Effects of Management Practices on Avian Abundance in Oregon Coast Range Forests
Advisor: Delbert Hutchison
Forestry in Oregon has traditionally used an industrial model aimed to maximize timber production and revenue, with little attention to the potentially negative affects on ecosystem health and biological diversity. However, some landowners have begun experimenting with more sustainable management practices. This study examines the affect of some of these innovative silvicultural techniques on avian abundance and diversity in the Oregon Coast Range. Data on bird number and species were collected across designated stops for three years with each stop characterized by forest type (predominantly Douglas fir, mixed, or predominantly a species other than Douglas fir), understory (woody shrub, fern, or herbaceous), and treatment (control, lightly thinned, thinned, or patch cut). Total number of birds, number of birds in certain foraging guilds, and four indicator songbird species were compared across stops. Because data were collected with no clear analysis in mind, not all combinations of stop characteristics could be considered. Data were analyzed using one and two-way ANOVA, and results were corrected using the Bonferroni correction. While no significant results were found related to forest type or understory, birds clearly preferred the lightly thinned treatment. Studies analyzing all combinations of forest characteristics and comparing sustainably managed forests to industry methods should be implemented to more thoroughly answer the question of what management practices maximize forest ecosystem health.
Commercial forestry has remained a predominant use of forestland in Oregon since the mid-1800s (Oregon Forest Resources Institute 2013). Forests cover about half of Oregon, and roughly 80% of that forestland supports growth of commercial-grade timber. The vast majority of management of working forests in Oregon uses an industrial model aimed to maximize timber production and revenue, with little attention to the potentially negative effects on ecosystem health and biological diversity (Jones et al. 2012, Oregon Forest Resources Institute 2013). Industrial forest management uses a rotational monoculture crop system, heavy application of herbicides, and large clear cuts to maximize production of Douglas fir, the main commercial forestry crop in the Northwest (Oregon wild 2012). A focus on maximizing production and profit has led to a decrease in structurally complex and biologically diverse ecosystems. With most of Oregon’s working forestland managed in this unsustainable way, old growth forest is one of the few ecologically healthy forest types remaining. With old growth in Oregon diminishing from 14.2 million acres in 1935 to only 4.9 million acres in 1992 (Bolsinger 2011), finding new ways to mindfully manage Oregon’s forestlands to maintain and create ecological complexity and vitality is essential.
Some Oregon landowners have begun to shift their focus from maximum wood production and economic gain to a model aimed to enhance ecosystem health while still supporting an economically viable business (Thomas et al. 2006). While these changes have primarily taken place on private lands with owners who feel a responsibility to manage their properties sustainably, in response to public pressure these innovative silvacultural techniques are beginning to be implemented on public lands that have been managed in a more traditional, industrial way in the past (Franklin and Johnson 2013). However, these changes in management on public land are recent and small-scale. For example, recent legislation has mandated that the O&C lands, approximately 2.4 million acres of forestland in western Oregon that was reclaimed from the Oregon & California Railroad by the Federal Government in 1916 and has been heavily logged ever since, will be managed using a variable retention method (Association of O&C Counties 2008, Franklin and Johnson 2013). Variable retention harvesting, a method modeled on natural patterns of disturbance, may help to maintain and rebuild healthy ecosystem functioning (Franklin and Johnson 2013). This recent shift in focus has led to management practices focused on diversifying forests in terms of age and species in order to support ecosystem health and biodiversity (Davis et al. 2007, Davis et al. 2009, Muir et al. 2002). These new management practices often include stand thinning, limited herbicide use, and encouragement of variation in forest species and age composition, all of which are beneficial in that they support ecosystem health and biodiversity (Betts et al. 2013, Cahall et al. 2013, Hagar et al. 2004, Linden et al. 2012, Yegorova et al. 2013).
