Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The status and distribution of terrestrial amphibians in old-growth forests and managed stands Dupuis, Linda A. 1993

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

Item Metadata


831-ubc_1993_fall_dupuis_linda.pdf [ 2.82MB ]
JSON: 831-1.0086212.json
JSON-LD: 831-1.0086212-ld.json
RDF/XML (Pretty): 831-1.0086212-rdf.xml
RDF/JSON: 831-1.0086212-rdf.json
Turtle: 831-1.0086212-turtle.txt
N-Triples: 831-1.0086212-rdf-ntriples.txt
Original Record: 831-1.0086212-source.json
Full Text

Full Text

THE STATUS AND DISTRIBUTION OF TERRESTRIAL AMPHIBIANS IN OLD-GROWTH FORESTS AND MANAGED STANDS by LINDA ANNE DUPUIS B.Sc., Carleton University, 1988 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Zoology)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA JULY 1993 © Linda Anne Dupuis, 1993  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  71^  The University of British Columbia Vancouver, Canada  Date  ^  DE-6 (2/88)  t  i7  19 ^/993  ABSTRACT  In this thesis, I compare the abundance of terrestrial amphibians in old-growth forests and young and mature post-harvest stands. I also examine the habitat features with which they are strongly associated, and contrast these between old growth and second growth to gain insight on the specific effects of logging operations. Terrestrial amphibian surveys were carried out by means of two techniques: 1) area-constrained searches (ACS), and 2) log surveys (LGS). ACS consisted of thoroughly searching a series of quadrats (3x3m in 1991, 1x2 m in 1992), which were randomly placed within study sites. LGS involved detailed searches of logs in early, intermediate and advanced stages of decay. My study demonstrated that clearcut harvesting reduces terrestrial amphibian populations by as much as 70%. I suggest that the mechanism behind this pattern is a reduction in the availability of moist microhabitats. In support of this, salamander densities in managed stands were similar to those in old growth within 10 m of streams. Moreover, managed stands lack large logs in intermediate and late stages of decomposition, and these are an important source of cover for salamanders, notably the western red-backed salamander,  a.  vehiculum. The proximity of old growth to managed stands may also be crucial. Second  growth habitats which were isolated from old growth, maintained lower densities of terrestrial amphibians than second-growth stands adjacent to old growth. In light of my results, I make several recommendations to forest managers, which consider the needs of terrestrial amphibians.  TABLE OF CONTENTS  Abstract^  ii  Table of Contents^  iii  List of Tables  ^v  List of Figures^  vi ^v i i  Acknowledgements CHAPTER I: General Introduction^  1  CHAPTER II: The relation of stand age and history to the abundance of terrestrial amphibians Introduction^ Study Area Site Selection and Habitat Description ^  5 ^6 7  Experimental Design  ^7  Methods  ^1 0  Amphibian Surveys^  10  Soil Moisture Samples  ^1 1  Habitat Measurements  ^1 1  Results^  12  Diversity and Abundance of Terrestrial Amphibians^1 2 The Effect of Moisture^  15  Habitat Attributes^  19  Landscape Characteristics ^  19  Discussion  ^2 3  The Status of Terrestrial Amphibians on Vancouver Island ^2 3 The Importance of Moisture ^  24  Habitat Attributes^  26  Landscape Considerations  ^2 6  Management Recommendations  ^2 8  CHAPTER III The importance of logs to the amphibian fauna of a coastal forest  Introduction^  30  Study Area, Site Selection and Experimental Design  ^31  Methods  ^3 1  Downed Wood Surveys  ^31  Amphibian Surveys  ^3 2  Results^  33  Wood Quality and Quantity in Forests and Stands  ^3 3  The Distribution of Amphibians on Downed Wood  ^3 6  Discussion  ^4 0  The Importance of Wood Quality to Salamanders Management Recommendations  ^4 0 ^4 4  CHAPTER IV General Conclusions The inhospitability of post-harvest stands to terrestrial amphibians  ^4 6  Future Research  ^4 7  A Comparison of Survey Techniques ^  48  Literature Cited^  51  Appendix^  57  iv  LIST OF TABLES Table  Page  2.1  Physical characteristics of each study site  8  2.2  Landscape characteristics of each study site  9  2.3  The condition of vehiculum in Douglas-fir old-growth forests and managed stands, in the springs of 1991 and 1992 ^  14  Average differences in vegetation structure and soil depth between old-growth forests, and young and mature managed stands ^  22  The availability of soil-interface refuges among logs of differing size classes in old-growth forests and managed second-growth stands ^  35  The distribution of terrestrial amphibians on various log sizes in 1992, in old growth and mature post-harvest second growth ^  41  2.4 3.1 3.2  a.  v  LIST OF FIGURES Figure  ^  Page  2.1^The average density of amphibians encountered in old growth and mature and young managed second growth, in the springs of 1991 and 1992 ... ^13 2.2^Terrestrial amphibian abundance and precipitation levels in the spring ^1 6 and summer of 1991 and 1992 2.3^The proportion of small, medium and large E. vehiculum in early and late spring, 1991 and 1992^  17  2.4^Soil moisture in old growth and managed stands in 1991-92^ 1 8 2.5^Seasonal use of cover objects in 1991 and 1992  ^2 0  2.6^The abundance of salamanders near (<10m) and away from (>30m) streams, in old growth and managed stands, in May 1992  ^21  3.1^The volume of wood in various stages of decay in old-growth forests, and in young and mature managed second growth stands  ^3 4  3.2^The total number of logs with sloughed bark encountered in old growth, and young and mature managed second growth ^3 7 3.3^Log positions adopted by terrestrial amphibians in old growth and mature managed stands  ^3 8  3.4^Seasonal use of logs in various stages of decay, in old growth and mature managed stands, in 1991 and 1992 ^3 9  vi  ACKNOWLEDGEMENTS  I would like to express my thanks to Jamie Smith, master of the red pen, for helping to improve my thesis by thoroughly and patiently editing all the rough drafts. I also owe a debt of gratitude to Fred Bunnell for his continued encouragement and enthusiasm about my project. Both members of my committee, as well as J.P. Savard, provided financial support which was much appreciated. I owe a special thanks to Elaine Floritto, Michele Evelyn, Chris Cheng, Colin Duffield and Elvira Harms, who were eager, reliable and enthusiastic assistants despite the swarms of no-see-urns, blackflies and mosquitoes, the tent floods, the salal, and the darn 'regen'. I am also greatful to Grègoire CM, Gabriella Matsha and Markus Laub, my wonderful and helpful volunteers. I'm sure they agree that the hot tub at the Port Alberni Community Centre deserves to be mentioned too. Thanks to the staff of MacMilland Bloedel for their cooperation. Ron McGlaughlin in particular, worked hard to help me succesfully complete my first field season. My biggest and most special thank-you goes to Wayne French, engineer for the M&B Sproat Lake Division, in Port Alberni. He bent over backwards to help me every step of the way, from field work to write-up.  vii  CHAPTER I  GENERAL INTRODUCTION  In recent years, there has been a worldwide increase in interest in conservation and the preservation of biological diversity. As human populations continue to grow, the impact of industrial and agricultural activities on the landscape and biosphere is increasing in parallel. Thus, there is a pressing need for resource managers to consider the needs of wildlife species. A first requirement of sound wildlife management is an inventory of species, information on their natural histories, and an assessment of their habitat needs and relative abundances (Burton et al. 1992; Beiswenger 1988; Bogan et al. 1988). Growing public and scientific concern over the potential effects of harvesting oldgrowth forests on wildlife has prompted considerable research on species' habitat requirements, and whether or not these needs are met by present day forestry practices. Post-harvest stands are usually managed to maximize future timber supplies. Current silvicultural practices include: 1) mechanical or chemical elimination of unwanted stems in young stands; 2) application of herbicides designed to suppress broadleaf vegetation growth and hasten reforestation of conifers (Morrison and Meslow 1984; Slagsvold 1977); 3) piling and often burning the wood debris left behind after harvesting to render a site plantable and to reduce fire hazard (Bartels et al. 1985); and 4) seeding and planting selected genetic stocks, in areas where natural regeneration is insufficient to meet the desired number of seedlings per hectare (Meslow and Wright 1975; Santillo et al. 1989). The removal of the trees and the multilayered canopy substantially alters the microclimate of a forest (Franklin 1988; Geiger 1971). For example, ground temperatures in a forest and clearcut can differ by as much as 15.6 °C on clear sunny days (e.g., Blymyer and McGinnes 1977). Such habitat alterations can affect the species composition of plant and animal communities (e.g., Eckert et al. 1991; Bruce et al. 1985; Hall et al. 1985; Titterington et al. 1979).  1  Most past work on wildlife has focused on a few game species (Jones 1988; Buhlmann et al. 1988). Recently, attention has shifted to include non-game birds and mammals (Gerell 1988; Helle and Jarvinen 1986; MacFarlane 1988; Raphael 1988). Research on amphibians and reptiles lags well behind (Gibbons 1988) due to: a) a lack of interest; b) their complex (aquatic and terrestrial) life histories; and c) because they are nocturnal and often hidden under moist cover objects to avoid wind and fluctuating temperatures (Beiswenger 1988). They are, however, numerically dominant in many habitats. Corn and Bury (1990) estimated a mean density of 400 salamanders/ha in California redwoods. Densities can reach 3,000 salamanders/ha in eastern deciduous forests (Burton and Likens 1975). Hairston (1987) states that the biomass of salamanders in the Southern Appalachians exceeds that of all other vertebrate predators combined. Despite their large numbers in favourable habitats, amphibian population declines are occurring worldwide (McDiarmid 1991), mainly because of habitat loss and degradation (Johnson 1992). The responses of species to habitat change are directly related to their range of behavioural and morphological adaptations (Bruce et al. 1985). This range is restricted in amphibians (Jaeger 1980; Keen 1979; Krzysik 1979; Maiorana 1978), because they have a low tolerance for moisture and temperature changes (Jaeger 1980; Maiorana 1978; Welsh 1990). Cutaneous respiration, their sole means of gaseous exchange, can not occur unless their semi-permeable skin is moist. This vulnerability may be linked to their worldwide demise, particularly in the face of large scale disturbance. Their semipermeable skin also makes them sensitive to airborne and waterborne pollutants, acid rain and pesticide runoff. Known effects of pollutants on amphibians include: paralysis, reduced growth rates, loss of balance, reduced levels of vital chemicals and vitamins during egg and larval development, and reduced tolerance to low temperature (Pawar et al. 1983; Marchal-Segault and Ramade 1981; Johnson 1980). Terrestrial amphibians are potentially valuable indicators of habitat condition because they (1) are physiologically constrained by their need for moisture, (2) occur in 2  large numbers, (3) may spend their entire life cycle in one area (K. Ovaska pers. corn.; T. Davis pers corn.), (4) have relatively stable populations (Hairston 1987), and (5) are long-lived. Hairston (1987) estimates a generation time of approximately 10 years for Plethodon jordani and P. alutinosus. Longevity is adaptive in species where environments are unpredictable (Maiorana 1978). Terrestrial amphibian abundance might therefore serve as an excellent index of habitat quality; reflecting the long-term impact of disturbances such as logging operations with considerable sensitivity. Recently the U.S. Forest Service (USFS) has focused on amphibian and reptile forest associations in Oregon, California, and Washington. In general, studies there have found that amphibian species are one and a half to five times more abundant in old growth than in managed stands (Bury and Corn 1988; Welsh and Lind 1988). Several species are strongly associated with old growth (Welsh 1990; Buhlmann et al. 1988; Bury and Corn 1988; Ratmonik and Scott 1988). Welsh (1990) suggested that some species may become extinct if forestry practices fail to consider their specific needs. To my knowledge, few studies have examined the association of amphibians and forest age and structure in Canada. The diversity of amphibian species is relatively low in Canada, as few species can tolerate cold climates. The species found in coastal forests of Vancouver Island are: western toad (Bufo boreas), pacific treefrog (Hyla regilla), red-legged frog (Rana aurora), northwest salamander (Ambystoma aracile), ensatina (Ensatina eschscholtzi), long-toed salamander (Ambystoma rnacrodactylum), rough-skinned newt (Taricha aranulosa), clouded salamander (Aneides ferreus), and western red-backed salamander (Plethodon vehiculum) (Cook 1984; Green and Campbell 1984; Nussbaum et al. 1983). The last three are endemic to the Pacific Northwest, and represent 25% of the Island's amphibian fauna. The island fauna also includes three introduced species, the: bullfrog (Rana catesbeiana), green frog (Rana clamitans), and northern leopard frog (Rana pipiens) (Nussbaum et al.1983; Orchard pers. com .). These species are at the northern limit of their range in Canada. With the reduced diversity of northern latitudes, one might expect them to occur in greater abundances than in the remainder of the Pacific NorthWest, particularly if they are 3  more hardy and able to withstand greater microclimatic fluctuations (Welsh 1990). Conversely, individuals at the northern edge of their range may be particularly sensitive to climate fluctuations, such that suitable microhabitats could be more limiting for them than for individuals near the centre of their ranges. In this thesis, I examine the status and distribution of terrestrial amphibians in natural Douglas fir/western hemlock forests on south-central Vancouver Island and in young and mature stands which have a history of clearcutting. My objectives are to: (1) compare diversity and relative abundance in old growth and managed stands; (2) determine with which habitat features amphibians are strongly associated, and contrast these features in old growth forests and managed stands; and (3) consider potential effects of logging operations on amphibians communities to recommend procedures for managing amphibian species in northern coastal forests.  