Forest Management in the Fort St. John Timber Supply Area
Transcript of Forest Management in the Fort St. John Timber Supply Area
Monitoring Birds for Sustainable
Forest Management in the Fort St.
John Timber Supply Area: Species-
Habitat Relationships, Trends, and
Species of Conservation Concern
Prepared for:
Canadian Forest Products Ltd. Swanson Lumber Road RR1, SITE 13, COMP 2 Fort St. John, BC V1J 4M6
Prepared by:
Stantec 11 – 2042 Mills Road West Sidney, BC V8L 5X4 Tel: (250) 656-7966
Stantec Project Number:
123210054 Forest Investment Account Number 8047001
March 2010
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
AUTHORSHIP
Michael Preston, M.Sc., R.P.Bio.
Pierre R. Vernier, M.Sc.
Joanna Preston, B.Sc., BIT.
Meghan O‘Neill, B.Sc.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
i
Table of Contents
1.0 INTRODUCTION ................................................................................................................ 1
1.1 PURPOSE AND OBJECTIVES ........................................................................................... 1
2.0 STUDY AREA ..................................................................................................................... 2
3.0 METHODS .......................................................................................................................... 2
3.1 BIRD SURVEYS ................................................................................................................. 3
3.1.1 Breeding Bird Survey ........................................................................................... 3
3.1.2 Directed Warbler Surveys ..................................................................................... 6
3.2 HABITAT SAMPLING ......................................................................................................... 8
3.3 DATA INPUT ....................................................................................................................... 8
3.4 DATA ANALYSIS: BREEDING BIRD SURVEY ................................................................... 9
3.4.1 General Indices .................................................................................................... 9
3.4.2 Species-Habitat Associations ............................................................................... 9
3.4.3 Power to Detect Trend.......................................................................................... 9
3.4.4 Trend Estimates ..................................................................................................11
3.5 DATA ANALYSIS: DIRECTED WARBLER SURVEYS .......................................................12
3.5.1 Habitat Models and Power Analysis ....................................................................12
3.5.2 Power and Precision ...........................................................................................13
4.0 RESULTS ..........................................................................................................................15
4.1 BREEDING BIRD SURVEYS .............................................................................................15
4.1.1 Sample Intensity and Representation ..................................................................15
4.1.2 Richness, Abundance and Diversity ....................................................................15
4.1.3 Patterns of Species Richness and Bird Abundance in Old and Mature Stands ....18
4.1.4 Species‘ Responses to Habitat Attributes ............................................................18
4.1.5 Power to Detect Trend.........................................................................................28
4.1.6 Species‘ Trend Estimates ....................................................................................33
4.1.6.1 Fort St. John TSA ................................................................................................33
4.1.6.2 Provincial Context ...............................................................................................34
4.2 DIRECTED WARBLER SURVEYS ....................................................................................43
4.2.1 Sample Intensity and Representation ..................................................................43
4.2.2 Summary of Occurrences ....................................................................................43
4.2.3 Effects of Habitat Change ....................................................................................45
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
ii
4.2.4 Habitat Models ....................................................................................................47
4.2.5 Power and Precision ...........................................................................................50
4.3 BIRD SPECIES OF CONSERVATION CONCERN ............................................................52
4.3.1 Summary .............................................................................................................52
5.0 SUMMARY AND CONCLUSION .......................................................................................63
5.1 BREEDING BIRD SURVEYS .............................................................................................63
5.1.1 Species-habitat relationships ...............................................................................63
5.1.2 Trends .................................................................................................................64
5.2 DIRECTED WARBLER SURVEYS ....................................................................................65
5.2.1 Habitat Models ....................................................................................................65
5.2.2 Power and Precision ...........................................................................................65
5.3 SPECIES OF CONSERVATION CONCERN .....................................................................66
5.4 CONCLUSION ...................................................................................................................67
6.0 ACKNOWLEDGEMENTS ..................................................................................................68
7.0 LITERATURE CITED .........................................................................................................69
8.0 APPENDICES ....................................................................................................................72
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
iii
List of Tables
Table 1 Classification Scheme used to Describe Habitat Attributes ................................. 8
Table 2 BBS Route Names, Numbers of Stations and Route Weightings used in the Power Analysis ............................................................................................10
Table 3 Sample Data for a Single Species Used in program TRIM. ................................11
Table 4 Description of Habitat Variables used in Logistic Regression Analysis ...............12
Table 5 Model Results of Species Responses to Habitat and Year Effects .....................20
Table 6 Power to Detect a Significant Decreasing Trend for 34 Bird Species Occurring in the Fort St. John TSA ....................................................................29
Table 7 Trend Results for 33 Regularly Occurring Species in the Fort St. John TSA, 2005–2009 ................................................................................................35
Table 8 Local and Provincial Comparison of Population Trends. ....................................42
Table 9 Percentage of Survey Stations Occupied by Listed Warbler Species from Breeding Bird Surveys and Directed Warbler Surveys .......................................44
Table 10 Abundance of Provincially-Listed Warbler Species from Breeding Bird Surveys and Directed Warbler Surveys ..............................................................44
Table 11 Logistic Regression Model Results of Three of the Five Listed Warbler Species ..............................................................................................................48
Table 12 Number of Bird Records for Species of Conservation Concern from Different Components of the Monitoring Program, 2005–2009 ...........................53
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
iv
List of Figures
Figure 1 Location of the Fort St. John Timber Supply Area in British Columbia ................. 2
Figure 2 Sample Orthophoto Datasheet ............................................................................ 4
Figure 3 Locations of Breeding Bird Survey Routes in the Fort St. John TSA ................... 5
Figure 4 Locations of Directed Warbler Survey Stations in the Fort St. John TSA ............. 7
Figure 5 Total Abundance of all Species having ≥100 Detections from Breeding Bird Surveys in the Fort St. John TSA from 268 Point Count Stations Sampled Annually for the Period 2005–2009 .....................................................16
Figure 6 Rank Abundance Curve for all Species Detected from Breeding Bird Surveys in the Fort St. John TSA, 2005–2009 ...................................................17
Figure 7 Annual and Cumulative Species Richness from 268 Breeding Bird Survey Stations Sampled Annually from 2005–2009 ......................................................17
Figure 8 Mean Species Richness and Abundance from 100 m Radius Point-count Stations in Class 4 Stand Types ........................................................................19
Figure 9 Effect of Percentage Shrub Cover on the Expected Number of Birds per Plot for Alder Flycatcher. ....................................................................................21
Figure 10 Effect of Percentage Softwoods and Percentage Shrub Cover, among Forest Classes on the Expected Number of Birds per Plot for Golden-crowned Kinglet and Lincoln‘s Sparrow, Respectively ........................................22
Figure 11 Effect of Shrub Height and Percentage Hardwoods , among Forest Classes on the Expected Number of Birds per Plot for Orange-crowned Warbler and Red-breasted Nuthatch, Respectively ............................................23
Figure 12 Effect of Forest Class and Percentage Hardwoods among Shrub Height Groups , on the Expected Number of Birds per Plot for Swainson‘s Thrush and Warbling Vireo, Respectively ..........................................................24
Figure 13 Effect of Percentage Hardwoods among Forest Classes on the Expected Number of Birds per Plot for Ovenbird ...............................................................25
Figure 14 Effect of Percentage Softwoods and Shrub Cover on the Expected Number of Birds per Plot for Ruby-crowned Kinglet ...........................................26
Figure 15 Effect of Percentage Softwoods and Shrub Cover, among Forest Classes, on the Expected Number of Birds per Plot for Yellow-rumped Warbler ..............27
Figure 16 Relationship between Species‘ Rank Abundance and Probability of Detecting a -3% Change in Annual Abundance for Three Time Intervals ...........32
Figure 17 Five-year Trend for Alder Flycatcher and American Redstart ............................36
Figure 18 Five-year Trend for Black-throated Green Warbler and Golden-crowned Kinglet ...............................................................................................................37
Figure 19 Five-year Trend for Fox Sparrow and Lincoln‘s Sparrow ...................................38
Figure 20 Five-year Trend for Red-breasted Nuthatch and Ruby-crowned Kinglet ............39
Figure 21 Five-year Trend for Warbling Vireo and White-throated Sparrow ......................40
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
v
Figure 22 Five-year Trend for Yellow-rumped Warbler .......................................................41
Figure 23 Effect of Partial or Complete Harvesting on Combined Abundance of Listed Warblers Compared to Non-harvested Sites Among Years. ....................46
Figure 24 Interaction Between ―Year‖ and ―Habitat‖ on Total Abundance of all Listed Warbler Species Combined ...............................................................................46
Figure 25 Sample Predicted Probability of Occurrence Map Reclassified into Four Broad Classes for Black-throated Green Warbler...............................................49
Figure 26 Precision of Estimates for the Breeding Bird Survey and Directed Warbler Sampling as Measured using the Width of 95% Confidence Interval for a given Proportion/Sample Size ............................................................................50
Figure 27 Sample Size Required to Detect Changes in Occupancy from Baseline Year to a Future Survey Year ............................................................................51
Figure 28 Locations of Barn Swallow Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods ...............................................54
Figure 29 Locations of Bay-breasted Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods .........................55
Figure 30 Locations of Black-throated Green Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods .................56
Figure 31 Locations of Broad-winged Hawk Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods .........................57
Figure 32 Locations of Canada Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods ........................................58
Figure 33 Locations of Cape May Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods ........................................59
Figure 34 Locations of Connecticut Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods .........................60
Figure 35 Locations of Le Conte‘s Sparrow Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods ........................................61
Figure 36 Locations of Olive-sided Flycatcher Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods .........................62
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
vi
List of Appendices
Appendix 1 2009/2010 Work Plan ........................................................................................73
Appendix 2 Alphabetical list of observed species (2005–2009).............................................83
Appendix 3 Species richness and abundance (2005–2009) .................................................87
Appendix 4 Species richness, abundance, and diversity (2009) ...........................................89
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
1
1.0 Introduction
This project is a continuation of work initiated in 2005 designed to address a range of questions
and objectives related to biodiversity conservation in relation to sustainable forest management
in the Fort St. John Pilot Project area. Throughout the duration of the project, Breeding Bird
Surveys (BBS) have been a constant fixture, with sampling representation increasing in 2006
and 2007. Other project components, such as Forest Interior surveys, and Directed Warbler
sampling, have been used to address questions related to the development of specific
components of the Sustainable Forest Management Plan (SFMP). In 2009, we continued the
BBS program, and completed a second year of the Directed Warbler sampling.
1.1 PURPOSE AND OBJECTIVES
The purpose of this study was to estimate species richness, abundance, and diversity of birds in
the Fort St. John Timber Supply Area (TSA), and to evaluate components of habitat
relationships and population trends for various songbird species. We also evaluate the
effectiveness of model-based sampling for provincially-listed warblers and report on the
occurrence of other species of conservation concern. The specific objectives defined in the
2009/2010 work plan (Appendix 1) include:
Analyzing species-habitat relationships using forest age and structural attributes, and
providing context by comparing results with other studies.
Updating 2008 habitat-based models for five warbler species of conservation concern
and using the models to project areas of low, medium, and high suitability.
Estimating occupancy rates for listed warblers with a specified level of precision, and
evaluating power to detect change in occupancy between two years (2008 and 2009).
Evaluating species‘ trends, and the power to detect trend, for a range of bird species
occurring on BBS routes for the period 2005–2009.
Providing a summary of birds of conservation concern as determined by the BC
Conservation Data, the Committee on the Status of Endangered Wildlife in Canada, and
the Species at Risk Act, Schedule 1.
The data and results from this study are intended to provide forest managers with baseline
information to support the development of a biodiversity conservation strategy and to aid in the
development of future monitoring programs. Data and results also contribute to the Species
Accounting System being developed concurrently by Dr. F.L. Bunnell (University of British
Columbia) in cooperation with Canadian Forest Products Ltd. and the B.C. Ministry of
Environment.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
2
2.0 Study Area
The study area is the Fort St. John Timber Supply Area in the Boreal Plains Ecoprovince of
northeast British Columbia (Figure 1). Details pertaining to topography, vegetation, and climate
are provided in Preston (2008).
Figure 1 Location of the Fort St. John Timber Supply Area in British Columbia
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
3
3.0 Methods
3.1 BIRD SURVEYS
3.1.1 Breeding Bird Survey
We used a modified version of the BBS method discussed by Sen (1981) and Bystrak (1981) to
sample a variety of bird species occurring in a broad range of forested and non-forested
habitats. The modification included the use of orthophotos (black-and-white aerial photos) with
50-, 100-, 150-, and 200-m radius concentric rings overlaid and printed individually as 8.5" x 11"
datasheets (Figure 2). We sampled 16 previously established BBS routes in the Fort St. John
TSA (Figure 3), with each BBS route being approximately 24.8 km in length and including 30
point count stations spaced approximately 800 m apart. In some instances, some routes had
stations that were > 800 m apart to accommodate site-specific surveys of areas of management
interest (e.g., Buick Creek, KobesHay). Initial route selection within the study area was
determined largely by road availability (location and length), and within those areas, some
routes were selected because of specified management interest (e.g., Wonowon).
An analysis of BBS habitat representation using biogeoclimatic zones is summarized in Preston
(2009). Briefly, BBS sampling is proportionately allocated in relation to the area of each of three
BEC zones (BWBS, ESSF, SWB), but it is unlikely that the current number of BBS stations in
the ESSF (n = 26) and SWB (n = 28) zones are adequate to address management questions for
species occurring in those areas.
All surveys, in all years, were conducted between May 28 and July 2, with individual surveys
commencing at sunrise and lasting approximately four hours (RIC 1999). In 2005 and 2006
sample duration of individual stations was three minutes, whereas in 2007 and onwards, sample
duration was five minutes (RIC 1999). The change in sample duration was necessary to
accommodate the use of orthophotos. By increasing the sample duration of individual stations,
the total number of stations per route was reduced from 50 to 30. To maximize the utility and
compatibility of data collected in 2005 and 2006, we indicated on datasheets from 2007–2009
which birds occurred in the latter two minutes of the 5-minute survey by underlining them on the
datasheet.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
4
Figure 2 Sample Orthophoto Datasheet
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
5
Figure 3 Locations of Breeding Bird Survey Routes in the Fort St. John TSA
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
6
3.1.2 Directed Warbler Surveys
We used point count surveys that were similar to those used for the BBS, except that instead of
point counts occurring continuously along a transect, the warbler survey stations were habitat
specific and less systematic in their spatial arrangement. Potential warbler sampling areas were
determined using preliminary habitat model predictions based on data gathered for the period
2005–2007. Using the predicted occurrence areas, a randomly selected pool of 1,000 non-
overlapping road-based point count stations within 150 km of Fort St. John was created. From
that pool, inspection of site accessibility and an evaluation of survey efficiency (i.e., distance
and/or time between stations) resulted in 104 Directed Warbler survey stations being used
(Figure 4). All Directed Warbler survey stations were located in the BWBS biogeoclimatic zone.
The survey protocol (e.g., survey period, time, duration) and corresponding data attributes were
the same as those used for the BBS. Similarly, all data were transferred to a geodatabase upon
completion of the field program. For additional detail on the Directed Warbler sampling methods,
see Vernier et al. (2009).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
7
Figure 4 Locations of Directed Warbler Survey Stations in the Fort St. John TSA
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
8
3.2 HABITAT SAMPLING
A protocol for recording habitat at road-based bird survey stations in the Fort St. John TSA was
established in 2007 and details are provided in a corresponding report (Preston 2008). Table 1
summarizes the kinds of habitat attributes documented at each bird survey station. The habitat
data described here meet requirements for assessing species-habitat relationships as described
by RIC (1999).