Birds, and specifically songbirds, function as indicators of ecosystem health due to their close associations with vegetation (Yegorova et al. 2013). These relationships between bird species and density and vegetation may function and be apparent in a variety of different ways. For example, certain species have been identified as especially good indicators due to their reliance on certain understory types for breeding habitat, the destruction of which is associated with declining bird populations (Canterbury et al. 2000, Stephens et al. 2011). Studies in forests of the Oregon Coast Range support the hypothesis that industrial practices including clear cutting, heavy herbicide use, and single age monoculture rotation, negatively affect avian abundance as compared to more varied and ecologically minded management techniques. For example, thinned, as opposed to unthinned, regeneration after clearcut promotes avian diversity (Cahall et al. 2013, Hagar et al. 2004, Hagar et al. 2009): changes in vegetation caused by thinning may mediate post-thinning bird response by affecting food sources such as invertebrate insects as well as nesting habitat for local birds (Hagar et al. 2012, Yegorova et al. 2013). Some have argued that recently clearcut areas may play an important role in initiating forest succession, an ecological process that provides diverse habitat to bird populations (Bosakowski 1997). However large clearcuts are highly destructive and may harm cavity nesting species as well as other birds that depend on trees and thick underbrush (Mahon et al. 2008). Clearcuts also vary in their value as avian habitat based on the amount of woody debris and snags retained (Bosakowski 1997). Herbicides applied in amounts typical of industrial forestry practices also negatively impact avian abundance as compared to sites with light or no herbicide application (Betts et al. 2013); while thinning promotes understory vegetation favorable to many bird species, herbicide use reduces understory vegetation (Betts et al. 2013, Hansen et al. 1995). Additionally, stand complexity in terms of tree age and species promotes avian diversity, while single-age monocultures significantly reduce avian diversity (Felton et al. 2011).
Located in the Oregon Coast Range, the owners of the family-owned and managed forestry business Hyla Woods have managed their forestlands to promote ecosystem health and biological diversity since 1986 (Hayes and Hayes 2013). Hyla Woods owns 825 acres divided between three locations in the Northern Oregon Coast Range: Manning (100 acres), Timber (173 acres), and Mt. Richmond (552 acres) (Fig. 1 – 4). Their management practices include varied thinning and patch cutting with a maximum clearcut size of two acres, limited, point specific herbicide use, and encouragement of multi-age, multi-species tree communities. Since 1986, the managers of Hyla Woods have worked to better understand the ecosystem functioning of their forestland through a variety of monitoring regimes. Through data collection and analysis they hope to answer three central questions: “what is the status of the health of these forests?; how is it changing over time?; and what can we understand about the causes of these changes, particularly the impacts of our actions on these changes?” (Hayes and Hayes 2012). Available data sets include water temperature, aquatic invertebrates, juvenile fish, spawning salmon, birds, owls, terrestrial amphibians, aquatic amphibians, large mammals, trees, and oak habitat. All data collection followed set protocols, but with no specific analysis in mind.
Because birds function as indicators for overall ecosystem health, and given the robust and long time-term nature of the Hyla Woods data set on birds, I used the bird data set to examine the relationship between forest treatment (unthinned, lightly thinned, thinned, or patch cut), understory, and forest health (Yegorova et al. 2013) as defined by the number and diversity of species supported by a given area. Due to the challenging nature of detecting trends in data collected with no clear study design, as was done in this project, one lesson from this analysis is the importance of clearly outlining study objectives and specific statistical analyses before beginning data collection.
Materials and Methods
Data were collected from three non-contiguous pieces of forestland (Timber, Manning, and Mt. Richmond sites), all within the western hemlock zone and between 400’ and 1200’ in elevation (Fig. 1). The three study areas consisted of predominantly Douglas fir ranging in age from recently planted to around 200 years old, and also contained a variety of other tree species including big leaf maple, western red cedar, grand fir, red alder, and Oregon white oak. The owners manage all three areas to promote biodiversity through a variety of harvest techniques
(thinning and patch-cutting), very limited herbicide use (spot treatments only), and encouragement of multi-age as well as multi-species tree communities.
At each forest, trails were created with multiple “stops” where data were collected. Bird and vegetation data were collected from 14 stops at Timber, 10 stops at Manning, and 24 stops at Mt. Richmond (Fig. 2-4). Three factors influenced stop determination. First, the combination of the stops at each separate forest is representative of the variation in that forest in terms of age, species, and understory; the owners tried to create enough stops to sample as many combinations of these variables as possible. Second, the route between stops must be accessible by foot on roads, trails, and skid trails. Lastly, stops were as evenly spaces as possible, while taking into account the first two considerations. Because of the size, shape, and accessibility of Mt Richmond, data collection occurred on two routes with 12 stops each (Fig. 4).
