt and Rob Butler for reviews that improved this
paper.
Literature Cited
Blackwell, B.A. and R.W. Gray. 2003. Developing a coarse
scale approach to the assessment of forest fuel
conditions in southern British Columbia. Report
produced for Natural Resources Canada, Canadian
Forest Service, Victoria, B.C. 32p.
British Columbia Parks. 1990. Master Plan for Okanagan
Mountain Provincial Park. B.C. Parks, Southern Interior
Region, Kamloops, B.C. 42p.
Central Okanagan Naturalist’s Club. 2001. Tracks, Trails and
Naturalists’ Tales: a history of the Central Okanagan
Naturalist’s Club 1962 to 2000. Central Okanagan
Naturalist’s Club, Kelowna, B.C. 154p.
Fire and bird abundance - Gyug
British Columbia Birds
24
Volume 23, 2013
COSEWIC. 2007. COSEWIC assessment and status report
on the Olive-sided Flycatcher Contopus cooperi in
Canada. Committee on the Status of Endangered Wildlife
in Canada. Ottawa. 25p. http://www.sararegistry.gc.ca/
document/default_e.cfm?documentID=1629. Accessed
2012 April.
Gyug, L.W. 2000. Timber-harvesting effects on riparian
wildlife and vegetation in the Okanagan Highlands of
British Columbia. Wildlife Bulletin B-97, B.C.
Environment, Victoria, B.C. 112p.
Hannon, S.J. and P. Drapeau. 2005. Bird responses to burning
and logging in the boreal forest of Canada. Studies in
Avian Biology 30:97-115.
Huff, M.H., N. E. Seavy, J.D. Alexander and C.J. Ralph. 2005.
Fire and birds in maritime Pacific Northwest. Studies in
Avian Biology 30:46-62.
Hutto, R.L. 1995. Composition of bird communities following stand
replacement fires in northern Rocky Mountain (U.S.A.)
conifer forests. Conservation Biology 9: 1041–1058.
National Audubon Society. 2012. Christmas Bird Count.
http://birds.audubon.org/christmas-bird-count.
Accessed 2012 April.
Saab, V.A. and H.D. Powell. 2005. Fire and avian ecology in
North America: process influencing pattern. Studies in
Avian Biology 30:1-13.
Saab, V.A., H.D. Powell, N.B, Kotliar and K.R. Newlon. 2005.
Variation in fire regimes of the Rocky Mountains:
implications for avian communities and fire management.
Studies in Avian Biology 30:76-96.
Smucker, K.M., RL. Hutto and B.M. Steele. 2005. Changes in bird
abundance after wildfire: importance of fire severity and time
since fire. Ecological Applications 15(5):1535-1549.
Appendix 1. Relative abundance (number tallied per 100 party-hours) of 75 bird species considered not
sufficiently common for meaningful statistical analyses (i.e. average <3 per year, or recorded on <60%
of counts either before or after fire) in Okanagan Mountain Provincial Park before and after the 2003 fire.
Species sorted by taxonomic order. Totals are those counted on all counts, summed over all years.
Before Fire
(1993‒2003)
After Fire
(2006‒2011)
Species
Total tallied on 11 counts
Number tallied /100 party-hours
Total tallied on 5 counts
Number tallied /100 party-hours
Gadwall 1 0.2 1 0.3American Wigeon 1 0.3 0 0.0Blue-winged Teal 8 1.8 0 0.0Cinnamon Teal 0 0.0 1 0.4Green-winged Teal 7 1.2 2 0.7Lesser Scaup 2 0.3 2 0.8Bufflehead 16 2.6 15 5.7Common Goldeneye 34 6.3 0 0.0Barrow's Goldeneye 152 28.4 34 16.3Hooded Merganser 3 0.5 3 1.2Ruddy Duck 8 1.5 28 12.5Chukar 1 0.1 1 0.3Ring-necked Pheasant 1 0.1 0 0.0Spruce Grouse 4 0.7 0 0.0Dusky Grouse 5 0.7 4 1.5Pacific Loon 1 0.1 0 0.0Common Loon 25 4.2 4 1.5Pied-billed Grebe 0 0.0 3 1.8Horned Grebe 1 0.1 0 0.0Red-necked Grebe 28 4.7 0 0.0Western Grebe 12 1.9 0 0.0Double-crested Cormorant 1 0.1 0 0.0Bald Eagle 30 5.0 12 4.4Northern Harrier 3 0.5 1 0.3Sharp-shinned Hawk 11 1.8 3 1.1Cooper's Hawk 13 2.2 6 2.3Northern Goshawk 2 0.3 3 1.2Swainson's Hawk 2 0.3 0 0.0Rough-legged Hawk 2 0.3 0 0.0Golden Eagle 5 0.9 1 0.3Merlin 5 0.8 5 1.8
►
Fire and bird abundance - Gyug
British Columbia Birds
25
Volume 23, 2013
◄ Appendix 1 Before Fire
(1993‒2003)
After Fire
(2006‒2011)
Species
Total tallied on 11 counts
Number tallied /100 party-hours
Total tallied on 5 counts
Number tallied /100 party-hours
Sora 7 1.0 5 2.1Killdeer 13 2.3 10 3.7Long-billed Curlew 1 0.2 0 0.0Wilson's Snipe 1 0.1 9 4.1Bonaparte's Gull 2 0.5 0 0.0Herring Gull 8 1.5 1 0.4Parasitic Jaeger 2 0.3 0 0.0Rock Pigeon 11 1.7 0 0.0Great-horned Owl 1 0.1 3 1.1Northern Pygmy-Owl 1 0.2 1 0.4Barred Owl 3 0.5 1 0.4Great gray Owl 1 0.1 0 0.0Common Nighthawk 3 0.5 31 14.6Common Poorwill 2 0.4 1 0.4Black Swift 10 1.7 2 0.7Black-chinned Hummingbird 2 0.5 0 0.0Belted Kingfisher 13 1.9 2 0.7Lewis' Woodpecker 4 0.8 0 0.0Willow Flycatcher 8 1.3 10 4.7Least Flycatcher 1 0.1 2 0.8Gray Flycatcher 3 0.4 0 0.0Western Kingbird 3 0.6 7 2.7Eastern Kingbird 5 0.9 14 6.3Red-eyed Vireo 20 3.1 11 3.9Cliff Swallow 1 0.1 7 2.4Boreal Chickadee 17 2.5 1 0.4Brown Creeper 18 3.3 0 0.0American Dipper 2 0.4 0 0.0Varied Thrush 24 5.0 0 0.0American Pipit 1 0.2 0 0.0Magnolia Warbler 3 0.5 0 0.0American Redstart 1 0.1 1 0.6Common Yellowthroat 3 0.5 8 3.6Clay-colored Sparrow 0 0.0 1 0.3Lark Sparrow 1 0.1 0 0.0Savannah Sparrow 3 0.6 1 0.3Golden-crowned Sparrow 1 0.1 0 0.0Black-headed Grosbeak 27 4.8 12 4.5Western Meadowlark 3 0.6 5 1.7Yellow-headed Blackbird 1 0.2 0 0.0Pine Grosbeak 20 3.7 5 1.9White-winged Crossbill 1 0.1 0 0.0House Sparrow 3 0.6 0 0.0
Appendix 2. Scientific names of birds mentioned in the text.