4  CHAPTER II THE RELATION OF STAND AGE AND HISTORY TO THE ABUNDANCE OF TERRESTRIAL AMPHIBIANS  INTRODUCTION Salamanders have an important role in food webs in forest ecosystems. Firstly, their small median size (3 g as adults as opposed to 20 g for birds and small mammals) enables them to exploit small prey (Feder 1983), and their slow metabolism enables them to use prey with low food value. They convert these items into biomass that is available to birds and small mammals (Pough 1983). Secondly, ectotherms have a conversion efficiency of 40-98%, as compared with only 0.5-1.22% for endotherms (Feder 1983; Pough 1983). Therefore, they can subsist on little food and use large fractions of their energy intake for growth and reproduction (Feder 1983). They are also prolific and their populations can reach high densities (Burton and Likens 1975), which accounts for significant biomass and secondary productivity. The use of skin as a respiratory organ renders terrestrial amphibians vulnerable to dessication because it has little resistance to evaporative water loss. Dehydration (Spotila 1972) accelerates the metabolic rate and decreases assimilation efficiency, and can lead to severe negative energy budgets (Maiorana 1978; Spotila 1972; Jaeger 1971). Thus, water loss governs the salamanders' exploitation of the habitat, and affects their foraging activities. Coastal old-growth forests are moist environments. They have a very large crown surface and the large trees occupy an extensive volume. They are effective at gleaning moisture from fog and clouds, and reducing soil evaporation rates (Franklin 1988). Under these conditions, terrestrial amphibian populations do not usually experience moistureinduced limits to growth and reproduction, except in open habitats, in the driest months, or after extended periods of activity during courtship aggression or egg defense (Feder 1983). Sustained adverse conditions should limit their populations, and even cause them to decline 5  or disappear. Stands developing after a clearcut harvest, for example, have reduced canopy cover, increased insolation, higher ground temperatures, and higher evaporative water loss (Geiger 1971). These conditions may restrict amphibians to a narrow window of activity in the wet spring and fall thereby limiting reproduction, as they remain in underground burrows when conditions are too dry. Studies of the effects of clearcut logging on terrestrial amphibians in the U.S. Pacific Northwest have shown that amphibian abundance is reduced in post-harvest stands (Corn and Bury 1991; Bury and Corn 1988; Raphael 1988; Pough et al. 1987). I shall compare terrestrial amphibian numbers in old growth and post-harvest second growth on central Vancouver Island, and investigate their microhabitat use within both environments. Species in Canada are nearer the northern limits of their ranges. They may be more sensitive to the climatic changes accompanying logging, than species near the centre of their ranges.  STUDY AREA The study area was in the Coastal Western Hemlock Biogeoclimatic Zone near Port Alberni (49.3'N latitude and 125.3'W longitude), central Vancouver Island. This region receives more than 250 cm of precipitation annually (mainly in April, May, October and November), has a mean summer temperature of less than 16° C, and a mean of 180 frostfree days per year (Green et al. 1988). The study area included two distinct biogeoclimatic subzones: the Windward Island Mountain (WIM) (Western Vancouver Island; Coast and Mountains Ecoprovince) and the Leeward Island Mountain (LIM) (Eastern Vancouver Island; Georgian Depression Ecoprovince) subzones (Green et al. 1988). The LIM has drier soils or lower soil water retention capacity. The study was conducted in a 1925 km area, varying in elevation from 150 to 550 m above sea level.  6  SITE SELECTION AND HABITAT DESCRIPTION I chose low to mid-slope, moist, nutrient rich, south-facing or flat sites. Some stands, however, were relatively dry. The physical site characteristics are summarized in Table 2.1. I sampled mixed Douglas-fir (Pseudotsuga menziesii)/western hemlock ("ama heterophylla) stands. Western redcedar (Thuja plicata) was also present, as a co-dominant. Age classes sampled in 1991 included: old-growth (undisturbed) forests and 54 to 72year-old (mature), managed second growth. In 1992, surveys were also conducted in 17 to 18-year-old (young) managed stands, and some work was carried out in 5-year-old clearcuts. All sites were a minimum of 150 m from lakes and rivers. One old-growth (Anderson), one mature (Mactush) and one younger (Summit) second-growth plot were dissected by permanent streams; study plots were 75 m from these streams. Study plots were at least 30 m from temporary creeks. Stream surveys were made within 10 m of the 3 permanent creeks, and within 10 m of temporary streams in one additional replicate of each age class: Nahmint old growth, Kanyon (54 years) and Cous (18 years). Large old-growth sites were selected to ensure that terrestrial amphibians were surveyed under the most natural conditions, but one old-growth site is a 90-ha island, surrounded by young, post-harvest seral stages. Managed stands were not contiguous with large old-growth tracts, except for one mature and one young managed stand. The sites' landscape characteristics are summarized in Table 2.2. The mature stands were burned within 5 years of harvesting, and had regenerated naturally. Douglas-fir was the only commercial species at the time of harvest, leaving large volumes of felled wood (hemlock and cedar) on site. The young stands were burned and planted, and subsequently mechanically spaced to remove alder and unwanted conifer stems.  EXPERIMENTAL DESIGN Three replicates were selected of each forest age class, and studies were carried out in three 2 ha plots (400 x 50 m), within each replicate. Due to time constraints, clearcut surveys were limited to 1 study plot per replicate. 7  Table 2.1. Physical characteristics of each study site Site Nahmint River Anderson Creek Cathedral Grove Kanyon Creek Mactush Creek Sproat Lake Lake Main Summit Cous Creek Upnah Canal  Age (years) 380+ 330+ 500+ 54 60 72 17 18 18  Aspect  5 5  South South Flat South Flat South Flat Flat Flat  Average Slope 29° 12° 4° 16° 6° 13° 0° 0° 0°  Nutrient Level rich rich rich rich med poor-med very rich rich poor  Moisture Regime very moist moist moist moist moist fresh very moist moist dry  South Flat  10° 0°  rich med  moist fresh  Tree Composition' Hw/Fd Fd/Hw Hw/Fd/Cw Hw/Fd/Cw Fd/Hw/Cw Hw/Fd Hw/Ba/Cw Fd/Hw/Cw Fd/Hw/Cw Hw/Fd Fd/Hw/Cw  a Fd=Douglas-fir (Pseudotsuga menziesii); Hw=western hemlock (Tsuga heterophylla); Cw= western redcedar (Thula plicata); Ba=balsam fir (Abies amabilis) written in, in order of dominance  Table 2.2: Landscape characteristics of each study site Stand Age^Site^Size (ha)^Contiguous Old-growth Old-growth^Nahmint River^>5,000^present Anderson Creek^>800^present Cathedral Grove^90^none 54-72 years Kanyon Creek^>1000^a few 20-200 ha patches within; mature^ scrubby, dry, 400-600 m strip on adjacent ridge top managed^Mactush River^131^300m wide corridor across class 3 stream; corridor links up with a 300 ha patch Sproat Lake^>2000^several 5-20 ha patches; some patches across class 1 river 15-20 years Lake Main^150^500-1000 ha young managed^Cous Creek^107^none Summit^50^500m strip midslope steep south-facing dry rocky ridge top 5-year^Upnah^58^large tract; some intermittent 30-50 ha clearcuts managed Canal^>500^none  METHODS Amphibian Surveys:  From mid May to mid June of 1991, area-constrained (quadrat) searches were accomplished by randomly placing ten 3x3 m quadrats within each study plot. The total area sampled was 270 m 2 per replicate. The 3x3 m quadrats were divided into a series of 1x1 m squares. Each square in which salamanders were encountered was recorded. Wiegert's method (see Krebs 1989: pp. 65-124) was employed to determine optimal quadrat size for assessing salamander abundance. This involved randomly combining squares into 1 x1, 1 x2, 1x3, 2x2, 2x3, and 3x3 m quadrats and calculating the mean and variance of the number of salamanders encountered. The relative minimum variance (variance of quadrat size x /minimum variance) was multiplied by the relative minimum set up costs (time to sample quadrat size x /minimum time). The minimal product of the two revealed an optimal quadrat size of 1 x2 m for Vancouver Island coastal Douglas-fir. The optimum quadrat size was verified using the TTLQV method described by Ludwig and Reynolds (1988). I therefore used 45 1x2 m quadrats randomly placed within each study site (270 m 2 per replicate), from mid April to late May in 1992 . In 1992, stream surveys were also conducted, by means of quadrat searches, to test for the effects of proximity to water. Two people searched for animals, while a third handled the captures and recorded the data. Searchers looked under all cover objects (logs, bark and rocks) and vegetation (ferns and moss), and probed all crevices to capture animals within reach. Captured salamanders were identified, sexed, and weighed with a digital OHAUS port-o-gram scale (to +/-0.05 g). Total length and snout-vent length were also measured. A body condition index was calculated from the observed weight divided by the expected weight (derived from a regression equation of weight against snout-vent length), as described by O'Donoghue (1991). I recorded the cover object and position occupied by the animal (in, on, under or near). In 1991, I estimated area covered by logs and by different forest strata, using visual estimation charts (Luttmerding et al. 1990).  1 0  The difference in species diversity between old-growth and mature managed stands was tested using a contingency table (Zar 1984). The relative abundances of P. vehiculum between forest age classes, years, and forest and stream environments were tested using the Kruskal-Wallis analysis of variance, or the Mann-Whitney U test when only two samples were involved (Zar 1984). Non-parametric tests were used as the frequency of amphibian encounters within quadrats was non-normal. Differences in condition between stand ages, salamander sizes, seasons or years were tested using t-tests or analysis of variance (Anova) depending on the number of variables (Zar 1984). To test whether or not the distribution of amphibians was correlated with forest strata and volume of downed wood, a principle component analysis was performed on these 4 variables. A Pearson correlation was then done on the presence/absence of salamanders and the principle components explaining > 20% of the variation in forest structure, thereby reducing the number of variables analyzed. Associations between levels of soil moisture and salamander abundance were determined using the Kendall coefficient of concordance (Conover 1971).  Soil Moisture Samples: In 1991, a soil sample was randomly taken from each of the thirty 3x3 m quadrats within each replicate. Samples of soil (32 cm 3 ) were placed in LT40 tin sample boxes. In 1992, the number of soil samples was doubled to decrease the variance, and randomly taken from the forest floor. Samples were weighed immediately (± .05 g). They were then ovendried for 48 hours at 105° C, and re-weighed. The difference between wet and dry weights gives a measure of soil moisture (grams of water /32 cm 3 ). Differences in soil moisture between forest age classes and seasons were tested using a multifactorial Anova (Zar 1984).  Habitat Measurements: The habitat of old-growth forests and managed stands was described by setting up twelve 10x10 m vegetation plots in each replicate (4 per study plot) and recording the 11  percent cover of each forest stratum (canopy, understory and herb/moss layer) and of downed logs. Soil depth was also recorded.  