UBC Undergraduate Research

Woodpeckers as a potential barrier to the mountain pine beetle (Dendroctonus ponderosae Hopkins) east… Sunter, Emily 2015-06

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


52966-Sunter_Emily_FRST_498_2015.pdf [ 1.01MB ]
JSON: 52966-1.0343442.json
JSON-LD: 52966-1.0343442-ld.json
RDF/XML (Pretty): 52966-1.0343442-rdf.xml
RDF/JSON: 52966-1.0343442-rdf.json
Turtle: 52966-1.0343442-turtle.txt
N-Triples: 52966-1.0343442-rdf-ntriples.txt
Original Record: 52966-1.0343442-source.json
Full Text

Full Text

   FRST 498 Undergraduate Thesis:  Woodpeckers as a potential barrier to the mountain pine beetle (Dendroctonus ponderosae Hopkins) east of the Rocky Mountains  Emily Sunter University of British Columbia Submitted June 10, 2015            1  Abstract Insect pest outbreaks have become substantial problems for forest managers as climate change and intensive land management practices affect the landscape. Woodpeckers have been suggested as a natural regulator of insect pest populations and outbreaks in combination with other management practices. The purpose of this study was to investigate the possibility of woodpeckers acting as a barrier to the invasive mountain pine beetle in the pine forests of Alberta. I hypothesized that woodpeckers would have the greatest effect on bark beetles where other food sources were abundant, such as wood borers, where nest sites were available, and where trees were larger and older. I also predicated feeding activity would differ by forest type. Where mountain pine beetle mass attacks were found, I predicated that woodpeckers would function as efficient predators and mortality agents. In four Alberta pine stands, two jack pine and two lodgepole pine stands, bark beetle population surveys were completed, along with the collection of prism cruise data and woodpecker foraging activity on bark beetle attacked trees. A binary logistic regression was used to determine which trees were most likely to have woodpecker foraging activity present, and Fisher’s exact test was used to determine whether woodpeckers fed on mountain pine beetle mass attacked trees preferentially over other trees containing other species of bark beetles. The results of the binary logistic regression indicate that wood borer presence, bordering coniferous forest, and large diameter are the most influential factors in determining whether woodpeckers feed on a tree with bark beetle activity. Tree species and nest site availability were not part of the final model but experimental set-up may have affected these results. Fisher’s Exact test results indicate woodpeckers do feed preferentially on mass attacked trees. These data along with supporting literature indicate woodpeckers could act as a potential barrier to mountain pine beetle expansion and future epidemics east of the Rocky Mountains. In order to promote woodpecker presence in a stand, timber management practices will have to manage for woodpecker habitat characteristics such as food sources during non-outbreak periods, nest sites, and habitat heterogeneity.      2  Introduction The mountain pine beetle (Dendroctonus ponderosae Hopkins [Coleoptera: Curculionidae, Scolytinae]) has been quoted as the worst-ever insect damage agent in North America when describing the most recent epidemic in BC (BC MFLNRO, n.d.; Mitton and Ferrenberg, 2012) Any option that could potentially mitigate this insect’s effect on pine forests of North America, especially where this beetle is non-native, should be investigated and if possible, implemented. Additionally, options that promote overall forest health and find a balance between the needs of forest operators and environmental objectives are most desirable. Using natural predators, such as woodpeckers, to control forest insect pests provides one such option.  Many woodpecker species feed heavily on sub-cortical insects, from wood borers to bark beetles to carpenter ants. In the pine forests of Alberta, Canada, a number of these species feed and thrive on this food source (Alberta ESRD, 2013). Woodpeckers are well-known to be able to respond to disturbance that creates influxes of short-term feeding opportunities (e.g. Duan et al., 2010; Hutto, 1995; Murphy and Lehnhausen, 1998). While much of the literature focuses on post-fire effects, another disturbance that creates the same conditions is bark beetle outbreaks, particularly in naïve habitats and during epidemic conditions. Currently, there is an emerging opportunity for woodpeckers able to take advantage of an influx in sub-cortical insect resources in the form the mountain pine beetle in Alberta, Canada. As the epidemic has progressed, the beetle has expanded its range latitudinally and elevationally within its historic range west of the Rocky Mountains, a process now attributed to climate effects (Carroll et al., 2003). Since 2006, its range has been expanding longitudinally east as well (NRC, 2015). The beetle crossed the Rocky Mountain divide into west-central Alberta in 2006 and since then has been moving steadily east across the province (Alberta ASRD, 2015; NRC 2015). In the lodgepole and jack pine forests of Alberta, the yellow-bellied sapsucker, northern flicker, and the downy, hairy, American three-toed, black-backed, and pileated woodpeckers are present and active foragers of many insect species (Alberta ESRD, 2013; Avibase, 2014). During fieldwork conducted in June, July, and August of 2014, most of these species were recorded. When the mountain pine beetle invades a naïve pine stand east of the Rockies, it is possible that some of these woodpecker species that depend on sub-cortical insects, such as wood borers and bark beetle, could inflict sufficient damage upon an increasing mountain pine beetle population as to actually limit the spread of the beetle. 