English name Scientific name English name Scientific name
Canada Goose Branta canadensis California Quail Callipepla californicaGadwall Anas strepera Chukar Alectoris chukarAmerican Wigeon Anas americana Ring-necked Pheasant Phasianus colchicusMallard Anas platyrhynchos Ruffed Grouse Bonasa umbellusBlue-winged Teal Anas discors Spruce Grouse Falcipennis canadensisCinnamon Teal Anas cyanoptera Dusky Grouse Dendragapus obscurusGreen-winged Teal Anas crecca Pacific Loon Gavia pacificaRing-necked Duck Aythya collaris Common Loon Gavia immerLesser Scaup Aythya affinis Pied-billed Grebe Podilymbus podicepsBufflehead Bucephala albeola Horned Grebe Podiceps auritusCommon Goldeneye Bucephala clangula Red-necked Grebe Podiceps grisegenaBarrow's Goldeneye Bucephala islandica Western Grebe Aechmophorus occidentalisHooded Merganser Lophodytes cucullatus Double-crested Cormorant Phalacrocorax auritusCommon Merganser Mergus merganser Turkey Vulture Cathartes auraRuddy Duck Oxyura jamaicensis Osprey Pandion haliaetus
►
Fire and bird abundance - Gyug
British Columbia Birds
26
Volume 23, 2013
◄ Appendix 2
English name Scientific name English name Scientific name
Bald Eagle Haliaeetus leucocephalus Barn Swallow Hirundo rustica
Northern Harrier Circus cyaneus Black-capped Chickadee Poecile atricapillus
Sharp-shinned Hawk Accipiter striatus Mountain Chickadee Poecile gambeli
Cooper's Hawk Accipiter cooperii Boreal Chickadee Poecile hudsonicus
Northern Goshawk Accipiter gentilis Red-breasted Nuthatch Sitta canadensis
Swainson's Hawk Buteo swainsoni White-breasted Nuthatch Sitta carolinensisRed-tailed Hawk Buteo jamaicensis Pygmy Nuthatch Sitta pygmaeaRough-legged Hawk Buteo lagopus Brown Creeper Certhia americanaGolden Eagle Aquila chrysaetos Rock Wren Salpinctes obsoletusAmerican Kestrel Falco sparverius Canyon Wren Catherpes mexicanusMerlin Falco columbarius House Wren Troglodytes aedonSora Porzana carolina Pacific Wren Troglodytes pacificusAmerican Coot Fulica americana American Dipper Cinclus mexicanusKilldeer Charadrius vociferus Golden-crowned Kinglet Regulus satrapaSpotted Sandpiper Actitis macularius Ruby-crowned Kinglet Regulus calendulaLong-billed Curlew Numenius americanus Western Bluebird Sialia mexicanaWilson's Snipe Gallinago delicata Mountain Bluebird Sialia currucoidesBonaparte's Gull Chroicocephalus philadelphia Townsend's Solitaire Myadestes townsendiRing-billed Gull Larus delawarensis Veery Catharus fuscescensHerring Gull Larus argentatus Swainson's Thrush Catharus ustulatusParasitic Jaeger Stercorarius parasiticus Hermit Thrush Catharus guttatusRock Pigeon Columba livia American Robin Turdus migratoriusMourning Dove Zenaida macroura Varied Thrush Ixoreus naeviusGreat Horned Owl Bubo virginianus Gray Catbird Dumetella carolinensisNorthern Pygmy-Owl Glaucidium gnoma European Starling Sturnus vulgarisBarred Owl Strix varia American Pipit Anthus rubescensGreat Gray Owl Strix nebulosa Cedar Waxwing Bombycilla cedrorumCommon Nighthawk Chordeiles minor Northern Waterthrush Parkesia noveboracensisCommon Poorwill Phalaenoptilus nuttallii Orange-crowned Warbler Oreothlypis celataBlack Swift Cypseloides niger Nashville Warbler Oreothlypis ruficapillaVaux's Swift Chaetura vauxi MacGillivray's Warbler Geothlypis tolmieiWhite-throated Swift Aeronautes saxatalis Common Yellowthroat Geothlypis trichasBlack-chinned Hummingbird Archilochus alexandri American Redstart Setophaga ruticillaCalliope Hummingbird Stellula calliope Magnolia Warbler Setophaga magnoliaRufous Hummingbird Selasphorus rufus Yellow Warbler Setophaga petechiaBelted Kingfisher Megaceryle alcyon Yellow-rumped Warbler Setophaga coronataLewis's Woodpecker Melanerpes lewis Townsend's Warbler Setophaga townsendiRed-naped Sapsucker Sphyrapicus nuchalis Wilson's Warbler Cardellina pusillaDowny Woodpecker Picoides pubescens Spotted Towhee Pipilo maculatusHairy Woodpecker Picoides villosus Chipping Sparrow Spizella passerinaAmerican Three-toed Woodpecker Picoides dorsalis Clay-colored Sparrow Spizella pallidaBlack-backed Woodpecker Picoides arcticus Vesper Sparrow Pooecetes gramineusNorthern Flicker Colaptes auratus Lark Sparrow Chondestes grammacusPileated Woodpecker Dryocopus pileatus Savannah Sparrow Passerculus sandwichensisOlive-sided Flycatcher Contopus cooperi Song Sparrow Melospiza melodiaWestern Wood-Pewee Contopus sordidulus Lincoln's Sparrow Melospiza lincolniiWillow Flycatcher Empidonax traillii White-crowned Sparrow Zonotrichia leucophrysLeast Flycatcher Empidonax minimus Golden-crowned Sparrow Zonotrichia atricapillaHammond's Flycatcher Empidonax hammondii Dark-eyed Junco Junco hyemalisGray Flycatcher Empidonax wrightii Western Tanager Piranga ludovicianaDusky Flycatcher Empidonax oberholseri Black-headed Grosbeak Pheucticus melanocephalusPacific-slope Flycatcher Empidonax difficilis Lazuli Bunting Passerina amoenaSay's Phoebe Sayornis saya Red-winged Blackbird Agelaius phoeniceusWestern Kingbird Tyrannus verticalis Western Meadowlark Sturnella neglectaEastern Kingbird Tyrannus tyrannus Yellow-headed Blackbird Xanthocephalus xanthocephalusCassin's Vireo Vireo cassinii Brewer's Blackbird Euphagus cyanocephalusWarbling Vireo Vireo gilvus Brown-headed Cowbird Molothrus aterRed-eyed Vireo Vireo olivaceus Bullock's Oriole Icterus bullockiiGray Jay Perisoreus canadensis Pine Grosbeak Pinicola enucleatorSteller's Jay Cyanocitta stelleri Cassin's Finch Carpodacus cassiniiClark's Nutcracker Nucifraga columbiana House Finch Carpodacus mexicanusBlack-billed Magpie Pica hudsonia Red Crossbill Loxia curvirostraAmerican Crow Corvus brachyrhynchos White-winged Crossbill Loxia leucopteraCommon Raven Corvus corax Pine Siskin Spinus pinusTree Swallow Tachycineta bicolor American Goldfinch Spinus tristisViolet-green Swallow Tachycineta thalassina Evening Grosbeak Coccothraustes vespertinusNorthern Rough-winged Swallow Stelgidopteryx serripennis House Sparrow Passer domesticusCliff Swallow Petrochelidon pyrrhonota
Fire and bird abundance - Gyug
British Columbia Birds
27
Volume 23, 2013
Introduction
British Columbia is experiencing climate warming trends
similar to those documented elsewhere in western North
America (Mote 2003, Taylor 2005). Some waterbirds in Brit-
ish Columbia have responded to changing climate by alter-
ing their arrival times inland after wintering at sea, their dura-
tion inland, the northward extension of their range and shift-
ing their relative abundance northward or southward
(Bunnell et al. 2008).
Bunnell et al. (2011c) documented and projected rates of
drying for wetlands in the central interior of British Colum-
bia. Drying rates change with wetland size and elevation.
Differential use of particular wetland sizes and elevations
make waterfowl species variably vulnerable to climate change.
Survey data collected by the Canadian Wildlife Service and
Ducks Unlimited Canada permit evaluation of relative use of
wetland size and elevation classes by different waterfowl
species. Our objectives are to: 1) describe influences of
wetland size and elevation on estimated waterfowl abun-
dance for the study area, 2) illustrate differences in responses
of waterfowl species to wetland size and elevation, and 3)
estimate relative vulnerabilities of waterfowl species to pro-
jected effects of climate change.
Data and methods
The study area was the Central Interior Ecoprovince
(CIE), an 11.1 million hectare region in central British Co-
lumbia (Figure 1). The CIE is divided into 12 Ecosections
reflecting landform and vegetation. It is a large and eco-
logically diverse region, incorporating 10 of the province’s
16 biogeoclimatic zones (Meidinger and Pojar 1991). Of
these, five forested zones predominate in the region:
Engelmann Spruce–Subalpine Fir, Sub-boreal Pine–Spruce,
Sub-boreal Spruce, Montane Spruce, Interior Douglas-fir
and Interior Cedar–Hemlock. Wetlands in the area support
large populations of waterfowl and other waterbirds (Breault
et al. 2007).
Data for waterfowl surveys were collected during May
by helicopter transect surveys in 2006, 2007 and 2008 (see
Breault et al. 2007 for details on survey methodology). Eight
of the 12 Ecosections of the Central Interior Ecoprovince
were surveyed; of those eight, two (Cariboo Plateau and
McGregor Plateau) were not sampled in 2007 and 2008. The
latter Ecosections were excluded from analyses of inter-an-
nual variability.
The standard, continent-wide method employed by the
US Fish and Wildlife Service and Canadian Wildlife Service
One size does not fit all: differential responses of waterfowl species
to impacts of climate change in central British Columbia
Fred L. Bunnell1,4, Ralph W. Wells1, Bruce Harrison2, and Andre Breault3
1 Faculty of Forestry, University of British Columbia, 2424 Main Mall, Vancouver, BC V6T 1Z4; email: [email protected] Ducks Unlimited Canada, 954A Laval Crescent, Kamloops, BC V2C 5P53 A. Breault, Canadian Wildlife Service, 5421 Robertson Road, RR 1, Delta, BC V4K 3N24 Corresponding author
Abstract: Wetlands in the central interior of British Columbia are experiencing increased drying that is expected to continue;
small wetlands at low elevations are most vulnerable to drying. Different waterfowl species concentrate their use of
wetlands at different elevations and over different wetland size classes, producing differential vulnerability to climate
change. In the central interior of British Columbia, the species currently most abundant are also among the more vulnerable
species, due to their general preference for smaller wetlands at low to moderate elevations. Historically, efforts at wetland
conservation have focussed on low elevations that will be most impacted by climate change. To effectively confront
consequences of climate change and encompass the entire range of waterfowl species, wetland management should
ensure that management efforts address differential vulnerabilities of both wetlands and waterfowl. In central British
Columbia, mid-elevations appear significantly less vulnerable than low elevations and may present greater opportunities
for conservation through advantages in water security, costs of conservation actions and ecosystem integrity.
Keywords: British Columbia, climate change, conservation actions, waterfowl, wetland drying
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
28
Volume 23, 2013
for estimating waterfowl breeding populations is the ‘total
indicated breeding’ population (Smith 1995). The method is
appropriate for continent-wide surveys. For the study area,
this method was refined by the Canadian Wildlife Service
and Ducks Unlimited Canada to reduce recorded numbers
of transients migrating through. That estimate, ‘indicated
breeding pairs’ or ‘total indicated pairs’ (TIP), is believed
to better reflect local productivity and is the measure used
here. It has been converted to density by dividing by
wetland size. At least some scoters and scaup may have
been migratory. Two measures of affinity for different size
and elevation classes of wetlands were employed: 1) the
proportion of observations of a species in a class, 2) the
density of species in a class. The former is strongly influ-
enced by the relative abundance of different size and el-
evation classes; the latter much less so.
Total wetlands evaluated were 1,573 in 2006, 2,226 in 2007 and
2,212 in 2008; wetlands for which counts were zero were included
in analyses. Elevations were extracted from the most recent 1:20
000 TRIM data (Terrain Resource Information Management Pro-
gram; archive.ilmb.gov.bc.ca/crgb/pba/trim/). Wetland size was
extracted from the British Columbia Freshwater Atlas that includes
waterbodies <1 ha (http://ilmbwww.gov.bc.ca/geobc/FWA_data).
In some instances, waterfowl are grouped as ‘divers’ and ‘dab-
blers’. That grouping follows the distinction between Anatinae
and Athyinae but encompasses all species based on their pri-
mary mode of foraging.
Treatment of relative vulnerabilities employs impacts of
climate change as projected by Bunnell et al. (2010a). Valid-
ity of that approach was evaluated using water depths for 33
wetlands measured over 11 years (1997 to 2007) in the Cen-
tral Interior Ecoprovince and for two small lakes over 20 years
(1983 to 2005) in the Southern Interior Ecoprovince (Bunnell
et al. 2010b, 2011c). Two climate variables were employed in
the drying index: annual precipitation as snow and summer
heat-moisture index, a combination of summer temperature
and precipitation (Bunnell et al. 2011b). These variables were
chosen because they were expected to have the greatest
effect on the water balance of wetlands in the study area.