Table 1 Classification Scheme used to Describe Habitat Attributes*
Forest Attribute
Attribute Description
Forest Class
FC1) trees ≤1.3 m in height
FC2) trees range from 1.4–3 m in height
FC3) trees range from 7.5–12.5 dbh
FC4) trees range from 13–40 dbh
FC5) trees are >40 dbh
Forest Type (FC2-FC5 only)
Hardwood (recorded to nearest 5%; ≥75% = hardwood stand)
Mixedwood (recorded to nearest 5%; ≥25% hardwoods and ≥25% = softwood) stand)
Softwood (recorded to nearest 5%; ≥75% softwood stand)
Tree Species (percentage composition of each tree species, min. 5% representation; FC2-FC5 only)
Canopy Cover (percentage cover of the forest canopy for trees >1.3 m); not recorded for FC1 or FC2
Shrubs (percentage cover and height)
Low (0.5–1.5 m in height)
Medium (1.5–2.5 m in height)
High (>2.5 m in height) * Attributes are described for uniform stations only. For example, habitat is scored for left and right sides of the road independently, using a 100-m radius semi-circle buffer comprised of a single forest class. We did not score bird survey stations comprised of more than one forest class per side. Left and right sides are determined on the basis of moving forward along a route from station 1 to 30. For Directed Warbler survey stations, cardinal direction indicators were used (N, E, W, S). Sites comprised of more than one forest class were documented as ―split‖.
3.3 DATA INPUT
Field data for birds (BBS and warblers) were transferred from orthophoto datasheets into a
geodatabase using ArcGIS 7.0. Each observation included a field for: route name, station
number, species, abundance, date, time (three or five minutes), and behaviour. Additional
weather attributes such as wind speed, temperature, precipitation, and cloud cover were also
transcribed as categorical variables for each survey station (RIC 1999; but see Approved
Variances in Appendix 1). Distances between sample stations and the associated birds
detected at those station was calculated in ArcGIS 7.0. Project subcomponents (i.e., BBS or
warblers) were identified in a separate column in the data table.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
9
Field data for habitat were transferred from individual datasheets to an Excel spreadsheet that
contains separate tabs for each year of data collection (2007–2009).
3.4 DATA ANALYSIS: BREEDING BIRD SURVEY
3.4.1 General Indices
Species richness, abundance and diversity were calculated using the program-group SPECIES
DIVERSITY (Krebs 1999) for BBS and Directed Warbler surveys separately. Results are
provided for the entire study area, and for individual BBS routes.
3.4.2 Species-Habitat Associations
To evaluate species-specific responses to structural habitat attributes we defined mixed Poisson
regression models1 using the statistical software package R (version 2.9.2). Poisson regression
was required because count data like ours violates the assumptions of normality of errors and
homogeneity of variance of regular Gaussian models. We defined model sets for eleven bird
species that had at least five occurrences in each forest class, and used Akaike‘s Information
Criterion (AIC) model selection methodology following Burnham and Anderson (1998) to choose
the best approximating model for each species. This was defined as the model with the smallest
number of predictors within two AIC units of the model with the lowest AIC. Fixed predictors
included forest class, hardwood/softwood gradient, shrub height, shrub cover, and year (2007,
2008, 2009). The random predictors were Station_ID and Station. Station was nested within
Station_ID to account for both temporal autocorrelation between plot counts in each of the three
years, and spatial autocorrelation between the often differing habitat attributes at each side of a
station. Data from Station_IDs with Forest Class 1 were excluded due to high correlation with
hardwood/softwood gradient (i.e., Class 1 forest types were always scored as zero for
hardwoods/softwoods). Therefore, modeled species responses to habitat attributes are not
applicable to clearcut sites (Forest Class 1). Overall model goodness-of-fit was determined
using an analysis of deviance test for significant differences between the best approximating
model and a null model with only random predictors (Crawley 2007). For each species, we
tested the assumption that bird counts were Poisson-distributed using a Pearson Chi-square
goodness-of-fit test of observed versus expected counts using the methodology described in
Crawley (2007).
3.4.3 Power to Detect Trend
The primary objective of most long-term bird monitoring programs is to detect changes in the
population size of target species. The detection of population decreases or increases can allow
managers to be more proactive in the conservation of bird populations, and can be a useful tool
for helping guide possible management options (e.g., salvage logging, retention harvest,
stubbing). Identifying a trend, however, often requires several years of continuous data
collection, and so a typical question might be: How many stations need to be surveyed to detect
a 30% decline in bird abundance over 10 years (i.e., 3% per year) with 80% power (accuracy)?
1 Using the glmer modeling function, family = poisson, in the lme4 package (Bates 2009)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
10
Conversely, we may want to know how long would be required to obtain adequate power using
current sampling intensity (5–25 transects; 30 or 50 stations per transect) with alpha and beta
set at 0.1 and 0.8 respectively for a particular target (e.g., 30% in 10 years, or 3% per year). In
this report we address the latter question using species data from 268 BBS stations sampled
annually in the Fort St. John TSA from 2005–2009. Where a particular species did not occur on
some survey routes, sample size was reduced accordingly. For the 2007-2009 dataset, we used
only those data which were recorded in the first three minutes of the survey for comparison with
data gathered from 2005-2006.
We used the program MONITOR (Gibbs 1995) to estimate the statistical power of population
change for several species. We ran three simulations for each species (10, 20 and 30 years)
and report the beta-value of that test (beta-value = power [0–1]), where values of zero have no
power to detect a trend, and values of one have perfect power to detect a trend, for five annual
decreasing population scenarios (-10, -5, -3, -2 and -1%). We report only the power to detect a
decreasing trend because it is primarily forest-dependent species that are at risk of population
decline in the presence of forest harvesting. As a caveat, however, the power to detect an
increase is generally similar to the power to detect a decrease.
The parameter settings used in program MONITOR include: number monitored (number of
routes the species occurred on), counts (1), initial values (mean, standard deviation, weight),
number conducted (10, 20 and 30 years), occasions (1,2,3…nconducted), type (exponential),
significance level (0.1), number of tails (2), trend coverage (partial), replications (500).
Remaining variables are kept at the default values provided. Route weightings were used to
account for the unequal number of sample stations among routes. Weights used in the power
analysis are provided in Table 2.
Table 2 BBS Route Names, Numbers of Stations and Route Weightings used in the Power Analysis
BBS Route Name Number of Stations Weight
Buick Creek 30 1.00
Haystack 24 1.20
KobesHay 20 1.33
PeeJay 1 30 1.00
PeeJay 2 30 1.00
PeeJay 3 16 1.47
Sikanni 28 1.07
Tommy Lakes 1 30 1.00
Tommy Lakes 2 30 1.00
Tommy Lakes 3 30 1.00
Total 268
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
11
3.4.4 Trend Estimates
We used program TRIM (Pannekoek and van Strien 2005) to estimate trend for 33 species in
the Fort St. John TSA. As specified in Section 3.4.3 for evaluating power to detect trend, we
used bird data from 268 stations sampled annually for the period 2005–2009, and used only
those data which were collected during the first three minutes of each point count survey. Data
imported into TRIM had the following parameter defaults for analyses: Number of Time Points =
5, Number of Covariates = 0, Missing Value = -1 and Weight Available = True. Route weights
used in the trend analysis are provided in Table 2, and a sample of the data structure is
provided in Table 3. Trend estimates were calculated using the Time Effects model with the
following parameters: Over Dispersion = true, Serial Correlation = False, Weighting = True,
Base Time 2005 = 1, Covariates = 0. Reporting of results include Goodness of Fit, Deviance
from Linear Trend, and Trend Pattern and Significance. For additional statistical details and
model assumptions relating to the trend analysis, see Pannekoek and van Strien (2005).
Table 3 Sample Data for a Single Species Used in program TRIM.
Route Year Abundance Weight
1 2005 0 1
1 2006 5 1
1 2007 2 1
1 2008 0 1
1 2009 2 1
2 2005 4 1.2
2 2006 9 1.2
2 2007 9 1.2
2 2008 5 1.2
2 2009 11 1.2
3 2005 5 1.33
3 2006 10 1.33
3 2007 9 1.33
3 2008 13 1.33
3 2009 9 1.33
4 2005 0 1
4 2006 18 1
4 2007 20 1
4 2008 17 1
4 2009 8 1
n…
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
12
3.5 DATA ANALYSIS: DIRECTED WARBLER SURVEYS
3.5.1 Habitat Models and Power Analysis
We developed habitat models for three of the five listed warbler species (Black-throated Green
Warbler, Canada Warbler, and Connecticut Warbler) using bird survey data and Vegetation
Resource Inventory (VRI) data collected in the Fort St. John TSA. There were not enough
detections of the other two target species (Bay-breasted Warbler and Cape May Warbler) to
develop reliable models. The bird survey data consisted of: 1) model-based surveys conducted
in 2008 and 2009; 2) forest interior surveys conducted in 2007; and 3) 16 standard roadside
transects surveyed from 2007-2009. We used the VRI data to measure habitat characteristics in
and adjacent to georeferenced bird detections. Habitat variables measured the broad forest
type, stand age, and interactions between age and forest type (Table 4).
Our modeling approach was based on relating species occurrences to habitat variables
measured from VRI data (Vernier and Bunnell 2008). Specifically, we used multiple logistic
regression to estimate the probability of occurrence of each bird species as a function of habitat
covariates:
pi = eηi / 1 + eηi
where pi is the detection probability (probability of occurrence at pixel i), and ηi = β0 + ∑βixi is
the linear predictor. The xi are the habitat covariates and the βi are the parameters to be
estimated. For each species, we started with a model that included all covariates and used AIC
to remove extraneous variables. We report model deviance (-2 times the log-likelihood of the
model), drop in deviance from a model with no covariates, and the area under the receiver
operating curve (ROC). The latter is a measure of the predictive accuracy of the model, with
values ranging from 0 to 1; values greater than 0.7 provide evidence of model reliability (Vernier
et al. 2008).
We linked the avian habitat models (i.e., the coefficients from logistic regression functions) to
ArcGIS to produce maps showing areas of varying habitat quality for portions of the Fort St.
John TSA. The functions were used to generate a probability of occurrence map for each bird
species which were reclassified into four habitat suitability classes: not suitable, low, medium,
and high suitability.
Table 4 Description of Habitat Variables used in Logistic Regression Analysis
Variable Description
RECENT Recently disturbed stand (≤30 years)
DECID Deciduous forest (≥75% hardwood species)
CONIF Conifer forest (≥75% conifer species)
MIXED Mixedwood forest (≥25% hardwood and conifer species)
AGE Stand age in years
DECID x AGE Interaction between deciduous forest and stand age
CONIF x AGE Interaction between conifer forest and stand age
MIXED x AGE Interaction between mixedwood forest and stand age
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
13
3.5.2 Power and Precision
In this section we explore the effectiveness and efficiency of two different monitoring scenarios,
each related to a specific management objective. In Scenario 1, our objective is to estimate the
status of warbler species in a region or habitat strata (e.g., a timber supply area or a BEC
variant). Specifically, we estimate warbler occupancy (measured as the proportion of stations
occupied) and are interested in how many samples are needed to achieve a specified level of
precision in estimating occupancy within a management region or habitat strata. In Scenario 2
the objective is to detect a change in occupancy between two different time periods—a baseline
year and a time in the future (e.g., five years post-baseline). We want to know, for example, how
many stations need to be monitored in order to detect a 30% change in occupancy with 80%
power.
Scenario 1: Estimating Warbler Occupancy with a Specified Level of Precision
A basic objective of the Pilot participant‘s bird monitoring program is to estimate the presence or
abundance of bird species within a management region (e.g., Fort St. John TSA) or a broad
habitat type (e.g., BWBSmw1). This information is useful for many purposes including
estimating population status at the regional level or validating assumptions present in
management guidelines and plans (e.g., Stand-level Management Guidelines for Selected
Forest Dwelling Species at Risk). In the case of the five warbler species at risk, estimating
occurrence or occupancy at the station level is more realistic than estimating abundance due to
the low number of detections. The problem, then, is to estimate the proportion of stations that
are occupied (occupancy) within a given management or habitat strata with a specified level of
precision. Schwartz (2009) suggests the following as survey precision guidelines:
For preliminary surveys, the 95% confidence interval should be ± 50% of the estimate
For management surveys, the 95% confidence interval should be ± 25% of the estimate
For scientific work, the 95% confidence interval should be ± 10% of the estimate
In Scenario 1 we then focus on questions at the management unit level (i.e., the Fort St. John
TSA) and note that the same approach can be applied to habitat strata. Specifically, we want to
know if the number of stations surveyed from 2008–2009 are sufficient to achieve one of the
three precision guidelines. If the desired precision levels are not achieved, we can use the
precision plots to determine what sample sizes would be needed to estimate occupancy within
10, 25 and 50% of their estimated true value at the 95% confidence level.
We estimated the proportion of stations occupied (occupancy) by the five warbler species at risk
for each of two survey types in the Fort St. John TSA: BBS (n = 480) and warbler sampling (n =
104) from 2008-2009 (Table 9). We calculated the width of the 95% confidence intervals for a
single proportion as follows:
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
14
Where n = the sample size (number of survey stations), p = the expected proportion of stations
with a detection, and W = width of confidence interval.
Scenario 2: Detecting Changes in Occupancy Between two Points in Time
Another important objective of the bird monitoring program is to detect changes in the
occurrence of bird species at two different points in time, for example from the start of the
monitoring program to the current year. The purpose is to detect whether a significant change,
requiring management action, has occurred. For example, a 30% decrease in Canada Warbler
occupancy over time could trigger management actions including more intensive surveys.
Alternatively, a 30% increase in Canada Warbler occupancy could trigger a re-evaluation of the
species status in British Columbia. The problem, in this case, is to estimate the proportion of
stations that are occupied (occupancy) in two different years with a specified level of precision
or to determine how many samples are needed to detect a change with a set amount power
(e.g., 80%).
In Scenario 2 we focus on questions at the management unit level (e.g., the Fort St. John TSA).
Specifically, we would like to know how many bird survey stations need to be sampled to detect
a 30% change (increase or decrease) in occupancy between two years with 80% power at the
90% confidence level. In other words, we would like to evaluate the relationship between
sample size and effect size.
Although we selected a 30% change as a reasonable management target to indicate the need
for action (e.g., more intensive surveys or setting aside critical habitat), for comparison, we also
examine a range of potential changes, from 20 to 80%.
Similar to scenario 1, we used the proportion of stations occupied (occupancy) by the five
warbler species at risk for each of two survey types (BBS and warbler sampling) in 2008–2009
as an estimate of the baseline occupancy (Table 9). We then compared baseline occupancy for
20 to 80% changes in a subsequent survey year to determine the required number of stations.
To illustrate, we compare the effectiveness of the two survey types using estimates from the
Fort St. John TSA (see Table 9). For a biologically important decline to be considered, a goal
equivalent to a 30% decline is set, at a confidence level of 90%. Assuming the same number of
sites is re-surveyed in a subsequent year, we use the following formula to calculate the required
sample size:
where n is the estimated sample size, Zα and Zβ are the Z coefficients for Type I and Type II
error rates, p1 is the occupancy rate for the baseline year, and p2 is the occupancy rate for the
future sampling period as determined by the magnitude of change to be detected. q1 and q2 are
equal to 1-p1 and 1-p2, respectively.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
15
4.0 Results
4.1 BREEDING BIRD SURVEYS
4.1.1 Sample Intensity and Representation
In 2009 473 of 480 BBS stations were surveyed, and the seven stations not surveyed was
because of a gate near the end of the Kobes route. Compared to 2008, sampling intensity and
representation of biogeoclimatic zones, landscape units, and operating areas did not change in
2009 (Preston 2009). Briefly, BBS stations were proportionately allocated among three
biogeoclimatic zones (BWBS = 426, ESSF = 26, and SWB = 28), and when combined with the
Directed Warbler survey (n = 584 survey stations), 7 of 11 landscape units were sampled, and
20 of 45 operating areas were sampled.