3  The goal of this thesis is to investigate whether bark-drilling insectivorous species of the Family Picidae, or woodpeckers, could act as an effective barrier to the invasive mountain pine beetle in the pine forests of Alberta and further east. This was investigated by examining woodpecker predation on the bark beetle family (Scolytinae) in general. Sub-questions include whether woodpecker foraging activity differed by forest type, whether they were associated with particular stand characteristics (density, DBH range, age, species composition, etc.), and finally whether woodpeckers were more likely to feed on bark beetles where wood boring beetles were also abundant. I predict that, wood borer activity, tree species, presence of nesting habitat, and older and larger trees and will be the most influential factors for predicting whether woodpeckers will feed on trees where bark beetles are also active. Additionally, I predict that where mountain pine beetle invades, woodpeckers will respond to any strip or mass attack and feed preferentially on these trees when compared to other similar trees without mountain pine beetle present. Methods Study Sites Four stands were surveyed between June and September, 2014 in Alberta, Canada (range = 5 – 8.65 ha, mean = 6.8 ha, total area = 27.25 ha). The Lac la Biche study sites, LLB1 and LLB2, were located at 54.98567°N, -112.02746°W and 55.12610°N, -111.98975°W, respectively (Figure 1). The Edson sites, EDS1 and EDS2, were located at 53.54484°N, -116.37310°W and 53.53589°N, -116.44962°W, respectively (Figure 1).  These two areas were chosen as both the eastern-most (Lac la Biche) and western-most (Edson) presence of mountain pine beetle populations in Alberta. These areas also coincide with genetically distinct forests, jack pine in the east and lodgepole pine in the west (Cullingham et al. 2012). In all stands that were surveyed in this study, there was evidence of mountain pine beetle arrival but not population persistence in an epidemic state. The four individual stands were picked from a limited number of lodgepole or jack pine dominated stands in the specified areas that had no mass attack mountain pine beetle damage upon initial observation, and were accessible by road. Pre-study beetle distribution and occurrence data were provided by the Environment and Sustainable Resource Development (ESRD) branch of the Alberta Government. 4          Figure 1 – Map illustrating location of study areas. Two sites were located north of Lac la Biche, Alberta and two sites south of Edson, Alberta (image courtesy of Vanellis, n.d.). Layout Procedure Stand layout consisted of a flagged baseline with transects extending perpendicularly every 25 meters. Along each transect, stations were marked every 50 meters. Layout was completed using a compass and chain. These transects were flagged consistently to ensure visibility from any position within the stand. Transects were laid until the stand edge was reached (i.e. where the stand was no longer dominated by lodgepole or jack pine or where an obstacle such as a campsite, road, or stream existed). Except in the first stand surveyed (LLB1), transect edges were at least 12.5 meters from stand edges to ensure stand characteristics were as consistent as possible anywhere within the layout. In LLB1, the first ~150 meters of the baseline was less than 12.5 meters from a road. Data Collection An initial 100% cruise was completed in order to assess the bark beetle community and woodpecker foraging damage in each stand before the flight period of the mountain pine beetle. Between June 20 and July 3, 2014, research assistants trained in Alberta bark beetle identification examined every pine tree in each stand for any current or recent (within the last 3 years) bark beetle damage by moving methodically along each transect and up to 12.5 meters around the border of each stand layout. Beetles were identified through examination of any fine, dark coloured frass, pitch tubes, larval galleries, visible beetles, or failing tree health.  Trees with suspected bark beetle presence or an unknown insect damage agent were flagged and then subsequently assessed by the site supervisor. Trees were only recorded if a bark beetle species was confirmed to be present or active within the previous three years. Each bark beetle species present 5  and the year of attack (where possible) was recorded along with the tree diameter breast height (DBH), height (using a Vertex III hypsometer), exact location in reference to the closest marked station, percent woodpecker feeding damage on the bole, tree injury codes (see Appendix A), and any other relevant information such as details on success of bark beetle attacks, tree foliage state, or the species of woodpecker that had caused foraging damage. Woodpecker foraging activity was recorded as a percent score by visually estimating the portion of the main stem that had been debarked; this method was roughly based on the work done by Bergvinson and Borden (1992). Bark beetle infested trees were assigned a number and labelled using flagging tape and log paint for future identification and re-examination. Between August 23 and September 1, 2014, another 100% cruise was completed after the flight time of mountain pine beetle to assess any new bark beetle activity. The same procedure was used as during the first 100% cruise and the same exact area was surveyed. Trees that had been flagged during the initial 100% cruise were re-examined by the site supervisor to search for any mountain pine beetle activity.  