Annual snowfall was expected to provide a primary water
source (input), while the heat-moisture index was expected
to provide an indication of drying trends (output).
Assessments of vulnerability to climate change were based
on two principles derived from projections of the wetland dry-
ing index as evaluated against measured water depths (Bunnell
et al. 2011c). First, small wetlands tend to be less deep and dry
faster, losing a greater proportion of habitat than do larger
wetlands. Second, the greater snowpack at higher elevations
tends to slow rates of drying. Species preferring small wetlands
at low elevations are thus most vulnerable to drying. Most
current drying is seasonal that sometimes affects potential
productivity. With continued warming, more drying will be-
come permanent with significant effects on productivity.
Our analyses included eight size classes and four eleva-
tion classes. The simplest index of vulnerability assumes the
lowest classes are most vulnerable and the highest least
vulnerable when these are ranked from smallest to greatest
size or lowest to greatest elevation. For vulnerability, affinity
was based on the proportions of each species in different
classes, rather than densities by class, because it is the vul-
nerability of wetlands as determined by size and elevation
that is critical. The density response of species may shift as
the relative distributions of available wetland sizes and el-
evations shift. For each species or group of species, we
summed the rank of each class multiplied by the percentage
of observations in each class. The total was inverted and
normalized from 0 to 1 across all species to provide a relative
ranking. Inversion equates the highest value with the great-
est vulnerability. Separate rankings were calculated for
wetland size class and elevation. We also combined the two
individual rankings additively or multiplicatively to incorpo-
rate both effects. Additive combination averages the two
effects; multiplicative combination accommodates the likeli-
hood that extreme rankings of either size or elevation can
have more effect than the two added or averaged. Combined
rankings also were normalized.
Statistical tests were analyses of variance evaluating year,
elevation and wetland size effects on estimated waterfowl
density (Systat 2000).
Results
Waterfowl density, wetland size and elevation
Of the two measures of waterfowl density, ‘total indicated
pairs’ is illustrated because it is believed to be the better meas-
ure of productivity. The two measures followed similar pat-
terns, but total indicated breeding birds was higher because
16
Figure 1. Central Interior Ecoprovince of British Colum-
bia.
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
29
Volume 23, 2013
fewer transients were eliminated. As assessed by total indi-
cated pairs (or total indicated birds), waterfowl numbers and
density varied with year, wetland size and elevation. The dis-
tribution of wetlands surveyed is summarized in Table 1.
The area had 155,672 wetlands covering 697,389 ha (hec-
tares). The large majority of wetlands were <2 ha (76.6 %;
Table 1). These, however, contributed only 9.6% of the total
wetland area. The two largest wetland size classes (>20 ha)
represented only 3% of all wetlands in the region, but 64% of
the wetland area. Wetland size classes were distributed simi-
larly across elevation (Table 1) and showed no statistical
association with elevation. During the three years of sur-
veys, the numbers of indicated breeding pairs over all
Ecosections were: 244,312 in 2006, 221,717 in 2007 and 220,238
in 2008. Mallard (Anas platyrhyncos) was the most frequently
encountered species (29.9% of all observations identified to
species or as scaup), followed by Bufflehead (Bucephala
albeola;18.8%) and Ring-necked Duck (Aythya collaris;
15.4%).
Although tested by analysis of variance, the nature of
interactions is best illustrated graphically. The year effect on
estimated abundance or wetland productivity was signifi-
cant for those wetlands sampled in all years (p < 0.01), and
interacted strongly with elevation class (Figure 2). The inter-
annual effect reduced the productivity measures of lower
elevation wetlands disproportionately to its effect on higher
elevation wetlands. When all waterfowl species are com-
bined, density response with elevation varied from being
significantly higher at low elevations in 2006 to no response
in 2008 (Figure 2). We found no evidence that snowfall in the
preceding year was related to measured waterfowl density.
The aggregate tendency for waterfowl density to be greater
at lower elevations in 2006 and 2007 is a definite response to
elevation rather than wetland size. The response is contrary
to the relative availability; <30% of wetlands were below
1000 m and size of wetlands was similarly distributed across
elevation (Table 1).
There was no significant inter-annual interaction between
waterfowl density and size class of wetlands. Over all years,
breeding densities of aggregated dabbling and diving ducks
showed near identical responses to wetland sizes, both be-
ing more abundant in smaller wetlands, but dabblers were
more abundant than divers (Figure 3).
Some wetlands were markedly more productive than oth-
ers. We examined estimated waterfowl density as a function
of wetland size and elevation for the most productive
wetlands. Data are shown for 2008 that had the highest
number of highly productive wetlands (>50 ducks/ha; n =
21; Figure 4). Data for 2006 and 2007 were nearly identical in
form.
Across all three years, the most productive wetlands were
the smallest, typically <2 ha. Each year had 1 or 2 wetlands of
larger sizes that hosted >50 ducks/ha, but these productive
larger wetlands differed between years. In each year, aggre-
gate density peaked about 1000 m, declining above and be-
low that elevation (2008 is illustrated in Figure 4). That is a
function of the overall density of wetlands themselves, and
particularly smaller wetlands, being centered around 1000 m
(Table 1).
Waterfowl species’ responses
Twenty individual species were recorded; Greater and
Lesser Scaup (Aythya marila and A. affinis) were not con-
sistently distinguishable from the air, so were combined as
one ‘species’ in the surveys and analyses following. We
used two measures of apparent selection by waterfowl: 1)
concentration of use relative to availability of wetland size
and elevation classes, 2) relative density within different size
and elevation classes. Both are illustrated. Table 2 summa-
rizes the size and elevation class in which species were most
commonly observed; values are the percentage of all obser-
vations for a species recorded from that size or elevation
class. Trends away from the most frequent size and elevation
classes reflect apparent preference rather than a response to
Table 1. Distribution of wetlands by size and elevation class in the central interior of British Columbia.
Elevation class
Size class (ha) 0-500m 500-1000m 1000-1500m 1500-2000 m Total Per cent
0-1 69 24187 58990 14195 97441 62.6
1-2 8 6845 13014 1870 21737 14.0
2-3 5 3248 5573 702 9528 6.1
3-5 5 3300 5379 619 9303 6.0
5-10 2 3078 4677 480 8237 5.3
10-20 12 1820 2636 279 4747 3.0
20-50 5 1179 1721 148 3053 2.0
50+ 2 645 920 59 1626 1.1
Total 108 44302 92910 18352 155672 100.0
Per cent <0.1 28.5 59.7 11.8 100.0
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
30
Volume 23, 2013
availability. Because wetlands were similarly distributed across
elevation belts (Table 1), apparent preferences for a particular
size or elevation class are independent of each other.
Wetlands 0–1 ha in size were the most frequent size class,
comprising 62.6 % of all wetlands (Table 1). The most com-
monly preferred elevation class across all species was 500 to
1000 m, followed by 1000 to 1500 m. Wetlands in these eleva-
tion classes comprised 28.5% and 59.7% of all wetlands, re-
spectively (Table 1). All 9 dabblers were observed more fre-
quently in the smallest size class; 7 of the 12 divers were re-
corded more commonly in the smallest size class, but other
species were recorded most often in larger size classes (Table
2). Among dabblers, use approximated availability. There was
little apparent preference for wetland size class, though Blue-
winged Teal (Anas discors) and Cinnamon Teal (Anas
cyanoptera) occurred in the smallest size class less than ex-
pected from availability. Most dabbler species used the eleva-
tion class 500 to 1000 m about twice as much as expected on
availability alone; occurrence in the elevation class 1000 to1500
m was as expected from relative availability of wetlands.
Departures from relative availability were stronger among
Figure 2. Waterfowl density of wetlands as a function of
elevation class in the Central Interior Ecoprovince of Brit-
ish Columbia (total indicated pairs, ducks / ha, mean +
SE). a) 2006 (n = 1573 wetlands), b) 2007 (n = 2226
wetlands, c) 2008 (n = 2212 wetlands). Elevation classes
are: 4 = 400 to 499 m, 5 = 500 to 599 m, 6 = 600 to 699
m, etc. Transect survey data of the Canadian Wildlife
Service and Ducks Unlimited Canada, 2006 to 2008.
Figure 3. Density of dabbling and diving duck species
on wetlands of different size classes in the Central Inte-
rior Ecoprovince of British Columbia (total indicated
pairs, ducks / ha, mean + SE). Transect survey data of
the Canadian Wildlife Service and Ducks Unlimited
Canada, 2006 to 2008.
a)
b)
c)
c)
17
4 5 6 7 8 9 10 11 12 13 14 15 16 17
0
2
4
6
8
10
Elevation Class
Ducks / ha
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
0
2
4
6
8
10
Elevation Class
Ducks / ha
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
0
2
4
6
8
10
Elevation Class
Ducks / ha
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
31
Volume 23, 2013
divers. Any species not occurring most commonly in the
smallest wetland size class is exhibiting preference for larger
size classes; 5 of the 12 species do so. Some of those that
occur more commonly in the smallest wetland size class do
so at values well below that expected from availability alone
(e.g. 49.5% versus 62.6% for Ruddy Duck). Whatever eleva-
tion class divers used most commonly, the value generally
was well above that based on relative availability of wetlands
(Table 2), suggesting strong preference. White-winged Scoter
(Melanitta fusca) was recorded in 2007 from a single wetland
>50 ha in the 500 to 1000 m elevation belt; Surf Scoter
(Melanitta perspicillata) was recorded in 2008 from a differ-
ent wetland with the same attributes. Breeding concentra-
tions of these species are well to the north and most are very
likely transients in the study area.
Relative densities of waterfowl in different size and el-
evation classes of wetlands are illustrated for selected spe-
cies in Figure 5 and 6. For most species, the density over all
wetlands, including all zero values, is much lower than that
for Mallard, but many show different responses. Species
Figure 4. Relations of
productivity in highly
productive wetlands
(>50 ducks / ha) with
a) wetland area or
size and b) elevation
for the Central Interior
Ecoprovince of British
Columbia in 2008.