RESULTS Diversity and Abundance of Terrestrial Amphibians:  In 1991 and 1992, I captured 571 individual amphibians of 6 species. Species richness did not differ significantly between old growth forests and managed stands (Contingency Table: X 2 =6.9, df=5, p<0.25). All species except Plethodon vehiculum were scarce (five Ensatina eschscholtzii, one Ambystoma gracile and one Rana aurora in old growth; two Anneides ferreus and one E. eschscholtzit in second growth). The abundance of E. vehiculum was six times greater in old-growth forests than in managed stands in 1991 (Mann-Whitney Test: U=81, df=1, p<0.001) and three times greater in 1992 (Kruskal-Wallis test: H=14.646, df=2, p=0.001, Figure 2.1). The quadrat searches yielded more animals in 1991 than in 1992, in the overlapping period of sampling (late May to early June). However, there was a doubling in the number of animals in second growth in 1992 (Mann-Whitney Test: U=0, df=1, p<0.05), and 70% ± 9 (s.e.) of the encounters occurred in early May, a sampling period which was not covered in 1991. In old growth, 39% ± 9 (s.e.) of the amphibians were caught in this time period. One E. vehiculum was found in the Upnah clearcut and none were found in the Canal clearcut.  Large E. vehiculum were more numerous than small and medium-sized ones in old growth, in both 1991 and 1992 (Goodness of Fit Test: X 2 =32.88, df=2, p<0.001). The same held true for Kanyon second growth (Goodness of Fit: X 2 =10.33, df=2, p<0.002), but not for the other second-growth replicates (Goodness of Fit: X 2 =1.55, df=2, p<0.1). In the mature managed stands, there was no significant difference in the abundance of large and medium-sized salamanders (Goodness of Fit Test: X 2 = 3.71, df=2, p<0.25).  E. vehiculum individuals were in significantly better condition in 1991 than in 1992 (t-test: t=-3.18, df=121, p<0.002, Table 2.3). Large salamanders were in slightly  12  200  100  0  Old growth  Mature stand  Young stand  Figure 2.1. The average density of amphibians encountered in old growth and mature and young managed second growth, in the springs of 1991 and 1992  13  Table 2.3. The conditions of P. vehiculum in Douglas-fir, old-growth forests and managed stands, in the springs of 1991 and 1992 Size (cm)  Year  Stand Age 1991  1992  Large (SVL > 3.5)  Old-growth Mature Managed  1.02 ± .02 (n=55) 1.12 ± .05 (n=13)  0.97 ± .04 (n=72) 1.09 ± .2 (n=5)  Medium (SVL = 2.6-3.5)  Old-growth Mature Managed  1.10 ± .08 (n=20) 1.20 ± .29 (n=3)  0.93 ± .05 (n=25) 1.10 ± .09 (n=12)  Small (SVL < 2.6)  Old-growth Mature Managed  1.06 ± .05 (n=8) 0.93 ± .13 (n=4)  0.80 ± .03 (n=17) 0.92 ± .05 (n=6)  a  Condition = an index calculated as the observed/expected ratio, derived from the relation between snout-vent length (mm) and weight  14  better condition than small salamanders (Anova Test: F=3.01, df=2, p<0.05, Table 2.3), and juveniles were in slightly better condition in managed stands than in old growth (Anova: F=3.12, df=2, p<0.046, Table 2.3). Amphibian numbers declined in late spring, presumably because animals burrowed into deeper moist areas (Figure 2.2-1991). Condition decreased in late spring, for both large (t-test: t=-19.47, df=125, p<0.001) and small salamanders (t-test: t=-9.482, df=33, p<0.001).  The Effect of Moisture: In 1991, salamander distribution was associated with soil moisture, in old growth (Kendall Tau: T=.3, df=70, p<.002), but not in second growth (J=Kendall Tau: T.-.1, df=78, p=.126) No moisture samples were taken within sampled quadrats in the drier spring of 1992. May 1992 received only 6.4 mm of rain, compared with 55.8 mm in 1991. Precipitation may thus have limited above-ground activity of terrestrial amphibians in 1992, but not in 1991 (Figure 2.2). Soil moisture declined from early to late spring in 1991 and 1992 (t-test: t=74.28, df=4, p<0.001). Salamander abundance peaked in the moist conditions of early spring (Figure 2.2-1992) and declined by mid June (Figure 2.2-1991). The abundance of small salamanders (yearlings) rose, then declined in late May (Figure 2.3). In the dry year of 1992, old-growth soils were slightly more moist in June than soils of mature stands, and much more moist than 20-year stands throughout the spring (Anova: F=12.37, df=4, p<0.001, Figure 2.4). The difference in moisture between oldgrowth and managed stands was not significant in the wetter year of 1991 (t-test: t=.734, df=1, p=0.46). The variation in  a. vehiculum densities in the young managed stands  reflects their nutrient/moisture regimes (Table 2.1). Lake was a very moist rich site, and had the largest abundance. Cous was very dry and supported no salamanders. The other drier sites, Cathedral Grove (old growth) and Sproat Lake (mature, managed) also  15  120  1991'  100  o ppt (mm) O No. of Sal.  E  E 80 ..._., c0 -47_, 60 CO  120 100 80  8:  20  20 0  0  MI  1992'  100  E  2 .  c  a •_ c.) a)^40  120  z  o ppt (mm) O No. of Sal.  80  120 100 80 60 40  20  20  0  0 April  ^  May  ^  June  ^  July  Figure 2.2. Terrestrial amphibian abundance and precipitation levels in the spring and summer of 1991 and 1992.  16  Figure 2.3. The proportion of small, medium and large spring, 1991 and 1992  17  a. vehiculum in early and late  0.7 ^  ii,1411991  0.6 0.5 0.4 0.3 0.2 Old-growth  0.1 -  Mature managed stand • I^ I Late April^Ealry May^Late May^Early June^Mid June^Late June  0.0  0.7 0.6 0.5 0.4 0.3 0.2 -  --Ø-- Old-growth —40— Mature managed stand  0.1 -  --a-- Young managed stand  0.0 ^ Late April^Early May^Late May^Early June  ^  • Mid June^Late June  Figure 2.4. Soil moisture in old-growth and managed stands in 1991-92  18  supported the lowest number of individuals within their age groups (Kruskal-Wallis Test: H=7.369, df=2, p=0.025). vehiculum makes use of a wide variety of cover objects, including ferns, moss, rocks and logs (Figure 2.5). Thirty percent of the individuals were associated with logs, and log use was constant throughout the season. The base of ferns also remained important throughout the season, particularly in late June. There was heterogeneity in the use of cover objects due to seasonal changes in use (Goodness of Fit: X 2 =71.96, df=20, p<0.001). Moss use decreased and bark and fern use increased (Figure 2.5). Moss was used primarily by juveniles (15 to 20% in 1991 and 1992). Terrestrial amphibian concentrations were markedly (4-8 times) higher along streams in both age classes of managed stands (Mann-Whitney Test: U=0, df=1, p=0.034), but not in old growth (Mann-Whitney Test: U=3, df=1, p=0.439, Figure 2.6). The young managed stand of Cous Creek had no salamanders at all. Salamander numbers along streams in managed stands were almost identical to those found in old-growth.  Habitat Attributes: There was less understory and ground cover in mature managed stands than in old growth, and the young stands had no canopy and more ground vegetation (Table 2.4). A principle component analysis on the 1991 3x3 m quadrat surveys revealed that once the variation in canopy and ground cover was removed, salamander distribution was correlated with understory and logs. This correlation was weak in old growth (Kendall Tau: T=0.08, n=90, df=1, p<0.008), and stronger in mature managed second growth (Kendall Tau: T=0.24, n=90, df=1, p<0.001).  Landscape Characteristics: The abundance of salamanders was greatest in the mature second growth adjacent to large tracts of old growth (Mactush Creek) and lowest in the mature stand which was  19  Figure 2.5. Seasonal use of cover objects in 1991 and 1992  20  10 Streamside Habitat la Terrestrial Habitat  8  6  4  2  0 Old growth  ^  Mature  ^  Young  Stand Age  Figure 2.6. The abundance of salamanders near (<10 m) and away (>30 m) from streams, in old growth and managed stands, in May 1992  21  Table 2.4. Average differences in vegetation structure and soil depth between old-growth forests, and young and mature managed stands* Habitat Attribute  Old-growth Forests  Percent Cover (%) 60-year Stands  20-year Stands  Canopy (Dominant Canopy)** Understory Ground Vegetation  41.6 6.4 24.9 68.7  ± 3.5 ± 1.2 ± 3.2 ± 4.2  41.4 ± 1.9 0 16.3 ± 2.3 43.3 ± 6.2  0 0 29.1 ± 2.0 30.3 ± 2.9  Soil Depth (cm)  66.0 ± 5.4  38.6 ± 3.6  31.7 ± 2.5  * Averages taken from 36 plots within each forest age class (12 per replicate) ** A dominant canopy is composed of trees that are taller than the main canopy  22  isolated from large patches of old growth (Sproat Lake). Only 17% of the Sproat Lake captures were mature adults, compared with approximately 40% adults in the other mature, managed stands contiguous with old-growth (Table 2.1). The same pattern is evident in young stands. The Lake Main study site (age 17 years) had salamander numbers equivalent to those found in mature managed stands (52 individual/ha). This site is contiguous with 500-1000 ha of old growth. Summit had an intermediate salamander density (22 individuals/ha) relative to the other young stands. A 500-m strip of midslope old-growth represented the only potential source of colonizers. Cous (age 18 years), a >500 ha cutblock belonging to a continuous stretch of harvested forest land, and abutting Alberni Inlet, had little colonizing potential, and no amphibians. Similarly, the single salamander encountered in clearcuts was in a site contiguous with 5,000 ha of old growth (Table 2.2).  DISCUSSION The Status of Terrestrial Amphibians on Vancouver Island : Amphibian faunal diversity and abundance is lower on Vancouver Island than elsewhere in the Pacific Northwest, excluding Alaska. For example, E. vehiculum densities in old growth, in my study area, were only 25% of densities observed in Californian oldgrowth redwoods (Corn and Bury 1990). This may be due to the cooler northern climate. There was a three to six fold difference in amphibian abundance between managed stands and old growth, with the lowest numbers in the youngest managed stands (Figure 2.1). These findings agree with other surveys conducted in the Pacific Northwest (Corn and Bury 1991; Raphael 1988; Welsh and Lind 1988; Pough et al. 1987). For example, Corn and Bury (1991) found 1.5 times more E. vehiculum in mesic old-growth forests of the Oregon Coast Range using pitfalls; the lowest abundances occurred in young managed stands. Differences in abundance between old growth and managed stands were two to four times greater on Vancouver Island than in the Oregon Coast Range. This greater disparity may be due to differences in climate, or because my study took place in flat areas and on slopes with 23  a southern aspect (Table 2.1). .E. vehiculum is known to be more abundant where the aspect is to the west or northwest (Corn and Bury 1991). As differences between old growth forests and managed stands may be less extreme on cooler north-facing slopes, future research on the effect of aspect on the abundance of amphibians would be worthwhile. Furthermore, the history of a stand must be considered, as areas which are not burned following harvest may have a smaller effect on salamander populations. Low terrestrial amphibian densities in managed stands indicate poor environmental conditions. These stands may offer less protection against evaporative water loss, which is discussed below.  The Importance of Moisture:  Terrestrial amphibians have a low tolerance for hot and dry conditions. Low soil moisture content, ambient temperature and the drying power of the air, cause dehydration (Spotila 1972; Heatwole 1962a; Hutchinson 1961). This can render hunting areas inaccessible to amphibians unless conditions are moist and wind-free (Feder 1983; Maiorana 1978; Spotila 1972; Jaeger 1971). In the spring of 1992, salamander numbers above ground were drastically reduced when precipitation levels approached zero (Figure 2.2). In the relatively moist year of 1991, amphibians were more numerous, and less affected by rainfall. This pattern confirms the important role moisture plays in salamander activity. Seasonal patterns in use of cover objects may also reflect the importance of moisture. For example, the moss Hylocomium splendens represents a good dispersal avenue. It adheres loosely to substrates, thereby serving as a crawl space and providing salamanders with constant cover from predators and adverse weather conditions. This moss dries up quickly in the spring (pers. obs.) and its use becomes restricted (Figure 2.5). Logs, the most heavily used cover object, are an important source of moisture throughout the season (Figure 2.5; Feder 1983). The drier sites in the windward maritime biogeoclimatic subzone (Green et al.1988), Cathedral Grove old growth, Sproat Lake mature second growth and Cous Creek 24  young second growth, supported the lowest number of amphibians within their respective age classes. In recent clearcuts there were virtually no salamanders. These findings support the idea that the moisture regime of a forest or stand affects abundance of individuals. It is noteworthy that Cathedral Grove old growth is flat, with very little talus, which may also contribute to the low numbers of E. vehiculum in this area. Cathedral Grove has a history of flooding and the silt and clay-dominated soils may limit burrowing activity. The soils of managed stands were less moist than those of old-growth forests (Figure 2.4), as suggested by Franklin (1988) and Geiger (1971). These workers proposed that increased wind and temperatures, and therefore evaporation, result from having smaller tree sizes and a less developed canopy. Additional support for the importance of moisture is that the concentration of salamanders within 10 m of a stream was much greater in second growth than in old growth (Figure 2.6). Indeed, there were four to six times more salamanders along the streambanks of managed stands, resulting in abundances comparable to those found at random in old growth. Body size in amphibians is directly related to re/dehydration rates (Spotila 1972; Ray 1958), and to the ability to withstand food deprivation. Maiorana (1978) found that California slender salamander (Batrachoseps attenuates) juveniles survived extended droughts less well than adults. The low abundance of small salamanders observed in June (Figure 2.3) and the small to large salamander ratio support this finding. The poorer condition of small salamanders in the dry spring of 1992 (Table 2.3) may be additional evidence that small salamanders are more vulnerable to dessication. However, animals in poorer condition may be more active than healthy animals in late spring, in a last effort to build up fat reserves before the onset of summer. That managed stands are generally drier (Figure 2.4), has implications for population sustainability in salamanders. Without a source of recolonizers, a very low juvenile survival rate can render populations non-viable (Pulliam 1988). Terrestrial amphibian eggs are particularly vulnerable to dehydration in areas permanently exposed to sun and surface winds, such as clearcuts. This could drastically 25  reduce numbers prior to stand re-establishment. The low numbers of individuals I observed in both young and mature second growth, may therefore be a result of past and present microclimatic conditions.  