4.1.2 Richness, Abundance and Diversity
In 2009 85 bird species were observed, and since inception of the study in 2005 two new
species were added, bringing cumulative species richness for BBS and Directed Warbler
surveys to 112 (see Appendix 2). Of 268 BBS stations surveyed annually for the period 2005–
2009 cumulative species richness for each BBS route has ranged from 42 (Sikanni) to 65 (Buick
Creek), and cumulative bird abundance has ranged from 588 (PeeJay 3) to 1,415 (Buick Creek)
(see Appendix 3). Additionally, 20 species have been recorded ≥100 times (Figure 5) with
Yellow-rumped Warbler (n = 1,235), Chipping Sparrow (n = 747), and Ruby-crowned Kinglet (n
= 691) being the most frequently detected species.
Using data only from BBS routes surveyed in 2009, species richness ranged from 20 (Graham
River 1) to 41 (KobesHay) and averaged 29.4 species per BBS route. Total bird abundance
ranged from 108 (Graham River 2) to 272 (KobesHay) and averaged 178 birds per BBS route.
Species diversity (H') ranged from 3.45 (Peejay 2) to 4.39 (Haystack) and averaged 3.92 per
BBS route. A summary of species richness, abundance, and diversity by BBS route for 2009 is
provided in Appendix 4. Rank abundance for species having a total abundance of ≥ 40 birds for
the period 2005–2009 is summarized in Figure 6. The species with the highest rank abundance
was Yellow-rumped Warbler, followed by Chipping Sparrow, Ruby-crowned Kinglet, Tennessee
Warbler, and Swainson's Thrush.
Among 268 BBS stations sampled annually since inception of the monitoring program in 2005,
average annual richness was 68 species (range: 61–72) (Figure 7). Total cumulative richness
for the same survey stations and period was 95 species (Figure 7).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
16
Figure 5 Total Abundance of all Species having ≥100 Detections from Breeding Bird Surveys in the Fort St. John TSA from 268 Point Count Stations Sampled Annually for the Period 2005–2009
Species codes with full English names can be found in Appendix 2.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
17
Figure 6 Rank Abundance Curve for all Species (≥40 birds) Detected from Breeding Bird Surveys in the Fort St. John TSA, 2005–2009
Abundance is based on 268 point count stations sampled annually for the period 2005–2009. Species codes with full English names can be found in Appendix 2.
Figure 7 Annual and Cumulative Species Richness from 268 Breeding Bird Survey Stations Sampled Annually from 2005–2009
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
18
4.1.3 Patterns of Species Richness and Bird Abundance in Old and Mature Stands
We compared species richness and abundance among broad stand types for forests classed as
old or mature (i.e., x stem dbh = 13-40 cm). Among the stands classed in 2008 (n = 99) and
2009 (n = 92), softwood (95–100% softwoods) and softwood leading (75–90% softwoods) had
the lowest mean species richness and abundance, compared to hardwood-leading (75–90%
hardwoods) stands that had the highest mean species richness and abundance in 2008, and
hardwood (95–100% hardwoods) and mixedwood (30–70% softwoods or hardwoods) stands
that had the highest mean species richness and abundance in 2009 (Figure 8). Generally, there
was little difference in mean species richness or abundance among mixedwood, hardwood-
leading, and hardwood stands, whereas softwood and softwood-leading stands were similar to
each other, but considerably lower than the other stand types. There was a notable decrease in
mean species richness and abundance in 2009, compared to 2008, among all stand types, but
hardwood-leading stands had the greatest overall decrease in both indices.
4.1.4 Species’ Responses to Habitat Attributes
Eleven species had data amenable to a detailed analysis of habitat and year effects, and all
species except Dark-eyed Junco had significant results (Table 5). Alder Flycatcher was the only
species with just one significant predictor variable, and within Forest Class 2 stands, increasing
percentage shrub cover was a significant predictor of abundance (Figure 9). Forest class was a
significant predictor for seven species, with Golden-crowned Kinglet, Ovenbird, Red-breasted
Nuthatch, Swainson‘s Thrush, and Yellow-rumped Warbler showing an increase in abundance
with increasing forest age, and Lincoln‘s Sparrow and Orange-crowned Warbler showing an
increase in abundance with decreasing forest age (see Figures 10 to 13). Abundance of
Golden-crowned Kinglet increased with both increasing forest age and percentage softwoods
whereas for Red-breasted Nuthatch and Ovenbird bird abundance increased with increasing
forest age and percentage hardwoods. Increasing percentage hardwoods and shrub height
were significant predictors of increasing abundance of Warbling Vireo (Figure 12 bottom),
whereas only increasing shrub height was a significant predictor of increasing Orange-crowned
Warbler abundance, especially in recently harvested stands. Decreasing percentage shrub
cover and increasing softwoods were significant predictors of increasing abundance of Ruby-
crowned Kinglet (Figure 14) and Yellow-rumped Warbler (Figure 15). A negative year-effect was
observed for Lincoln‘s Sparrow, Ruby-crowned Kinglet, and Warbling Vireo, and a positive year-
effect was observed for Swainson‘s Thrush (Table 5).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
19
Figure 8 Mean Species Richness (top) and Abundance (bottom) from 100 m Radius Point-count Stations in Class 4 Stand Types
Percentage cut-off values for each stand type are: Hardwood (95–100% hardwood), Hardwood-leading (75–90% hardwood), Mixedwood (30–70% hardwood or softwood), Softwood-leading (75–90% softwood), and Softwood (95–100% softwood). Bars represent standard error; n represents number of sites sampled.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
20
Table 5 Model Results of Species Responses to Habitat and Year Effects
Species Model Parameter
β p Model
Deviance Model AIC Model ω
ALFL 1 Intercept -1.767 <0.001 327.7 335.7 0.295
Shrub Cover 0.017 <0.001
DEJU
2 Intercept -2.013 <0.001 1,357.0 1,365.0 0.064
% Softwoods 0.002 0.213
GCKI Intercept -11.214 <0.001 473.2 483.2 0.077
Forest Class 1.791 <0.001
% Softwoods 0.015 0.008
LISP Intercept -1.143 0.005 686.8 698.8 0.307
Forest Class -0.726 <0.001
Shrub Cover 0.023 <0.001
Year -0.307 0.005
OCWA Intercept -0.944 0.218 413.9 423.9 0.259
Forest Class -1.487 <0.001
Shrub Height 0.761 <0.001
OVEN Intercept -4.61 <0.001 326.2 336.2 0.218
Forest Class 0.548 0.098
% Hardwoods 0.035 <0.001
RBNU Intercept -7.061 <0.001 385.1 395.1 0.199
Forest Class 0.985 0.006
% Hardwoods 0.012 0.044
RCKI Intercept -2.966 <0.001 969.7 981.7 0.142
Shrub Cover -0.014 <0.001
% Softwoods 0.017 <0.001
Year -0.243 0.002
SWTH Intercept -3.546 <0.001 1,002.0 1,012.0 0.082
Forest Class 0.360 <0.001
Year 0.245 0.001
WAVI Intercept -3.772 0.041 1,021.0 1,033.0 0.138
Shrub Height 0.448 <0.001
% Hardwoods 0.022 <0.001
Year -0.147 0.040
YRWA
3 Intercept -1.884 <0.001 1,535.0 1,547.0 0.657
Forest Class 0.304 <0.001
Shrub Cover -0.006 0.001
% Softwoods 0.006 <0.001
Species codes with full English names are in Appendix 2.
1 Model using data subset Forest Class = 2 because counts of Forest Class 2, 3, 4 significantly different from
expected Poisson (p ≤ 0.001). 2 Model not significantly different from null (p = 0.192).
3 Counts significantly different from expected Poisson (p ≤ 0.001). Could not be resolved by using a data subset.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
21
Figure 9 Effect of Percentage Shrub Cover on the Expected Number of Birds per Plot for Alder Flycatcher.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
22
Figure 10 Effect of Percentage Softwoods (top) and Percentage Shrub Cover (bottom), among Forest Classes (ForClass) on the Expected Number of Birds per Plot for Golden-crowned Kinglet and Lincoln’s Sparrow, Respectively
Golden-crowned Kinglet
Lincoln’s Sparrow
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
23
Figure 11 Effect of Shrub Height (top) and Percentage Hardwoods (bottom), among Forest Classes (ForClass) on the Expected Number of Birds per Plot for Orange-crowned Warbler and Red-breasted Nuthatch, Respectively
Orange-crowned Warbler
Red-breasted Nuthatch
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
24
Figure 12 Effect of Forest Class (ForClass; top) and Percentage Hardwoods among Shrub Height Groups (bottom), on the Expected Number of Birds per Plot for Swainson’s Thrush and Warbling Vireo, Respectively
Swainson’s Thrush
Warbling Vireo
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
25
Figure 13 Effect of Percentage Hardwoods among Forest Classes (ForClass) on the Expected Number of Birds per Plot for Ovenbird
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
26
Figure 14 Effect of Percentage Softwoods and Shrub Cover on the Expected Number of Birds per Plot for Ruby-crowned Kinglet
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
27
Figure 15 Effect of Percentage Softwoods and Shrub Cover, among Forest Classes (ForClass), on the Expected Number of Birds per Plot for Yellow-rumped Warbler
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
28
4.1.5 Power to Detect Trend
Power to detect a significant trend was completed for 34 regularly occurring species that were
adequately represented in the BBS data. Species not included in the analysis, although fairly
common, were Gray Jay, Pine Siskin, and White-winged Crossbill because they were either: a)
generally non-vocal and likely underrepresented, or b) generally erratic in annual abundance
and not likely to show a meaningful trend over the short term.
Table 6 summarizes results of the power analysis for three time periods (10, 20 and 30 years),
and five levels of change (-10, -5, -3, -2 and -1%), for 34 species. Values > 0.8 are considered
adequate for reliably predicting the specified trend over the specified time interval, given
constant sampling effort (i.e., the same 268 BBS stations sampled annually) (Pannekoek and
van Strien 2005). Using the relationship between a species‘ rank abundance and the power (β)
to detect a -3% per year change in annual abundance, Figure 16 shows that the current
sampling efforts will prove inadequate for most species over a 10-year period. Only Yellow-
rumped Warbler and Chipping Sparrow are likely to provide evidence of a significant trend, if in
fact one exists. At the 20-year period, approximately half of all species assessed are likely to
show a significant trend if one exists, and at 30 years, virtually all species assessed will be
adequate for identifying a significant trend if one exists.
To reduce the number of years (e.g., from 30 to 10) required to detect a significant trend, one
option is to increase the number of samples in the area of interest. Thus, if we expanded our
analyses to accommodate the additional 212 BBS stations that are established, but not sampled
annually, the power to detect a significant trend earlier is expected to improve. Unfortunately,
statistical methods for conducting a power analysis using datasets with unequal sampling
among years are currently poorly developed. A second option for increasing statistical power is
to use a larger trend estimate (e.g., -5 or -10%). The risk with this option is that if the trend is
real, the species‘ population decrease is substantially greater, which and may hinder, or make
more expensive, the efforts necessary for recovery.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
29
Table 6 Power to Detect a Significant Decreasing Trend (if one exists) for 34 Bird Species Occurring in the Fort St. John TSA. Values in bold (> 0.8) are considered adequate for reliably predicting a trend for the specified period and rate of change.