During the same time period as the second 100% cruise, a prism cruise was completed to record stand level characteristics at each site. Prism cruises were completed using the variable plot method. One prism plot/hectare was installed, with an additional plot if additional partial hectares were cruised (e.g. LLB1, 8.65 ha = 9 prism plots). In order to select prism plots, stands were divided into 1 ha units and one station that did not fall on the stand edge was chosen per hectare using randomized systematic selection. It is worth noting that in LLB1, four plots were reselected as they fell in open areas with no trees; the original selection was used to estimate stand density, and additional selections were used to calculate tree age, DBH, and species statistics. The basal area factor (BAF) was chosen in each stand to ensure six to ten trees were selected in each plot. In each prism plot, the DBH, height, injury codes, and notes for all selected trees were collected using the same procedures and equipment as during the 100% cruise. Ages of the largest and smallest diameter living trees were also collected (if a rotten core was encountered, the next largest/smallest tree was selected).  Throughout both cruises, the author of this paper also recorded every woodpecker sighted while surveying each stand down to species where possible. 6  Data Analysis Data were analyzed using SAS 9.4 (SAS Institute, Cary NC), Microsoft Excel 2010, and R version 3.1.1. The original 100% cruise data was processed in order to analyse using logistic regression and to perform descriptive statistics: - woodpecker damage recorded as pileated woodpecker only was eliminated due to the fact that their presence is most often controlled by the presence of carpenter ants (Beckwith and Bull, 1985; Bull et al., 1992)  - any trees with current or past mass attack damage were removed from the logistic regression since this study focused on stands that ideally would not contain any mass attacks; this allowed the regression to specifically analyze on endemic beetle levels; the removed data was analyzed using the Fisher’s Exact test as discussed below For the descriptive statistics, the data from the initial 100% cruise was used to calculate average/median woodpecker foraging activity by stand, and prism cruise data was used to calculate the averages and ranges of stand age, height, DBH (including values excluded from stem/ha calculation). Prism cruise data was also used to calculate tree species composition and stems per hectare. Two-tailed t-tests assuming unequal variance were used to distinguish between stands when a difference was detected in one-way ANOVA tests.  In order to address the hypotheses and determine which factors were most influential in determining presence of woodpecker foraging damage on bark beetle infested trees, a number of independent variables were tested using logistic regression. The variables are described in Table 1. The dependant variable was woodpecker presence or absence determined by the presence of woodpecker foraging damage. This dependant variable was labelled WP. Binary logistic regression was used to investigate the relationship between woodpecker presence and absence on bark beetle infested trees and stand and tree characteristics. The significance level was set at 0.05 and the models were assessed based on model convergence, significance (Likelihood Ratio, Score, and Wald Chi Square tests), the AIC, the average of R² and rescaled R², percent concordant, discordant, and tied, as well as Somers D (UCLA: Statistical Consulting Group, 2015; A. Marciniak, personal communication, February 12, 2015).   7  Table 1 – Independent variables tested using binary logistic regression in order to determine which, if any, were successful in predicting which bark beetle infested trees also had woodpecker foraging activity in jack and lodgepole pine stand studied in Alberta, Canada, in 2014. Code Description D DBH in cm  W Wood borer presence/absence; recorded in field by injury code 9 during initial 100% cruise T Tree species; either jack or lodgepole pine as defined by Cullingham et al. (2012) and Farrar (1995) H Proportion hardwood stems; calculated by dividing total basal area of hardwoods in the prism plots by total basal area of all prism plots (BCMOF, 2009) G Proportion dead stems; calculated in same manner as H, using trees noted as dead  S Stems per hectare; all trees above the DBH limit of 7.5 cm were tallied (BCMOF, 2009)  AD Adjacent deciduous; ‘0’ indicated that the stand was bounded by a change in conifer tree species, ‘1’ indicates the stand was bounded by deciduous trees  Where the AIC did not change more than 2 points with the removal of a variable, that variable was assumed to not contribute significantly to the model. Where the R² and rescaled R² for the model were higher when considered in combination, the model was assessed to be better. And finally, where the percent concordant was higher, the model was assessed as improved and the same went for a lower percent discordant, lower percent tied, and higher Somer’s D (UCLA: Statistical Consulting Group, 2015; A. Marciniak, personal communication, February 12, 2015) The initial model contained all variables outlined in Table 1. They were eliminated systematically and individually based on their significance values and according to the procedure outlined above. Once the best model had been identified, all permutations of the three final variables were tested to ensure a simpler model could not improve the chosen model. Fisher’s Exact Test of Independence Mass attack