Habitat Attributes: Plant cover indirectly affects the local distribution of amphibians, by altering the moisture and temperature of cover objects (Pough et al. 1987; Heatwole 1962b). For example, shaded logs can be three and half to five times more moist and up to five times cooler than exposed ones (Heatwole 1962b). Managed conifer stands offer less shelter because the ground vegetation and shrub component are reduced (Hall et al. 1985; Table 2.4). These stands also have smaller wood volumes (Maser and Trappe 1984; Triska and Cromack 1979; Figure 3.0), yet downed wood is an important cover object for salamanders (Figure 2.5, Chapter III). Consequently, a relationship between salamander distribution, and the amount of understory and logs might be expected. I did in fact find a weak positive association. Corn and Bury (1991) found a strong association of plethodontids with downed wood volume. Thus vegetation structure and wood availability may also help to explain why mature stands harbour low salamander numbers.  Landscape Considerations: Heavy cutting coupled with intensive replanting has increased the area (ha) of young stands (Harris 1984), and rapidly reduced low to mid-elevation old-growth forests (Hammond 1991; Norse 1990; Robinson 1988). Indeed, it was difficult to find sites on Vancouver Island in which to conduct this study, as only 9% of watersheds remain unlogged (Hammond 1991). There is no readily available information on the landscape-level status of low-lying old growth, which hampers researchers' attempts at providing wildlife management recommendations. If it is assumed that 9% of low-lying old growth remained, and if young and mature managed stands now host 17 and 27% of old growth terrestrial amphibians densities respectively (Figure 2.1), salamander densities on Vancouver Island 26  are now 24 to 33% of what they once were. If amphibian populations in managed stands are sensitive to the presence of neighbouring old growth, as this study suggests, this estimated loss is conservative. Since only 5% of Vancouver Island's forested land is currently protected in parks and reserves, the maintenance of terrestrial amphibian populations may be critically dependent on how we manage our forests in the future. Clearcutting and its aftermath can reduce amphibian populations severely, and numbers can be sparse even 54 to 72 years after logging (Figure 2.2). In addition, managed stands may be "sinks" with very low local reproductive success and poor survival. The short rotations periods (60 to 120 years) currently planned may prevent the reestablishment of forests with sufficient moisture and cover, and immigration from old growth may be necessary to sustain terrestrial amphibian populations in post-harvest second growth. The decreased salamander densities observed in young and mature stands with increasing isolation from old growth, confirms the role of old growth as a source of colonists. For example, Sproat Lake, a roughly 2,000-ha managed stand, is isolated from old growth, except for a dozen 5-20 ha patches (Table 2.2). It is surrounded on all sides by wide (Stamp and Somass) rivers and a large (Sproat) lake. There are some old growth patches on the other side of these rivers (e.g., Stamp Falls Provincial Park), from which salamanders could raft. This stand had the lowest abundance of salamanders of the three managed stands examined here. The higher proportion of adults found in stands contiguous with old growth suggests that recolonization by adults is important. Thus, the position of study sites in relation to old growth may be critical and should be considered in future studies on the abundance of animals in logged and unlogged stands. In conclusion, clearcutting decreases numbers of terrestrial amphibians, probably by reducing the availability of moist microhabitats and by limiting foraging and reproductive opportunities. Seventy percent of the salamanders found in second growth were active in the warm and rainy part of early May (compared to 40% in old growth),  suggesting a narrower window of activity, which in turn reflects poorer habitat quality. In fact, my study suggests that the heavily managed stands of south-central Vancouver Island 27  now contain roughly 30% of the densities present in the old growth they have replaced. The continued removal of old growth will likely lower abundances of  a. vehiculum even further,  and may endanger the rarer species. As salamanders can represent an important part of forest food webs, landscape level management for their continued survival could be crucial to ecosystem integrity. The tolerance and resiliency of populations to rapid environmental deterioration or changes may already have been compromised in Canada, where species are challenged by the harsh conditions at their northern range limits.  MANAGEMENT RECOMMENDATIONS: Given that amphibian populations are already significantly reduced on Vancouver Island, some management recommendations are appropriate.  1. Shrubs and ground vegetation should be maintained in clumps throughout a cutblock, to provide shade. I recommend no burning and no treatment of broadleaf vegetation with herbicides, particularly along streams, as amphibians are very sensitive to chemical pollutants (Pawar et al. 1983; Marchal-Segault and Ramade 1981; Johnson 1980).  2. As the canopy is removed, ground temperatures rise and waterloss inevitably increases. I propose enhanced streamside protection, where high densities occur in managed stands. A buffer of no less than 20-30 m should be left untouched on either side of all streams. In the past, there has been no concern for small (class III and IV) streams (Hammond 1991). Buffer strips were maintained only along commercial fish-bearing (class I and II) streams (B.C. Ministry of Forests and B.C. Ministry of Environment 1992). Small streams are now also considered, but their management is geared towards channel and slope stabilization rather than to wildlife (B.C. Ministry of Forests et al. 1992). For small permanent streams which influence downstream fish-bearing waters, and steep streams which have the ability to transport debris, understory vegetation and large wood are retained to maintain 28  streambank integrity (B.C. Ministry of Forests et al. 1992; B.C. Ministry of Forests and B.C. Ministry of Environment 1992). Small low gradient streams, which do not feed into fish spawning grounds, are ignored. Cross-stream yarding of small low gradient streams should be minimized. Their banks represent a potential stronghold for terrestrial amphibians in the face of large scale disturbances like clearcutting.  3. Large old growth tracts adjacent to second growth should be protected because they can serve as species pools for the recolonization of post-harvest stands. If a circular reserve smaller than 100 ha contains no true forest interior (Wilcove et al. 1986), reserves should contain at least several hundred hectares of uninterrupted forest to provide amphibians with large microclimatically stable areas. Large reserves also increase the likelihood of maintaining viable amphibian populations in the region, by reducing their susceptability to local demographic, genetic or environmental stochastic extinction (Gilpin and Soule 1986). Moreover, large reserves contain more species than small ones (Wilcove et al. 1986). Where it is impossible to maintain large areas of old growth, smaller tracts could be interconnected with corridors, or maintained next to existing protected areas (Harris 1984). Recently, biodiversity guidelines have incorporated forest ecosystem networks (FENs) to maintain some connection between old-growth forests and other natural areas (B.C. Min. of Forests and B.C. Min. of Environment 1992). FENs include: wetlands, unmerchantable or inoperable forests, environmentally sensitive areas (e.g., forests with unstable slopes and fish and wildlife riparian corridors), mature second growth, and parks and reserves. Terrestrial salamanders avoid wet habitats (Bury and Corn 1988; Welsh and Lind 1988), and will not use exposed ridge tops or forests that exceed 1250 m in elevation (Nussbaum et al. 1983). For FENs to help maintain terrestrial amphibian populations, they must include relatively moist forested habitats (i.e., deer winter ranges, wildlife corridors, parks, unstable slopes or mature stands) to be suitable dispersal agents, and these must link old-growth forests.  29  CHAPTER III THE IMPORTANCE OF LOGS TO THE AMPHIBIAN FAUNA OF A COASTAL FOREST  INTRODUCTION Until recently, downed wood has been considered useless debris, and an impediment to reforestation, as it affects forest management by creating a fire hazard and interfering with tree planting (Bartels et al. 1985). It also provides more breeding sites for insects, such as bark beetles, which harm the trees. In addition, it may enhance local populations of lagomorphs and rodents, which reduce tree regeneration by feeding on seedlings and saplings when recent slash foliage is gone (T.P. Sullivan pers. com .). Thus, wood is normally removed by piling and burning, and low levels are maintained by spacing young stands and thinning mature ones. In addition, there is an increasing interest in using downed wood as an alternate source of energy (Bartels et al. 1985). Such issues may conflict with wildlife needs, and thus the removal of woody debris from managed stands should be carefully examined. Fallen logs play several important ecological roles. They are important in nutrient cycling (Means et al. 1992; Spies et al. 1988), as a feeding or breeding substrate for wildlife (Norse 1990; Harmon et al. 1986; Bartels et al. 1985), and as terrestrial and stream habitat diversifiers (Norse 1990; Franklin 1988; Triska and Cromack 1979). They accumulate randomly over time after natural disturbances such as windthrow, insect outbreaks and lightning strikes (Spies et al. 1988; Maser and Trappe 1984; Triska and Cromack 1979). Old-growth forests in the Pacific Northwest (300+ years) contain large volumes of downed wood (Triska and Cromack 1979). A single tree with a diameter at breast height of 100 cm can weigh 10 metric tons (Triska and Cromack 1979). The quality of logs is important, as different stages of wood decomposition attract different wildlife species or life stages (Norse 1990; Harmon et al. 1986; Maser 1990).  30  For example, a fresh fallen log may be a suitable substrate for a wood boring beetle, but tree seedlings can only become established once the wood has softened (Maser 1990). Terrestrial amphibians appear to be strongly associated with fallen logs, which afford a moist refuge. The mean density of plethodontid salamanders in wood can reach 744/m 3 /ha in old-growth forests of the U.S. (c.f. Corn and Bury 1990). In this chapter, I explore the strength of their association with downed wood. My specific objectives are to: 1) look at the abundance and distribution of terrestrial amphibians on various types of logs; and 2) assess wood volumes in various stages of decay, within old-growth forests and managed stands of south-central Vancouver Island. I expected lower volumes of wood and a poorer representation of wood quality in second growth, as a function of age and forestry practices, and a correspondingly low abundance of amphibians.  STUDY AREA, SITE SELECTION AND EXPERIMENTAL DESIGN The study took place in the Coastal Hemlock Biogeoclimatic Zone, near Port Alberni, Vancouver Island. Amphibian and wood surveys were conducted in 3 replicates of Douglas fir (Pseudotsuga jnenziesii/western hemlock (Lucia heterophylla), and of mature post-harvest second growth. Wood quantity and quality were also recorded in three young managed stands. Each replicate had 3 study plots of two hectares. See Chapter II for details.  METHODS Downed wood surveys: Wood volumes were assessed along three 50-m transect lines at 120° angles, starting from the sampling area's centre, as described by Trowbridge et al. (1987) and modified by Beese (1992). Two such inverted triangles were located systematically within each forest plot (6/replicate). The diameters of downed wood were measured along each sample line (see tallying rules in Van Wagner 1968), using 80-cm calipers or a standard calibrated measuring tape for very large pieces. Volume (m 3 /ha) was calculated using Van Wagner's (1968) formula: 31  V = ( 71 2 ED 2 ) / 8L where^D = piece diameter (cm) L = sample line length (m) The quality of wood (decay class, presence of bark, state of suspension) was recorded for each piece >6 cm in diameter. Logs were grouped within 3 decay categories: A(early), B(intermediate) and C (advanced). These categories are simplified from the five-class scale described by Maser and Trappe (1984). Class A logs are intact with more or less complete bark cover. Class B logs have decaying sapwood, and the bark is beginning to slough off. Class C logs have little to no bark, and are disintegrating. Analysis of variance (Anova) was used to assess differences in volume, amount of bark and proportion of suspended logs, in all three decay classes, between old growth and second growth. I used a goodness of fit test to determine whether or not the amount of bark corresponded to a particular stage of decay.  