Species Routes (Stations)
Trend Period
-10% -5% -3% -2% -1%
Alder Flycatcher 10 (268) 10 years 0.998 0.836 0.570 0.296 0.168
10 (268) 20 years 1.000 1.000 0.996 0.950 0.508
10 (268) 30 years 1.000 1.000 1.000 1.000 0.884
American Redstart 6 (164) 10 years 0.854 0.480 0.270 0.180 0.144
6 (164) 20 years 1.000 0.980 0.802 0.602 0.270
6 (164) 30 years 1.000 1.000 0.992 0.910 0.534
American Robin 10 (268) 10 years 0.986 0.696 0.394 0.268 0.156
10 (268) 20 years 1.000 1.000 0.980 0.780 0.396
10 (268) 30 years 1.000 1.000 1.000 0.990 0.754
Black-and-white Warbler 6 (150) 10 years 0.520 0.266 0.188 0.148 0.112
6 (150) 20 years 0.972 0.758 0.488 0.304 0.152
6 (150) 30 years 0.996 0.948 0.828 0.644 0.284
Black-capped Chickadee 7 (180) 10 years 0.742 0.348 0.224 0.164 0.114
7 (180) 20 years 0.998 0.864 0.654 0.486 0.218
7 (180) 30 years 1.000 0.998 0.946 0.792 0.412
Blackpoll Warbler 6 (164) 10 years 0.636 0.330 0.212 0.162 0.084
6 (164) 20 years 0.968 0.822 0.550 0.386 0.162
6 (164) 30 years 0.998 0.980 0.850 0.660 0.324
Black-throated Green Warbler 6 (164) 10 years 0.598 0.308 0.204 0.146 0.120
6 (164) 20 years 0.952 0.812 0.604 0.356 0.156
6 (164) 30 years 0.984 0.938 0.820 0.654 0.338
Boreal Chickadee 9 (244) 10 years 0.682 0.300 0.178 0.138 0.100
9 (244) 20 years 0.996 0.854 0.592 0.366 0.212
9 (244) 30 years 1.000 0.996 0.914 0.720 0.330
Chipping Sparrow 10 (268) 10 years 1.000 0.980 0.782 0.546 0.238
10 (268) 20 years 1.000 1.000 1.000 0.998 0.770
10 (268) 30 years 1.000 1.000 1.000 1.000 0.990
Common Yellowthroat 8 (208) 10 years 0.772 0.392 0.256 0.210 0.114
8 (208) 20 years 0.992 0.926 0.698 0.476 0.222
8 (208) 30 years 1.000 0.996 0.958 0.834 0.444
Dark-eyed Junco 10 (268) 10 years 0.996 0.816 0.486 0.276 0.132
10 (268) 20 years 1.000 1.000 0.990 0.908 0.454
10 (268) 30 years 1.000 1.000 1.000 0.994 0.838
Fox Sparrow 8 (224) 10 years 0.708 0.454 0.252 0.148 0.116
8 (224) 20 years 0.978 0.876 0.666 0.488 0.218
8 (224) 30 years 0.996 0.986 0.928 0.818 0.468
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
30
Species Routes (Stations)
Trend Period
-10% -5% -3% -2% -1%
Golden-crowned Kinglet 9 (238) 10 years 0.958 0.654 0.302 0.220 0.160
9 (238) 20 years 1.000 1.000 0.942 0.722 0.380
9 (238) 30 years 1.000 1.000 1.000 0.984 0.650
Hermit Thrush 10 (268) 10 years 0.996 0.766 0.458 0.290 0.138
10 (268) 20 years 1.000 1.000 0.984 0.850 0.418
10 (268) 30 years 1.000 1.000 1.000 0.998 0.794
Least Flycatcher 7 (180) 10 years 0.748 0.424 0.270 0.184 0.146
7 (180) 20 years 0.990 0.892 0.708 0.506 0.238
7 (180) 30 years 1.000 0.992 0.948 0.814 0.454
Lincoln's Sparrow 10 (268) 10 years 1.000 0.904 0.634 0.412 0.174
10 (268) 20 years 1.000 1.000 1.000 0.942 0.626
10 (268) 30 years 1.000 1.000 1.000 1.000 0.934
Magnolia Warbler 8 (210) 10 years 0.848 0.420 0.232 0.166 0.110
8 (210) 20 years 0.998 0.958 0.758 0.542 0.258
8 (210) 30 years 1.000 1.000 0.982 0.882 0.494
Northern Waterthrush 6 (164) 10 years 0.600 0.304 0.172 0.106 0.064
6 (164) 20 years 0.954 0.738 0.524 0.346 0.168
6 (164) 30 years 0.994 0.954 0.832 0.612 0.324
Olive-sided Flycatcher 7 (186) 10 years 0.468 0.278 0.182 0.128 0.136
7 (186) 20 years 0.742 0.640 0.452 0.322 0.214
7 (186) 30 years 0.832 0.770 0.712 0.568 0.298
Orange-crowned Warbler 9 (238) 10 years 0.978 0.734 0.410 0.276 0.134
9 (238) 20 years 1.000 1.000 0.976 0.808 0.386
9 (238) 30 years 1.000 1.000 1.000 0.996 0.792
Ovenbird 6 (150) 10 years 0.926 0.584 0.386 0.204 0.130
6 (150) 20 years 1.000 0.986 0.874 0.672 0.310
6 (150) 30 years 1.000 1.000 0.998 0.960 0.598
Red-breasted Nuthatch 9 (240) 10 years 0.824 0.438 0.280 0.146 0.098
9 (240) 20 years 1.000 0.978 0.792 0.504 0.258
9 (240) 30 years 1.000 1.000 0.986 0.856 0.484
Red-eyed Vireo 8 (210) 10 years 0.806 0.430 0.266 0.162 0.088
8 (210) 20 years 0.994 0.950 0.718 0.524 0.212
8 (210) 30 years 1.000 1.000 0.972 0.864 0.416
Rose-breasted Grosbeak 8 (208) 10 years 0.762 0.394 0.280 0.190 0.106
8 (208) 20 years 0.992 0.934 0.740 0.536 0.248
8 (208) 30 years 1.000 0.994 0.960 0.838 0.444
Ruby-crowned Kinglet 10 (268) 10 years 1.000 0.870 0.550 0.354 0.168
10 (268) 20 years 1.000 1.000 0.998 0.962 0.576
10 (268) 30 years 1.000 1.000 1.000 1.000 0.948
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
31
Species Routes (Stations)
Trend Period
-10% -5% -3% -2% -1%
Swainson's Thrush 10 (268) 10 years 1.000 0.930 0.660 0.390 0.216
10 (268) 20 years 1.000 1.000 0.998 0.990 0.634
10 (268) 30 years 1.000 1.000 1.000 1.000 0.950
Tennessee Warbler 9 (238) 10 years 1.000 0.922 0.662 0.462 0.178
9 (238) 20 years 1.000 1.000 0.998 0.976 0.660
9 (238) 30 years 1.000 1.000 1.000 1.000 0.960
Varied Thrush 8 (222) 10 years 0.764 0.426 0.234 0.176 0.120
8 (222) 20 years 0.996 0.930 0.712 0.506 0.210
8 (222) 30 years 1.000 0.998 0.954 0.828 0.448
Warbling Vireo 9 (238) 10 years 1.000 0.944 0.696 0.480 0.256
9 (238) 20 years 1.000 1.000 1.000 0.980 0.712
9 (238) 30 years 1.000 1.000 1.000 1.000 0.968
Western Tanager 7 (180) 10 years 0.760 0.406 0.238 0.164 0.134
7 (180) 20 years 0.996 0.928 0.738 0.502 0.228
7 (180) 30 years 1.000 0.996 0.968 0.866 0.452
White-throated Sparrow 10 (268) 10 years 1.000 0.934 0.680 0.432 0.192
10 (268) 20 years 1.000 1.000 0.996 0.962 0.590
10 (268) 30 years 1.000 1.000 1.000 1.000 0.954
Wilson's Warbler 7 (178) 10 years 0.666 0.334 0.238 0.146 0.106
7 (178) 20 years 0.968 0.844 0.598 0.390 0.176
7 (178) 30 years 0.994 0.958 0.902 0.748 0.388
Yellow-bellied Sapsucker 9 (238) 10 years 0.912 0.626 0.314 0.170 0.138
9 (238) 20 years 1.000 0.988 0.896 0.662 0.330
9 (238) 30 years 1.000 1.000 1.000 0.966 0.642
Yellow-rumped Warbler 10 (268) 10 years 1.000 0.992 0.882 0.652 0.298
10 (268) 20 years 1.000 1.000 1.000 1.000 0.862
10 (268) 30 years 1.000 1.000 1.000 1.000 0.998
Based on data collected annually from 2005–2009.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
32
Figure 16 Relationship between Species’ Rank Abundance and Probability of Detecting a -3% Change in Annual Abundance for Three Time Intervals
All values above the dashed line are considered reliable trend estimates (β ≥ 0.8). Note: not all ranks are represented because some species are not conducive to power analysis given current survey methods (e.g., Gray Jay, White-
winged Crossbill, Pine Siskin).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
33
4.1.6 Species’ Trend Estimates
4.1.6.1 Fort St. John TSA
Trend was evaluated for 33 species and results are summarized in Table 7. A significant
change in population was observed for 11 species, with 6 showing an increase, and 5 showing
a decrease. Among those increasing, Alder Flycatcher and American Redstart had strong
trends, and Black-throated Green Warbler, Red-breasted Nuthatch, Warbling Vireo, and Yellow-
rumped Warbler had moderate trends. Observed values for Alder Flycatcher, however, did not
fit the expected model ( Goodness of Fit p <0.05; Table 7) and may be less reliable than other
species‘ models. Among those species decreasing, Ruby-crowned Kinglet and White-throated
Sparrow had strong trends, and Fox Sparrow, Golden-crowned Kinglet, and Lincoln‘s Sparrow
had moderate trends. Observed values for Fox Sparrow and Ruby-crowned Kinglet; however,
appeared not to fit the expected models ( Goodness of Fit p <0.05; Table 7) and thus should
be interpreted cautiously. Species with significant trends are illustrated in Figure 17 to Figure
22.
In the Fort St. John TSA, 14 species showed stable populations, and for eight species trend
could not be assessed with certainty (Table 7). Among those species with a significant trend,
American Redstart and Red-breasted Nuthatch had strong peaks in abundance in 2007 and
2008, respectively. All remaining species showed relatively little inter-annual variation in their
respective indices of abundance, although it is notable that Black-throated Green Warbler
increased nearly four times in 2007 from 2005–06 levels, and has since remained at least two
times greater than 2005–06 levels.
By comparing those species currently showing a significant trend with the results of their
respective power analysis (see Table 6), two inferences are possible: 1) all species‘ populations
except Yellow-rumped Warbler are changing at rates > ±3%/yr assuming the current trend
(based on 2005–2009 data) persists over a 10-year period, or 2) all species‘ populations are
changing at rates = ± 3% per year, but over a much greater trend period (e.g., 20 or 30 years).
Additional inaccuracies in trend results may arise from natural fluctuations in a species‘ annual
abundance over the short term. For example, for a species with large fluctuations in short-term
annual abundance (e.g., Red-breasted Nuthatch), the resulting short-term trend may not be
indicative of long-term patterns. Trend estimates for species with considerably less variation in
annual abundance (e.g., Yellow-rumped Warbler) are more likely to show a significant trend
earlier, rather than later.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
34
4.1.6.2 Provincial Context
We compared species‘ trends observed in the Fort St. John TSA to those observed in the rest of
British Columbia for regional context and to identify potential local anomalies. Among the 25
species for which a significant trend in the Fort St. John TSA was observed, 6 species had
trends consistent with those at the provincial scale (Sauer et al. 2008; Table 8). Two species
present in the Fort St. John TSA, Red-breasted Nuthatch and Warbling Vireo, had local and
provincial increases, and Least Flycatcher, Magnolia Warbler, Northern Waterthrush, and
Varied Thrush had stable populations at the local and provincial scale. Among the five species
that may have significant declines in the Fort St. John TSA, at the provincial scale two are
considered stable and three have uncertain trends. Additionally, Olive-sided Flycatcher,
Orange-crowned Warbler, Wilson‘s Warbler, and Red-eyed Vireo appear to be stable in the Fort
St. John TSA, but appear to be declining provincially (Table 8). Nine species for which data from
the Fort St. John TSA were amenable to a reliable trend prediction had uncertain or unknown
trends at the provincial scale (Sauer et al. 2008), including three species showing a decrease
(Golden-crowned Kinglet, Lincoln‘s Sparrow, and White-throated Sparrow), and one species
showing an increase (Black-throated Green Warbler) (Table 8).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
35
Table 7 Trend Results for 33 Regularly Occurring Species in the Fort St. John TSA, 2005–2009. Values in bold indicate significance (p < 0.05)
Goodness of Fit Deviance from Linear Trend
Trend Pattern and Significance
Species P Wald ( ) P Trend p
Alder Flycatcher 1 62.13 <0.005 24.88 <0.001 Strong increase <0.05
American Redstart 16.64 0.676 32.60 <0.001 Strong increase <0.05
Black-and-white Warbler 13.89 0.836 5.59 0.133 Stable Black-capped Chickadee 37.82 0.036 7.33 0.062 Uncertain Blackpoll Warbler 18.55 0.551 4.49 0.213 Stable Black-throated Green Warbler 13.74 0.843 9.22 0.027 Moderate increase <0.05
Boreal Chickadee 51.03 0.018 4.62 0.202 Uncertain Chipping Sparrow 53.74 0.029 3.64 0.303 Uncertain Common Yellowthroat 27.52 0.490 6.35 0.096 Stable Dark-eyed Junco 97.89 <0.001 17.30 <0.001 Uncertain Fox Sparrow
1 47.10 0.013 5.66 0.129 Moderate decline <0.05
Golden-crowned Kinglet 27.01 0.718 17.81 <0.001 Moderate decline <0.05
Hermit Thrush 64.14 <0.003 5.90 0.117 Uncertain Least Flycatcher 26.26 0.340 0.15 0.985 Stable Lincoln's Sparrow 50.43 0.056 4.55 0.208 Moderate decline <0.01
Magnolia Warbler 27.68 0.482 5.28 0.153 Stable Northern Waterthrush 28.42 0.099 1.02 0.796 Stable Olive-sided Flycatcher 26.65 0.321 0.04 0.946 Stable Orange-crowned Warbler 35.32 0.314 6.70 0.082 Stable Ovenbird 20.02 0.457 5.35 0.147 Stable Red-breasted Nuthatch 27.62 0.688 36.35 <0.001 Moderate increase <0.05
Red-eyed Vireo 35.52 0.155 6.49 0.090 Stable Rose-breasted Grosbeak 51.11 0.005 1.41 0.704 Uncertain Ruby-crowned Kinglet
1 65.68 0.002 7.59 0.055 Strong decline <0.01
Swainson's Thrush 69.21 <0.001 8.41 0.038 Uncertain Tennessee Warbler 49.93 0.023 5.24 0.155 Uncertain Varied Thrush 36.48 0.131 2.22 0.528 Stable Warbling Vireo 30.64 0.535 9.39 0.025 Moderate increase <0.01
Western Tanager 33.24 0.099 1.82 0.610 Stable White-throated Sparrow 22.11 0.966 5.36 0.148 Strong decline <0.01
Wilson's Warbler 23.82 0.472 4.12 0.249 Stable Yellow-bellied Sapsucker 44.19 0.074 6.79 0.079 Stable Yellow-rumped Warbler 48.47 0.080 0.87 0.833 Moderate increase <0.05
1 although the trend is significant, the overall model has poor Goodness of Fit (p >0.05).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
36
Figure 17 Five-year Trend for Alder Flycatcher (top) and American Redstart (bottom)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
37
Figure 18 Five-year Trend for Black-throated Green Warbler (top) and Golden-crowned Kinglet (bottom)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
38
Figure 19 Five-year Trend for Fox Sparrow (top) and Lincoln’s Sparrow (bottom)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
39
Figure 20 Five-year Trend for Red-breasted Nuthatch (top) and Ruby-crowned Kinglet (bottom)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
40
Figure 21 Five-year Trend for Warbling Vireo (top) and White-throated Sparrow (bottom)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
41
Figure 22 Five-year Trend for Yellow-rumped Warbler
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
42
Table 8 Local (2005–2009; this study) and Provincial (1980–2007; Sauer et al. 2008) Comparison of Population Trends.
Species
Trend Estimate
Fort St. John (2005–2009)*
British Columbia (1980–2007)*
Alder Flycatcher Increasing 1 Stable
1
Red-breasted Nuthatch Increasing 2 Increasing
1
Warbling Vireo Increasing 2 Increasing
1
Black-throated Green Warbler Increasing 2 Not available
American Redstart Increasing 2 Stable
1
Yellow-rumped Warbler Increasing 2 Stable
1
Fox Sparrow Decreasing 1 Stable
1
Ruby-crowned Kinglet Decreasing 1 Stable
1
Golden-crowned Kinglet Decreasing 2 Uncertain
2
Lincoln's Sparrow Decreasing 2 Uncertain
2
White-throated Sparrow Decreasing 2 Uncertain
2
Olive-sided Flycatcher Stable 1 Decreasing
1
Orange-crowned Warbler Stable 1 Decreasing
1
Wilson's Warbler Stable 1 Decreasing
1
Red-eyed Vireo Stable 1 Decreasing
2
Common Yellowthroat Stable 1 Increasing
1
Black-and-white Warbler Stable 1 Not available
Least Flycatcher Stable 1 Stable
1
Magnolia Warbler Stable 1 Stable
1
Northern Waterthrush Stable 1 Stable
1
Varied Thrush Stable 1 Stable
1
Blackpoll Warbler Stable 1 Uncertain
2
Western Tanager Stable 1 Uncertain
2
Ovenbird Stable 1 Uncertain
3
Yellow-bellied Sapsucker Stable 1 Uncertain
3
* Species in bold have a statistically significant trend 1 Data has few deficiencies (e.g., goodness of fit satisfied, good abundance and representation)
2 Data has deficiencies (e.g., poor goodness of fit, too little data, poor route representation)
3 Data has important deficiencies (e.g., very low route abundance, very low route representation)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
43
4.2 DIRECTED WARBLER SURVEYS
4.2.1 Sample Intensity and Representation
All Directed Warbler sample stations were located in the BWBS biogeoclimatic zone. The
majority of directed Warbler survey stations were located in the Blueberry operating area,
followed by Lower Beatton, Kobes, and Tommy Lakes. When combined with the BBS data (n =
584 survey stations), 7 of 11 landscape units were sampled, and 20 of 45 operating areas were
sampled.