data removed from binary logistic regression data were analyzed using Fisher’s Exact test. This test was chosen since the sample size was small and in order to analyze whether woodpecker foraging activity differed significantly between mass attacked trees and non-mass attacked trees (McDonald, 2014). Non-mass attacked trees were those where there were only bark beetles other than mountain pine beetle present, or where mountain pine beetle attacks were sparse on the bole. A Microsoft Excel 2010 spreadsheet sourced from McDonald (2014) was used to analyze data. 8  Results Descriptive Statistics ANOVA tests for both height and stand data from the prism cruise provided P values < 0.001 and subsequent t-tests distinguished which differences existed between the stands. LLB1 stood out as distinct from the other stands surveyed. Table 2 shows that this jack pine stand had significantly lower average height and DBH measurements compared to the other stands, as well as the smallest individual trees picked up in the prism cruise (minimum 3.1 cm DBH). Although ANOVA tests were not possible to distinguish between stems/ha and proportion of dead and hardwood stems due to the nature of the experimental set-up, LLB1 also shows the lowest overall stand density, and highest proportion of dead stems.  LLB1 and LLB2 were more similar in terms of both height and DBH, and neither stand had any hardwood stems surveyed in the prism cruise, indicating distinct stand edges rather than gradual transitions to other forest types (Table 2). LLB2 and EDS2, despite being different tree species, showed similarities in DBH ranges and average DBH measurements, having the largest trees surveyed (up to 60.9 cm DBH). It is worth noting that LLB2 had no dead stems surveyed in the prism cruise. EDS1 and EDS2, the two lodgepole pine stands, also showed significant similarity in their mean stand heights and DBH, as well as the inclusion of hardwood stems within the stands and low proportions of dead stems (Table 2). Finally, EDS2 was overall much denser than any other stand at 1317.3 stems/ha while also including a number of large stems (11 trees surveyed > 30 cm DBH).  Table 2 – Descriptive statistics using prism cruise data from four pine stands in Alberta, Canada, surveyed in 2014. LLB1/LLB2 were the jack pine stands and EDS1/EDS2 were lodgepole pine stands. Where t-tests were used (α = 0.05), significant values are indicated by different letters (i.e. ‘a’ differs from ‘b’, ‘ab’ does not differ from ‘a’ or ‘b’).  Stand Mean Stand Age Mean Stand Height (m) Mean stand DBH (cm) Range of DBH (cm) Stems/ha Proportion dead stems Proportion hardwood stems LLB1 80 15.6𝑎 20.7𝑎 3.1-38.2 445.2 0.27 0 LLB2 63 19.0𝑏 23.5𝑎𝑐 12.2-60.9 969.1 0 0 EDS1 106 24.6𝑐 28.5𝑏 12.3-44.5 924.2 0.04 0.03 EDS2 100 23.8𝑐 25.9𝑏𝑐 10.1-59.5 1317.3 0.07 0.08  9  The 100% cruise allows an analysis of woodpecker foraging patterns. LLB1 once again stood out as distinct from the other groups with significantly higher mean foraging, larger range of foraging, higher number of total trees with foraging activity, and higher proportion of trees with wood borer activity (Table 3). EDS2 had the lowest mean foraging values (although it did not differ significantly from LLB2 or EDS1) smallest range of total foraging on the tree boles, fewest trees with foraging activity, and lowest proportion of trees with wood borer activity (Table 3). My comparing Table 2 and Table 3, it is evident that DBH ranges were all smaller for bark beetle attacked trees than the stand as a whole, and the range of mean DBH measurements amoung the four stands was also narrower (stands ranged from 20.7- 28.5 cm, whereas bark beetle attacked trees mean DBH ranged from 21.3 to 26.1 cm). Table 3 – Descriptive statistics for 100% cruise data, i.e. all Pinus spp. with recent (≤ 3 years) bark beetle activity from four pine stands in Alberta, Canada, surveyed in 2014. ‘WP’ indicates woodpecker and ‘DBH’ indicates diameter breast height. Where t-tests were used (α = 0.05), significant values are indicated by different letters (i.e. ‘a’ differs from ‘b’, ‘ab’ does not differ from ‘a’ or ‘b’). Stand Mean WP Foraging (%) Range of WP foraging (%) Total trees with WP foraging Mean DBH (cm) Range of DBH (cm) Trees with wood borers (%) LLB1 5.6𝑎 0 - 60 25 23.4𝑎𝑏𝑐 11.4 – 42.4 0.54 LLB2 1.7𝑏 0 – 30 6 26.1𝑎𝑏 8.9 – 49.7 0.21 EDS1 1.5𝑏 0 – 40 6 24.1𝑎 14.6 – 37.4 0.17 EDS2 0.3𝑏 0 – 5  4 21.3𝑐 12.3 – 33.8 0.17  Logistic Regression  The step by step selection process for the binary logistic regression is displayed in Table 4. Following the first step, proportions of dead and proportion of hardwood stems were removed by the SAS program as they were found to be linear combinations of other variables. Stems/ha was eliminated next as it had the highest P value at 0.4940. No change in any of the model parameters was noted so this variable was considered to not add to the model. Next, the tree species variable was eliminated as it had the highest remaining P value at 0.5203. This removal did not alter the AIC or R² and rescaled R² values, increased the percent convergent, and decreased the percent discordant and percent tied to this variable was considered to take away from model efficiency. Finally, the three remaining variables were tested in pairs and individually to ensure no simpler model better predicted woodpecker foraging activity. The final chosen model showed that increased DBH, wood borer presence, and adjacent coniferous stands contributed to increased woodpecker foraging. This combination of these three variables were able to predict presence of woodpecker foraging activity 83.5% of the time (see Somers