Table 2. Observations (%) of waterfowl species in the most frequently occupied wetland size and elevation class and
availability (%) of that class in the central interior of British Columbia 2006 through 2008.
Most Frequent Size Class Most Frequent Elevation Class
Species Class (ha) % Obs % Avail Class (m) % Obs % Avail
Dabblers
American Wigeon 0-1 64.9 62.6 1000-1500 55.4 59.7
Blue-winged Teal 0-1 48.3 62.6 500-1000 57.2 28.5
Canada Goose 0-1 51.3 62.6 500-1000 50.6 28.5
Cinnamon Teal 0-1 41.5 62.6 500-1000 52.9 28.5
Gadwall 0-1 55.1 62.6 500-1000 62.0 28.5
Green-winged Teal 0-1 73.7 62.6 500-1000 51.0 28.5
Mallard 0-1 72.5 62.6 500-1000 50.4 28.5
Northern Pintail 0-1 57.4 62.6 1000-1500 61.0 59.7
Northern Shoveler 0-1 62.2 62.6 500-1000 58.2 28.5
Divers
Barrow's Goldeneye 0-1 62.2 62.6 500-1000 60.0 28.5
Bufflehead 0-1 66.6 62.6 500-1000 59.3 28.5
Canvasback 0-1 71.0 62.6 500-1000 94.8 28.5
Common Goldeneye 3-5 68.1 6.0 500-1000 100.0 28.5
Common Merganser 50+ 61.6 1.1 500-1000 86.3 28.5
Hooded Merganser 0-1 66.5 62.6 500-1000 74.3 28.5
Redhead 5-10 44.4 5.3 500-1000 84.4 28.5
Ring-necked Duck 0-1 55.6 62.6 500-1000 63.1 28.5
Ruddy Duck 0-1 49.5 62.6 500-1000 91.1 28.5
Scaup 0-1 54.8 62.6 1000-1500 61.1 59.7
Surf Scoter 50+ 100.0 1.1 500-1000 100.0 28.5
White-winged Scoter 50+ 100.0 1.1 500-1000 100.0 28.5
19
0 500 1000 1500 2000
Elevation (m)
0
100
200
300
400
Ducks/Ha
0 10 20 30
Wetland Area (ha)
0
100
200
300
400
Ducks/Ha
a) b)
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
32
Volume 23, 2013
Figure 5. Distributions of observed densities by wetland size class for selected waterfowl species in the Central Interior
Ecoprovince of British Columbia, 2006 to 2008. Note that total observations differed greatly among species; the density
scales differ across species.
Wetland size class (ha)Wetland size class (ha)
illustrated were selected to expose the variability in apparent
choice. Several species most commonly observed in the
smallest wetlands (<1 ha in size), including Mallard (the most
abundant species) and Green-winged Teal (Anas crecca),
also attained their highest densities in the smallest wetlands.
Species showing similar responses included American
Wigeon (Anas americana), Barrow’s Goldeneye (Bucephala
islandica), Bufflehead, Canada Goose (Branta candensis),
Gadwall (Anas strepera), Hooded Merganser (Lophodytes
cucullatus), Northern Shoveler (Anas clypeata), Ring-necked
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
33
Volume 23, 2013
Duck and scaup (greater and lesser combined). For other
species, including Blue-winged Teal, Cinnamon Teal,
Canvasback (Aythya valisineria), Northern Pintail (Anas
acuta) and Ruddy Duck (Oxyura jamaicensis), reasonably
high densities were attained over a broader range of size
classes, up to 3 ha in size (Figure 5). A third group of species
including Common Merganser (Mergus merganser), Com-
mon Goldeneye (Bucephala clangula), Surf Scoter and
White-winged Scoter were largely limited to the largest
wetlands during the survey period (Figure 5; Table 2).
There were four broad patterns of association with el-
evation. Mallard was most dense at the lowest elevations, as
was Northern Shoveler (Figure 6) and to a lesser extent Ring-
necked Duck. For a second group, including Blue-winged
Teal, Canvasback, Cinnamon Teal, Common Merganser,
Gadwall, Hooded Merganser, Redhead (Aythya americana),
Ruddy Duck and scaup, highest densities were concentrated
or limited to elevations of 500 to 1500 m. A third group showed
a tendency towards higher densities at higher elevations,
including Barrow’s Goldeneye, Canada Goose, Northern
Pintail and, to a lesser degree, Bufflehead and Green-winged
Teal. The remaining small group included species recorded
from only a single elevation class, including Common
Goldeneye, Surf Scoter and White-winged Scoter. In all cases
they were observed only in wetlands from 500 to 1000 m in
elevation that comprised 28.5% of all wetlands.
The two broad measures (frequency of use and relative
density) are independent—a species can express high or low
density in a particular size or elevation class of wetland, re-
gardless of how common that size or age class is. For example,
the numbers of Northern Pintail using different wetland size
and elevation classes appear to follow relative availability (el-
evation class 1000 to 1500 m represented 59.7% of wetlands
and 61% of pintails were found in that class; Table 2). Al-
though most of the Northern Pintail population was found at
1000 to1500 m, the highest densities occurred in wetlands at
1500 to 2000 m (Figure 6). Across species, there were few con-
sistent associations between the most commonly or densely
used wetland size and elevation, but the dabbling ducks most
common in the smallest wetland sizes were usually at the low-
est elevations as well (Table 2; Figure 6).
Relative vulnerability
Wetland size and elevation influence the relative vulner-
ability to drying. Size because smaller wetlands generally are
less deep and elevation because higher elevations receive
more snow fall to replenish moisture than do lower eleva-
tions (Bunnell et al. 2010b, 2011c). Four indices of relative
Normalized relative vulnerability
Size class Elevation Class Multiplicative Additive
Dabblers
Northern Pintail 0.900 0.000 0.000 0.000
Northern Shoveler 0.948 0.595 0.652 0.669
Cinnamon Teal 0.882 0.523 0.534 0.526
Gadwall 0.914 0.615 0.650 0.655
Canada Goose 0.886 0.488 0.500 0.493
Blue-winged Teal 0.877 0.570 0.577 0.569
American Wigeon 0.961 0.434 0.482 0.516
Green-winged Teal 1.000 0.494 0.571 0.618
Mallard 0.987 0.534 0.609 0.646
Mean 0.928 0.473 0.508 0.521
Divers
Ruddy Duck 0.887 0.910 0.934 0.934
Canvasback 0.913 0.948 1.000 1.000
Redhead 0.709 0.843 0.691 0.679
Scaup 0.888 0.385 0.395 0.388
Hooded Merganser 0.984 0.741 0.843 0.859
Common Goldeneye 0.481 1.000 0.557 0.605
Common Merganser 0.253 0.863 0.252 0.224
Barrow’s Goldeneye 0.936 0.589 0.638 0.651
Ring-necked Duck 0.927 0.633 0.678 0.687
Bufflehead 0.980 0.586 0.664 0.694
Surf Scoter 0.000 1.000 0.000 0.104
White-winged scoter 0.000 1.000 0.000 0.104
Mean 0.663 0.791 0.554 0.577
Table 3. Relative vulnerability of waterfowl species as conferred by wetland size, elevation and both combined in the
central interior of British Columbia; normalized 0 (least vulnerable) to 1 (most vulnerable) across all species.
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
34
Volume 23, 2013
Figure 6. Distributions of observed densities by elevation class for selected waterfowl species in the Central
Interior Ecoprovince of British Columbia, 2006 to 2008. Note that total observations differed greatly among spe-
cies; the density scales differ across species.
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
35
Volume 23, 2013
vulnerability were derived for the study area: 1) size class
effects, 2) elevation effects, 3) combined effects additive,
and 4) combined effects multiplicative. In each case, smaller
and lower wetlands were assumed more vulnerable to drying
than larger or higher wetlands (see methods). Derived indi-
ces are reported by species in Table 3.
Vulnerability indices do not mimic the patterns of den-
sity because vulnerability reflects the physical distribution
of wetlands (Table 1) plus aggregate waterfowl response to
that distribution (e.g. Table 2). Realized densities within the
distribution classes may be a better reflection of preference,
but not of use. In terms of effects of size class, divers’ more
common use of larger size classes made them much less vul-
nerable to climate change on average (Table 3). As a group,
divers were generally more vulnerable to the effects of el-
evation and potentially diminished snowpack. Indices ranged
widely among diver species and part of their greater vulner-
ability to elevation is due to three species (Common
Goldeneye, Surf Scoter and White-winged Scoter) observed
at very low frequency and restricted to 500 to 1000 m eleva-
tion. Eliminating those species yields an average vulnerabil-
ity of 0.722 based on elevation, still markedly higher than for
dabblers when averaged across species.
Whether the two main effects were combined additively
or multiplicatively had little effect; average combined indi-
ces for the two groups were about equal. Both elevation and
size class were influential, but the dominant effect was that
of dabblers occurring more frequently in smaller wetlands,
thus making them more vulnerable. Removing the two scoter
species observed in only one year, changed the averages to
0.435 and 0.465 for multiplicative and additive, respectively.
Discussion
Waterfowl density, wetland size and elevation
Although some species attained highest densities at
higher elevations (Figure 6), size of wetlands had the greater
effect on estimated density or productivity, with smaller
wetlands generally being the most productive (Figure 3).
There are two potential reasons for small wetlands to have
higher densities of waterfowl, one geometric and the other
ecological. The geometric reason is that if waterfowl are con-
centrated at the margins or perimeter of wetlands, the esti-
mate of ducks/ha necessarily will be larger for smaller
wetlands. The ecological reason is that smaller wetlands
should exhibit greater primary and secondary productivity,
thus more forage. That follows from the sample of 33 wetlands
from the Central Interior Ecoprovince analyzed by Bunnell et
al. (2011c). They found that size and water depth were sig-
nificantly correlated, with water depth increasing with wetland
size. Shallower wetlands will be warmer, encouraging greater
primary and secondary productivity, including waterfowl.
We expect both reasons to be acting.