Amphibian surveys: Amphibians were surveyed in the spring of 1991 and 1992, in old growth and mature managed stands, using a technique of Corn and Bury (1990). I systematically searched 90 logs (30 per study plot) greater than 7 cm in diameter, in both replicates of mature managed stands and old-growth forests. Within each plot, I selected 10 logs in each of the 3 decay classes (A,B and C) described above. Each log was searched for a maximum of 20 person-minutes. A transect was set up through each study plot, 50 to 75 m from forest edges. Logs were selected by a systematic sampling scheme, along a designated path; that is, I searched every first class A, every second class B and every third class C log encountered. This approach enabled me to cover a large proportion of each study plot. When a log was selected, I determined the decay class, and measured the maximum width and length, the portion sampled (length) and the tree species when known. In 1992, I  32  recorded whether or not a log was suspended, and the presence/absence and location of bark. These factors may be critical to cover-seeking animals. A survey began with a search of the log's surface, including the bark in adjacent areas. This was followed by the removal of bark on the log, and the tearing of the decayed wood layer by layer, using crow bars and potato rakes. Where very large logs were concerned, I concentrated my efforts on a portion of the log (salamander densities are based on the volume of wood actually searched), and surveyed length was recorded. Salamanders were identified, sexed, measured and weighed. Logs were rolled to their original positions when possible, and large slabs of bark were replaced to reduce habitat damage. The density of log-associated salamanders was calculated by dividing the number of salamanders by the volume of wood surveyed (Dcr 2 x surveyed length, where r=radius) and multiplying the product by the volume of wood per hectare calculated above. Differences in the abundance of salamanders on logs of the 3 decay classes was tested using the Kruskal-Wallis test. The Goodness of Fit test was used to see if positions adopted by salamanders differed with stage of decomposition, and stand age. Differences in the frequency of occurrence of salamanders on various log sizes in second growth and old growth was tested with using an Anova.  RESULTS Wood Quality and Quantity in Forests and Stands: I surveyed a total of 2,549 logs. Volumes of wood increase with forest age, in all stages of decay (Anova: F=20.59, df=2, p<.001): old growth had nearly twice as much wood as mature stands, and five times more wood than young stands. The volume of intermediately decayed wood was especially low in managed stands (Figure 3.1). There was a large representation of big logs in old growth (Table 3.1). The younger the stand, the smaller the proportion of large logs (Anova: F=33.79, df=8, p<.001). The smaller the log was, the more likely it was to be suspended in managed stands (Anova: F=45.31, df=23, p<.001, Table 3.1), but not in old growth (Anova: F=.8, df=11, p=.433). 33  Figure 3.1. The volume of wood in various stages of decay in old-growth forests, and in young and mature managed second-growth stands  34  Table 3.1. The availability of soil-interface refuges' among logs of differing size classes, in old-growth forests and managed second-growth stands Sample Size  % Logs with % Occurrence % Suspended % Suspended w/ no soil interface sloughed bark  Stand Age  Log size  Old growth 300+ years  7-10 cm 10.1-20 cm 20.1-30 cm >30 cm  123 236 191 389  13 25 20 42  41 28 25 33  2 3 8 27  39 25 17 6  Second growth 54-72 years  7-10 cm 10.1-20 cm 20.1-30 cm >30 cm  154 303 229 197  18 31 25 26  65 39 21 0  0 1 1 4  65 38 20 0  Second growth 17-18 years  7-10 cm 10.1-20 cm 20.1-30 cm >30 cm  97 257 95 71  19 49 18 14  54 38 26 17  1 2 4 11  53 36 22 6  a Soil interface can be provided by grounded logs as well as by suspended logs with sloughed bark  However, there was no difference in the proportion of suspended logs between old growth and managed stands (Anova: F=.492, df=8, p=.634). About one third (35.3% ± 1.5) of all logs were suspended by at least 75% of their length, and the majority of these were fresh: 80% new logs, 50% intermediate logs, and 6% very decayed logs. Sloughed bark was most commonly associated with logs in the intermediate stages of decomposition (Goodness of Fit: G=114.85, df=2, p<.001), and was scarce in managed stands (Anova: F=61.423, df=8, p<.001)(Figure 3.2). Small logs rarely had sloughed bark (Anova: F=14.18, df=8, p=.001).  The Distribution of Amphibians on Downed Wood: In the springs of 1991 and 1992, I found amphibians associated with 29 % of logs (129 of 450) in old growth. Ninety four percent of the encounters were with E. vehiculum. The scarcer species included 4 E. eschscholtzii, 3 A. ferreus, 1 Rana aurora, 2 Ambystoma aracile and 2 Taricha aranulosa. In mature managed stands, 9.3% of logs (42 of 450) were occupied, 92% by  a. vehiculum. Four A. ferreus were also captured from the three  replicates. In 1991 the densities of salamanders associated with logs, were 217/m 3 /ha ± 55 in old growth and 86/m 3 /ha ± 26 in second growth. In 1992 density estimates were 198 ± 34 and 95 ± 35, respectively. There were significant differences in the log positions occupied by salamanders (Goodness of fit: G=188.095, df=4, p<.001). Amphibians relied primarily on the log or bark/soil interface for cover (Figure 3.3). In managed stands, most were found under logs (Goodness of Fit: G=36.5, df=1, p<.001). In old growth, E. vehiculum used logs in the three stages of decay differentially (Kruskal-Wallis: H=6.49, df=2, n=246, p=.039), making most use of logs in intermediate and advanced stages of decay(Figure 3.4). In the mature stands, there was no significant difference in the use of various stages of decay (Kruskal-Wallis: H=5.402, df=2, n=61, p=.067), and the use of all decay classes was low in early June (Figure 3.4).  36  Early  ^  Intermediate  ^  Advanced  Decay Category Figure 3.2. The total number of logs with sloughed bark encountered in old growth, and young and mature managed second growh  37  Figure 3.3. Log positions adopted by terrestrial amphibians in old growth and mature managed stands  38  Early April Early May^Late May  Early June  II Early Mature managed stand  0 Intermediate Advanced  Late April^Early May Late May^Early June  Figure 3.4. Seasonal use of logs in various stages of decay, in old growth and maturer managed stands, in 1991 and 1992  39  Large logs housed more species (up to 6) than small ones (1-3). In old growth, given an equal search volume for each log size category, the frequency of occurrence of salamanders differed with log size (Kruskal-Wallis: H=8.4, df=2, p=.039), large logs being used more often (Table 3.2). There were no significant differences in use of various log sizes in managed stands (Kruskal-Wallis: H=2.4, df=2, p=.49). The use of grounded and suspended logs differed in old growth (Kruskal-Wallis: H=9, df=1, p<.046),  a. vehiculum rarely occurring on suspended logs when these did not  provide a sloughed bark/soil interface. There was no significant difference in the use of suspended and grounded logs in mature stands (Kruskal-Wallis: H=8, df=1, p=.127). Aside from one A. ferreus encounter in a crack on a log, cedar was not utilized by terrestrial amphibians; nor were deciduous trees unless these were in contact with the ground. Douglas-fir was most commonly used by  a. vehiculum.  DISCUSSION The Importance of Wood Quality to Salamanders: In coastal Douglas-fir forests on Vancouver Island, 30% of P. vehiculum are associated with logs (see Chapter II), and this species was found in association with approximately 30% of logs. Corn and Bury (1991) obtained identical results in the Oregon Coast Range old growth, although an additional 30% of logs in Oregon were occupied by A.  ferreus and  a. eschscholtzii as compared with only 6% on Vancouver Island. This  observation illustrates the importance of downed wood for terrestrial amphibians. Indeed, it provides a stable moist refuge which is crucial to their survival (see Chapter II). Wood volumes in all stages of decomposition were larger in old growth (Bury and Corn 1988; Figure 3.1), as expected if volume accumulates over time (Maser and Trappe 1984; Triska and Cromack 1979). As greater quantities of wood increase the carrying capacity of many species (Norse 1990), this may contribute to the higher densities of amphibians associated with logs in old growth. Despite the lower volumes in second growth, and the three-fold reduction in the number of logs housing individuals (Figure 3.4), the 40  Table 3.2. The distribution of terrestrial amphibians on various log sizes in 1992, in old growth and mature post-harvest second growth Log Size Class  300+ years undisturbed  7-10 cm 10.1-20 cm 20.1-30 cm >30 cm  13 25 20 42 (n=940)  1 13 26 60 (n=99)  0.08 0.52 1.3 1.43  54-72 years managed  7-10 cm 10.1-20 cm 20.1-30 cm >30 cm  17 30 29 24 (n=974)  4 38 29 29 (n=21)  0.24 1.27 1 1.21  a  % Occurrence % Occupancy  Index of Use  Stand Age  Index = % occurence of a given log size class/% occupancy by salamanders  41  densities of salamanders associated with wood were on average only half that in old growth. This observation suggests that animals concentrated on fewer logs and that wood, may therefore be a limiting resource in managed stands. In fact, wood volumes were particularly low in the windward drier stands of Sproat Lake and Cathedral Grove (Figure 3.1), and these sites supported the lowest density of log-occupying salamanders (13 ± 7 in 1992) in their respective age class. Corn and Bury (1991) predicted that the clouded salamander, A. ferreus, will be rare in intensively managed stands where downed wood is reduced. This  may apply to E. vehiculum as well. Heatwole (1962b) found that salamanders sought shaded logs because they were more moist than exposed ones. Similarly, intermediately and very decayed logs were favourable for E. vehiculum (Figure 3.4) and E. eschscholtzii (Corn and Bury 1991; Bury and Corn 1988) probably because they were moist due to the increased microbial activity and water-holding capacity of the woody substrate as cell walls break down (Triska and Cromack 1979). Moreover, logs become more closely embedded in the soil as they decay, buffering them against fluctuating temperatures. Together these observations confirm the importance of wood quality to wildlife (Norse 1990; Harmon et al. 1986; Maser and Trappe 1984). Wood quality is likely to be important in managed stands as well, although my sample sizes were too small to detect such patterns. Wood in post-harvest stands consists of very decayed logs carried over from primeval forests, and new and small logs added during self-thinning (Maser and Trappe 1984). Consequently, intermediately decayed logs are relatively uncommon. They were over-represented in my mature stands because in the early part of the century, only Douglas-fir and high grades of other species were selected (MacMillan Bloedel engineer, pers. corn.). Thus large sound logs were left on site, and due to their long residence time they are in intermediate stages of decay in current mature stands. Large logs are no longer left on site, as can be seen in my 17 to 18 year-old sites (Figure 3.1). If we base amphibian density on the volume of the preferred intermediately decayed wood, these young stands only have one quarter of old growth densities. As small logs decay relatively quickly, 42  volumes will be further reduced in these stands in the future. Consequently, salamander densities may continue to decline, particularly if these stands are harvested at 60 to 120year intervals. Terrestrial amphibians use large logs disproportionately (Bury and Corn 1988; Gordon et al. 1987; Table 3.2), which offer more protection and a more stable environment (Maser and Trappe 1984). This pattern was not significant in managed stands, but sample sizes were small (42 salamanders in three replicates) and logging causes the distribution of logs to be non-random. Large logs are severely reduced by current forestry practices (Corn and Bury 1991; Spies et al. 1988), and are replaced by small logs (Table 3.1), which usually lack the critical log/soil interface because they are suspended, and are more susceptible to drying out. As seasonal fluctuations in water content increase as decay progresses (Triska and Cromack 1979), small and very decayed logs may be particularly unsuitable in dry seasons. Such seasonal trends may explain the low encounter rates for salamanders in managed stands in June (Figure 3.4). Sloughed bark is characteristic of intermediate decay. In natural forests, 35% of .E. vehiculum were encountered under bark on the ground (Figure 3.3), in agreement with the findings of Aubry and Hall (1988). Small logs have thin bark, which maintains the shape of the tree when it is shed. This is particularly true of deciduous species (pers. obs.). Thin curled bark does not provide a stable and moist micro-environment. Bark is scarce in managed stands (Figure 3.2) due to the low volumes of intermediately decayed wood and to the high proportion of small logs. This absence of bark further reduces the availability of moist refuges.  