4.2.2 Summary of Occurrences
In 2008 and 2009 the combined BBS and Directed Warbler sampling total of listed warbler
occurrences was 222. Of these, 173 (77.9%) occurrences were from the directed warbler
sampling component of the study. Relative to survey effort (i.e., number of stations sampled),
the percentage of occupied stations was higher on average for all species using the directed
sampling method compared to the BBS method (Table 9). Compared to 2008, the percentage of
directed warbler sampling stations occupied by a listed warbler was lower in 2009, although a
large proportion of the decrease appears to be attributed to habitat change at 12 stations (see
Section 4.2.3 below). A similar pattern between methods is observed for total warbler
abundance (Table 10).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
44
Table 9 Percentage of Survey Stations Occupied by Listed Warbler Species from Breeding Bird Surveys and Directed Warbler Surveys (n = number of survey stations)
Species Percentage of Occupied Stations (%)
Breeding Bird Surveys (n=480) Directed Warbler Sampling (n=104)
2008 2009 Average 2008 2009 Average
BAYW 0.21 0.21 0.21 0.00 0.96 0.48
BTNW 3.33 2.29 2.81 42.31 32.70 37.51
CAWA 0.21 0.21 0.21 23.08 13.46 18.27
CMWA 0.83 0.63 0.73 1.92 0.96 1.44
COWA 0.83 0.63 0.73 10.58 8.65 9.62
Occupied plots, all species combined
4.58 3.75 x = 4.17 65.38 47.11 x = 56.25
*Species codes and full names are provided in Appendix 2
Table 10 Abundance of Listed Warbler Species from Breeding Bird Surveys and Directed Warbler Surveys (n = number of survey stations)
Species Abundance
Breeding Bird Surveys (n=480) Directed Warbler Sampling (n=104)
2008 2009 Total 2008 2009 Total
BAYW 1 1 2 0 1 1
BTNW 18 12 30 58 44 102
CAWA 1 1 2 29 17 46
CMWA 4 3 7 2 1 3
COWA 4 3 7 11 10 21
Totals 28 20 48 100 73 173
Birds/station 0.058 0.042 x = 0.05 0.96 0.70 x = 0.83
*Species codes and full names are provided in Appendix 2
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
45
4.2.3 Effects of Habitat Change
Among the 104 Directed Warbler survey stations sampled, 8 were partially harvested and 4
were completely harvested after completion of the 2008 survey, but prior to the 2009 survey
(Figure 23). A mixed log-linear model was defined, which used ―Habitat Change‖ and ―Year‖ as
fixed predictors of warbler abundance. Using ―Survey Station‖ as a random variable accounted
for temporal autocorrelation between sampling stations in successive years. The effect of
―Habitat Change‖ on listed warbler abundance was significant (p = 0.01), with abundance being
significantly lower after harvest. There was also a significant interaction between ―Year‖ and
―Habitat Change‖ (p = 0.026) indicating a significant difference in warbler abundance between
years, even after accounting for change in habitat. Figure 24 illustrates the modeled effect of
―Habitat Change‖ and ―Year‖ on warbler abundance, and illustrates the relatively small ―Year‖
effect compared to the larger ―Habitat Change‖ effect.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
46
Figure 23 Effect of Partial or Complete Harvesting (i.e., changed) on Combined Abundance of Listed Warblers Compared to Non-harvested (i.e., unchanged) Sites Among Years.
Figure 24 Interaction Between “Year” and “Habitat” (i.e., Changed or Unchanged) on Total Abundance of all Listed Warbler Species Combined
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
47
4.2.4 Habitat Models
We developed habitat models for three of the five listed warbler species: Black-throated Green
Warbler, Canada Warbler and Connecticut Warbler (Table 11). The other two species (Bay-
breasted Warbler and Cape May Warbler) had insufficient sample sizes. The predictive
accuracy of the models was fairly good, with ROC values ranging from 0.77–0.88 (Table 11),
indicating that these models are able to discriminate adequately between occupied and
unoccupied sites. Each model included five to seven habitat covariates, all of which were
significant at the p <0.1 level, and all except two were significant at the p <0.05 level. For all
three species, the interaction between forest type and stand age were important predictors of
species occupancy indicating that the species were more likely to be present in older deciduous,
coniferous, and mixedwood forests than younger forests of the same type. Figure 25 provides a
sample habitat suitability map for Black-throated Green Warbler that was generated by linking
the logistic regression coefficients to ArcGIS.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
48
Table 11 Logistic Regression Model Results of Three of the Five Listed Warbler Species
Species* Model Parameter β p Model
Deviance Drop in
Deviance ROC Area
BTNW Intercept -4.566 <0.001 1142.0 166.4 0.771
RECENT 2.680 <0.001
DECID -1.628 0.075
MIXED 1.581 0.017
AGE -0.075 0.043
MIXED x AGE 0.083 0.021
CONIF x AGE 0.083 0.023
DECID x AGE 0.111 0.003
CAWA Intercept -4.185 <0.001 469.9 51.3 0.775
DECID -1.892 0.019
CONIF -3.287 <0.001
DECID x AGE 0.028 <0.001
CONIF x AGE 0.012 0.038
MIXED x AGE 0.004 0.062
COWA Intercept -6.326 <0.001 295.0 76.6 0.883
RECENT 2.346 0.021
DECID 4.061 <0.001
CONIF 4.428 <0.001
AGE -0.065 0.007
DECID x AGE 0.078 0.001
MIXED x AGE 0.057 0.021
* Species codes with full English names can be found in Appendix 2.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
49
Figure 25 Sample Predicted Probability of Occurrence Map Reclassified into Four Broad Classes for Black-throated Green Warbler
white: not suitable; increasing shades of grey indicate increasing suitability
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
50
4.2.5 Power and Precision
Scenario 1–Estimating warbler occupancy with a specified level of precision
Figure 26 summarizes the precision of the estimates for the BBS and Warbler sampling (model-
based approach) as measured using the width of the 95% confidence interval for a given
proportion/sample size. The desired width for the 95% confidence interval for scientific (±10%),
management (±25%), and preliminary (±50%) surveys are indicated using horizontal dashed
lines. Vertical lines indicate the number of stations sampled in 2008–2009.
200 400 600 800 1000
01
00
20
03
00
40
0
Number of bird survey stations
Re
lative
SE
of e
stim
ate
d o
ccu
pa
ncy (
%)
Fort
St
John s
am
ple
siz
e (
2008-2
009)
Preliminary surv ey s
Management surv ey s
Scientif ic surv ey s
Precision of Roadside Surveys
Species (prevalence)
BAYW (0.002)BTNW (0.028)CAWA (0.002)
CMWA (0.007)COWA (0.007)
100 200 300 400 500
05
01
00
15
02
00
Number of bird survey stations
Re
lative
SE
of e
stim
ate
d o
ccu
pa
ncy (
%)
Fort
St
John s
am
ple
siz
e (
2008-2
009)
Preliminary surv ey s
Management surv ey s
Scientif ic surv ey s
Precision of Model-based Surveys
Species (prevalence)
BAYW (0.005)BTNW (0.375)CAWA (0.186)
CMWA (0.014)COWA (0.096)
Figure 26 Precision of Estimates for the Breeding Bird Survey and Directed Warbler Sampling as Measured using the Width of 95% Confidence Interval for a given Proportion/Sample Size
Desired width of 95% confidence interval of ±10, or 25 to 50%, are indicated using horizontal dashed lines while sample size in 2008-2009 is shown as vertical lines. Note that BAYW and CAWA lines, and CMWA and COWA lines, overlap in the left graph.
Given the extremely low proportion of stations in which warblers at risk were detected using the
BBS, no species except for Black-throated Green Warbler (preliminary survey) were close to the
target precision for either of the three survey targets (i.e., preliminary, management, or
scientific). Many more stations would need to be surveyed to approach a reasonable level of
precision for the other four species. In contrast, the prevalence of four of the warbler species
(except for BAYW) was much higher using Directed Warbler Sampling compared to either the
BBS or the Forest Interior Surveys. Black-throated Green Warbler prevalence (37.5%) was
approximately twice that of Canada Warbler (18.27%). Sample size was sufficient for both
species to estimate occupancy within ± 50% of the true estimate. Black-throated Green Warbler
also achieved the target for a management type survey, the only combination or species/survey
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
51
type to do so. Connecticut Warbler and Cape May Warbler did not achieve the minimum
precision target level of preliminary surveys, although in the case of the former species only 20
more stations would be needed.
Scenario 2–Detecting changes in occupancy between two points in time
Figure 27 depicts the relationship, for Black-throated Green Warbler, between the number of
survey stations and the proportion change in occupancy from a baseline year that can be
detected. Clearly, this number is prohibitive using the BBS, with approximately 4,000 stations
needing to be sampled in two different years to detect a 30% change. The number of BBS
stations required for the other four listed warbler species is even greater. Conversely, using the
Directed Warbler Sampling approach, just over 200 stations would need to be surveyed to
detect a 30% change in Black-throated Green Warbler occupancy between two points in time.
Similarly, approximately 550 and 1,100 stations, respectively, would need to be surveyed to
detect a 30% change in the occupancy of Canada Warbler and Connecticut Warbler (not shown
in graph). As discussed in other sections of this report, fewer stations of either survey type
would be needed to detect downward or upward trends in occupancy if the surveys were
conducted annually.
Figure 27 Sample Size Required to Detect Changes in Occupancy from Baseline Year to a Future Survey Year
Intersection of dashed lines indicates sample size required to detect a 30% change in occupancy. Empty cells indicate that species was not detected using a given survey type in a given region.
0.2 0.3 0.4 0.5 0.6 0.7 0.8
10
02
00
30
04
00
50
0
Proportion change in occupancy from baseline
Nu
mb
er
of b
ird
su
rve
y s
tatio
ns
Model-based Survey - Black-throated Green Warbler
Baseline occupancy:
0.375
0.2 0.3 0.4 0.5 0.6 0.7 0.8
02
00
04
00
06
00
08
00
01
00
00
Proportion change in occupancy from baseline
Nu
mb
er
of b
ird
su
rve
y s
tatio
ns
Roadside Survey - Black-throated Green Warbler
Baseline occupancy:
0.028
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
52
4.3 BIRD SPECIES OF CONSERVATION CONCERN
4.3.1 Summary
The British Columbia Conservation Data Centre (CDC) recognizes 20 species as either red- or
blue-listed in the Peace Forest District, which includes the Fort St. John TSA (CDC 2010). The
Committee on the Status of Endangered Wildlife in Canada (COSEWIC) recognizes four of
those species as Special Concern and two of those species as Threatened (COSEWIC 2009).
In addition, COSEWIC also recognizes Common Nighthawk as Threatened. Among the 21
species of conservation concern, we have documented 353 occurrences of nine species for the
period 2005–2009 (Table 12). In 2009, 37 observations were from the BBS, and 76
observations were from the Directed Warbler surveys.
Since completion of the 2008 annual report (Preston 2009), the following changes have
occurred to species listings in the Peace Forest District:
Brant (Branta bernicla): added to the Peace Forest District on the basis of a single
occurrence record during autumn migration in 1986 (Campbell et al. 1990)
Rough-legged Hawk (Buteo lagopus): added to the Peace Forest District; there are no
confirmed breeding records in British Columbia (Campbell et al. 1990). In the Peace
Forest District it migrates through the area and occasionally overwinters (Campbell et al.
1990)
Sandhill Crane (Grus canadensis): status downgraded to ―yellow‖
Common Nighthawk (Chordeiles minor): added to COSEWIC rankings table as
―Threatened‖
Olive-sided Flycatcher (Contopus cooperi): added to the CDC rankings table as ―blue‖
Rusty Blackbird (Euphagus carolinus): added to the CDC rankings table as ―blue‖
Details of individual records, including GPS coordinates and dates of observation, are included
in data files submitted to Canadian Forest Products Ltd., and the Wildlife Species Inventory
(WSI) database, as per the 2009/2010 work plan (see Appendix 1). Maps depicting the locations
of all species of conservation concern are provided in Figure 28 to Figure 36.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
53
Table 12 Number of Bird Records for Species of Conservation Concern from Different Components of the Monitoring Program, 2005–2009.
Species
Provincial/Federal Status Number of Records
CDC Rank COSEWIC
Rank BBS Interior Warblers Total
American Bittern Blue 0 0 0 0
Barn Swallow Blue 7 0 0 7
Bay-breasted Warbler Red 5 1 1 7
Black-throated Green Warbler Blue 57 8 102 167
Brant Blue 0 0 0 0
Broad-winged Hawk Blue 1 0 0 1
Canada Warbler Blue Threatened 6 0 46 52
Cape May Warbler Red 24 2 3 29
Common Nighthawk Yellow Threatened 0 0 0 0
Connecticut Warbler Red 9 0 21 30
Le Conte's Sparrow Blue 16 0 2 18
Nelson's Sharp-tailed Sparrow Red Not At Risk 0 0 0 0
Olive-sided Flycatcher Blue Threatened 40 0 2 42
Peregrine Falcon Red Special Concern
0 0 0 0
Rough-legged Hawk Blue 0 0 0 0
Rusty Blackbird Blue Special Concern
0 0 0 0
Short-eared Owl Blue Special Concern
0 0 0 0
Surf Scoter Blue 0 0 0 0
Swainson's Hawk Red 0 0 0 0
Upland Sandpiper Red 0 0 0 0
Yellow Rail Red Special Concern
0 0 0 0
Totals 165 11 177 353
BBS = Breeding Bird Survey, Interior = Interior forest surveys conducted in 2007, Warblers = directed warbler sampling conducted in 2008 and 2009. Species sorted alphabetically.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
54
Figure 28 Locations of Barn Swallow Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
55
Figure 29 Locations of Bay-breasted Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
56
Figure 30 Locations of Black-throated Green Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
57
Figure 31 Locations of Broad-winged Hawk Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
58
Figure 32 Locations of Canada Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
59
Figure 33 Locations of Cape May Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
60
Figure 34 Locations of Connecticut Warbler Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
61
Figure 35 Locations of Le Conte’s Sparrow Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
62
Figure 36 Locations of Olive-sided Flycatcher Occurrences as Detected from Breeding Bird Survey and Directed Warbler Sampling Methods
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
63
5.0 Summary and Conclusion
5.1 BREEDING BIRD SURVEYS
5.1.1 Species-habitat relationships
Stand-level habitat models were created for all species where data were suitable for testing. In
general, forest class and/or tree composition were important predictors for 10 of 11 species.
Species typically associated with old and mature forest stands, such as Golden-crowned
Kinglet, Ovenbird, Red-breasted Nuthatch, Swainson‘s Thrush, and Yellow-rumped Warbler
showed strong selection for FC3, FC4, and FC5 stand types, which agrees with general findings
from other studies (Van Horn and Donovan 1994; Ingold and Galati 1997; Hunt and Flaspohler
1998; Ghalambor and Martin 1999; Mack and Yong 2000). Species significantly associated with
younger stands types (e.g., FC1 and FC2) included Lincoln‘s Sparrow and Orange-crowned
Warbler, which is in general agreement with other studies (Sogge et al. 1994; Ammon 1995).
Tree composition (i.e., softwoods or hardwoods) was an important predictor for seven species,
and in general, hardwood-leading and mixedwood stands supported the highest average
species richness and abundance, followed by hardwood, softwood-leading, and softwood
stands. This is consistent with other studies that show higher bird richness and abundance in
mixedwood stands and lower species richness and abundance in coniferous stands (James and
Wamer 1982).
Shrub cover and shrub height were important predictors for several species, and were included
in species‘ models that represented all forest classes and composition types (e.g., hardwood,
mixedwood). Alder Flycatcher was the only species where shrub cover was the only significant
habitat variable, and although the analysis was constrained to Forest Class 1, the observed
habitat specificity is likely real. In Nova Scotia, peak abundance of Alder Flycatcher in 3 to 8
year-old clearcuts or burns corresponded with peak shrub stem density and foliage cover
(Morgan and Freedman 1986). The strong dependency on very high shrub cover (80–100%) in
young stands by Alder Flycatcher suggests that shrub control or management (e.g., range
management activities, brushing, herbicide) would likely have a negative effect on Alder
Flycatcher site occupancy. Also within young stands, Lincoln‘s Sparrow and Orange-crowned
Warbler responded positively to increasing shrub cover, and shrub height, respectively. In older
stands, particularly in softwood and softwood-leading, a lack of shrub cover was an important
predictor of increasing site occupancy by Ruby-crowned Kinglet and Yellow-rumped Warbler.
For species showing significant responses to shrub cover and shrub height, our findings are
consistent with those reported elsewhere (e.g., Sogge et al. 1994, Ammon 1995, Lowther 1999).
Among the species tested, strong responses to specific forest attributes including age,
composition, and structure suggests that forest management activities can be tailored to meet
objectives for species of management interest.
Generally, multi-year analyses were consistent with single-year analyses (see Preston 2008,
2009), although significant year-effects were observed in some species models (Lincoln‘s
Sparrow, Ruby-crowned Kinglet, Swainson‘s Thrush, and Warbling Vireo). The year effect
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
64
suggests that factors related to variation in annual conditions (e.g., climate, over-winter survival)
can result in significantly different levels of habitat occupancy for some species.