The distribution of wetland sizes over elevation was
broadly similar, with all size classes being best represented
from 1000 to 1500 m followed by 500 to 1000 m (Table 2).
Because wetlands are similarly distributed across elevation,
the year effect is most apparent with elevation (Figure 2) and
manifests itself as a strong decrease in waterfowl density in
wetlands below 1000 m elevation in 2008. Densities at higher
elevations remained largely unchanged. Both realized and
projected climate effects on wetland drying are more pro-
nounced at lower elevations (Dawson et al. 2008, Bunnell et
al. 2011c, Werner 2011). If the year effect is a response to
regional climate, it is worrisome that this is evident in sur-
veys conducted in May because that implies still greater
drying as the season progresses. Although worrisome, it is
not surprising. Several species of waterfowl already had ex-
tended their breeding ranges significantly northwards be-
tween the 1960s and 1990s (Bunnell et al. 2008, 2013). Data
illustrated here may simply expose dynamics of a much
broader, longer-term process.
Waterfowl species’ responses
Table 2 summarizes the availability of wetlands by size
and elevation class, plus the proportion of total numbers
observed in those classes by species. If species were re-
sponding directly to wetland availability, they would be most
frequently observed in the smallest size class (62.6%
wetlands) and between elevations of 500 to 1500 m (88.2% of
wetlands; Table 2). That is generally true for dabblers; for
divers, about half of the species were more frequently ob-
served in larger wetlands. Apparent preference for larger
wetlands may simply reflect that within the study area, larger
wetlands tend to be deeper (Bunnell et al. 2011c) and likely
provide better habitat for typical diver foods, such as aquatic
macro-invertebrates and fish. There was no apparent differ-
ence between all dabblers and all divers in the elevation
class in which they were most frequently recorded (Table 2).
The restriction of three species (Common Goldeneye, Surf
Scoter and White-winged Scoter) to 500 to 1000 m (Table 2)
could simply be happenstance; none were common in the
surveys. Unlike Common Goldeneye, both scoters were re-
stricted to the largest wetland size class. Although White-
winged Scoter breeds in the area, it is an uncommon breeder
and both scoter species are likely to be transient moving
northward. That also is suggested by large groups in a sin-
gle wetland.
Density and frequency of occurrence show somewhat
different patterns across size and elevation classes. The geo-
metric effect enhances relative density in the smaller wetland
classes, but across all species, we find some attaining their
highest densities at either end of the range in size or eleva-
tion (Figures 5 and 6). A species’ vulnerability to effects of
climate change is more directly related to the attributes of
individual wetlands than to current density, particularly
wetland size and elevation. We based relative vulnerability
of the species on the frequency at which they were observed
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
36
Volume 23, 2013
in different size and elevation classes rather than on density.
Availability is largely fixed while relative density is a flexible
response to availability. For many species the two measures
are closely similar.
Relative vulnerability
Where topography is rugged and long-term weather sta-
tions are sparse, as in much of British Columbia, it is difficult
to project changes in climate with confidence (e.g. Hamann
and Wang 2005, Mbogga et al. 2010). Tests of the drying
index by Bunnell et al. (2010b,2011c) provide confidence in
the direction of change, but not the rate. The indices of rela-
tive vulnerability used here are related to broad physical
features for which we have confidence in their effect and
that are not dependent on rate of climate change.
Apparent affinities by species for particular size and el-
evation classes of wetlands make them differentially vulner-
able to climate change. Small wetlands at lower elevations
are particularly likely to dry up (Bunnell et al. 2010b, 2011c).
Moreover, warming and drying trends are expected to con-
tinue in the region (Dawson et al. 2008, Mbogga et al. 2010,
Werner 2011). Bunnell et al. (2011b) observed that “For
wetland species, management will struggle with the concept
of a real-world triage – allocating conservation efforts where
they are most likely to succeed and have the most benefit.”
They found that climate is creating a natural triage for
wetlands in British Columbia. Some regions will have too
little water to sustain smaller wetlands no matter what is
done; wetlands in other regions will not be strongly impacted
for decades; through action now, still other wetlands can be
maintained for a period during which some conditions may
change for the better. The invocation of triage in assigning
conservation effort is exactly analogous to that of a war-time
medic. Like the medic, we can attempt to reduce effects of
the wounds, and focus our efforts and limited resources
where they will achieve their greatest gain. Findings here
suggest that the triage is not consistent across species, and
some species (e.g. Canvasback, Ruddy Duck and Hooded
Merganser; Table 3) are particularly vulnerable.
Generally, the nature and distribution of wetlands in the
area augur poorly for species preferring small, low-elevation
wetlands. A substantial majority (62.6%) of wetlands are in
the smallest size class; about 28.5% of elevations are below
1000 m elevation (Table 1). During testing of the drying in-
dex, using empirical depth measurements, Bunnell et al.
(2011c) found that, on average, wetlands <2 ha in area lost
16% more water depth than larger wetlands. Under projected
climate change scenarios, the greatest impacts of drying
occurred at the lowest elevations where temperatures were
greater and snowpack was least (Bunnell 2010a, Mbogga et
al. 2010, Werner 2011).
We expected the multiplicative form of the combined index
to show a broader range of impact than the additive form. For
example, if normalized vulnerability indices for size and eleva-
tion were each 0.5, the additive index would be 0.5 and the
multiplicative index 0.25 before normalizing. For specific
wetlands, the multiplicative form is more revealing (Bunnell et
al. 2011c) and that appears broadly true of species as well. For
example, Northern Pintail and Green-winged Teal are highly
vulnerable based on their preference for small wetlands, but
this vulnerability is greatly reduced by their use of higher
wetlands. Addressing size and elevation effects separately
(Table 3) allows evaluation of likely outcomes of different re-
gional climates and distributions of wetlands.
Relative use and relative density reveal plasticity within
species. For example, Northern Pintail occurs most commonly
at elevations of 1000 to 1500 m (61% of observations), but
observations at 1500 to 2000 m (19% of total observations)
yielded densities about 4 times higher than densities at 1000
to 1500 m (Figure 6). The fact that realized densities do not
necessarily follow the same pattern as the frequency at which
the species occurs in wetlands of different sizes or eleva-
tions illustrates plasticity in wetland use (compare Figures 5
and 6 with Table 2). Moreover, Bunnell et al. (2008, 2013)
documented significant, and sometimes dramatic, shifts in
range, relative density, arrival and departure times, amount
of overwintering and reproductive measures among
waterbirds in British Columbia during the ‘climate normal’
period of the Intergovernmental Panel on Climate Change
(1961 to 1990).
The size and elevation of a wetland determines the vul-
nerability of that wetland to climate change. The flexibility of
the species determines the impact of the loss of wetlands
from a particular size or elevation class. Findings here illus-
trate both flexibility and constraints. For example, Northern
Pintail occurs most frequently in wetlands 1000 to 1500 m
elevation that are more susceptible to drying than are higher
elevations. Northern Pintail attains markedly higher densi-
ties in wetlands at 1500 to 2000 m elevation, but those
wetlands comprise <12% of all wetlands and only 19% of
pintail observations already occur there. Northern Pintail is
certainly flexible enough to exploit elevations higher than
those it uses most commonly and may even prefer them, but
its opportunities for using them are limited.
Historically, conservation efforts have concentrated on
lower elevations (<1000 m) that comprise less than 30% of
wetlands in the region (Table 1), but are the most frequently
used (Table 2). In the region, use patterns may already be
changing in response to climate (compare 2006 and 2008 in
Figure 2). Small wetlands appear to be the most favoured
and the highest waterfowl densities occur at elevations of
800 to 1200 m (Figure 4). The most favourable elevation is
likely to rise with increased drying. Mid-elevations currently
exhibit high waterfowl productivity (Figure 4), moderate dry-
ing (Bunnell et al. 2010a), good potential to intercept and
store water (snowpack is still present) and intact forest cover
to potentially buffer wetlands. They also experience less
demand for water than do lower elevations. Conversely, there
is less opportunity for successful conservation of small
wetlands at lower elevations where there is little snowpack
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
37
Volume 23, 2013
and drought already is frequently expressed.
Lower elevation wetlands are more subject to drying than
are higher elevation wetlands, so waterfowl use is likely to
shift towards higher elevations than those used presently. A
shift in conservation effort to higher elevations is not as
expensive as it would be if concentrated in low-lying, well-
populated areas, but is potentially simplistic. The response
of waterfowl density (Figures 5 and 6) and wide range in
relative vulnerability (Table 3) suggests that conservation
efforts must somehow encompass a diversity of areas.
Most wetlands in the region, and many elsewhere, re-
ceive much of their input of water as ground water or from
small streams. Current regulations in British Columbia pro-
vide for no buffers around smaller, non-fish-bearing streams
or around wetlands. The dramatic outbreak of mountain pine
beetle (Dendroctonus ponderosae) in the province focused
both retrospective and new studies on the impacts of
streamside forestry on amounts and temperature (thus evapo-
ration) of water. Review of those studies suggests that re-
gardless of elevation, buffering of small streams would have
beneficial effects on sustaining wetlands (Bunnell et al.
2011a).
Conclusions
Findings here reveal that wetland conservation efforts
cannot adopt a ‘one size fits all’ approach, even within lim-
ited areas. They illustrate that some waterfowl species ap-
pear far more vulnerable to climate change than do others,
but that there is considerable flexibility within species (Fig-
ures 5 and 6). Given the apparent inter-annual variation in
effects of weather on wetland productivity, and the apparent
variability within species, data over three years are insuffi-
cient to provide detailed guidance. They do indicate that a
variety of conservation measures are likely necessary to
maintain the entire diversity of waterfowl species.
Acknowledgements
The BC Forest Sciences program, Ducks Unlimited
Canada and Environment Canada provided support. We
thank Neil Bourne and Andy Buhler for helpful comments.
Literature cited
Breault, A., B. Harrison, D. Kroeker, S. Shisko and P. Watts.
2007. Waterfowl breeding population survey of the
Central Interior Plateau of British Columbia. Canadian
Wildlife Service Report. Delta, B.C.