a. vehiculum were strongly associated with Douglas-fir, and were never present on cedar, possibly due to its high tannin concentrations (L. Jones pers. com .). Thus tree species composition affects a log's role to wildlife. So does the location of wood; because salamanders concentrated along streamsides in second growth (Figure 2.5), and logs are a primary cover object (Bury and Corn 1988; Figure 5.2), fallen logs near streams may serve as a stronghold for terrestrial amphibian survival. Finally, the distribution of wood 43  should also receive consideration. The more even its distribution on the forest floor (excluding streams), the greater should be its habitat value for wildlife (Maser 1990). When forests are logged by typical modern methods, all felled trees are brought to the roadside, to be sorted prior to loading operations, which results in roadside woodpiles, and large areas devoid of large fallen logs. In summary, my study has shown a reduced availability of moist log/soil interfact microhabitats in managed second growth (Aubry and Hall 1988, Figure 3.3). This lack of microhabitat may increase the mortality rate associated with extreme dehydration, particularly of the eggs of terrestrial species, which are deposited in logs (Peacock and Nussbaum 1973). Pough (1983) believed that the recovery of this soil interface microhabitat is a prerequisite for salamanders to re-populate a site, which is a major concern, if these stands are to be recolonized when low-lying old growth is scarce or absent. Size, state of decay, species, physical distribution and location of logs may all contribute to the abundance and distribution of terrestrial amphibians, and should be considered during management decisions. The U.S. Forest Service has investigated these properties of downed wood in old-growth forests and various seral stages of second growth, and is now developing guidelines for the conservation of wood following harvesting. Research on downed wood lags behind in Canada, and has focused primarily on volume and decomposition rate (Trofymo and Beese 1991; Prescott et al. 1989). This lack of knowledge is reflected in the recent biodiversity guidelines, which manage downed wood for quantity but not quality (B.C. Ministry of Forests et al. 1992).  MANAGEMENT RECOMMENDATIONS: 1. I suggest that log size, species, location and state of decay be managed by retaining suppressed and large trees and snags of the dominant canopy species on 10% of the area clearcut (c.f. Maser and Trappe 1984). Patches of snags and trees can be left on the outer perimeter of cutblocks, and within the centre of cable configurations. Such trees and snags will provide future recruitment to the forest floor and ensure a continual supply of logs in 44  all stages of decay. Where possible, single trees and snags with accompanying understory should also be retained to avoid too much clumping of future downed wood supplies.  2.  An even distribution of wood in an area can be attained by on-site selection of  economically feasible logs, by trained choker operators rather than by roadside yarder operators; undesired felled logs would be left distributed within the cutblock.  3. Downed wood should be managed primarily by retaining large felled trees and snags. When possible, these can be in the form of spiral grain and highly knotted trees. Their importance is recognized in the recent biodiversity guidelines (B.C. Min. of Forests and B.C. Min. of Environment 1992), but should be considered a priority. ^Large logs can accommodate more wildlife and are especially important for  .a. vehiculum. Their long  residence time on the forest floor also ensures a continual wood supply until the new stand is re-established.  4. Cutting cycles of 300 years should be incorporated within standard rotations to assure the presence of large logs at all times (Maser and Trappe1984).  5. Large reserves of old growth and moist forested corridors linking patches of old growth, should be set aside (see Chapter II), to allow continued recolonization of these marginal post-harvest second-growth habitats.  Downed wood, and the soil it protects, reduces the need for fertilizer and maintains a microorganism species pool (decomposers and nitrogen fixers) which enhances future productivity (Robinson 1988). Large decomposing logs also contain a lot of moisture and can reduce the risk of fire if left behind (Beese 1992). Thus downed wood management can benefit forest regeneration and enhance amphibian wildlife simultaneously, and should no longer be referred to as "debris". 45  CHAPTER IV  GENERAL CONCLUSION  The inhospitability of post-harvest stands to terrestrial amphibians: A look at terrestrial amphibian abundances and their key habitat requirements has revealed that post-harvest stands are marginal habitats for them. Densities in these stands are currently about one third of those in old growth and can be expected to decrease further, as isolation from old growth continues to increase. Factors contributing to the poor quality of these younger stands include increased evaporative water loss and reduced availability of microenvironments suitable for cover, maintenance of water-balance and successful reproduction. These factors are discussed below. Even-age management and short rotations produce small trees, a tight single-layered canopy, and sparse ground vegetation and shrub layers (Hall et al. 1985). All of these factors decrease microclimatic stability (Geiger 1971). In particular, changes in wind profile resulting from the reduction of secondary vegetation increase evaporative water loss (Franklin 1988). Soil compaction during harvesting also causes soils to loose their water retention capabilities (Robinson 1988). Ultimately, our forest practices may prevent the re-establishment of species such as terrestrial amphibians, whose successful establishment requires humid conditions, or persistent soil moisture. My findings that activity of amphibians is reduced when monthly precipitation levels approach zero (Figure 2.2), and that terrestrial amphibians are confined to streambanks in managed stands (Figure 2.6), support this suggestion. The importance of streambanks to terrestrial amphibians in managed stands should be investigated thoroughly. If streambanks are strongholds for the survival of salamanders following clearcut harvesting, this has major management implications. At present, small streams are often a source of lift and deflection for cable logging systems in coastal British Columbia, and all standing trees must be removed to allow cross-stream yarding to occur (MacMillan Bloedel engineer, pers. corn.). Moreover, small streams which influence commercial fish habitats downstream, are protected by the mere retention of large woody 'debris' and shrubby vegetation (Ministry of Forests and Ministry 46  of Environment 1992). Along larger streams with commercial fish, buffers are often too narrow (app. 5 m) to provide shelter against climatic fluctuations. Logs may be important in maintaining salamander populations in Douglas-fir forests (Aubry et al. 1988; Bury and Corn 1988), particularly when they are large (Table 3.2; Gordon et a. 1987) and decaying (Figure 3.4; Corn and Bury 1991). As the log or bark/soil interface is the most important moisture refuge for  a. v ehiculu m and E.  eschscholtzii (Figure 3.3; Aubry et al. 1988), logs in contact with the forest floor, and/or  with sloughed bark on the ground, are particularly important buffers against climatic fluctuations (Maser and Trappe 1984). In managed stands, these favourable wood characteristics are severely reduced (Table 3.1; Table 4.2), and volumes of downed and standing wood are substantially lower than in old growth (Figure 3.1; Bury and Corn 1988). Maintaining even moderate numbers of large logs in managed stands will require modifications of current harvesting and silvicultural practices (Spies et al. 1988; Harmon et al. 1986). Extensive short rotations (<100 years), in particular, may make it impossible to manage large wood for terrestrial amphibians.  Future Research:  Landscape considerations may provide additional insight on the status of terrestrial amphibians. The local diversity and abundance of amphibians in managed second growths may include dispersing 'surplus' animals and thus be strongly influenced by the size and proximity of old growth. Van Horne (1983) suggested that density estimates alone are misleading indicators of habitat quality, as abundance can be a function of continued immigration from another source. Similarly, Pulliam (1988) suggested that investigators could easily be misled about the habitat requirements of a species, because surplus individuals with no breeding opportunity may migrate to a poorer environment and breed in less suitable microhabitats. Populations in managed stands may not be viable without this movement of individuals from adjacent sources. Thus, survival, reproduction and migration of individuals should be considered in future assessments of habitat quality. Further studies 47  should investigate: 1) the survival of juveniles to reproductive maturity, and of adults (intensive demographic studies of salamanders are difficult due to their secretive nature); 2) whether or not migration from old growth is occurring, and to what extent; and 3) the effect of proximity of old growth on amphibian densities in managed stands, to determine with more certainty whether or not managed stands represent sink habitats. These studies are pressing, as much of the Vancouver Island's landscape is already dominated by managed stands. The elimination of the remaining old growth could lead to local amphibian extirpations. In conclusion, my study has shown that the growing public and scientific concern over the effects of clearcutting on wildlife is justified. Terrestrial amphibians are severely affected by the elimination of old growth, as are other groups of species (e.g., Eckert et al. 1991; Titterington et al. 1979). As terrestrial amphibians are sensitive and functionally important elements in forest ecosystems (McDiarmid 1991), they warrant effective management and conservation. In coastal British Columbia, where logging affects a large proportion of the landscape, there is an opportunity for management changes and improvements. If logging companies and provincial policy makers recognize this, they may be able to prevent the mistake of many European communities: the loss of any ability to manage for wildlife.  A Comparison of techniques: Similar numbers of salamanders were encountered in both the area-constrained searches (ACS-Chapter II) and log surveys (LGS-Chapter Ill). Both techniques uncovered differences in relative abundance between old growth and managed stands. However, each method revealed different information. LGS was a better technique than ACS for detecting the diversity of species in coastal Douglas-fir forests on Vancouver Island, although a time-constrained search (TCS) would have been equally effective at determining species diversity and would have involved less destructive sampling (Corn and Bury 1990). 48  ACS enabled me to gain some insight on microhabitat requirements. Small  a.  vehiculum rely on moss and ferns for thermal and predatory cover and adults are strongly  associated with logs, which may be a good egg-laying substrate. Although logs are an important source of cover, amphibians do not rely on them as heavily in managed stands. This difference could have important management implications. With ACS, I also observed differences in the proportion of adult and juvenile salamanders. These differences could lead to some interesting questions regarding the impact of logging on different age groups. As terrestrial amphibians do not have an aquatic larval stage, needs of all stages of their life histories must be considered in forest management. ACS enabled me to detect small female biases in old growth populations of  a.  vehiculum, which were not observed in second growth. This biased sex ratio may be  attributable to male territoriality at higher densities. Although the moist cool conditions of spring render food readily available and territoriality uneconomical (Ovaska 1988),  a.  vehiculum may switch to territorial behaviour when dry conditions restrict foraging  movements (Jaeger 1979). Ovaska's (1988) observation that  a. vehiculum is never  territorial is based on a study which was not temporally or spatially replicated in late spring and early summer, when conditions become dry enough to restrict amphibians. In conclusion, the ACS technique allowed me to estimate the density of terrestrial amphibians in old forests and managed stands; estimates that included both sexes, all age groups, and individuals in all parts of the forest ecosystem. It also gave insight on microhabitat needs and associations in natural and post-harvest managed second growth. LGS supplemented this by providing more detailed information on downed wood, which was the primary cover object in old-growth forests. However, it under-represented juveniles, which make extensive use of vegetation. An important factor to consider when designing a monitoring program for amphibians is the type of information desired: (1) presence-absence, species richness, (2) relative abundance, (3) absolute abundance, (4) macro/microhabitat associations or 49  (5) demographic data. LGS can reveal the relative abundance of species, but only areaconstrained searches can assess the absolute abundance. All sampling methods yield macrohabitat data but only intensive area-constrained searches yield microhabitat data. Welsh (1987) stated that time-constrained searches yield data on microhabitats, but these techniques have large experimenter biases. As the most useful information one can provide to a manager is information on habitat needs, all monitoring programs should initially include area-constrained searches to estimate the densities of terrestrial amphibians and assess the relative importance of various cover objects. One can then do stratified random sampling of all habitats, using a quick and efficient technique such as a time-constrained search, to effectively assess the status and distribution of terrestrial amphibians. Demographic data are also needed, but are hard to obtain for terrestrial salamanders unless a long-term extensive mark-recapture program is initiated during peak activity periods, to assess population viability in managed stands.  