At each forest, observers trained in bird identification by sight and sound collected bird data at all stops for three consecutive years: at Manning from 2007 through 2009, at Timber from 2010 through 2012, and at Mt. Richmond in 2013 only. Data collection at Mt. Richmond will continue in 2014 and 2015. Yearly, observers recorded data at each stop on three days, each approximately a week apart during late May and early June. Observers did not collect data on days with rain or strong wind. Observers began walking the loop at 5:00 am (around first light). Once reaching a stop, they stood silently for approximately one minute before beginning a three minute period in which they recorded the species and abundance of all birds detected by sight or sound. Observers followed this point-count protocol at all stops and finished the route between 7:00 and 7:30 am (Huff et al. 2000).
Habitat data were collected once in 2013 at all stops at Timber, Manning, and Mt. Richmond. At each stop four subplots were sampled, each with a radius of ten meters, and located a randomly determined distance between ten and 50 meters directly north, east, south, and west of the stop center. An ocular estimate of percent vegetation cover within the categories woody shrubs, ferns, and herbaceous species at two height classes (<1.5 m and >1.5 m) for each subplot was also made.
Due to the varied species, understory, and treatment history between stops, each stop was generalized into categories of “forest type”, “understory” and “treatment”. Forest type was determined by summing the numbers of individual trees in distinct species across the four subplots. Based on total number of trees of each species, each stop was classified as predominantly Douglas fir, mixed, predominantly a species other than Douglas fir, or no trees. A stop was considered predominantly a single species if half again more of one species was present than other species. To determine the majority understory present at each stop, ocular estimates for three categories (woody shrub, fern, and herbaceous) were summed across the four subplots. Each stop was then classified within one of these three understory categories based on the most common understory type. Finally, stops were characterized by treatment groups as control (unthinned), lightly thinned, thinned, or patch cut. A number of stops contained variation between the four subplots and therefore received two treatment categories.
Data on bird species and total numbers collected across years on the three observation days within each year for each stop were summed and then standardized assuming nine total stops in order to make all stops comparable. Bird species were then divided into foraging guilds and the total number of birds was calculated in each of seven foraging guilds, taking into consideration the standardized totals. Bird species were divided as classified by Bryce’s guild system (Bryce 2006), and those not included in this guild system were assigned based on their foraging habits (Poole 2005, Sibley 2003, Wilson 1999). Foraging guilds included ground gleaner (GG), foliage gleaner (FG), bark gleaner (BG), hawker (HA), hover feeder (HOV), aerial feeder (AER), and aerial patrol (AP). In addition to grouping species into guilds, four songbird species, Macgillivray’s warbler (MGWA), orange-crowned warbler (OCWA), Swainson’s thrush (SWTH), and Wilson’s warbler (WIWA), known as especially good indicators for forest biodiversity and health (Stephens et al. 2011) were used to analyze songbird association across forest type, understory, and treatment.
Although the three areas were studied at different times, their similarities in region, management, and community composition across the three forests meant they could be combined into one data set for analysis. Thus, when combined across forests the bird and habitat data sets included 48 stops. Because the data were collected without any specific analytical techniques in mind, there are gaps in sampling which limit statistical analyses. However, it was possible to “mine” the data for suitable combinations of variables that did lend themselves to one- and two-way ANOVAs (Rstudio, Minitab).
The unbalanced design of this study limited the number and type of possible statistical analyses. With the data available the effect of forest type, understory, and treatment on total number of birds, numbers of specific indicator species, and prevalence of guild types was analyzed. Data for all combinations of factors (forest type, understory, and treatment) were not available, and therefore results are specific to certain groupings of stand characteristics. Because twenty-two statistical tests were performed, alpha values were corrected using the Bonferroni correction (Rice 1989) to account for the fact that at an alpha level of 0.05, and given twenty-two tests, statistically one test would be expected to yield falsely positive results.
All results were deemed significant by the Bonferroni correction unless otherwise indicated by the appropriate p-value.
Using a one-way ANOVA, total number of birds was compared across three forest types (predominantly mixed, predominantly Douglas fir, and predominantly a species other than Douglas fir) at stops with herbaceous understory. At an alpha level of 0.10, the mean number of birds found in stands with a predominantly Douglas fir forest type with herbaceous understory was significantly higher than the mean number of birds found in mixed species stands with herbaceous understory (F = 49.85, d.f. = 1, p = 0.09, r2 = 6.32, Fig. 5). Total number of birds was also compared across the three forest types (predominantly mixed, predominantly Douglas fir, and predominantly a species other than Douglas fir) at stops with herbaceous understory and a control treatment. Although the results were not significant, a graphical representation shows a trend in which bird number was highest in stands with a predominantly Douglas fir forest type, followed by mixed stands, and lowest in stands with no trees (Fig. 6).