D 10  value in Table 4), with the model explaining between 35.1% and 56.8% of the data’s variability (see R² and rescaled R² values in Table 4). Table 4 – Binary logistic regression results systematically testing seven different predictor variables for woodpecker foraging presence on bark beetle affected Pinus trees in Alberta, Canada in order to find the most efficient model. Variable names and explanations are outlined in Table 1. For each model tested, the assumption of convergence was satisfied and passed SAS significance tests (Likelihood Ratio, Score, and Wald Chi Square tests). The final model is highlighted in bold. Variables AIC Rsq Rescaled R % Convergent % Discordant % Tied Somers D D,W,T,H,G,S,AD 208.18 0.354 0.573 91.3 8.4 0.3 0.829 D,W,T,AD 208.18 0.353 0.57 91.3 8.4 0.3 0.829 D,W,AD 208.18 0.351 0.568 91.6 8.2 0.2 0.835 D,W 208.18 0.287 0.465 87.1 12.4 0.5 0.746 D, AD 208.18 0.239 0.387 84.4 15.2 0.4 0.692 AD, W 209.41 0.274 0.4464 80.8 6.2 13 0.746 AD 209.82 0.14 0.23 52.3 6.1 41.6 0.462 W 209.41 0.23 0.37 64.8 3.7 31.5 0.611 D 208.18 0.085 0.139 71.2 27.9 0.9 0.433  Fisher’s Exact Test Fisher’s exact test revealed a significant P value of <0.001 when testing whether woodpeckers chose mass attacked trees preferentially, indicating there was a difference in foraging activity on mountain pine beetle attacked trees compared to other bark beetle attacked trees (Table 5).  Table 5 – Fisher’s Exact Test data for woodpecker foraging presence on bark beetle affected Pinus trees in Alberta, Canada surveyed in 2014. “WP” represents woodpecker and mass attacks refer to mountain pine beetle mass attacks. Category Name WP Activity No WP activity Mass attacked 11 6 No mass attacks 41 190  11  Discussion Predicting Woodpecker Foraging Activity  Initially, I predicted that the presence of deciduous trees within a stand or on stand borders would provide increased opportunities for woodpecker populations and mean an increased level of foraging activity (Bergvinson and Borden, 1992; Hutto, 1995; Drever and Martin, 2010). The logistic regression results from this study indicate the opposite and therefore refute this section of the hypothesis. Trees within stands lacking the adjacent deciduous forest were found to be more likely to have woodpecker foraging activity present than those with bordering aspen stands. This result can be explained in a few different ways. Out of 214 bark beetles known to exist in Canada, 156 exist solely in coniferous species, with an additional 6 species known to occur in both coniferous and deciduous host trees (Bright, 1976). It is then possible that a stand bordered by coniferous forest, like LLB1, will provide more continuous feeding opportunities than other stands bordered by deciduous trees. Since the forest surrounding LLB1 provided more feeding opportunities, this factor may have outweighed the effect of having immediately adjacent deciduous nesting habitat. This conclusion is supported by Bonnot et al. (2009) who found that food availability was the best predictor of black-backed woodpecker habitat selection when considering the territory scale and nesting opportunities only became important when considering the nest-tree scale. It is also possible that the scale of the study was too limited to take the size of woodpecker territories into account. For example, the black-backed woodpecker, Picoides arcticus, has been found to have territories as large at 766 ha (Dudley and Saab, 2007). In LLB1, there were deciduous patches of forest within approximately 2 kilometers meaning nesting habitat was available in the area if not within the stand itself. In future work, analyses could be improved by mapping all nearby areas of deciduous forest and creating an additional variable at the tree level stating shortest distance to deciduous trees or forest.  The second and third sections of my initial hypothesis were both supported. Both larger diameter trees and trees with wood borer activity were more likely to have woodpecker foraging activity. Since larger diameter trees had increased amounts of phloem area for bark beetles to inhabit, as well as increased volumes of wood for wood boring insects, it follows that this increase in possible food resources will mean an increase in woodpecker foraging activity. LLB1 had the smallest range of DBH measurements and lowest mean DBH overall, but the model shows that within the stand, it was larger individual trees that had the most woodpecker presence. Hutto (1995) found the same result and attributed this trend to the similar trends in occurrence of bark beetle larvae. This rule is not absolute as the largest diameter trees in LLB2, EDS1, and EDS2 had very limited woodpecker foraging activity 12  besides from the pileated woodpecker which seemed to focus feeding cavities on its preferred food source, carpenter ants (Torgersen and Bull, 1995; Bull et al., 1992).  At the point where DBH is very large and bark thickness becomes a barrier for bark beetles and woodpeckers, foraging activity of bark drilling birds should drop off. Finally, the presence of wood borers causing the increased occurrence of woodpecker foraging activity is well supported in the literature and was supported by this study as well (Bonnot et al., 2009; Murphy and Lehnhausen, 1998; Saab et al., 2014). The comparatively high fraction of dead stems in LLB1 could have contributed to the increased level of wood borer activity, then causing the increase in woodpecker activity in the stand in general.   