50  LITERATURE CITED  Aubry, K. B., J. L.L.C. & P. A. Hall (1988). Use of woody debris. In D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals, (pp. 171-185). Flagstaff, Arizona: USDA Forest Service. Pacific Northwest Region. Bartels, R., J. D. Dell, R. L. Knight & G. Schaefer 1985. Dead and Down Woody Material. Pp. 171-185 in. Management of Wildlife Fish Habitats in Forests of Western Oregon and Washington. Part 1 - Chapter Narratives. Portland, USDA Forest Service. Pacific Northwest Region. Beese, W. J. (1992). Third-year assessment of prescribed burning on forest productivity of some coastal British Columbia sites. In Canada-British Columbia partership agreement on forest resource development: FRDA II, FRDA Report 181 (pp. 10-26). Nanaimo: MacMillan Bloedel Ltd. Woodlands Services Division. Beiswenger, R. E. (1988). Integrating anuran amphibian species into environmental assessment programs. In R. Szaro Severson, K. and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, General Technical Report RM-166 (pp. 159-165). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Blymyer, M. J. & B. S. McGinnes 1977. Observations on possible effects of clearcutting on terrestrial amphibians. Bull. of Maryland Herp. Soc. 13: 79-83. Bogan, M. A., R. B. Finley Jr. & S. J. Petersburg (1988). The importance of biological surveys in managing public lands in the western United States. In R. Szaro Severson, K., and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, Gen. Tech. Report RM-166 . Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. British Columbia Ministry of Forests, British Columbia Ministry of Environment, Lands and Parks, Federal Department of Fisheries and Oceans & Council of Forest Industries 1992. British Columbia Coastal Fisheries/Forestry Guidelines (3rd Ed.). B. C. Ministry of Forests, Victoria. British Columbia Ministry of Forests, British Columbia Ministry of Environment, Lands and Parks 1992. Draft guidelines to maintain biological diversity in coastal forests. B.C. Ministry of Forests, Victoria. Bruce, C., D. Edwards, K. Mellen, A. McMillan, T. Owens & H. Sturgis 1985. Wildlife Relationships to Plant Communities and Stand Conditions. Pp. 33-55 in Management of Wildlife and Fish Habitats in Forests of Western Oregon and Washington. Part 1Chapter Narratives. Portland, USDA Forest Service. Pacific Northwest Region. Buhlmann, B. A., C. A. Pague, J. C. Mitchell & R. B. Glasgow (1988). Forestry operations and terrestrial salamanders: techniques in a study of the Cob Knob salamander, Plethodon punctatus. In R. Szaro Severson, K., and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, Gen. Tech. Report RM166 (pp. 38-44). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Burton, P.J., A.C. Balinsky, L.P. Coward, S.G. Summing, D.D. Kreeshaw 1992. The value of managing for biodiversity. The Forestry Chronicle 68: 222-224.  51  Burton, T. M. & G. E. Likens 1975. Salamander populations and biomass in the Hubbard Brook Experimental Forest, New Hampshire. Copeia 1975: 541-546. Bury, B. R. & P. S. Corn (1988). Douglas-fir forests in the Oregon and Washington Cascades: relation of the herpetofauna to stand age and moisture. In R. Szaro Severson, K. and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, Gen. Tech. Rep. RM-166 (pp. 11-22). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Conover, W. J. 1971. Practical Nonparametric Statistics (2nd ed. ed.). Toronto, John Wiley & Sons. Cook, F. 1984. Introduction to the Amphibians and Reptiles. Ottawa, D.W. Friesen & Sons Ltd. Corn, P. S. & B. R. Bury (1990). Sampling methods for terrestrial amphibians and reptiles. InGen. Tech. Rep. PNW-GTR-256 (pp. 34). Portland, OR: USDA Forest Service, Pacific Northwest Research Station. Corn, P. S. & B. R. Bury (1991). Terrestrial Amphibian Communities in the Oregon Coast Range. In K. B. Aubry Brookes, M.H., Agee, J.K., Anthony, R.G., Franklin, J.F., Noon, B.R., Raphael, M.G., Storm, R.M. and J. Verner (Ed.), Wildlife and Vegetation of Unmanaged Douglas-fir Forests, Gen. Tech. Rep. PNW-GTR-285 (pp. 305-318). Portland, OR: USDA, Forest Service, Pacific Northwest Research Station. Eckert, C. D., J. P. Savard & R. McLaughlin (1991). Breeding Bird Communities in Managed second-growth Stands. Riparian Areas. and Old-growth on Northeastern Vancouver Island Canadian Wildlife Service. Interim Report Feder, M. E. 1983. Integrating the ecology and physiology of plethodontid salamanders. Herpetologica 39: 291-310. Franklin, J. F. 1988. Structural and functional diversity in temperate forests. Pp. 166175 in Biodiversity (Wilson, E.O., Ed.). Washington, D.C., National Academy Press. Geiger, R. 1971. The Climate near the Ground. Cambridge, Massachusetts, Harvard University Press. Gerell, R. 1988. Faunal diversity and vegetation structure of some deciduous forests in southern Sweden. Holarctic Ecology 11: 87-95. Gibbons, J. W. (1988). The need for an environmental attitude. In R. Szaro Severson, K., and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, Gen. Tech. Report RM-166 (pp. 4-10). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Gilpin, M. E. & M.E. Soule. 1986. Minimum viable populations: processes of species extinction. Pp. 19-34 in Conervation Biology. The Science of Scarcity and Diversity (Soule, M.E., Ed). Sunderland, MA, Sinauer Associates, Inc. Gordon, R. E., J. A. MacMahon & J. Glanville 1987. Relative abundance, microhabitat and behavioue of some southern Appalachian salamanders. Zoologica 47: 9-14. Green, D. M. & W. R. Campbell (1984). The amphibians of British Columbia No. Handbook 45). British Columbia Provincial Museum, Victoria. 101 pp. 52  Green, R. N., P. J. Courtin, K. Klinka, R. J. Slaco & C. A. Ray (Ed.). (1988). Site Diagnosis. Tree Species Selection. and Slashburning Guidelines for the Vancouver Forest Region. B.C. Ministry of Forests, B.C. Forest Products Ltd., and University of B.C., Vancouver. Hairston, N. G. 1987. Growth, survival and reproduction of Plethodon jordani: trade offs between selective pressures. Copeia 4: 1024-1035. Hall, F. C., L. W. Brewer, J. F. Franklin & R. L. Werner 1985. Plant Communities and Stand Conditions. Pp. 17-31 in Management of Wildlife Fish Habitats in Forests of Western Oregon and Washington. Part 1 - Chapter Narratives. Portland, USDA Forest Service. Pacific Northwest Region. Hammond, H. 1991. Seeing the Forest Among the Trees. A Case for Wholistic Forest Use. Vancouver, Polestar Press Ltd. Harmon, M. E., J. F. Franklin, F. J. Swanson, P. Sollins, P. Gregory, S. V. Lattin, J. D. Anderson, N. H. Cline, S. P. Aumen, J. R. Sedell, G. W. Lienkaemper, K. Cromack Jr. & K. W. Cummins 1986. Ecology of coarse of woody debris in temperate ecosystems. Advances in Ecological Research 15: 133-302. Harris, L. D. 1984. The Fragmented Forest. Island Biogeography Theory and the Preservation of Biotic Diversity. Chicago, The University of Chicago Press. Heatwole, H. 1962a. Relation of substrate moisture to absorption and loss of water by the salamander Plethodon cinereus. Ecology 42: 814-819. Heatwole, H. 1962b. Environmental factors influencing local distribution and activity of the salamander, Plethodon cinereus. Ecology 43: 460-472. Helle, P. & O. Jarvinen 1986. Population trends of northern Finnish land birds in relation to their habitat selection and changes in forest structure. Oikos 46: 107-115. Hutchinson, V. H. 1961. Critical thermal maxima in salamanders. Physiol. Zool. 34: 92125 . Jaeger, R. G. 1971. Moisture as a factor influencing the distribution of two species of terrestrial salamanders. Oecologia (Berl.) 6: 191-207. Jaeger, R. G. 1979. Seasonal spatial distributions of the terrestrial salamander Plethodon cinereus. Herpetologica 35: 90-93. Jaeger, R. G. 1980. Fluctuation in prey availability and food limitation for a terrestrial salamander. Oecologia 44: 335-341. Johnson, B. (1992). Habitat loss and declining amphibian populations. In C. A. Bishop and K. E. Pettit (Ed.), Declines in Canadian amphibian populations: designing a national monitoring strateay, Occasional Paper No. 76 (pp. 71-75). Burlington: Canadian Wildlife Service. Johnson, C. R. 1980. The effects of five organophosphorous insecticides on thermal stress in tadpoles of the Pacific Tree Frog, WI Regina. Zool. J. Linn. Soc. 69: 143-147. Jones, K. B. (1988). Comparisons of herpetofauna of a natural and altered riparian ecosystem. In R. Szaro Severson, K. and D. Patton (Ed.), Management of Amphibians, 53  Reptiles and Small Mammals in North America, Gen. Tech. Rep. RM-166 (pp. 222227). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Keen, W. H. 1979. Feeding and activity patterns in the salamander Desmognathus ochrophaeus. J. of Herpetology 13: 461-467. Krebs, C. J. 1989. Estimating Abundance: Quadrat Counts and Line Transects. Pp. 64-124 in Ecological Methodology (Eds., Eds.). New York, Harper & Row Publishers. Krzysik, A. J. 1979. Resource allocation, coexistence and the niche structure of a stream bank salamander community. Ecol. Mon. 49: 173-194. Ludwig, J. A. & J. F. Reynolds 1988. Statistical Ecology. Toronto, John Wiley & Sons. Luttmerding, H. A., D. A. Demarchi, E. C. Lea, D. V. Meidinger & T. Void (Ed.). (1990). Describing Ecosystems in the Field (2nd ed.). Ministry of Environment, Victoria. MacFarlane, M. A. 1988. Mammal populations in mountain ash (Eucalyptus regnans) forests of various ages in the central highlands of Victoria. Australia Forestry 51: 14-27. Maiorana, V. C. 1978. Differences in diet as an epiphenomenon: space regulates salamanders. Can. J. of Zool. 56: 1017-1025. Marchal-Segault, D. & F. Ramade 1981. The effect of lindane, an insecticide, on hatching and postembryonic development of Xenopus laevis (Daudin) anuran amphibian. Environ. Res. 24: 250-258. Maser, C. 1990. The Redesigned Forest. Toronto, Stoddart Publ. Co. Ltd. Maser, C. & J. M. Trappe 1984. The seen and unseen world of the fallen tree. Portland OR, USDA Forest Service, Pacific Northwest Forest and Range Experiment Station, in cooperation with BLM,USDI. McDiarmid, R. W. (1991). Standard methods for measuring and monitoring biological diversity of amphibians. In A. a. K. E. P. Bishop (Ed.), Declines in Canadian amphibian populations: designing a national monitoring strategy, (pp. 80-86). Burlington, Ont.: Canadian Wildlife Service. Means, J. E., P. C. MacMillan & K. Cromack Jr. 1992. Biomass and nutrient content of Douglas-fir logs and other detrital pools in an old-growth forest, Oregon U.S.A. Can. J. For. Res. 22: 1536-1546. Meslow, E. C. & H. M. Wright (1975). Avifauna and succession in Douglas-fir forests of the Pacific Northwest. In D. R. Smith (Ed.), Proceedings of the Symposium on Management of Forest and Range Habitats for nongame birds, Gen. Tech. Rep. WO-1 (pp. 266-271). Tuscon, AZ: USDA Forest Service. Morrison, M. L. & E. C. Meslow 1984. Responses of avian communities to herbicide induced vegetation changes. J. Wild. Manage. 48: 14-22. Norse, E. A. 1990. Ancient Forests of the Pacific Northwest. Covelo, CA, Island Press. Nussbaum, R. A., E. D. J. Brodie & R. M. Storm 1983. Amphibians & Reptiles of the Pacific Northwest. Moscow, ID, University of Idaho Press. 54  O'Donoghue, M. (1991) Reproductiion. Juvenile Survival and Movements of Snowshoe Hares at a Cuclic Population Peak. MSc, University of British Columbia, Vancouver, B.C. Ovaska, K. 1988. Spacing and movements of the salamander Plethodon vehiculum. Herpetologica 44: 377-386. Pawar, K. R., H. V. Ghate & M. Katdare 1983. Effect of malathion on embryonic development of the frog Microhyla ornata (Dumerial and Bibron). Bull. Environ. Contam. Toxicol. 31: 170-176. Peacock, R. L. & R. A. Nussbaum 1973. Reproductive biology and bopulation btructure of the bestern red-backed salamander, Plethodon vehiculum (Cooper). J. of Herpetology 7: 215-224. Pough, H. F. 1983. Amphibians and reptiles as low energy systems. Pp. 141-188 in Behavioral energetics: the cost of survival in vertebrates. Columbus, Ohio, Ohio University Press. Pough, F. H., E. M. Smith, D. H. Rhodes & A. Collazo 1987. The abundance of salamanders in forest stands with different histories of disturbance. For. Ecol. Manage. 