Using a two-way ANOVA, total birds were compared across forest types by treatment at stops with herbaceous understory. Two-way ANOVA was used to compare total number of birds designated as a combination of the foraging groups ground gleaner and foliage gleaner across the three forest types (predominantly Douglas fir, mixed, and predominantly a species other than Douglas fir) at stops with herbaceous understory, but no significant results were found. Finally, total birds were compared across forest types at stops with herbaceous understory with the foraging guilds ground gleaners and foliage gleaners evaluated separately. No significant results were found for any of these tests. Given available data, no other statistical tests were possible using the variable forest type.
Using one-way ANOVA, total number of birds was compared between stops with woody shrub and herbaceous understory in stands with a predominantly Douglas fir forest type. No significant results were found. A series of one-way ANOVAs were used to analyze the distribution of four indicator species (MacGillivray’s warbler, orange-crowned warbler, Swainson’s thrush, and Wilson’s warbler) across understory types. None of these indicator species were significantly more prevalent in one understory type over another.
Sufficient data were available to perform a one-way ANOVA of total birds summed across the four indicator species across all four treatments (control, lightly thinned, thinned, and patch cut). The total number of birds was significantly higher in lightly thinned stands (an intermediate between control and patch cut) than in control, thinned, or patch cut stands (F = 5.96, d.f. = 3, p = 0.001, r2 = 59.97, Fig. 7). Using one-way ANOVA, total number of birds was compared between the treatments thinned and patch cut in stands with a predominantly Douglas fir forest type. No significant results were found. Total number of birds was also compared between control and thinned stands at stops with herbaceous understory and a predominantly Douglas fir or mixed forest type, as well as across treatments at stops with a predominantly Douglas fir forest type and herbaceous understory. Neither test delivered significant results. Two-way ANOVA was used to compare total birds between thinned and patch cut treatments by understory (only including woody shrub and herbaceous) in stands with a predominantly Douglas fir forest type. While no significant results were found, the graphical representation of total birds across treatment and understory in stands with a predominantly Douglas fir forest type revealed a striking interaction effect in which herbaceous understory was most common in patch cut stands and least common in thinned stands (Fig. 9).
Two-way ANOVA was used to compare total birds across treatments by the four indicator bird species. There was an interaction effect in which the summed number of birds across all indicator species (but especially orange-crowned warblers and Swainson’s thrushes) was higher in lightly thinned stands than in control, thinned, or patch cut stands (F = 2.06, d.f. = 9, p = 0.035, r2 = 59.97, Fig. 11). A series of one-way ANOVAs were used to determine the distribution of the four indicator bird species across treatments. Orange-crowned warblers were present in significantly higher numbers in lightly thinned stands than in control, thinned, or patch cut stands (F = 9.09, d.f. = 3, p < 0.0001, r2 = 38.26, Fig. 8), while the other three indicator species did not vary significantly across treatments.
Total number of birds was compared between the foraging guilds ground gleaners and foliage gleaners at stops with herbaceous understory, as well as across the four indicator species, using a series of one-way ANOVAs. There were significantly more birds in the foliage gleaner group than in the ground gleaner group across all stops (F = 195.23, d.f. = 1, p < 0.0001, r2 = 67.50, Fig. 10). The number of birds also varied significantly between indicator species with significantly more MacGillivray’s warblers than orange-crowned warblers, and significantly more orange-crowned warblers than Swainson’s thrushes and Wilson’s warblers (F = 48.13, d.f. = 3, p < 0.0001, r2 = 59.57, Fig. 11).
Overall bird numbers were higher in stands with a predominantly Douglas fir forest type and herbaceous understory than in mixed stands with an herbaceous understory (not significant under the Bonferroni correction). This result was counter to expectations that more birds would be found in mixed stands due to the wider variety of habitat and availability of associated niches (Molles 2008). This pattern of more birds in Douglas fir stands than mixed stands may have differed in stands with a fern or woody shrub understory type. When comparing total number of birds across forest type at stops with an herbaceous understory and control treatment the results were not significant, however the overall trend suggests that bird density is highest in predominantly Douglas fir stands, followed by mixed, with lowest bird density in stands with no trees. This trend agrees with the results found when comparing total bird number between predominantly Douglas fir and mixed stands with an herbaceous understory.