These results, in conjunction with other similar studies, indicate that the bird survey data and
associated rapid habitat assessment method is a viable approach for describing species-habitat
relationships in the absence of landscape-level data. Moreover, the habitat assessment used in
this study includes stand structure attributes such as shrub cover and shrub height that are not
readily available in VRI datasets. Our rapid-habitat assessment approach also accommodates
current conditions, and in a relatively short period of time (i.e., three years) is capable of
identifying site-specific changes and subsequent species‘ responses. Thus, a source of error
related to increasingly out-dated information from other datasets may be substantially reduced
by including this type of rapid habitat assessment on an annual basis (see example in Preston
2009, Table 10, page 69).
5.1.2 Trends
Monitoring of bird populations is an important component of conservation biology because it
facilitates the early identification of conservation problems and encourages the development of
possible solutions (Goldsmith 1991; James et al. 1996). Thus, trends are useful for identifying
those species for which decreases in population size may be occurring, which in turn can
provide guidance for determining which species require causal investigation. Following the
identification of causation, and the implementation of management plans, continued monitoring
can provide an appraisal of management action. Two possible monitoring options include:
1. Future trend estimates based on expected results of management action may be
modeled indirectly using predictions of habitat availability and known species-habitat
relationships at various spatial scales (see Vernier 2010)
2. Long-term monitoring can provide real-time results that can be used in conjunction with
model predictions for the purpose of model validation and refinement
Another component of trend monitoring is the capacity to identify and understand differences
between local (i.e., Fort St. John TSA) and regional (i.e., provincial or national) trends, which
can be an important tool for evaluating and implementing forest management options. For
example, if population trends for a particular species are declining provincially (based on results
from the North American BBS; Sauer et al. 2008), but not locally (e.g., Olive-sided Flycatcher),
management efforts directed at that species may be undervalued in terms of return on
investment. A common characteristic of the volunteer-based North American BBS is that many
of the routes occur on well-maintained roads at low elevations, often through rural or urban
landscapes and agriculture, which are not necessarily suitable habitats for many species
considered at risk. Forested regions, especially those that are extensive and managed for
forestry, are less well-represented by the North American BBS program, and concern has been
raised on the efficacy of these surveys to accurately predict trends of some forest-dependent
species (see Campbell et al. 2007). For example, in eastern Canada, where forests are
fragmented and gradually converted to agriculture or urban areas, declines in forest songbirds
largely correlate with habitat loss, but in managed forests of western boreal regions, rates and
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
65
direction of population change and the corresponding causes are more difficult to ascertain
(Marzluff and Sallabanks 1998; George and Dobkin 2002). In the absence of reliable provincial
trend information, identification of significant and locally changing population trends for a
species can allow forest managers to be adaptive. In the Fort St. John TSA we have identified
at least five species with significant declining population trends (Fox Sparrow, Golden-crowned
Kinglet, Lincoln‘s Sparrow, Ruby-crowned Kinglet, and White-throated Sparrow), for which
reliable provincial trends are not significant. Furthermore, we note that several other species
that currently do not have reliable provincial trend data, because of low sampling effort or patchy
distribution of the species‘ range (e.g., Cape May Warbler, Canada Warbler, Ovenbird, Yellow-
bellied Flycatcher, Alder flycatcher, Rose-breasted Grosbeak), may be better served by a local
monitoring program in the Fort St. John TSA (also see Campbell et al. 2007). The survey effort
that has been invested in the Fort St. John TSA from 2005–2009 has removed much uncertainty
in trend estimates for many species, including provincially-listed species.
5.2 DIRECTED WARBLER SURVEYS
5.2.1 Habitat Models
In general, habitat models developed for three of the five listed warbler species were consistent
with known habitat associations. Each model was comprised of five to seven habitat covariates
including, in each case, the interaction between forest type and stand age. This indicates that
the warblers, depending on the species, were more likely to be present in older deciduous,
coniferous, and mixedwood forests than younger forests of the same type – thus confirming
known natural history. We were not able to develop models for the other two species (Bay-
breasted Warbler and Cape May Warbler), primarily because of insufficient sample sizes. At
least four options are available to rectify this problem: 1) collect more data over time, 2)
combine data from nearby study areas, 3) use models developed in similar geographic areas
(e.g., Alberta boreal forest), or 4) use expert-based habitat models (e.g., HSI).
A common criticism of habitat-based models is that they are rarely validated (Vernier et al.
2008). As a first step, we evaluated the predictive accuracy of the three warbler models using
"in-sample" data and found them to be suitable for use in a management setting. However, we
caution against their use beyond the region from which the surveys were conducted. Moreover,
we suggest that independent tests of the models using "out-of-sample" data be conducted. For
example, the models could be tested using data from a different geographic region (e.g., TFL
48) or for a different time period (e.g., a year ahead). For this iteration of model development,
we attempted to simplify the models without losing predictive ability. To achieve this, we used
habitat variables directly from VRI attribute tables rather than first converting the attributes to
grids and generating "neighbourhood" variables using moving window analyses. This approach
worked well for the three listed warbler species, with the advantage that the models are
significantly easier to apply from a management perspective.
5.2.2 Power and Precision
We evaluated the usefulness of two types of bird monitoring surveys used in the Fort St. John
TSA to: 1) determine the overall level of occupancy for each species in a region or habitat strata
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
66
and, 2) detect changes in the level of occupancy between two years within a region. Results
were aimed at guiding refinements to the monitoring program by providing recommendations to
managers on the intensity and frequency of sampling in relation to their power and precision.
For each management scenario, the model-based surveys proved to be the most efficient
survey type for warbler species at risk. This is not surprising since such surveys specifically
target high quality habitat for those species. However, it is important to remember that although
the model-based surveys enable managers to detect key species more effectively, such surveys
are not representative of the larger landscape. The relatively low rate of detection of warblers
along the standard roadside surveys is likely due to the fact that those surveys are not efficient
at sampling habitat types that are sought by certain species, including the five warbler species.
Such habitats consist in large part of late seral deciduous, coniferous, and mixedwood forests –
the same forest types that are the focus of harvesting along logging roads.
Clear, specific and measurable management and sampling objectives are needed in order to
evaluate the efficiency of the different survey types. Management objectives such as detecting
changes in occupancy over time need to be matched to more specific sampling objectives.
Specifically, biologically relevant effect sizes or confidence intervals need to be specified -
whether as a single threshold or a range of acceptable conditions. In scenario 2, for example,
we specified a 30% change in occupancy between two time periods. Such decisions need to
balance biological and management importance and will likely result in a range of values rather
than a single threshold. Similarly, some thought needs to be given to the desired level of
precision that is required for decision making. Narrow confidence intervals will lead to
unrealistically high sample sizes for the five listed warbler species. Conversely, wide confidence
intervals will not be useful for guiding management actions for those species. Stratifying by
habitat type would likely reduce variability in occupancy for each species – however this would
be partially offset by reducing sample size of each stratum. This, in part, is the approach used in
model-based surveys. Applying it to the standard roadside surveys would be very difficult given
the current logging road networks and would require substantial flexibility in the number and
placement of stations within transects.
5.3 SPECIES OF CONSERVATION CONCERN
Since inception of the bird monitoring program in 2005, 353 observations of 9 species of
conservation concern (i.e., CDC, COSEWIC, SARA) have been recorded. All observations have
been submitted to the Wildlife Species Inventory database, which are subsequently available to
the BC Conservation Data Centre. Many of the observations represent new information on the
distribution and timing of occurrence in British Columbia. In particular, data resulting from the
warbler sampling project has resulted in several new observations of Black-throated Green
Warbler, Connecticut Warbler, Cape May Warbler, and Canada Warbler for areas where
information was previously lacking.
We did not expect to observe several species of conservation concern in the Peace Forest
District for two main reasons. First, some species are very rare and unlikely to be seen in the
area (e.g., Brant, Swainson‘s Hawk), and second, some species are restricted to habitat types
that are not sampled adequately (e.g., Nelson‘s Sharp-tailed Sparrow, Yellow Rail, American
Bittern) using methods designed to survey forest-dependent species.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
67
5.4 CONCLUSION
The collective database for birds, since inception of the program in 2005, includes 15,563
records from the BBS (2005–2009), 1,990 records from Directed Warbler surveys (2008–2009),
469 records from Interior Forest Surveys (2007), and 26 records from woodpecker call-playback
surveys (2006), for a total of 18,058 records. The information has been used to:
Evaluate and identify species-habitat relationships at stand and landscape scales using
different sets of habitat data (e.g., rapid habitat assessment and VRI)
Evaluate alternate survey methodologies for sampling species of interest:
o Listed warblers: passive BBS versus model-based sampling
o Woodpeckers: passive BBS versus call-playback
o On-road versus off-road sampling; effects on sampling of forest-interior species
Evaluate species‘ trends, including:
o Power to detect trend over different time intervals
o Power to detect trend using different sampling intensities
o Preliminary trend predictions for select species
Create a species inventory of predominantly forest-dependent species for the Fort St.
John Timber Supply Area
Estimate local and regional measures of bird diversity
Estimate local and regional measures of relative bird abundance
Inform forest managers and public stakeholders of important findings that may support
and facilitate decision-making processes that affect forest management for birds
Provide a better understanding of the distribution and occurrence of species of
conservation concern
In addition to the above-mentioned points, the reports and papers that have resulted from this
work have been used as reference material by other firms and industries to assist in the
assessment of other local developments both within the Fort St. John TSA, and in other parts of
the Peace Region.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
68
6.0 Acknowledgements
Thanks to Andrew Tyrrell (Canadian Forest Products Ltd.) for his assistance with planning,
reporting, and safety, and for providing review comments that improved an earlier draft of this
report. Funding for this project was provided by the Forest Investment Account and the Forest
Sciences Program, British Columbia Ministry of Forestry. Additional support was provided by
participants of the Fort St. John Pilot Project.
Thanks to Harry Williams (Stantec–Project Manager), Colleen Bryden (Stantec–Senior Review),
Lindsay Sherman (Stantec–GIS), Ryan Stohmann (Stantec–GIS), and Sue King (Stantec–
Project Management Assistant) for help and support throughout this project.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
69
7.0 Literature Cited
Ammon, E.M. 1995. Lincoln's Sparrow (Melospiza lincolnii), The Birds of North America Online
(A. Poole, Ed.). Cornell Lab of Ornithology, Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/191. [Accessed: 12
January 2010].
B.C. Conservation Data Centre [CDC]. 2010. BC Species and Ecosystems Explorer. B.C.
Ministry of Environment, Victoria, BC. URL: http://a100.gov.bc.ca/pub/eswp/ [Accessed
25 January 2010].
Bates, D. 2009. Linear mixed-effects models using S4 classes. URL: http://lme4.r-forge.r-
project.org/. [Accessed: 5 November 2009].
Burnham, K.P., and D.R. Anderson. 1998. Model selection and inference: a practical
information-theoretic approach. Springer-Verlag, New York, NY.
Bystrak, D. 1981. The North American Breeding Bird Survey. In Studies in Avian Biology 6:252-
261.
Campbell, R.W., N.K. Dawe, I. McTaggart-Cowan, J.M. Cooper, G.W. Kaiser, and M.C.E.
McNall. 1990. The birds of British Columbia, Volume 2 (Nonpasserines: diurnal birds of
prey through woodpeckers). Royal British Columbia Museum, Victoria, BC. 636 pp.
Campbell, R.W., M.I. Preston, M. Phinney, C. Siddle, and J. Deal. 2007. Featured Species –
Canada Warbler. Wildlife Afield 4:95-160.
Canadian Forest Products [Canfor]. 2004. Fort St. John pilot project–sustainable forest
management plan. 271 pp.
Committee on the Status of Endangered Wildlife in Canada [COSEWIC]. 2009. Wildlife species
search. URL: http://www.cosewic.gc.ca/eng/sct1/searchresult_e.cfm. [Accessed: 25
October 2009].
Crawley, M. 2007. The R book. Wiley, Chichester, England.
George, T. L., and D. Dobkin. 2002. The effects of habitat fragmentation on western bird
populations. Studies in Avian Biology 25:4-7.
Ghalambor, C.K., and T.E. Martin. 1999. Red-breasted Nuthatch (Sitta canadensis), The Birds
of North America Online (A. Poole, Ed.). Cornell Lab of Ornithology, Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/459. [Accessed: 12
January 2010].
Gibbs, J.P. 1995. MONITOR 7.0–Software for estimating the statistical power of population
monitoring programs. URL: http://www.mbr-pwrc.usgs.gov/software/monitor.html.
[Accessed: 4 November 2009].
Goldsmith, F.B. (ed.). 1991. Monitoring and Conservation Ecology. Chapman and Hall, London,
England. 292 pp.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
70
Hunt, P.D., and D.J. Flaspohler. 1998. Yellow-rumped Warbler (Dendroica coronata), The Birds
of North America Online (A. Poole, Ed.). Cornell Lab of Ornithology, Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/376. [Accessed: 12
January 2010].
Ingold, J.L., and R. Galati. 1997. Golden-crowned Kinglet (Regulus satrapa), The Birds of North
America Online (A. Poole, Ed.). Cornell Lab of Ornithology, Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/301. [Accessed: 12
January 2010].
James, F.C., and N.O. Wamer. 1982. Relationships between temperate forest bird communities
and vegetation structure. Ecology 63: 159-171.
James, F.C., C.E. McCulloch, D.A. Wiedenfeld. 1996. New approaches to the analysis of
population trends in Land Birds. Ecology 77:13-27.
Krebs, C. 1999. Ecological methodology. Second Edition. Addison Wesley Longman, Inc. Menlo
Park, CA. 620 pp.
Lowther, Peter E. 1999. Alder Flycatcher (Empidonax alnorum), The Birds of North America
Online (A. Poole, Ed.). Cornell Lab of Ornithology, Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/446. [Accessed: 12
January 2010].
Mack, D.E., and W. Yong. 2000. Swainson's Thrush (Catharus ustulatus), The Birds of North
America Online (A. Poole, Ed.). Cornell Lab of Ornithology, Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/540. [Accessed: 12
January 2010].
Marzluff, J.F., and R. Sallabanks. 1998. Avian Conservations: Research and Management.
Island Press, Washington D.C. 563 pp.
Morgan, K., and B. Freedman. 1986. Breeding bird communities in a hardwood forest
succession in Nova Scotia. Canadian Field-Naturalist 100:506-519.
Pannekoek, J., and A. van Strien. 2005. TRIM 3 manual–trends and indices for monitoring data.
Statistics Netherlands. 57 pp.
Preston, M.I. 2008. Monitoring birds for sustainable forest management in the Fort St. John
Timber Supply Area: relating species occurrences to forest structure. Final Report.
Canadian Forest Products Ltd., Fort St. John, BC. 82 pp.
Preston, M.I. 2009. Monitoring Birds for Sustainable Forest Management: Species-Habitat
Associations in the Fort St. John Timber Supply Area. Canadian Forest Products Ltd.,
Fort St. John, BC. 87 pp.
Resource Inventory Committee [RIC]. 1999. Inventory methods for forest and grassland
songbirds - standards for components of British Columbia‘s biodiversity #15, ver. 2.0,
March 1999. BC Ministry of Environment, Lands and Parks, Resources Inventory
Branch, Victoria, BC. 49 pp.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
71
Sauer, J.R., J.E. Hines, and J. Fallon. 2008. The North American Breeding Bird Survey, Results
and Analysis 1966–2007. Version 5.15.2008. USGS Patuxent Wildlife Research Center,
Laurel, MD.
Schwartz, C. 2009. Chapter 4 – Survey Sampling. Course Notes for Beginning and Intermediate
Statistics. URL: http://www.stat.sfu.ca/~cschwarz/CourseNotes.html. [Accessed: 12
January 2010].
Sen, A.R. 1981. Methodological Studies of Breeding Bird Surveys in North America. In Studies
in Avian Biology 6:496-501.