The final prediction that tree species would be an important factor in determining level was refuted. The binary logistic regression did not include trees species in the best model. The descriptive statistics indicate that the stand with the highest overall level of woodpecker activity was a jack pine stand: LLB1. When considering the life histories of North American bark beetles, this makes sense as most species that attack one species of pine will attack other species as well and often many other conifer species (Amman and Cole, 1983; Bright, 1976). Since woodpeckers most likely do not differentiate between particular insect species, instead feeding on particular groups or families, estimating the level of woodpecker activity in stands where bark beetles are present based simply on trees species is not logical. Surveying resource opportunities and nesting habitat will be a better predictor of woodpecker presence and foraging activity (Bergvinson and Borden, 1992; Bonnet et al., 2009; Drever and Martin, 2010; Flower et al, 2014; Hutto, 1995; Murphy and Lehnhausen, 1998; Saab et al., 2014). Woodpeckers as a potential barrier to invasive bark beetles  The results of Fisher’s exact test indicate that woodpeckers chose mass attacked trees preferentially over non-mass attacked trees with a high level of significance. Although this information is not necessarily novel when considering the work of Bonnot et al. (2009), Murphy and Lehnhausen (1998), Flower et al. (2009), and others, it is still necessary to stress its importance. If woodpeckers prove to be viable regulators of potentially eruptive insect pests, then this regulatory role should be encouraged by forest managers.   Reviewing the research on woodpeckers as regulating agents of insect populations presents authors who are either confident or unsure of bark-drilling insectivores impact as species specific predators (e.g. Fayt et al., 2005; Flower at al., 2014; Machmer and Steeger, 1995). For instance, Machmer and Steeger (1995), in a comprehensive review of invertebrate and vertebrate tree-users 13  effects on many beetle species, state that vertebrate predators simply buffer insect population growth and decline and have a limited effect on both endemic and epidemic activity. Conversely, Fayt et al. (2005) focus on explaining what they call the overlooked potential of woodpeckers as regulating agents of problem insect populations in their review of spruce bark beetles and woodpeckers. They present evidence of the woodpeckers’ effects on this eruptive species and state that these avian predators most likely have a stabilizing effect on their prey. It is interesting to note that the second and third authors of the Fayt et al. (2005) paper are the same as those in Machmer and Steeger (1995). It appears a decade of additional research contributed to the change in conclusions of these authors on woodpeckers as regulating predators. Additionally, woodpeckers have been found to be significant predators of the invasive emerald ash borer in eastern North America (Flower et al., 2014; Duan et al., 2010) and the southern pine beetle in the southern United States of America (Kroll, 1980).   As for the woodpeckers direct effect on mountain pine beetle populations, evidence exists for their role in maintaining low population numbers during both endemic (Korol, 1985) and epidemic (Saab et al., 2014) population phases. Amman (1984) found that woodpeckers inflicted the most damage overall on the mountain pine beetle populations with predation percentages highest during epidemics. Amman’s (1984) work did not address woodpecker effects on incipient infestations, the stage just prior to an outbreak. In the jack pine stands surveyed in Alberta in this study, small patches of pines with high mountain pine beetle attack densities were almost entirely predated by woodpeckers. In these cases, the area of the bole debarked may have underestimated predatory effects as after investigation of bole areas where pitch tubes were present but debarking was not, it was noted that the beetle galleries were either non-existent or unsuccessful. This pre-eruption population phase may be where woodpeckers will have the greatest impact on beetle populations. Murphy and Lehhausen (1998) found that following fire, three-toed woodpecker relied primarily on Scolytid beetles, despite this group’s comparatively small size. This research presents a similar scenario to an incipient infestation, where a large number of trees in a relatively small area are attacked by bark beetles and the woodpeckers are able to respond quickly and efficiently to the resource opportunity. The unprecedented mountain pine beetle outbreak that has occurred in BC over the last two decades may not provide an ideal situation to gage the effectiveness of woodpeckers as control agents since the sheer expansiveness of the outbreak would mean woodpeckers would not have a chance of controlling the population, especially considering they are conservative breeders and have unaltered fecundities during times of resource abundance (Edworthy et al., 2011). The continuing spread of the 14  beetle across central and eastern North America presents a novel study arena to examine woodpecker-mountain pine beetle interactions, especially where limited outbreaks occur on the landscape. Management Implications Drever et al. (2009) found that the richness of insectivorous birds increased with the mountain pine beetle outbreak in the stands they studied in BC. Hutto (1995) also found that this same group, especially woodpeckers, were most abundant and able to take advantage of sudden resource opportunities following stand replacing fires. The literature indicates that if woodpeckers are present on the landscape, they will respond to any abundant resource. This means