Bunnell, F.L., M.I. Preston and A.C.M. Farr. 2008. Avian
response to climate change in British Columbia –
toward a general model. p. 9–27 in F. Dallmeier, A.
Fenech, D. MacIver and R. Szaro (eds.). Climate
change, biodiversityand sustainability in the
Americas. Smithsonian Institution Scholarly Press,
Washington, DC.
Bunnell, F.L., R. Wells, A. Moy, A. Breault and B. Harrison.
2010a. Vulnerability of wetlands in the Central Interior
Ecoprovince to climate change. Report to Canadian
Wildlife Service, Delta, B.C.
Bunnell, F.L., A. Moy and T. Northcote, 2010b. Evaluating
the drying index for the Southern Interior Ecoprovince.
Report to Canadian Wildlife Service, Delta, B.C.
Bunnell, F.L., L.L. Kremsater and I.Houde. 2011a. Mountain
pine beetle: A synthesis of the ecological consequences
of large-scale disturbances on sustainable forest
management, with emphasis on biodiversity.
Information Report BC-X-426. Canadian Forest Service,
Pacific Forestry Centre, Victoria, B.C.
Bunnell, F.L., L.L. Kremsater and R.W. Wells. 2011b. Global
weirding in British Columbia – climate change and the
habitat of terrestrial vertebrates. BC Journal of
Ecosystems and Management 12(2):21–38. http://
jem.forrex.org/index.php/jem/article/view/74/81.
Accessed 3 March 2012.
Bunnell, F.L., A. Moy, R. Wells and A. Breault. 2011c. Using
measured wetland depths to evaluate climate
influences on wetlands. Canadian Wildlife Service
Technical Report. Delta, B.C. (in press).
Bunnell, F.L., A. Moy, M. I. Preston and R. W. Wells. 2013.
Bird distribution and climate change in British Columbia.
British Columbia Birds 23: in press.
Dawson, R.J., A.T. Werner and T.Q. Murdock. 2008. Preliminary
analysis of climate change in the Cariboo-Chilcotin area
of British Columbia. Pacific Climate Impacts Consortium
report. University of Victoria. Victoria, B.C.
Hamann, A. and T.L. Wang. 2005. Models of climatic normals
for genecology and climate change studies in British
Columbia. Agricultural and Forest Meteorology
128:211-221.
Meidinger, D. and J. Pojar (compilers and editors). 1991.
Ecosystems of British Columbia. Special Report Series
No. 6. B. C. Ministry of Forests, Research Branch,
Victoria, B.C. http://www.for.gov.bc.ca/hfd/pubs/Docs/
Srs/Srs06/front.pdf . Accessed 15 September 2011.
Mbogga, M.S., X. Wang and A. Hamann. 2010. Bioclimate
envelope model predictions for natural resource
management predictions: dealing with uncertainty.
Journal of Applied Ecology 47:731-740.
Mote, P.W. 2003. Trends in temperature and precipitation in
the Pacific Northwest during the twentieth century.
Northwest Science 77:271-282.
Smith, G.W. 1995. A critical review of the aerial and ground
surveys of breeding waterfowl in North America.
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
38
Volume 23, 2013
Biological Science Report 5, US Department of the
Interior, Washington, D.C.
Taylor, B. 2005. Climate change and variability. p. 4-12 in
Implications of Climate Change in British Columbia’s
southern Interior Forests. Columbia Mountains Institute of
Applied Ecology. http://www.env.gov.bc.ca/cas/pdfs/
impact_wshp_sforest_bc.pdf . Accessed 15 September 2011.
Systat. 2000. Systat 10 for Microsoft Windows. SPSS
Inc. Chicago, Ill.
Werner, A.T. 2011. BCSD downscaled transient
climate projections for eight select GCMs over
Bri t i sh Columbia, Canada . Paci f ic Cl imate
Impacts Consortium, Universi ty of Victoria,
Victoria, B.C.
Waterfowl & Climate Change - Bunnell et al.
British Columbia Birds
39
Volume 23, 2013
Bird observations by Dr. J.E.H. Kelso in the West Kootenay area of
British Columbia, 1913–1932
Bill Merilees
3205 Granite Park, Nanaimo, B.C., V9T 3C8 [email protected]
Abstract: The contribution of Dr. J.E.H. Kelso to the early understanding of the bird fauna of the Lower Arrow Lake area of
the West Kootenay region is presented for the period of his residence, 1913 to 1932. In addition to his published account
(Kelso 1926), additional observations and information contained in an untitled, unpublished manuscript are presented.
From these materials 89 new species are documented as occurring in the West Kootenay. When added to the observations
of John Macoun and party in 1890, and to those of William Spreadborough in 1902, the known West Kootenay bird fauna
was 186 species as of 1932.
Key Words: J.E.H. Kelso, historic bird records, West Kootenay, British Columbia.
Dr. John Edward Harry Kelso (Figure 1) arrived in
Edgewood, on Lower Arrow Lake, in the spring of 1913. He
was born in Madras, India, received his M.D. from Edinburgh
University and practised
medicine in India, Mo-
rocco, England and
Scotland before accept-
ing the position as
Medical Health Officer at
Edgewood, B.C.
(Anonymous 1932a)
From early boyhood
his hobby was orni-
thology, an interest that
continued right up to
his death in August of
1932 (Anonymous
1932b). Before arriving
in Canada he had writ-
ten one book and many
articles on birds and was
a Member of the prestigious British Ornithologist’s Union.
In the Preface to his book Notes on Some Common and Rare
British Birds (Kelso 1912), Kelso stated that he strongly
believed in publishing the results of his bird observations
“in order to show what a vast amount of instruction and
recreation can be got from this pursuit, even if an observer
be placed in unfavourable surroundings”. At Edgewood he
immediately began recording bird observations and build-
ing a collection of bird skins, bird nests and photographs.
His B.C. bird skin collection, numbering 400 specimens of
147 species, along with a small number of nests (Anony-
mous 1933, James 1976) was given to the Royal Ontario
Museum through Allan Brooks, Sr. (Anonymous 1933). In
addition, according to Mrs. K. Johnson (Johnson 1951), “Part
of his fine collection of mounted birds may be seen in the
Jordan ranch home” (near Edgewood, B.C.). It is unknown to
me where or whether these specimens still exist.
In 1926 Kelso published Birds of the Arrow Lakes, West
Kootenay District, British Columbia (Kelso 1926). In order to
substantiate his findings, “Major Allan Brooks and Mr. F.
Kermode, Curator of the Victoria Museum [sic], have very kindly
named my specimens when there was any doubt as to their iden-
tity” (Kelso 1926). The order and nomenclature followed the
American Ornithologist’s Union (AOU) Checklist in vogue as of
1925.
This publication is a summation of his information gath-
ered up to 1925. It mentions 187 bird species. Of these, 69
had been reported earlier for the West Kootenay area by
Macoun (Merilees 2011) and a further 20 by Spreadborough
(Merilees 2012). Of the remaining 98 species, 62 were docu-
mented by specimens, nine were documented by dates and
locations and 27 were mentioned without specific support-
ing details (see Kelso 1926).
In addition to the data published in 1926, in 1925 Kelso
began writing a more detailed account of the Edgewood bird
fauna. At the time of his death, this tome had reached 290
single-spaced typewritten pages covering the species, listed
in AOU Checklist order. Beginning with Western Grebe, it
concludes, but was not completed, with American Goldfinch.
Figure 1. Dr. J.E.H. Kelso.
Kelso bird observations - Merilees
British Columbia Birds
40
Volume 23, 2013
This manuscript, to which I give the title “Kelso’s Bird Notes,
1913 to 1932” (Kelso 1932), contains a wealth of additional
information on occurrence, nesting and natural history. De-
tails, adequately documenting a further 11 species in Kelso
(1926), were located here, leaving the following 15 species
without specific dates or location documentation (names
used by Kelso where they differed from the present AOU
Checklist [2011] are indicated in parenthesis):
• Jaeger sp. (Stercorarius sp.)
• Gadwall
• Greater Scaup (Marila marila)
• Northern Bobwhite (Bob white [sic])
• Gray Partridge (Hungarian Partridge)
• Ptarmigan sp. (Lagopus subsp?)
• Swainson’s Hawk (Swainson Hawk)
• Peregrine Falcon (Duck Hawk)
• Western Screech Owl (MacFarlane Screech Owl)
• Snowy Owl
• Boreal Owl (Richardson Owl)
• Black-chinned Hummingbird
• Chestnut-backed Chickadee
• Orange-crowned Warbler
• Townsend’s Warbler (Townsend Warbler)
Hand written on the last page of Kelso’s personal reprint
of his 1926 publication, under the title “Additional Speci-
mens Obtained”, are listed Barn Swallow and Bullock Oriole.
These species are neither listed as being among his speci-
mens (Anonymous 1933) nor in his “Bird Notes” manuscript
(Kelso 1932). These two species are therefore considered
insufficiently documented to be accepted. As a result, 82
species, based on the information presented in Kelso (1926,
1932) are considered sufficiently well documented to be
added to the known West Kootenay bird fauna of that time.
With publication however, Dr. Kelso did not stop his birding
activities. From 1926 to 1932, the following seven species are
listed in his specimen collection inventory (Anonymous
1933) or in his “Bird Notes” (Kelso 1932) that are additions
to those in his publication:
• Long-tailed Duck or Old Squaw – page 66: “I know of
only one instance of these ducks being seen on the
[Arrow] Lakes. Mr. Colgrave with a right and a left
killed amale and female together. There is no doubt
about the identity of these birds.”
• American Golden Plover - specimen - August 12, 1929
• California Gull - two specimens - October 1, 1928 and
May 17, 1931
• Short-eared Owl - specimen - December 3, 1926
• Bobolink - specimen - June 30, 1926
• Red-winged Blackbird - specimens - March 24, 1928
and June 27, 1929
In total, Dr. Kelso’s activities documented 89 species new
to the West Kootenay area, bringing the total to 186 species
as of 1932.
Acknowledgements
Thanks to: Audrey Viken; Dr. R.D. James, Royal Ontario
Museum, Toronto; Jack & Madeline Eselmont; Leslie
Kennes & Mike McNall, Royal B.C. Museum, Victoria;
Michele Gosselin, Canadian Museum of Nature, Ottawa;
Andy Buhler and Art Martell for their assistance during the
preparation and review of this publication.