Sogge, M.K., W.M. Gilbert and C. Van Riper III. 1994. Orange-crowned Warbler (Vermivora
celata), The Birds of North America Online (A. Poole, Ed.). Cornell Lab of Ornithology,
Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/101. [Accessed: 12
January 2010].
Van Horn, M.A., and T.M. Donovan. 1994. Ovenbird (Seiurus aurocapilla), The Birds of North
America Online (A. Poole, Ed.). Cornell Lab of Ornithology, Ithaca, NY. URL:
http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/088. [Accessed: 12
January 2010].
Vernier, P.R. 2010. Biodiversity Model Explorer: TFL 39 Study Area. URL:
http://biod.forestry.ubc.ca/sfm/tfl39/index.php. [Accessed: 13 January 2010].
Vernier, P.R., and Bunnell, F.L. 2008. Warbler Species at Risk in Northeastern British Columbia:
Model Development, Validation, and Directed Sampling. Report Prepared for Forest
Science Program Project Y081137. UBC Centre for Applied Conservation Research.
Vancouver BC.
Vernier, P.R., M.I. Preston, F.L. Bunnell, and A. Tyrrell. 2009. Adaptive monitoring framework
for warblers at risk in northeast BC: using habitat models and expert opinion to refine
monitoring. In Press
Vernier, P.R., F.K.A. Schmiegelow, S. Hannon, and S.G. Cumming. 2008. Generalizability of
songbird habitat models in boreal mixedwood forests of Alberta. Ecological Modelling
211:191-201.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
72
8.0 Appendices
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
73
APPENDIX 1 2009/2010 Work Plan
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
74
Objectives
Continue monitoring birds from 16 currently established Breeding Bird Survey routes in the Fort St. John Timber Supply Area (FSJ TSA).
Continue monitoring 104 listed warbler survey stations established in 2008 to refine variance estimates and habitat-based models.
Confirm habitat conditions for each BBS station with data collected in 2007 and 2008; update the habitat data for those stations that have changed, or that were not previously described.
Analyze and report on species richness, abundance, and diversity by broad habitat (i.e., stand age and composition–see below).
Analyze species-habitat relationships using broad habitat type, and specific structural attributes (e.g., shrub cover, canopy cover) and provide context by assessing agreement/disagreement of results with other studies.
Where species data are adequately sampled, conduct a power analysis using data from all Breeding Bird Survey stations sampled for the period 2005–2009, with a tentative target of being able to detect a significant trend of from 0 to ±10%/year change over a 20-year period.
Revise habitat-based models for warbler species (where sample size permits) and use models to project areas of low, medium, and high suitability.
Estimate listed warbler occupancy with a specified level of precision and power to detect changes in occupancy between two points in time.
Provide an observation summary of all bird species listed by the BC Conservation Data, the Species at Risk Act, Schedule 1, and the Committee on the Status of Endangered Wildlife in Canada.
Work Schedule
Field work–June 2009
Data compilation and report writing–late July through December 2009
Delivery of draft and final reports–March 2010
Background
Breeding Bird Survey (BBS) routes are established in the Fort St. John Timber Supply Area (FSJ TSA): seven routes were established in 2005, four routes were established in 2006, and one route was established in 2007. In 2007, all BBS routes were modified from the typical 39.2 km long with 50 stations each sampled for 3 minutes, to BBS routes that are 23.2 km long with 30 stations each sampled for 5 minutes. The reason for the modification was to accommodate the extra time needed to use orthophotos for increasing the spatial resolution in the bird/habitat relationship data. The total number of routes, using 2007 methods, is 16. The total number of BBS stations sampled each year for the period 2005-2008 is 268.
All BBS routes have been surveyed and reported on by Preston et al. (2006, 2007) and Preston (2008, 2009). The BBS routes are intended to provide a sample of bird occurrences from which links to a broad range of forest types and structural attributes may be made over a range of spatial scales by a variety of different interest groups (e.g., Centre for Applied Conservation Research, University of British Columbia). Specifically, the BBS is part of a biodiversity monitoring program that provides data and results that may be used to refine future updates of the Fort St. John Pilot Project Sustainable Forest Management Plan (Canfor 2004). The data and results will also contribute to the Species Accounting System concurrently being developed by Dr. F.L. Bunnell at the University of British Columbia in cooperation with Canfor and BC Ministry of
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
75
Environment. The annual collection of new BBS data is intended to complement existing BBS data for the purpose of ensuring that:
1) As many species and habitats as possible are accounted for in the Species Accounting System;
2) Estimates of species variability, and probability of association within and among specific habitat types, are properly identified and statistically defensible;
3) Important habitat features for certain species are identified (e.g., shrub cover); and,
4) Through long-term monitoring, species' trend information may be evaluated, which will include effects of annual variability in species abundance.
An additional value to this monitoring program is that it provides new information on listed species for which there is considerable interest from multiple stakeholders (e.g., Ministry of Environment, Committee on the Status of Endangered Wildlife in Canada, Canadian Wildlife Service, Partners in Flight). Specifically, among the listed species, data for forest-dwelling Black-throated Green Warbler, Connecticut Warbler, Canada Warbler, Bay-breasted Warbler, and Cape May Warbler will be used in a BC Forest Sciences Program grant entitled "Developing and Validating Habitat-based Management Models for Species 'at risk' in Northeastern BC‖ (Y092137). This program began in 2008 and will continue through this project.
Methods
Breeding Bird Survey (BBS) routes (Table 1), and ―listed warbler‖ stations will be surveyed using standardized methods for forest songbirds (RISC 1999), with approved variances
Breeding Bird Survey routes
Each BBS route has 30 stations, and each station will be sampled for 5 minutes, following RISC (1999) standards. Each BBS route, and its corresponding stations, will be surveyed once during the survey period following an approved variance granted in 2009 (see below for Approved Variance 1). Orthophoto datasheets (where available), with an accompanying 50-m radius buffer overlay showing intervals up to 200-m will be used to map all bird detections (RISC 1999). Each station will be sampled for 5 minutes, with individual records demarcated as occurring within the 0-3 min and 3-5 min intervals (RISC 1999). Weather data will be collected at each station (e.g., wind speed, temperature, precipitation, cloud cover) following RISC (1999) standards and an approved variance granted in 2009 (see below for Approved Variances 2, 3 and 4).
Table 1 Names of existing BBS routes to be surveyed in 2009
Route Name Route Name
PeeJay1 Pink Mountain
PeeJay2 Haystack
PeeJay3 Kobes
PeeJay4 KobeHay
Buick 1 Sikanni
Tommy Lakes 1 Graham River 1
Tommy Lakes 2 Graham River 2
Tommy Lakes 3 Wonowon
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
76
Listed Warbler Surveys
In 2008, 104 ―listed warbler‖ survey stations were established on the basis of preliminary habitat model predictions. In 2009, the same 104 stations will be sampled again, using the same sampling procedures as for the Breeding Bird Surveys (i.e., duration of survey, data sheets, weather, etc., as per RISC 1999 and Approved Variances). Because ―listed warbler‖ survey stations are not spatially arranged as per Breeding Bird Survey stations (i.e., 800-m apart), spatial rules were applied as per RISC (1999) guidelines. Resultantly, all ―listed warbler‖ survey stations are ≥ 200-m apart (i.e., minimum of 400 m centre-to-centre). In addition to the warblers of interest, all other birds will be documented, so as to increase the potential species-habitat relationship model sample size for other species. Where sample size is sufficient, the survey data will be used in conjunction with VRI data to develop habitat-based models. Important habitat variables, model coefficients and standard errors will be presented in a table; model coefficients will be linked to the VRI data to produce maps showing areas of varying habitat quality for portions of the Fort St. John TSA. The data from 2008 and 2009 will also be used to evaluate the precision and power of the survey for estimating occupancy and detecting changes in occupancy between two points in time. Results will be presented in graphical or tabular form.
Habitat Classification
Habitat data, using techniques developed in collaboration with the University of British Columbia, Forest Sciences (Fred Bunnell, Laurie Kremsater) was gathered from most BBS stations in 2007 and 2008. In 2009, habitat data will be updated for those BBS and ―listed warbler‖ stations where major changes have occurred (e.g., a mature stand is now harvested). Among the BBS and ―listed warbler‖ survey stations, habitat data will only be gathered from stations that have uniform (i.e., single stand type) habitat conditions within 100-m of the station centre (see Table 2 for habitat classification scheme). Habitat conditions will be described for each side of the road independently, and linked to the bird data that is similarly described. In special circumstances, whereby one of the listed warblers is detected in a forest patch that does not conform to the classification rules as described above, habitat will be described for the patch. The issue of contrasting adjacent habitats within the survey protocol is expected to be addressed in the FSP project using VRI data. Species-habitat relationships using bird and habitat data collected in 2009 will be presented in the final report for this project using appropriate statistical techniques (RISC 1999).
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
77
Table 2 Classification scheme used to describe broad habitat attributes at BBS stations*.
Forest Attribute Attribute Description
Forest Class
FC1) trees ≤ 1.3 m in height
FC2) trees range from 1.4–3.0 m in height
FC3) trees range from 7.5–12.5 dbh
FC4) trees range from 13–40 dbh
FC5) trees are >40 dbh
Forest Type
Hardwood (recorded to nearest 5%; ≥ 75 % = hardwood stand)
Mixedwood (recorded to nearest 5 %; ≥ 25% hardwoods and ≥ 25 % = softwood) stand) Softwood (recorded to nearest 5 %; ≥ 75 % softwood stand)
Tree Species (percentage composition of each tree species, for species having ≥ 5% representation)
Canopy Cover (percentage cover of the forest canopy for trees >1.3m); not recorded for FC1
Shrubs (percentage cover and height)
Low (0.5–1.5 m in height)
Medium (1.5–2.5 m in height)
High (>2.5 m in height)
* attributes are described for uniform stations only. For example, habitat will be scored for left and right sides of the road independently, using a 100-m radius semi-circle buffer comprised of a single forest class. We do not score BBS stations comprised of more than one forest class per side. The habitat data described here meets the requirements for assessing species-habitat relationships as described by RISC (1999).
Data Analysis and Summaries
Analysis of BBS and "listed warbler" data collected in 2009 will be summarized using standard techniques for diversity (H'), abundance (total birds counted), and richness (total number of species) (Krebs 1999). Data analysis will focus on species-habitat relationships using appropriate statistical methods (e.g., resource selection tests, multivariate logistic regression). Additional reporting may include an analysis of BBS sampling intensity by habitat type (e.g., BEC, forest class), analysis of survey efficiency (i.e., BBS sampling versus ―listed-warbler‖ sampling), comparison of species occurrence and abundance among forest types, and an overview of all "listed species" occurrences.
Data Submission
Data, and a copy of the final report, will be provided to the Fort St. John Pilot Project Participants and the Ministry of Environment WSI database upon completion of the project. The data files will include a spreadsheet (Excel, Access, WSI Template) of species observations with associated species behaviours, station attributes, and habitat and weather attributes. A spatial file (.shp) of all survey stations and species observations will also be provided.
Standards
This project will meet and follow the mandatory standards defined and described by the BC Ministry of Environment, Environmental Stewardship Division: Terrestrial Biological and Physical Monitoring, and Wildlife Species Inventory (WSI). Specifically, this project will meet and follow:
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
78
Mandatory Standards
Species Inventory Fundamentals: Standards for Components of British Columbia's Biodiversity, No. 1., including Errata 2 (version 2, November 1998).
Species and Ecosystems at Risk LBIP Standard (see Appendix 1).
Land Based Investment Program Forest Investment Account Data Management Standard, version 2.0 (June 2008)
Inventory Methods for Forest and Grassland Songbirds: Standards for Components of British Columbia's Biodiversity No. 15 (version 2.0, March 1999). See RISC 1999 in Literature Cited.)
Approved Variances and Exemption
The following variances were approved on 12 May 2009 by the Investment Manager of the Forest Investment Account, PricewaterhouseCoopers LLP. The details within each variance are meant to cover more than one project in more than one Canfor Division. Therefore, not all of the items discussed are relevant to the objectives of this work plan.
Variance 1: Roadside Breeding Bird Points
In 2006, 2007, and 2008, we were granted a variance from RISC standards to survey the BBS routes only once instead of three - four times as the standard suggests. We would like the same variance granted for 2009 for Fort Nelson and Peace Forest Districts.
As context, it is important to appreciate that one innovative approach we introduce to the Breeding Bird Surveys (BBS) is to assess habitat at each station. We also are attempting to assess both habitat use and trend. As well, we are attempting to provide regional assessments of the effects of management practices (effectiveness monitoring). The specification of three times by RISC is to attempt to ensure precision at that station. UBC and others have analyzed the power of the test (precision) of BBS routes and found:
We gain more precision about predicting habitat use by increasing the number of routes and increasing the number of samples of a particular habitat than by repeatedly sampling the same spot within a year (e.g., Ralph et al. 1995).
The assessment of trend is most dependent on repeated annual sampling than the number of assessments at a particular station (e.g., Dunn 2002; Huggard 2002). That is, there is a clear cost trade-off between sustaining the annual sampling to acquire an accurate trend and sampling the same station more frequently in a specific year.
For effectiveness monitoring we seek trend by habitat class (the class may be a particular forest treatment or habitat structure class). That is, the trend applies generally to that habitat type over a large area, rather than to a particular example of the habitat type sampled more intensively. Our analysis of data collected over past years in both coastal and interior British Columbia indicates that we gain much greater generality of application, as well as precision by habitat type, by sampling more examples of the habitat type (e.g., Vernier and Bunnell 2005). Moreover, we experience no loss of ability to detect trends by sampling each station only once (it is for this reason that COSEWIC, thus SARA, and Partners in Flight rely on a single annual sample). The latter finding occurs only when the same effort (funding) is allocated over about 3 times the stations rather than one-third of the stations sampled three times. That is our intent.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
79
Analyses indicate, and we believe, that sampling individual stations once serves to increase the utility of effectiveness monitoring in three significant ways:
1) greater generality of findings to guide management actions;
2) greater likelihood of detecting trends because annual cost is less; and
3) much improved ability to link trend to habitat type.
In addition, a similar FIA-funded project conducted by Tembec near Cranbrook used the same methodology.
Additional information on the NE project can be found on the following website: http://3614b.forestry.ubc.ca/nebc/ <http://3614b.forestry.ubc.ca/nebc/>
References
Dunn, E.H. 2002. Using decline in bird populations to identify needs for conservation action. Conservation Biology 16:1632–1637.
Huggard, D. 2002. Precision of the Breeding Bird Surveys for monitoring on Weyerhaeuser's Coastal BC Tenure. Analysis based on pilot study results from Mike Preston and Wayne Campbell. Report to Weyerhaeuser, Nanaimo.
Ralph, C.J, J.R. Sauer, and S. Droege. 1995. Monitoring bird populations by point counts. USDA Forest Service General Technical Report PSW-GTR-149:1-181, Portland, OR.
Vernier, P.R. and F.L. Bunnell. 2005. Forest songbird-habitat relationships in northeast BC. Report to Canadian Forest Products. 16 pp.
Variance 2: Listed Warbler Surveys
As stated above, in past years, we successfully requested a variance from RISC standards to monitor breeding bird survey routes (road transects) once a year instead of three times a year. The variance enables us to more effectively meet our objectives of monitoring to detect population trends, develop habitat associations, and evaluate management effectiveness at large spatial scales. We would like to extend that request to include the ―listed warbler‖ survey stations in the Peace Forest District.