that if foresters wish to proactively manage potential mountain pine beetle damage both in western North America and further east, they can manage for woodpeckers. To ensure bark-drilling woodpecker presence in a stand where there is also the potential for bark beetle damage, managers can consider the factors indicated to be important by the logistic regression in this study and from the literature. By leaving a few larger living conifers and leaving behind standing or fallen dead wood for wood borer activity, they will provide woodpeckers with adequate resources for periods of low food availability (Kroll, 1980; Machmer and Steeger, 1995). This means that if an incipient infestation occurs with the possibility of becoming an epidemic, the woodpecker community may exist in sufficient numbers to control these small outbreaks and prevent population explosion.    LLB1 had the most woodpecker presence with higher woodpecker measurement in all categories: mean overall woodpecker foraging, range of bark removal percentages, and total trees with any foraging activity. This stand has the necessary characteristics for supporting a healthy woodpecker community most likely as a result of the abundant wood borers, high number of dead stems, and structural heterogeneous habitat although this stand did not appear to have high yield in terms of basal area and merchantable timber. Fayt et al. (2005) also noted the importance of stand and landscape structure in their review, specifically that woodpeckers seem to have greater predatory potential in more open stands where bark beetles may develop more successfully as a result of warmer temperatures. Drever and Martin (2010) also found that woodpecker species richness declined with increasing densities of lodgepole pines. Managers must find the balance between short-term profit and long-term forest health and protection, ensuring a stand and the landscape as a whole can respond and rebound from an invasive or eruptive pest. Specific management actions that will favour woodpecker presence on the landscape include leaving behind some dead stems to supplement woodpecker diets, ensuring some larger living conifer 15  and deciduous trees for feeding and nesting opportunities, as well as timber management and site preparation practice, such as limiting clear-cut size and favouring irregular shapes, as well as using winter burning instead of machine brush removal to prepare for planters (Kroll, 1980). Chan-McCleod and Bunnel (2003) discuss similar recommendations to those listed above, and also stress the importance of dead stems both standing and downed, as well dying trees. This means managers should leave these trees on the landscape where possible. Conclusion Woodpeckers can act as significant predators of bark beetles, but the level of significance is still being debated, both for specific woodpecker species and bark beetle species. The mountain pine beetle epidemic in BC was too vast to be effectively controlled by woodpeckers, but the beetle’s slower spread across central North America may produce very different ecological results. As long as stand characteristics exist to support woodpecker populations, these avian predators may be able to act as a barrier to small eruptions or even pre-epidemic conditions by inflicting sufficient damage upon an increasing population as to actually limit the spread of the beetle. Woodpecker habitat management in combination with responsible timber management practices may make mountain pine beetle control in this novel environment a real possibility.             16  References: Alberta Environment and Sustainable Resource Development. (2015). History of infestations. Retrieved from http://esrd.alberta.ca/lands-forests/mountain-pine-beetle/beetle-facts/history-of-infestations.aspx Alberta Environment and Sustainable Resource Development. (2013). The general status of Alberta wild species 2005: Species at risk. Retrieved from http://esrd.alberta.ca/fish-wildlife/species-at-risk/albertas-species-at-risk-strategy/documents/GeneralStatusAlbertaWildSpecies2005-Feb25-2013.pdf  Amman, G. D. (1984). Mountain pine beetle (Coleoptera: Scolytidae) mortality in three types of infestations. Environmental Entomology, 13, 184 – 191. Amman, G. D. & Cole, W. E. (1983). Mountain pine beetle dynamics in lodgepole pine forests part II: population dynamics (USDA Forest Service, General Technical Report INT-145). Ogden, UT: USDA. Avibase (2014). Avibase - bird checklists of the world: Alberta. Retrieved from http://avibase.bsc-eoc.org/checklist. jsp?region=caab&list=howardmoore  Beckwith, R. C., & Bull, E. L. (1985). Scat analysis of the arthropod component of pileated woodpecker diet. The Murrelet, 66(3), 90-92. Bergvinson, D. J. & Borden, J. H. (1992). Enhanced woodpecker predation on the mountain pine beetle, Dendroctonus Ponderosae Hopk., in glyphosate-treated lodgepole pines. Canadian Entomology, 124, 159-165. Bonnot, T. W., Millspaugh, J. J., & Rumble, M. A. (2009). Multi-scale nest-site selection by black-backed woodpeckers in outbreaks of mountain pine beetles. Forest Ecology and Management, 259, 220-228.  