Literature Cited
American Ornithologist’s Union. 2011. Checklist of North
American Birds. http://www.aou.org/checklist/north/
full.php.
*Anonymous. 1932a. Death Certificate of John Edward Harry
Kelso. Province of British Columbia, B.C. Archives
Microfilm Number B13144.
*Anonymous. 1932b. Obituary, Dr. J.E.A. [sic]. Kelso, Arrow
Lakes News, August 11th, 1932, p.1.
*Anonymous. 1933. The Dr. J.E.H. Kelso [Bird] Collection.
Donor Mrs. Kelso thru Allan Brooks. Royal Ontario
Museum, Toronto. 11p., unpublished.
James, R.D. 1976. personal correspondence, Royal Ontario,
Museum, Toronto.
Johnson, K. 1951. Pioneer Days of Nakusp and the Arrow
Lakes, to commemorate Nakusp’s Diamond Jubilee, 1892-
1952. Self-published.
Kelso, J.E.H. 1912. Notes on some common and rare British birds.
J. & J. Bennett Ltd., Century Press, London, U.K. 420p.
*Kelso, J.E.H. 1926. Birds of the Arrow Lakes, West Kootenay
District, British Columbia. The Ibis 68:689-723, Plate XIV.
(Note: Kelso’s personal, annotated copy)
*Kelso, J.E.H. 1932. “Kelso’s Bird Notes, 1913 to 1932”.
Unfinished manuscript.
Merilees, B. 2011. Bird observations of John Macoun and
party in the West Kootenay area of British Columbia.
June - July, 1890. British Columbia Birds 21:2-8.
Merilees, B. 2012. Bird observations by William
Spreadborough in the West Kootenay area of British
Columbia, May – June, 1902. British Columbia Birds
22:5-7.
* Copies of these references are being donated to the
Library at Selkirk College, Castlegar, B.C.
Kelso bird observations - Merilees
British Columbia Birds
41
Volume 23, 2013
Changes in the abundance of wintering waterbirds along the
shoreline of Stanley Park, Vancouver, British Columbia, between
2001/2002 and 2010/2011.
Robyn Worcester
Stanley Park Ecology Society, PO Box 5167, Vancouver, B.C. V6B 4B2; e-mail: [email protected]
Abstract: Waterbird counts were conducted weekly from October through April in 2001/2002 and 2010/2011 along the
seawall of Stanley Park. Overall abundance and peak numbers of most groups of waterbirds were lower in 2010/2011 than
in 2001/2002. The difference was most dramatic for loons, grebes including Western Grebe (Aechmophorus occidentalis),
and Pigeon Guillemot (Cepphus columba). Barrow’s Goldeneye (Bucephala islandica) also showed a decrease in numbers
between years but Surf Scoter (Melanitta perspicillata) did not. The changes observed at Stanley Park are, in general,
consistent with those observed in the Strait of Georgia over the same period. This study confirms local knowledge held by
naturalists and bird watchers, that there have been declines in many species of wintering waterbirds using Stanley Park’s
marine habitat. The changes in abundance of many wintering waterbirds indicate that the ecological integrity of this IBA may
be threatened and that habitat conservation and measures to reduce human disturbance are warranted.
Key Words: Stanley Park, waterbirds, English Bay & Burrard Inlet Important Bird Area, Surf Scoter, Melanitta perspicillata,
Barrow’s Goldeneye, Bucephala islandica, Western Grebe, Aechmophorus occidentalis, Pigeon Guillemot, Cepphus
columba.
Introduction
Stanley Park is a 405 ha peninsula of forest, gardens, fresh-
water lakes and intertidal shorelines and is one of the largest
urban parks in North America. Along the outer edge of Stanley
Park is an 8.85 km seawall which is used extensively for recrea-
tion and also provides an ideal location for viewing marine birds.
The upper limit of the intertidal area is largely defined by the
seawall and the low tide mark ranges from 30 m (near the Lions
Gate Bridge) to 200 m (near Second and Third Beach) offshore.
The intertidal areas of the Park include rocky, cobble, and sand
beaches with some kelp beds slightly offshore particularly in
the protected waters of Burrard Inlet’s middle and inner har-
bours. The diversity of habitats within the study area supports
many species of wintering marine birds. The rocky shoreline
provides haul-out rocks as well as a variety of foods for
both dabbling and diving ducks. Tide levels dictate when
marine birds can access the rich food resources - mussels,
barnacles, clams, and other invertebrates of the intertidal
area. Extensive beds of Blue Mussel (Mytilus edulis) occur
on the western side of the Stanley Park foreshore. Large
numbers of wintering Surf Scoter (Melanitta perspicillata)
and Barrow’s Goldeneye (Bucephala islandica) feed on
this resource in nearshore waters.
Stanley Park lies within the English Bay & Burrard Inlet
Important Bird Area (IBA) which was designated because it
supports large concentrations of overwintering waterbirds
including globally significant numbers of Surf Scoter, Bar-
row’s Goldeneye and Western Grebe (Aechmophorus
occidentalis), and nationally significant numbers of Great
Blue Heron (Ardea herodias) (IBA Canada 2011). In 1999,
winter waterbird monitoring of the intertidal areas off Stanley
Park was started as a partnership between the Canadian
Wildlife Service and the BC Institute of Technology‘s Fish,
Wildlife, and Recreation program (BCIT FWR). For the past
several years, students undertaking the winter-long survey
have also worked in partnership with the Stanley Park Ecol-
ogy Society (SPES) to receive bird identification and survey
training and to contribute data to SPES’s ongoing bird moni-
toring programs.
For several years, local naturalists and birders have been
raising alarm bells about the declining trends in winter
waterbirds in English Bay. This was first documented in a
short paper by Price (2010) who suggested that many spe-
cies of waterbirds, especially small fish-eating species such
as loons, grebes, and terns have seen huge declines since
the 1980s. This information was considered anecdotal as no
monitoring data had been collected to analyse trends in the
numbers of birds specifically for this area. However, a recent
Stanley Park waterbirds - Worcester
British Columbia Birds
42
Volume 23, 2013
analysis of winter waterbird counts from the B.C. Coastal
Waterbird Surveys, coordinated by Bird Studies Canada, has
noted similar declines for several species of waterbirds in
the Strait of Georgia (Crewe et. al. 2012).
This report evaluates weekly bird monitoring data col-
lected from standardised surveys at Stanley Park to identify
whether waterbird declines have been occurring as sug-
gested by local knowledge. Most winters for the past 12
years, BCIT FWR students have surveyed the seawall for
winter birds, focusing on Barrow’s Goldeneye and Surf Scoter,
but tracking all species observed in the study area.
Methods
This report compares data from similar dates across two
seasons; October 2001 through April 2002 (Boisclair-Joly
and Worcester 2002) and October 2010 through April 2011
(La Fond and Thomas 2011). These time periods were com-
pared because the data were available in raw form, the same
project supervisors were in place, and the students followed
nearly identical survey methods.
The survey area was broken down into twenty-two sur-
vey zones along the Stanley Park seawall from Coal Harbour
to the end of Second Beach (see Worcester 2011). Bird sur-
veys were conducted using the same methods in both 2001/
2002 and 2010/2011. Surveys were done approximately once
per week from October through April and took from 3-5 h to
complete. Beginning at approximately 10:00 Pacific Standard
Time, two or more observers walked a circle route around
Stanley Park along the seawall, alternating starting points
each week to reduce potential bias caused by the time of
day. The observers recorded the number of all marine birds,
and paid special attention to large groups of Barrow’s
Goldeneye and Surf Scoter by recording both age and sex of
these species. On each survey day they recorded the time,
temperature, weather conditions, sea state, visibility, and tide
level. Data were recorded for every bird sighted between the
shoreline and 1 km offshore for each of the 22 zone poly-
gons. A spotting scope was used to identify distant birds.
All birds were counted when possible and large flocks were
estimated using standard techniques. To avoid duplicate
counts, birds observed flying into the area yet to be sur-
veyed were not counted. Birds seen taking off from or land-
ing in the zone being surveyed and flying towards the area
already surveyed were counted.
In 2010, the students received training from SPES Con-
servation Programs Manager Robyn Worcester, who also
took part in the 2001/2002 survey. Care was taken in 2010/
2011 to ensure the students conducted their survey using
the same methodology as the earlier study so that data
comparisons could be made between years. In 2010/2011,
large flocks of Surf Scoter were also photographed and
counted to confirm that initial estimates were accurate.
In order to avoid any potential bias due to variation in
bird identification skills, data were grouped from the spe-
cies level into higher level groups of consistently recog-
nisable species, with the exception of Barrow’s Goldeneye,
Surf Scoter, Western Grebe, Pigeon Guillemot (Cepphus
columba), Canada Goose (Branta canadensis) and Great
Blue Heron, which were consistently identified to species.
Surveys dates were compared if they were similar between
both years (i.e. the first week of December was surveyed in
both years so it was used but the third week of January was
not, so it was dropped).
Results
Overall abundance (the total number of individuals
counted during the survey period) and peak numbers
(the greatest number of birds counted on one survey
day) of most groups of waterbirds were lower in 2010/
2011 than in 2001/2002 (Table1). The difference was most
dramatic for loons, grebes and Pigeon Guillemot. Loons
were present throughout the winter in 2001/2002 but only
a single bird was seen in 2010/2002 (in November). Grebes,
other than Western Grebe, occurred throughout the win-
ter in 2001/2002 but only a few individuals were seen
October-January 2010/2011. Western Grebe were present
October to April in 2001/2002 but only four birds were
seen in 2010/2011 (in late-October and early-November).
Pigeon Guillemot occurred throughout the winter of 2001/
2002 but only a single bird was seen in 2010/2011 (in
March). All of these species feed on small fish in this
area in winter.
Differences in abundance and peak numbers of Bar-
row’s Goldeneye and Surf Scoter were not as dramatic
between years (Table 1). Barrow’s Goldeneye numbers
were noticeably greater in October-November 2001/2002
than in 2010/2011 (Fig. 1) while, in contrast, Surf Scoter
numbers were greater in October-November 2010/2011
than in 2001/2002 (Fig. 2).