The ―listed warbler‖ stations are established throughout the Peace Forest District (primarily the southern and eastern portions of the Fort St. John Timber Supply Area). All ―listed warbler‖ stations are at least 200 m (600 m centre-to-centre) from each other to ensure independence. The ―listed warbler‖ surveys permit us to:
Increase the representation of Black-throated Green Warbler, Connecticut Warbler, Bay-breasted Warbler, Canada Warbler, and Cape May Warbler;
Determine if detections of these species differs substantially from comparable data collected from the Breeding Bird Survey stations;
Pool data for the analysis of habitat associations;
Evaluate the effectiveness of management activities for five species of conservation concern (one of which is of national concern); and
Increase the sample size of some rare species by combining surveys.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
80
Because the ―listed warbler‖ surveys are meant to complement and enhance the quality of information obtained from large scale breeding bird surveys, the reasons for requesting this variance are similar to an earlier request (in 2007) whereby interior forest surveys were conducted for the purpose of comparing on-road versus off-road sampling effectiveness. Specifically, we are attempting to assess both habitat use and trend, as well as attempting to provide regional assessments of the effects of management practices (i.e., effectiveness monitoring).
Variance 3: Ambient Temperature and Wind
Canfor‘s breeding bird and ―listed warbler‖ surveys are conducted mainly in very remote areas, and require field crews to be onsite for extended periods of time. RISC (1999) standards are being used for these FIA-funded projects, combined with BBS protocols used throughout North America by Environment Canada and the USFWS. The BBS protocol requires surveys to be conducted between 28 May and 7 July in north-eastern BC (which fall within RISC standard dates).
Past experience in the northeast suggests that weather standards as stated in RISC (1999) may limit the numbers of ―acceptable‖ days to such a low number that survey programs designed to monitor breeding songbirds across Forest Districts are not feasible. The issues are:
1) Time of day restrictions (see Section 3.1.4 in RISC 1999) require surveys to be conducted in the hours between sunrise and 4 hrs after sunrise. In NE BC, weather conditions are often within the ―unacceptable‖ range (see Table 2), during this period. Temperature may often be 0-3°C during the first 1-2 hrs of the survey period, and winds may often pick up to exceed Beaufort 2 during the later 2 hrs of the survey period.
2) Study design requires that the full 4 hrs be available each survey day to complete the necessary number of survey points in each transect.
3) An unreasonably large crew would be required to complete the survey program on days that are entirely ―acceptable‖, and many person-days (with associated large monetary costs) would be lost on ―unacceptable‖ days as crews wait for ideal conditions.
Given that:
1) Canfor‘s Inventory Programs have established breeding songbird survey transects throughout various Forest Districts in north-eastern BC.
2) These surveys are conducted following RISC and BBS standards.
3) Temperature and wind are often in the ―unacceptable‖ range according to RISC (1999) for at least part of the early morning 4 hr survey period.
4) Our experience suggests that forest songbirds in NE BC sing and are as active when temperatures are 0-3°C and winds are at Beaufort 3 as they are when conditions are more ―acceptable‖ according to the standard.
Canfor requests a variance to the RISC standard to allow surveys during days when temperatures are ≥0°C and winds are <Beaufort 4. This variance will apply to Fort Nelson and Peace Forest Districts (Fort Nelson and Fort St. John TSAs, and TFL 48 in Chetwynd).
Variance 4. Precipitation
We also request a variance to be able to conduct Breeding Bird Survey work in conditions described as "very light" rain. This variance was approved in 2007 and 2008.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
81
Exemption from Quality Assurance Standard
An exemption to the ‗Quality Assurance Standards for Wildlife Inventory Projects‘ was granted by the Investment Manager of the Forest Investment Account, PricewaterhouseCoopers LLP on June 8, 2009.
Deliverables
Prepare a draft report to summarize findings of the field data. The "draft" will be submitted to Canadian Forest Products Ltd. for review by 30 December 2009. The "final report", with review comments and a copy of this work plan, will be submitted by 5 March 2009.
A section of the report will describe methods and results specific to the listed warbler species analyses including descriptions of habitat models (including appropriate tables and/or figures), and estimates of precision and power.
Bird and habitat data will be provided in MS Excel or MS Access format, to Canadian Forest Products Ltd., and to the Ministry of Environment WSI database using their WSI spreadsheet template.
ARCVIEW 3.x/ARCMAP 9.x shapefiles or MXDs of all bird detections and sampling stations will be provided to Canfor and the Ministry of Environment WSI database.
Literature Cited
Dunn, E.H. 2002. Using decline in bird populations to identify needs for conservation action. Conservation Biology 16:1632–1637.
Fort St. John Pilot Project Participants. 2004. Fort St. John pilot project–sustainable forest management plan. 392 pp.
Huggard, D. 2002. Precision of the breeding bird surveys for monitoring on Weyerhaeuser‘s Coastal BC Tenure. Unpublished Report to Weyerhaeuser Company Ltd., Nanaimo, BC. 24 pp.
Krebs, C.J. 1999. Ecological methodology. Second Edition. Addison Wesley Longman, Inc. Menlo Park, CA.
Preston, M.I. 2009. Monitoring birds for sustainable forest management: species-habitat associations in the Fort St. John Timber Supply Area. Final Report. Canadian Forest Products Ltd., Fort St. John, BC. 87 pp.
Preston, M.I. 2008. Monitoring birds for sustainable forest management in the Fort St. John Timber Supply Area: relating species occurrences to forest structure. Final Report. Canadian Forest Products Ltd., Fort St. John, BC. 82 pp.
Preston, M.I., P. Vernier, and R.W. Campbell. 2006. A four-year summary of breeding bird surveys in TFL 48 and the Fort St. John Timber Supply Area. Progress Report. Canadian Forest Products Ltd. Chetwynd, BC. 34 pp.
Preston, M.I., P. Vernier, and R.W. Campbell. 2007. Using birds for species accounting and effectiveness monitoring in Tree Farm License 48 and the Fort St. John Timber Supply area. Progress Report. Canadian Forest Products Ltd., Chetwynd and Fort St. John, BC., and Environment Canada, National Wildlife Research Centre, Ottawa, ON. 58 pp.
Ralph, C.J, J.R. Sauer, and S. Droege. 1995. Monitoring bird populations by point counts. USDA Forest Service General Technical Report PSW-GTR-149:1-181, Portland, OR.
Resource Inventory Committee (RISC). 1999. Inventory methods for forest and grassland songbirds - standards for components of British Columbia‘s biodiversity #15, ver. 2.0, March 1999. BC Ministry of Environment, Lands and Parks, Resources Inventory Branch, Victoria, BC. 49 pp.
Vernier, P.R. and F.L. Bunnell. 2005. Forest songbird-habitat relationships in northeast BC. Report to Canadian Forest Products. 16 pp.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
82
Forest Investment Account
Species and Ecosystems at Risk LBIP Standard (applicable to various activities)
Based on information collected during forest management planning, if the Recipient deems that the work conducted during an LBIP funded activity could affect a species or ecosystem at risk, or an identified wildlife (IW) species, they must include documentation in their FIRS submission on where the following information is documented/available (all available sources related to the specific species/ecosystem):
The goals, objectives and strategies of appropriate species recovery plans; and or
The designations and management practices under the Identified Wildlife Management Strategy; and or
Identified and or established Ungulate Winter Ranges (UWR) and objectives; and or
Other legislated and planning requirements for fish, wildlife and habitat.
Recovery Plans and other existing documents may contain specific standards that should apply to the FIA funded activity, and such standards must be adhered to as part of the FIA project implementation.
Within one month of the project being approved, the Recipient must notify the following contacts and provide them with a copy of the project submission, if requested:
Recovery Team chair (or designate) (see http://www.env.gov.bc.ca/wld/recoveryplans/rcvry1.htm for a list of chairs); or
[email protected] if there is no Recovery Team in place (or contact for Recovery Team chair is not known), or for IW species.
For further information on species and ecosystems at risk, see http://www.env.gov.bc.ca/atrisk/index.html. For further information on identified wildlife, see http://www.env.gov.bc.ca/wld/frpa/iwms/index.html. For information on UWR, or if the Recipient has not historically operated in the area where the FIA funded activity is being implemented, please discuss with your contacts at the MoE Regional Office.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
83
APPENDIX 2 Alphabetical list of observed species (2005–2009)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
84
Common Name Taxonomic Name CODE
Alder Flycatcher Empidonax alnorum ALFL
American Crow Corvus brachyrhynchos AMCR
American Kestrel Falco sparverius AMKE
American Pipit Anthus rubescens AMPI
American Redstart Setophaga ruticilla AMRE
American Robin Turdus migratorius AMRO
American Three-toed Woodpecker Picoides dorsalis ATTW
Barrow's Goldeneye Bucephala islandica BAGO
Barn Swallow Hirundo rustica BASW
Black-and-white Warbler Mniotilta varia BAWW
Bay-breasted Warbler Dendroica castanea BAYW
Blackburnian Warbler Dendroica fusca BBNW
Black-capped Chickadee Poecile atricapillus BCCH
Brown-headed Cowbird Molothrus ater BHCO
Blue-headed Vireo Vireo solitarius BHVI
Blackpoll Warbler Dendroica striata BKPW
Boreal Chickadee Poecile hudsonicus BOCH
Bonaparte's Gull Chroicocephalus philadelphia BOGU
Bohemian Waxwing Bombycilla garrulus BOWA
Brewer's Blackbird Euphagus cyanocephalus BRBL
Brown Creeper Certhia americana BRCR
Black-throated Green Warbler Dendroica virens BTNW
Broad-winged Hawk Buteo platypterus BWHA
Canada Goose Branta canadensis CAGO
Canada Warbler Wilsonia canadensis CAWA
Clay-colored Sparrow Spizella pallida CCSP
Cedar Waxwing Bombycilla cedrorum CEWA
Chipping Sparrow Spizella passerina CHSP
Cape May Warbler Dendroica tigrina CMWA
Common Raven Corvus corax CORA
Connecticut Warbler Oporornis agilis COWA
Common Yellowthroat Geothlypis trichas COYE
Dark-eyed Junco Junco hyemalis DEJU
Downy Woodpecker Picoides pubescens DOWO
Dusky Flycatcher Empidonax oberholseri DUFL
Dusky Grouse Dendragapus obscurus DUGR
Eastern Phoebe Sayornis phoebe EAPH
Evening Grosbeak Coccothraustes vespertinus EVGR
Fox Sparrow Passerella iliaca FOSP
Golden-crowned Kinglet Regulus satrapa GCKI
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
85
Common Name Taxonomic Name CODE
Golden-crowned Sparrow Zonotrichia atricapilla GCSP
Great Horned Owl Bubo virginianus GHOW
Gray Jay Perisoreus canadensis GRJA
Greater Yellowlegs Tringa melanoleuca GRYE
Hammond's Flycatcher Empidonax hammondii HAFL
Hairy Woodpecker Picoides villosus HAWO
Hermit Thrush Catharus guttatus HETH
Horned Lark Eremophila alpestris HOLA
House Wren Troglodytes aedon HOWR
Le Conte's Sparrow Ammodramus leconteii LCSP
Least Flycatcher Empidonax minimus LEFL
Lesser Yellowlegs Tringa flavipes LEYE
Lincoln's Sparrow Melospiza lincolnii LISP
MacGillivray's Warbler Oporornis tolmiei MACW
Merlin Falco columbarius MERL
Magnolia Warbler Dendroica magnolia MGNW
Mountain Bluebird Sialia currucoides MOBL
Mourning Warbler Oporornis philadelphia MOWA
Northern Hawk Owl Surnia ulula NHOW
Northern Flicker Colaptes auratus NOFL
Northern Harrier Circus cyaneus NOHA
Northern Shoveler Anas clypeata NOSL
Northern Waterthrush Seiurus noveboracensis NOWA
Northern Pygmy-Owl Glaucidium gnoma NPOW
Orange-crowned Warbler Vermivora celata OCWA
Olive-sided Flycatcher Contopus cooperi OSFL
Ovenbird Seiurus aurocapilla OVEN
Palm Warbler Dendroica palmarum PAWA
Philadelphia Vireo Vireo philadelphicus PHVI
Pine Grosbeak Pinicola enucleator PIGR
Pine Siskin Spinus pinus PISI
Pileated Woodpecker Dryocopus pileatus PIWO
Pacific-slope Flycatcher Empidonax difficilis PSFL
Purple Finch Carpodacus purpureus PUFI
Rose-breasted Grosbeak Pheucticus ludovicianus RBGR
Red-breasted Nuthatch Sitta canadensis RBNU
Ruby-crowned Kinglet Regulus calendula RCKI
Red Crossbill Loxia curvirostra RECR
Red-eyed Vireo Vireo olivaceus REVI
Ring-necked Duck Aythya collaris RNDU
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
86
Common Name Taxonomic Name CODE
Rock Ptarmigan Lagopus muta ROPT
Red-tailed Hawk Buteo jamaicensis RTHA
Ruffed Grouse Bonasa umbellus RUGR
Red-winged Blackbird Agelaius phoeniceus RWBL
Savannah Sparrow Passerculus sandwichensis SAVS
Solitary Sandpiper Tringa solitaria SOSA
Song Sparrow Melospiza melodia SOSP
Spruce Grouse Falcipennis canadensis SPGR
Spotted Sandpiper Actitis macularius SPSA
Sharp-shinned Hawk Accipiter striatus SSHA
Swamp Sparrow Melospiza georgiana SWSP
Swainson's Thrush Catharus ustulatus SWTH
Tennessee Warbler Vermivora peregrina TEWA
Townsend's Solitaire Myadestes townsendi TOSO
Townsend's Warbler Dendroica townsendi TOWA
Tree Swallow Tachycineta bicolor TRSW
Varied Thrush Ixoreus naevius VATH
Violet-green Swallow Tachycineta thalassina VGSW
Warbling Vireo Vireo gilvus WAVI
White-crowned Sparrow Zonotrichia leucophrys WCSP
Western Tanager Piranga ludoviciana WETA
Willow Ptarmigan Lagopus lagopus WIPT
Wilson's Snipe Gallinago delicata WISN
Wilson's Warbler Wilsonia pusilla WIWA
Winter Wren Troglodytes troglodytes WIWR
White-throated Sparrow Zonotrichia albicollis WTSP
White-winged Crossbill Loxia leucoptera WWCR
Western Wood-Pewee Contopus sordidulus WWPE
Yellow-bellied Flycatcher Empidonax flaviventris YBFL
Yellow-bellied Sapsucker Sphyrapicus varius YBSA
Yellow Warbler Dendroica petechia YEWA
Yellow-rumped Warbler Dendroica coronata YRWA
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
87
APPENDIX 3 Species richness and abundance ( 2005–2009)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
88
BBS Route # of
Stations
Cumulative
Richness
Cumulative
Abundance
Buick Creek 30 65 1,415
Haystack 24 53 1,059
KobesHay 20 60 966
PeeJay 1 30 54 1,040
PeeJay 2 30 36 779
PeeJay 3 16 50 588
Sikanni 28 42 820
Tommy Lakes 1 30 48 1,140
Tommy Lakes 2 30 53 1,019
Tommy Lakes 3 30 51 1,042
* based on 268 BBS stations sampled annually.
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
89
APPENDIX 4 Species richness, abundance, and diversity (2009)
MONITORING BIRDS FOR SUSTAINABLE FOREST MANAGEMENT
March 25, 2010
90
Route # of Stations Richness Abundance Diversity
Buick Creek 30 35 226 3.88
Graham River 1 30 20 151 3.45
Graham River 2 30 22 108 3.54
Haystack 30 36 225 4.39
Kobes 23 25 141 3.83
KobesHay 30 41 272 4.57
PeeJay 1 30 34 159 4.38
PeeJay 2 30 25 163 3.32
PeeJay 3 30 35 206 4.29
PeeJay 4 30 34 196 4.16
Pink Mountain 30 27 120 4.05
Sikanni 30 21 130 3.47
Tommy Lakes 1 30 24 206 3.92
Tommy Lakes 2 30 24 149 3.48
Tommy Lakes 3 30 29 186 3.81
Wonowon 30 38 210 4.25