Bright, D. E. Jr. (1976). The insects and arachnids of Canada part 2: the bark beetles of Canada and Alaska. Ottawa, ON: Agriculture Canada. British Columbia Ministry of Forests. (2009). Vegetation resources inventory: Sample data compilation process. Retrieved from http://www.for.gov.bc.ca/hts/vri/standards/compiler/vri _complier_mar09.pdf British Columbia Ministry of Forests, Lands, and Natural Resource Operations. (n.d.). Mountain pine beetle in B.C. Retrieved from https://www.for.gov.bc.ca/hfp/mountain_pine_beetle/bbbrochure.htm Bull, E. L., Beckwith, R. C., & Holthausen, R. S. (1992). Arthropod diet of pileated woodpeckers in northeastern Oregon. Journal of Wildlife Management, 73(2), 42-45. Carroll, A. L., Taylor, S. W., Regniere, J., & Safranyik, L. (2003). Effect of climate change on range expansion by the mountain pine beetle in British Columbia. In T. L. Shore, J. E. Brooks, & J. E. Stone (Eds.), Mountain pine beetle symposium: challenges and solutions (223 – 232). Victoria, BC: Natural Resources Canada. Chan-McLeod, A. A., & Bunnell, F. (2003). Potential approaches to integrating silvicultural control of mountain pine beetle with wildlife and sustainable management objectives. . In T. L. Shore, J. E. Brooks, & J. E. Stone (Eds.), Mountain pine beetle symposium: challenges and solutions (267 – 277). Victoria, BC: Natural Resources Canada. Cullingham, C. I., James, P. M. A., Cooke, J. E. K., & Coltman, D. W. (2012). Characterizing the physical and genetic structure of the lodgepole pine - jack pine hybrid zone: mosaic structure and differential introgression. Evolutionary Applications, 5, 879–891. 17  Drever, M. C, & Martin, K. (2010). Response of woodpeckers to changes in forest health and harvest: Implications for conservation of avian biodiversity. Forest Ecology and Management, 259, 958-966.  Drever, M. C., Goheen, J. R., & Martin, K. (2009). Species-energy theory, pulsed resources, and regulation of avian richness during a mountain pine beetle outbreak. Ecology, 90(4), 1095 – 1105. Duan, J. J., Ulyshen, M. D., Bauer, L.S., Gould, J., & van Driesche, R. (2010). Measuring the impact of biotic factors on populations of immature emerald ash borers (Coleoptera: Buprestidae). Environmental Entomology, 39(5), 1513-1522.  Dudley, J. G., & Saab, V. A. (2007). Home range size of black-backed woodpeckers in burned forests of southwestern Idaho. Western North American Naturalist, 67(4), 593–600. Edworthy, A. B., Drever, M. C., & Martin, K. (2011). Woodpeckers increase in abundance but maintain fecundity in response to an outbreak of mountain pine bark beetles. Forest Ecology and Management, 261, 203-210. Farrar, J. L. (1995). Trees in Canada. Markham, ON: Fitzhenry & Whiteside Ltd. Fayt, P., Machmer, M. M., & Steeger, C. (2005). Regulation of spruce bark beetles by woodpeckers – a literature review. Forest Ecology and Management, 206, 1-14. Flower, C. E., Long, L. C., Knight, K. S., Rebbeck, J., Brown, J. S., Gonzalez, M. A., & Whelan, C. J. (2014). Native bark-foraging birds preferentially forage infected ash (Fraxinus spp.) and prove effective predators of the invasive emerald ash borer (Agrilus planipennis Fairmaire). Forest Ecology and Management, 313, 300-306. Hutto, R. L. (1995). Composition of bird communities following stand-replacement fires in northern Rocky Mountain (U.S.A.) conifer forests. Conservation Biology, 9(5), 1041-1058. Korol, J. J. (1985). A simulation of predation by non-game birds on the mountain pine beetle (Dendroctonus ponderosae Hopkins) (Unpublished master’s thesis). University of British Columbia. Kroll, J., Conner, R. N., & Fleet, R. R. (1980). Southern pine beetle handbook: woodpeckers and the southern mountain pine beetle (Agricultural Handbook No. 564). Washington, DC: U.S. Government Printing Office. Machmer, M. M. & Steeger, C. (1995).  The ecological role of wildlife tree users in forest ecosystems (Land Management Handbook No. 35). Victoria, BC: Research Branch B.C. Ministry of Forests. Mitton, J. B., & Ferrenberg, S. M. (2012). Mountain pine beetle develops an unprecedented summer generation in response to climate warming. The American Naturalist, 179(5), 163 – 171. Murphy, E. C. & Lehnhausen, W. A. (1998). Density and foraging ecology of woodpeckers following a stand-replacement fire. The Journal of Wildlife Management, 62(4), 1359-1372. Natural Resources Canada. (2015). Mountain pine beetle (factsheet). Retrieved from http://www.nrcan.gc.ca/forests/insects-diseases/13397 Saab, V. A, Latif, Q. S., Rowland, M. M., Johnson, T. N., Chalfoun, A. D., Buskirk, S. W., Heyward, J. E., & Dresser, M. A. (2014). Ecological consequences of mountain pine beetle outbreaks for wildlife in western North American forests. Forest Science, 60(3), 539-559.  SAS Institute Inc. (2014). SAS® 9.4 Statements: Reference, Third Edition. Cary, NC: SAS Institute Inc. Torgersen, T. R., and Bull, E. L. (1995). Down logs as habitat for forest-dwelling ants- the primary prey of pileated woodpeckers in northeastern Oregon. Northwest Science, 69(4), 294 – 303. 18  UCLA: Statistical Consulting Group. (2015). SAS Annotated Output Ordered Logistic Regression. Retrieved from http://www.ats.ucla.edu/stat/sas/output/sas_ologit_output.htm Vanellis, S. (n.d.). Blank map of Canada [online image]. Retrieved from http://chatt .hdsb.ca /~vanelliss/FOV1-000B9892/FOV1-000B9932/FOV1-001056BE/?OpenItemURL=S0A05B61E                       19  Appendix A Codes below were ascribed to each bark beetle infested tree where applicable. Trees were also noted as “dead” where applicable. Code Description 1 Broken top 2 Near windfall; next to large group of fallen trees where probably cause is wind damage 3 Scarred 4 Thin crown 5 Leaning  6 Suppressed; understory tree definitively not part of upper canopy 7 Mistletoe  8 Pitch moth; large masses of pitch noted and/or body of insect found 9 Wood borer presence 10 Bear damaged 11 Root rot 12 Other disease 13 Other; any damage not attributable to any other category; description required 14 Old beetle damage; damage older than 5 years 15 Fork or crook; definitive fork or crook anywhere on main stem           


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items