Discussion
The changes observed at Stanley Park are, in general,
consistent with those observed in the Strait of Georgia over
the same period, including the noticeable decline in spe-
cies that feed on small fish (Crewe et al. 2012). Western
Grebe showed a significant decline throughout the Strait of
Georgia, although this may have been partly due to a pos-
sible southern shift in wintering areas. In contrast to the
decrease observed at Stanley Park, Pigeon Guillemot showed
a strongly increasing trend in the Strait of Georgia. Surf
Scoter did not show a significant trend in the Strait of Geor-
Stanley Park waterbirds - Worcester
British Columbia Birds
43
Volume 23, 2013
gia but Barrow’s Goldeneye showed a significant decrease,
as was seen at Stanley Park. The eastern shore of the Strait
of Georgia, including English Bay and Burrard Inlet, are
particularly important wintering areas for Barrow’s
Goldeneye (Crewe et al. 2012).
Comparing data between survey years provides evidence
for the change in the abundance of species of wintering
birds using the Stanley Park foreshore that have been ob-
served by local birders and naturalists. The declines are even
more notable when we consider that Western Grebe num-
bers in English Bay & Burrard Inlet IBA reached 15 000 in
1970 and Barrow’s Goldeneye numbers reached 7126 in 1990
but since 2000 numbers have occurred at significantly lower
levels and have not exceeded the 1% global threshold (high-
est counts: Western Grebe, 1029 in January 2002; Barrow’s
Goldeneye, 1901 in November 2000) (IBA Canada 2011). In
contrast, Surf Scoter numbers in the IBA have been regu-
larly greater than the 1% global threshold, with peak counts
of 7030 to 10 011.
Large scale threats that may influence wintering bird
populations in Burrard Inlet include: industrial pollution
including tanker ballast and oil spills (exports of petro-
leum and canola), overfishing, habitat degradation, urban
encroachment, and the negative effects of climate change
such as changes in mussel bed distribution and fish
spawning habitat. Local threats that have been docu-
mented to negatively affect birds using the shoreline in-
clude: direct disturbance by people and off-leash dogs
using the beaches as well as by personal watercraft, such
as jet skis, kayaks and paddleboards, in intertidal areas.
These disturbances have been observed to influence the
resting and feeding habits of shorebirds, gulls and dab-
bling ducks using the shoreline as well as the large flocks
of Surf Scoter, Barrow’s Goldeneye and other waterfowl
that gather here in large numbers in winter (Pers. comm.,
Peter Woods, Vancouver, B.C., 2012).
In summary, the results from winter waterbird surveys
demonstrate that fewer waterbirds are using the nearshore
���������� ���� ����������������
���������������
��������������������� ��������� ��������� ���������
������� �� �! "� #$ �$� ""
%����������� &�� "#" "��� �'�" (#
)��*)� +�� "�!' $,�� #'&" �$�$' "�
���� -.�� �����/� ##� ,�' &&�$ $"$" (,�
�+������������� �!� �,� �&�' �,�� (�'
0��������� "� �� �&� �,' ($$
1 �� �! � �$, � (���
2��+�������� �� " !$ $ (#$
�+��������� �� " ��# �� (#!
� � ���+� && &" #,# ��!" �&
����+����3�� � ' " �� �� (,,
)� ������� &! $' $#� �$� (,�
���� ������� + # � $� � (#&
Table 1. Comparison of peak numbers and overall abundance of birds observed from the Stanley Park
seawall during the winters of 2001/2002 and 2010/2011.
Figure 1: Barrow’s Goldeneye abundance along the
Stanley Park seawall between October and April 2001/
2002 (dark grey) and 2010/2011 (light grey).
Figure 2: Surf Scoter abundance along the Stanley Park
seawall between October and April 2001/2002 (dark
grey) and 2010/2011 (light grey).
Stanley Park waterbirds - Worcester
�����
�����
�����
���
���
���
���
��� �
��� �
��� �
��� �
�����
�����
�����
�����
��� �
��� �
��� �
��� �
�����
�����
�����
�����
�����
�����
�
���
���
���
���
����
� ��������
�
��
��
��
� �
!"
�����
�����
�����
���
���
���
���
��� �
��� �
��� �
��� �
�����
�����
�����
�����
��� �
��� �
��� �
��� �
�����
�����
�����
�����
�����
�����
�
����
����
����
����
����
� ��������
��� ������� !
British Columbia Birds
44
Volume 23, 2013
habitat of Stanley Park during the overwintering season,
therefore supporting earlier observations made by natural-
ists. Whatever the reasons for declines (true popula-
tion decline or re-distribution), conservation of the
nearshore habitat and a reduction of human disturbance
will be beneficial for waterbirds. There is no question
that the English Bay & Burrard Inlet IBA, located in a
highly urbanized landscape, is heavily impacted by
human use and that birds in the area are under stress.
Stanley Park Ecology Society is committed to provid-
ing ongoing public education, monitoring and steward-
ship of the areas in and around Stanley Park, yet all of
the local shorelines of Burrard Inlet need attention. Al-
though some of the impacts on these birds are beyond
SPES’s control, there are many ways we can help in
their conservation. Future monitoring will be important
to documenting further changes in waterbird abundance
and habitat use patterns.
Acknowledgments
I would like to recognise the work of students and
faculty of BCIT’s Fish, Wildlife, and Recreation pro-
gram who contributed the time to collect the data in
this report. This project was possible thanks to BCIT
Instructor Daniel J. Catt and Dr. Sean Boyd of the Ca-
nadian Wildlife Service who supervised the student
projects and the Stanley Park Ecology Society who pro-
vided the venue for the data analysis to take place. I
would also like to thank local naturalists Peter Woods
and Michael Price for their input as well as Danny Catt,
Karen Barry and Art Martell for their help reviewing
and editing this paper.
Literature Cited
Boisclair-Joly, A. and R. Worcester. 2002. Stanley Park
Barrow’s Goldeneye survey 2001-2002. Fish, Wildlife
and Recreation projects course final report. British
Columbia Institute of Technology, Burnaby, B.C.
Unpublished report, 54 p. [Copy in Stanley Park Ecology
Society Office]
Crewe, T., K. Barry, P. Davidson and D. Lepage. 2012. Coastal
waterbird population trends in the Strait of Georgia 1999-
2011: Results from the first 12 years of the British Columbia
Coastal Waterbird Survey. British Columbia Birds 22:8-35.
IBA Canada. 2011. English Bay & Burrard Inlet IBA site
summary. http://ibacanada.ca/site.jsp?siteID=BC020&
lang=EN. Accessed 2011 November 20.
La Fond, K. and M. Thomas. 2011. Wintering marine birds of
the Stanley Park foreshore 2010-2011. Fish, Wildlife and
Recreation projects course final report. British Columbia
Institute of Technology, Burnaby, B.C. Unpublished
report, 47 p. [Copy in Stanley Park Ecology Society Office]
Price, M. 2009. Decline in waterbird populations around
Stanley Park (1980s to present). Appendix 11 in R.
Worcester. State of the park report for the ecological
integrity of Stanley Park. http://stanleyparkecology.ca/
wp-content/uploads/downloads/2012/02/SOPEI-
Seabird-species-declines-around-Stanley-Park.pdf .
Accessed 2010 October 5.
Worcester, R. 2011. Trends in the abundance of wintering
waterbirds along the Stanley Park shoreline between
2001/2002 and 2010/2011. Stanley Park Ecology
Society, Vancouver, B.C. Unpublished report, 17 p. http:/
/ s tan leyparkecology.ca /wp-content /uploads /
downloads/2012/02/SPES_-Winter-Waterbird-Trend-
Report-6-Dec-2011.pdf. Accessed 2012 October 24.
Stanley Park waterbirds - Worcester
Acknowledgements & editor’s comments
This issue of British Columbia Birds presents several pa-
pers on changes in bird populations in British Columbia. The
knowledge gained through these contributions is from both
careful observation by birders and naturalists as well as from
scientific studies. We need both for effective bird conservation
and both are welcome in British Columbia Birds.
The quality of all of the papers is enhanced by our Edito-
rial Board: Neil Bourne, Andy Buhler, Rob Butler, Mark Phinney
and Mary Taitt. Thanks go to them as well as to the external
reviewers of the papers, all of whom have given willingly of
their time and thought to help deliver this issue of British
Columbia Birds. Neil Dawe again has done a splendid job of
producing the journal and of placing the papers on the website.
My greatest appreciation goes to the authors who have
submitted manuscripts; without their commitment to write
up their observations BCFO would not have a journal. The
regular submission of manuscripts over the past year has
ensured the publication of Volume 23 in a timely manner. We
do not have any additional submissions, but we need a
steady flow if we are to have British Columbia Birds pub-
lished annually. All members are encouraged to submit manu-
scripts and to encourage friends and colleagues to do like-
wise. This is your journal, and it has room for a diversity of
papers on wild birds in British Columbia. – Art Martell
Back cover: The Ponderosa Pine Biogeoclimatic Zone of Okanagan Mountain Provincial Park from Boulder Trail, 30 May 2009,
nearly six years after the Okanagan Mountain fire (see page 16). All the live trees in the foreground are ponderosa pine (Pinus
ponderosa) but dead trees include Douglas-fir (Pseudotsuga menziesii) as well. Shrubs are predominantly red-stem ceanothus
(Ceanothus sanguineus) that have sprouted since the fire. The ground layer is dominated by bunchgrasses including fescues
(Festuca spp.) and bluebunch wheatgrass (Pseudoroegneria spicata) but also contains arrow-leaved balsamroot (Balsamorhiza
sagittata) and pinegrass (Calamagrostis rubescens). Photograph by Les Gyug.
Photo essay
An Osprey vying for a coveted perch
along the Okanagan River channel,
but being rebuffed by a conspecific
and l eav ing unsuccess fu l .
Okanagan Falls, 31 August 2012.
Photos by Laure W. Neish.
British Columbia BirdsJournal of the British Columbia Field Ornithologists
Volume 23 • 2013