british columbia birds - wordpress.combritish columbia birds is published annually ....

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


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.


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)



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.



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.



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.



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.



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.



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.



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)



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-



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.



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



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.


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.



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.



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


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.



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.



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.



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.



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


◄ 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.


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).



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.



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


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


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



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

►



25

Volume 23, 2013

◄ Appendix 1 Before Fire

(1993‒2003)

After Fire

(2006‒2011)

Species





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

►



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



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.


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.



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



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



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)



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



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.



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.



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



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



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.


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.



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.



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


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


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


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-



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).


��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

��

��

��

��

� ��

�

��

��

��

� �

!"

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

��

��

��

��

� ��

�� !


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.


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

british columbia birds - wordpress.combritish columbia birds is published annually ....

Documents