No significant results were found when using understory as the independent variable. Due to data limitations, a comparison between understory types within all forest types and treatments was not possible. Performing more tests using understory may have produced more significant results.
When the total number of birds was summed across the four indicator species and compared across treatments, more birds were present in lightly thinned stands than in control, thinned, or patch cut stands. Additionally more orange-crowned warblers were found in lightly thinned stands than in control, thinned, or patch cut stands. These differences in bird density between treatments may result from available nesting habitat, prevalence of certain types of understory, or the amount of bare ground or open space (Parrish and Hepinstall-Cymerman 2012). Because birds appear to prefer lightly thinned stands to other treatments, the majority of birds likely forage in vegetation associated with lightly thinned stands and do not require the open space provided by patch cut stands.
A non-significant trend suggests that herbaceous understory is most common in patch cut stands and least common in thinned stands. While data were not available to determine the most common understory type in lightly thinned stands, it may be extrapolated that lightly thinned stands either have more fern or woody shrub understory types or more bare ground. The association between bird number and treatment is clearly visible in the higher density of birds in lightly thinned stands than stands with other treatments, and is likely explainable by the association between understory and treatment. Sufficient data were not available to accurately analyze this connection. These results on bird distribution across treatments do not include an analysis of species distribution beyond looking into foraging guilds and four specific indicator species. Analyzing species distribution would be valuable in determining whether the trends apparent in this study result from many individuals belonging to a small number of species or fewer individuals belonging to many species. While light thinning may be the best harvest method for the greatest number of birds, light thinning may not promote species diversity because certain species may rely on habitat not available in lightly thinned stands.
Comparing total number of birds across the two foraging guilds, ground gleaner and foliage gleaner, significantly more foliage gleaners were present across all stops. This abundance of foliage gleaners is likely explained by overall species composition in the area. However, an analysis of species composition within the stops monitored would determine whether this result is due to many distinct foliage gleaning species or many individuals belonging to a small number of foliage gleaning species. Because foliage gleaners are present at higher densities than ground gleaners across all stops, and the total number of birds summed across the four indicator species is highest in lightly thinned stands, there may be more foliage present in lightly thinned stands than in other treatments. To accurately determine this association, data on total bird density across all species would have to be analyzed. Due to the presence of foliage at all stops analyzed, this abundance of foliage gleaners is expected. An analysis of industrially managed, clearcut land may result in more ground gleaning species.
Significantly more MacGillivray’s warblers than orange-crowned warblers, and significantly more orange-crowned warblers than Swainson’s thrush or Wilson’s warblers were present across all stops. Similar to the distribution of birds across foraging guilds, these differences may simply indicate general abundance in the area. It is also possible that Hyla Woods creates habitat that better provides for the requirements of certain indicator species than for those of other indicator species. These results are not directly associated with foraging guild type because all indicator species used in this study belong to the foliage gleaning guild.
The most significant trends from this data are that birds prefer stands with a predominantly Douglas fir forest type, and birds prefer lightly thinned stands to other treatments. Because the data on forest type only includes stops with an herbaceous understory and is not significant using the Bonferroni correction, additional research should be done before drawing any major conclusions regarding forest type. Because similar results emerged from a variety of tests on treatment, the conclusion that birds prefer lightly thinned stands seems fairly robust. In creating avian habitat, landowners should implement the harvest method of light thinning. However, because different species require a variety of habitat (Parrish and Hepinstall-Cymerman 2012), and total number of birds does not necessarily correlate with total number of species, maintaining a variety of habitat through varied treatment methods remains important.
More significant differences between stops may not have been found due to limitations in data. It is also possible, however, that the amount birds use these different stops does not vary significantly. Because the properties owned by Hyla Woods are relatively small, birds may travel throughout each property, using a variety of habitat types. Because the Hyla Woods properties are primarily adjoined to land managed in an industrial method relying on large clearcuts, heavy herbicide application, and single-age monoculture rotation, birds may pack into Hyla woods properties from these surrounding areas (Hejl 1992). Birds may also gravitate toward the edge habitat between forest and clearcut (Bosakowski 1997).