20: 1-9. Prescott, C. E., J. P. Corbin & D. Parkinson 1989. Input, accumulation, and residence times of carbon, nitrogen, and phosphorus in four Rocky Mountain coniferous forests. Can. J. For. Res. 19: 489-498. Pulliam, H. R. 1988. Sources, sinks, and population regulation. Am. Nat. 132: 652-661. Raphael, M. (1988). Long-term trends in abundance of amphibians, reptiles and mammals in Douglas-fir forests of Northwestern California. In R. Szaro Severson, K. and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, Gen. Tech. Rep. RM-166 (pp. 23-31). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Raphael, M. G. & R. H. Barrett 1981. Methodologies for a comprehensive wildlife survey and habitat analysis in old-growth Douglas-fir forests. Cal-Nev Wildlife Transactions 1981: 106-121. Ratmonik, C. A. & N. J. J. Scott (1988). Habitat requirements of New Mexico's endangered salamanders. In R. Szaro Severson, K. and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, Gen. Tech. Rep. RM166 (pp. 54-63). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Ray, C. 1958. Vital limits and rates of dessication in salamanders. Ecology 39: 75-83. Robinson, G. 1988. The Forest and the Trees. A Guide to Excellent Forestry. Covelo, CA, Island Press. Santillo, D. J., P. W. Brown & D. M. Leslie 1989. Response of songbirds to glyphosateinduced habitat changes on clearcuts. J. of Wildi. Manage. 53: 64-71. Slagsvold, T. 1977. Bird population changes after clearance of deciduous scrub. Biol. Cons. 1 2: 229-243. Spies, T. A., J. F. Franklin & T. B. Thomas 1988. Coarse woody debris in Douglas-fir forests of western Oregon and Washington. Ecology 69: 1689-1702. 55  Spotila, J. R. 1972. Role of temperature and water in the ecology of lungless salamanders. Ecol. Monogr. 42: 95-125. Titterington, R. W., H. S. Crawford & B. N. Burgason 1979. Songbird responses to commercial clear-cutting in Maine spruce-fir forests. J. Wildlife Manage. 43: 602-606. Triska, F. J. & K. J. Cromack (1979). The role of woody debris in forests and streams. In Forests: Fresh Perspectives from Ecosystem Analysis. Proceedings of the 40th Annual Biology Colloquium, (pp. 171-189). Corvalis: State University Press. Trofymo, J. A. & W. J. Beese (1991). Quantities of coarse woody debris in old-growth forests. In Old-Growth Strateay Background Papers, (pp. 21). Victoria: Ministry of Forests. Trowbridge, R., B. Hawkes, A. Macadam & J. Parminter 1987. Field Handbook for Prescribed Fire Assessments in British Columbia: Logging Slash Fuels. Victoria. B.C. Ministry of Forests and Lands and Candian Forestry Service. pp. 19-29. Van Horne, B. 1983. Density as a misleading indicator of habitat quality. J. of WiIdl. Manage. 47: 894-901. Van Wagner, C. E. 1968. The line intersect method in forest fuel sampling. Forest Science 14: 20-26. Welsh, H., Jr. 1987. Monitoring herpetofauna in woodland habitats of northwestern California and southwestern Oregon: a comprehensive strategy. Berkeley, CA, USDA Forest Service, Pacific Southwest Forest and Range Experimentation Station. Gen. Tech. Rep. PSW-100. Welsh, H. H., Jr. 1990. Relictual amphibians and old-growth forests. Conservation Biology 4: 309-319. Welsh, H. H., Jr. & A. Lind (1988). Old growth forests and the distribution of the terrestrial herpetofauna. In R. Szaro Severson, K. and D. Patton (Ed.), Management of Amphibians. Reptiles and Small Mammals in North America, Gen. Tech. Rep. RM166 (pp. 439-455). Flagstaff, AZ: USDA Forest Service, Pacific Northwest Region. Wilcove, D. S., C. H. McLellan & A. P. Dobson 1986. Habitat fragmentation in the temperate zone. Pp. 237-256 in Conservation Biology. The Science of Scarcity and Diversity (Soul, M.E., Ed.). Sunderland, MA, Sinauer Associates, Inc.Zar, J. H. 1984. Biostatistical Analysis (2nd ed.). New Jersey, Prentice Hall. Zar, J. H. 1984. Biostatistical Analysis (2nd ed.). New Jersey, Prentice Hall.  56  APPENDIX SNAG USE BY TERRESTRIAL AMPHIBIANS  INTRODUCTION Snags (wildlife trees) provide sites for perching, feeding, over-wintering, roosting, and breeding in natural forests. They may also provide refuges from predators and inclement weather (Maser 1990; Norse 1990; Neitro et al. 1985; Cline 1980). Snags are formed by insect outbreaks, fungal infections, wind storms, suppression, and breakage from falling neighbour trees (Neitro et al. 1985). Their subsequent deterioration includes sloughing of bark, branch breakage, and the softening of the heartwood (Neitro et al. 1985), which enhance their value to wildlife. In natural forests, snags of all types and sizes are well-represented due to long-term exposure of trees to the various mortality factors. Thus, they can house an array of wildlife. Terrestrial amphibians can be abundant in forests, and may play an important role in energy flow between trophic levels (Feder 1983; Pough 1983). They are long-lived, have stable populations, and spend their entire lives in one area. These factors make them valuable indicators of habitat deterioration. In this study I assess the strength of their association with snags, and determine which snag characteristics are important to terrestrial amphibians. My aim is to contribute to better snag management and to an understanding of amphibian needs.  STUDY AREA AND EXPERIMENTAL DESIGN The research was conducted from May 26-28 1992, in the in the vicinity of Port Alberni (49.3°N latitude and 125.3°W longitude), central Vancouver Island. Two coastal Douglas-fir (Pseudotsuga menziesii)/hemlock (Tsuga heterophylla) old-growth forests (Cathedral Grove and Nahmint River), and two managed stands: Kanyon Creek (age 51 years) and Sproat Lake (age 59 years) were selected. The managed stands had been clearcut logged, burned, and allowed to regenerate naturally. The study sites ranged over an area of 15-20 km by 10 km (see Chapter II). 57  METHODS Snag Surveys: Snag species, height, diameter at breast height (dbh), and state of decomposition, were measured in four randomly located 20 x20 m plots, within each of the 4 study sites. Snags were grouped into 3 stages of decay: A (early), B (intermediate), and C (advanced). Refer to Chapter III for more information on decomposition.  Amphibian Surveys: Snags were searched for amphibians by turning over sloughed bark and slash, and exploring hollows and cracks up to 1 m from the ground. Amphibians were identified and their position relative to the snag recorded. They were released when the search was completed. Searches did not involve destructive sampling. After a maximum of 20 minutes, searchers moved to the nearest neighbouring snag. This was repeated until a total of 30 snags were sampled. For each snag, diameter at breast height (dbh)(cm), height (m), state of decay (A, B and C), and species composition was recorded. Snags were classified as being smaller or greater than 30 cm in dbh.  RESULTS Snag Density and Quality in Forests and Stands: In Nahmint old-growth, 73% of snags were Douglas-fir; the remainder were other conifers. Eighty-seven percent of the snags were > 30 cm in dbh (mean = 60 cm). In Cathedral old-growth, 33% of snags were Douglas-fir, the other 67% represented other conifers. 20% of them were > 30 cm in dbh (mean = 72 cm). Densities of small and large snags are summarized in Table 4.1. In second-growth, it was difficult to determine the species composition of small snags; 7-9% were deciduous. Two large (remnant) snags were found, a hemlock in Kanyon and a Douglas-fir in Sproat (Table 4.1). The mean dbh was 16.6 cm for Kanyon and 21 cm for Sproat. 58  Fresh snags are characterized by tight bark (Table 4.2). Sloughed bark is characteristic of intermediately and very decayed snags. 17% of snags in managed stands had sloughed bark, as compared with 47% in old-growth. 15% of the snags in these stands were very decayed, compared with 37% in old-growth.  The Distribution of Amphibians on Snags: Twenty-six and 30 salamanders were encountered on samples of 60 snags within the two old growth replicates (Figure 4.1). Their distribution was clumped (26 salamanders on 9 snags in Nahmint, 30 salamanders on 10 snags in Cathedral); all were associated with snags > 30 cm in dbh (Table 4.1). Species included the western red-backed salamander (Plethodon vehiculum), the clouded salamander (Anneides ferreus), the ensatina (Ensatina eschscholtzii), and the rough-skinned newt (Taricha aranulosa). A mean of one salamander was found on a total of 60 snags in second-growth. One E. vehiculum was found at the base of an old-growth remnant Douglas-fir, and one E. vehiculum at the base of a snag <30 cm wide. The abundance of salamanders relative to the frequency of occurrence of each snag size class is summarized in Table 4.1. Most salamanders (85%) were associated with snags having sloughed bark (Figure 4.2). Of these, 80% were found under bark on the ground. The dominant snag species, Douglas fir, housed 77-100% of salamanders encountered.  59  Table 1. The number of salamanders associated with small and large snags, in old-growth forests and mature managed second growth stands, in May 1992 # Snags/ha  Relative Abundance of Salamanders*  Stand Age  Dbh(cm)  Mean dbh  Old-growth  >30  75.9  11 ± 6  21  <30  51.2  9.5 ± 5.5  0  20.5 ± .5  21  1±0  0  26 ± 1  1  28 ± 1  1  Total 60 years  >30  Managed  <30 Total  20.7 17.6  * (The no. of salamanders/60 searched snags)x no. of snags/ha  60  Table 2. The frequency of occurrence (%) of snags in different stages of decay, with and without bark, in old growth and mature managed second growth Decay Class  Old growth  Second growth  17  24  0 100 100  0 100 0  46  61  No bark Bark on Sloughed bark  7 30 63  3 69 28 15  No bark Bark on Sloughed bark  37 36 14 50 0.5  Bark Status  Early No bark Bark on Sloughed bark Intermediate  Advanced  61  33 33 33  Figure 1. The number of salamanders encountered on snags in old growth and in mature managed second growth, in the spring of 1992  62  ^ Snag with no bark  El Snag with bark on  g  Snag with bark on and on ground  M Snag with bark on ground M Snag with wood chunks on ground  Snag Type  Figure 2. The use of various types of snags by salamanders in old growth, in May of 1992  DISCUSSION Thirty percent of snags were occupied by terrestrial amphibians in both old-growth replicates, which suggests that they are important to terrestrial amphibians during the spring season. Their clumped distribution (56 salamanders on 19 snags) may imply that all snags are not equally suitable. Snag types (species, sizes) are variable in old-growth (Table 4.1), as a function of time and chance. At Kanyon and Sproat Lake, clearcutting and burning drastically reduced the number of large snags (> 30 cm in dbh). Clearcutting also decreases the occurrence of inte4 workers and create plantable sites. After  64  harvesting, there is a '3 m knock-down' process which involves felling anything left standing, to reduce the risk of lightning induced fires. This clearcutting practice is often followed by prescribed burns to reduce wood volumes to lower the risk of fire spreading (W. Beese, pers. corn.; W. French, pers. corn.). I recommend that: (1) large snags (>30 cm dbh) should be retained, particularly in the intermediate stages of decay; (2) hard snags and suppressed trees should be retained for future snag recruitment; and (3) patches of windfirm snags and trees, with accompanying understory vegetation, be left untouched during felling operations, around landing, outside the guyline circle, between yarder cable configurations, and in riparian zones.  65  LITERATURE CITED  Cline, S. P.,Berg, A. B. & Wight, H. M. 1980. Snag characteristics and dynamics in Douglas-fir forests, western Oregon. J. of Wildl. Manage., 44, 773-786. Conner, R. N.,Miller, O. K., Jr. & Adkisson, C. S. 1976. Woodpecker dependence on trees infected by fungal heart rots. Wilson Bull.. 88, 575-581. Feder, M. E. 1983. Integrating the ecology and physiology of plethodontid salamanders. Herpetologica. 39, 291-310. Gano, R. D., Jr. & Mosher, J. A. 1983. Artificial cavity construction-an alternative to nest boxes. Wildl. Soc. Bull.. 11, 74-76. Mannan, R. W.,Meslow, E. C. & Wight, H. M. 1980. Use of snags by birds in Douglas-fir forests, western Oregon. J. of Wildl. Manage.. 44, 787-797. Maser, C. 1990. The Redesigned Forest. Toronto: Stoddart Publ. Co. Ltd. McClelland, B. R.,Frissell, S. S. & Fischer, W. C. 1975. Identifying forest snags useful for holenesting birds in forests of western larch and Douglas-fir. J. For.. 77, 480-483. Neitro, W.A., Binkley, V.W., Cline, S.P., Mannan, R.W., Marcot, B.G., Taylor, D. & Wagner, F.F. 1985. Snags (wildlife trees)(pp. 129-169) 10.: Management of Widlife and Fish Habitats in Forets of Western Oregon and Washington. Part 1- Chapter Narratives. U.S. Forest Service. Pacific Northwest Region. Publ. R6&F&WL-192-1985. Norse, E. A. 1990. Ancient Forests of the Pacific Northwest. Covelo, CA: Island Press. Peterson, A. W. & Grubb, T. C., Jr. 1983. Artificial trees as a cavity substrate for woodpeckers. J. Wildl. Manage., 47, 790-798. Pough, H. F. 1983. Amphibians and reptiles as low energy systems. In: Behavioral energetics: the cost of survival in vertebrates. (Ed. by W. P. Aspey & L. S.I.), pp. 141-188. Columbus, Ohio: Ohio University Press. Raphael, M. G. & White, M. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildl. Monogr.. 86, 66. Sedgewick, J. A. & Knopf, F. L. 1990. Habitat relationships and nest site characteristics of cavity nesting birds in cottonwood floodplains. J. Wildl. Manage., 54, 112-124. Zarnowitz, J. E. & Manuwal, D. A. 1985. The effects of forest management on cavity-nesting birds in northwestern Washington. J. Wild. Manage.,49, 255-263.  66  


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