Future research should be performed to more completely answer the question of how the different forest characteristics present on Hyla Woods property impact avian populations. A study design similar to the one used in this paper should be implemented with specific research questions and analysis methods in mind. The set up should include stops with all combinations of the factors investigated as well as replicates of those combinations. Stops should be located in the center of stands demonstrating the desired characteristics to avoid stops with intermediate characteristics. Stops should also be located away from roads and major trails to avoid the impact of these open spaces on bird behavior as well as habitat characteristics. In addition to analyzing differences between certain characteristics within Hyla Wood’s property, stops should be added on neighboring properties. These additional stops would provide data to test the packing hypothesis, that birds may disproportionately utilize the habitat created by Hyla Woods. Two of the three main factors that differ between Hyla Woods management and that of industrially managed forests were investigated in this study, that of forest type and harvest method. The third factor, herbicide use, could also be investigated with the addition of stops on neighboring property.
While avian abundance may be used as a proxy for forest health, it may also be important to analyze avian diversity. As many factors as possible should be used in determining whether a forest ecosystem is “healthy”. Forest health may be difficult to quantify in a landscape so heavily influenced by human presence. Hyla Woods has the appropriate data on birds as well as many other species to perform a study comparing species abundance and diversity on their properties to that of industrially managed forests. These additional studies may shed more light on the beneficial impacts of alternative silvicultural practices on forest ecosystems. More complete research may in turn influence policy associated with managing state forestland.
Thank you to Pam and Peter Hayes, owners and managers of Hyla Woods, who supplied data and oversaw the research portion of this project. I would also like to acknowledge the birders who helped with bird identification and data collection, Steve Engel, Nate Richardson, and Lars Norgren with a special thanks to Char Corkran and Lori Hennings who donated their time and expertise as lead birders. Thank you to Linda Craig, Ken Chamberlain, Kahler Martinson, and Charlie Graham for helping with count timing, recording of bird counts, and bird identification. Thank you to Joan Hagar who supplied her habitat monitoring protocol for this project. Finally, thank you to Delbert Hutchison, Associate Professor of Biology at Whitman College, for his support and guidance as the advisor to this project.
Association of O&C Counties. 2008. Oregon and California Revested Grantlands. http://www.oandccounties.com. April 7, 2014.
Bosakowski, T. 1997. Breeding bird abundance and habitat relationships on a private industrial forest in the Western Washington Cascades. Northwest Science. 71(3): 244-253.
Betts, M.G., Verschuyl, J., Biovanini, J., Stokely, T., and Kroll, A.J. 2013. Initial experimental effects of intensive forest management on avian abundance. Forest Ecology and Management.
Bolsinger, C. 2011. Analysis of a pre-existing condition: the Northwest’s old-growth forests. The Oregonian. December, 2, 2013.
Bryce, S.A. 2006. Development of a bird integrity index: measuring avian response to disturbance in the blue mountains of Oregon, USA. Environmental Management 38(3): 470-486.
Cahall, E.J., Hayes J.P., and Betts, M.G. 2013. Will they come? Long-term response by forest birds to experimental thinning supports the “field of dreams” hypothesis. Forest Ecology and Management. 304: 137-149.
Canterbury, G.E., Martin, T.E., Petit, D.R., Petit, L.J., and Bradford, D.F. 2000. Bird communities and habitat as ecological indicators of forest condition in regional monitoring. Conservation Biology. 14(2): 544-558.
Davis, L.R. and Puettmann, K.J. 2009. Initial response of understory vegetation to three alternative thinning treatments. Journal of Sustainable Forestry. 28: 904-934.
Davis, L.R., Puettmann, K.J., and Tucker, G.F. 2007. Overstory response to alternative thinning treatments in young douglas-fir forests of western Oregon. Northwest Science. 81: 1-14.
Felton, A., Andersson, E., Ventorp, D., and Lindbladh, M. 2011. A comparison of avian diversity in spruce monocultures and spruce-birch polycultures in Southern Sweden. Silva Fennica. 45(5): 1143-1150.
Franklin, J.F. and Johnson, K.N. 2013. Alternative harvesting method provides foundation for Wyden O&C plan. http://www.oregonlive.com/opinion/index.ssf/2013/11/alternative_har vesting_method.html. April 7, 2014.
Hagar, J., and Friesen, C. 2009. Young stand thinning and diversity study: response of songbird community one decade post-treatment. U.S. Geological Survey Open- File, Report 2009-1253.
Hagar, J., Howlin, S., and Ganio, L. 2004. Short-term response of songbirds to experimental thinning of young Douglas-fir forests in the Oregon cascades. Forest Ecology Management. 199: 333-347.
Hagar, J. C., Li, J., Sobota, J., and Jenkins, S. 2012. Arthropod prey for riparian associated birds in headwater forests of the Oregon Coast Range. Forest Ecology and Management. 285: 213-226.
Hansen, A.J., McComb, W.C.,Vega, R., Raphael, M.G., and Hunter, M. 1995. Bird habitat relationships in natural and managed forests in the west Cascades of Oregon. Ecological Applications. 5(3): 555-569.
Hayes, P.S., and Hayes, P.S. 2012. Hyla woods experimental forest – forest assessment. 1-10.
Hayes, P.S., and Hayes, P.S. 2013. Hyla Woods. hylawoods.com. April 7, 2014.
Hejl, S.J. 1992. The importance of landscape patterns to bird diversity: a perspective from the Northern Rocky Mountains. The Northwest Environmental Journal. 8: 119-137.
Huff, M.H., Bettinger, K.A., Ferguson, H.L., Brown, M.J., and Altman, B. 2000. A habitat-based point-count protocol for terrestrial birds, emphasizing Washington and Oregon. General Technical Report: PNW-GTR-501. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 39.
Jones J.E., Kroll, A.J., Giovanini, J., Duke, S.D., Ellis, T.M., and Betts, M.G. 2012. Avian species richness in relation to intensive forest management practices in early seral tree plantations. Plos ONE. 7(8): e43290.
Linden, D.W., Roloff, G.J., and Kroll, A.J. 2012. Conserving avian richness through structure retention in managed forests of the Pacific Northwest, USA. Forest Ecology and Management. 284: 174-184.
Mahon, L.C., Steventon, D.J., and Martin, K. 2008. Cavity and bark nesting bird responses to partial cutting in northern conifer forests. Forest Ecology and Management. 256: 2145-2153.
Molles, M.C. 2008. Ecology: Concepts & Applications, Fourth Edition. McGraw-Hill Higher Education. New York, NY.
Muir, P.S., R.L. Mattingly, J.C. Tappeiner, J.D. Bailey, W.E. Elliott, J.C. Hagar, J.C. Miller, E.B. Peterson, and E.E. Starkey. 2002. Managing for biodiversity in young Douglas-fir forests of western Oregon. U.S. Geological Survey, Biological Resources Division, Biological Science Report USGS/BRD/BSR–2002-0006. 76.
Oregon Forest Resources Institute. More to learn. OregonForests.Org. November, 30, 2013.
Oregon Forest Resources Institute. 2013. Oregon forest facts & figures 2013. Oregon: Oregon Forest Resources Institute.
Oregon wild. 2012. BLM practices vs. Oregon forest practices act. Oregon wild. December, 2, 2013.
Parrish, M.C. and Hepinstall-Cymerman, J. 2012. Associations between multiscale landscape characteristics and breeding bird abundance and diversity across urban-rural gradients in Northeastern Georgia, USA. Urban Ecosystems. 15(3): 559-580.
Poole, A. (Editor). 2005. The birds of North America online. Cornell Laboratory of Ornithology. November, 30, 2013.
Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution. 43: 223-225.
Sibley, D. 2003. The Sibley field guide to birds of western North America. New York: Knopf.
Stephens, J. L., K. Kreitinger, C. J. Ralph, and M. T. Green (editors). 2011. Informing ecosystem management: science and process for landbird conservation in the western United States.
U.S. Department of Interior, Fish and Wildlife Service, Biological Technical Publication, FWS/ BTP-R1014- 2011, Washington, D.C.
Wilson, J.B. 1999. Guilds, functional types and ecological groups. Oikos 86: 507-522.
Thomas, J.W., Franklin, J.F., Gordon, J., and Johnson, K.N. 2006. The Northwest forest plan: origins, components, implementation experience, and suggestions for change.
Yegorova, S., Betts, M.G., Hagar, J., and Peuttmann, K.J. 2013. Bird-vegetation associations in thinned and unthinned young Douglas-fir forests 10 years after thinning. Forest Ecology Management. 1-14.