This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain.

11 The Effects of Disturbance K EVIN S. McKELVEY and Succession on Wildlife Habitat and Animal Communities

his chapter discusses the study of disturbance and al l affect both the postdisturbance wildlife community Tsuccession as they relate to wildlife. As such, the and, more importantly, the trajectory of the postdistur­ discussion is confined to those disturbance processes bance community. Even for well-studied species, co­ that change the physical attributes of habitat, leading herent understandings of their relationships to distur­ to a postdisturbance trajectory. However, even with bance across time and space are therefore often vague. this narrowing of the scope of disturbances discussed, Commonly, we look at successional changes in habi­ there remain formidable obstacles prior to any co­ tat quality by using spatial samples of different ages as if herent discussion of disturbance. The first, and most they were a temporal series, which implies spatiotem­ fundamental, is definitional: what constitutes distur­ poral constancy in successional dynamics. This assump­ bance and succession, and what is habitat? The con­ tion has served wildlife research well, but in the face of cepts of disturbance, habitat, and succession are highly directional climate change and the nearly continuous scale-dependent; disturbances at one scale become addition of exotic species, this approach is becoming part of continuous processes at a larger scale, and ideas increasingly untenable. We need to embrace the idea associated with succession require assumptions of con­ that postdisturbance succession is increasingly unlikely stancy, which become highly problematic as spatiotem­ to produce communities similar to those that the dis­ poral scales increase. Literature on the effects of distur­ turbance altered: short-term successional patterns are bance and succession on wildlife, however, focuses on a likely to be influenced by the large and dynamic pool of narrow range of spatial scales, primarily occurs within exotic plants and animals and longer-term succession a narrow temporal window immediately following dis­ by directional climate change. turbance, and seldom includes interactions between Together, these observations indicate a need for areas within the disturbed patch and the landscape studies of disturbance and succession at larger spatia­ that surrounds it. While these largely descriptive stud­ temporal scales. It is, however, important to assess ies undoubtedly have great local value, more general the feasibility of scaling up studies in time and space. information about the relationships between organ­ Clearly, studies that expand both the spatial and tem­ isms and environments shaped by disturbance and poral dimensions of data collection quadratically in­ succession is remarkably limited. Multiple small-scale crease costs; longer time frames do not fit into current descriptive studies of the immediate postdisturbance competitive funding structures and contain a variety environment do not appear to coherently aggregate of negatives such as the potential loss of data. Broad­ into larger understandings of the effects of disturbance scale targeted monitoring, however, can provide a and succession on wildlife. Context is important: the framework allowing acquisition of data at these scales conditions at the time of the disturbance, in adjacent and can be designed to produce both immediate and undisturbed patches, and within the broader landscape longer-term results. 144 RESEARCH AND CONSERVATION

and other habitat features). Temporal habitat patterns Disturbance, Succession, and Questions are as critical for conservation as are spatial patterns, of Scale but much harder to study. In many cases, the events Bormann and Likens (1979) defined disturbance as dis­ that structure landscapes and define species ranges are ruption of the pattern of the ecosystem, principally by rare- often singular. For example, the genetic popula­ external physical forces. This idea, however, assumes tion structures of many species are strongly associated that ecosystems function as idealized Newtonian sys­ with glacial vicariance that occurred during the Pleis­ tems, in stasis until external energy is applied. But few tocene (see Shafer et al. 2010 for a review). To study ecosystems exist in equilibrium (Sousa 1984 ). Further, these rare events, we primarily look to the past to gain Rykiel (1985) noted that whpt±rer disturbance is viewed insight into their frequency, size, and postevent succes­ as changing the state of an ecosystem or being part of sional and evolutionary trajectories. Further, for prac­ that state is entirely a function of scale. At one scale, a tical reasons, we frequently use habitats of different tree-fall is a state-altering disturbance; at a larger scale, ages as surrogates for the passage of time, making the it is part of a continuous process that creates and main· tacit assumption that if we were to project one area tains the state of an old-forest ecosystem. The same forward (or backward), it would be similar enough to a thing is true for larger disturbances; at one scale, fire in sunogate area that we can infer its future or past state western US forests represents a significant disturbance by studying that surrogate. I refer to this approach as to ecological function, radically altering wildlife habi­ "trading space for time." tats. At a larger scale, fire is a part of the ecosystem­ The study of temporal patterns of habitat over time, many ecosystems are dependent on fire to mruntai.n and specifically the validity of using spatial surrogates to the presence, patterns, and juxtapositions of plant and infer temporal patterns, is tightly linked to the concept animal species (Habeck and Mutch 1973· Covington of succession. Here I use this term in a neo-Clemensian I and Moore 1994; Nowacki and Abrams 2008). fashion (Clements 1916; Daubenmire 1952): suc- Similar-scale dependencies are associated with cession assumes a pattern of orderly and predictable the effects of disturbance on wildlife habitat. Habitat changes in species presence and abundance that occur is often thought of with a particular scale in mind: over time after disturbance, leading to a fairly stable perhaps a forest stand or home-range area evaluated terminal state, or climax. As Gleason (1927) noted, the across a year or the lifespan of an organism. If succes­ idea of succession is very appealing: if understood, it sional thinking is applied, habitat may be defined as allows us to predict the future and to see into the past existing within a specific sere (e.g., a species may be without needing any data other than what we collect considered to be associated with early seral or late sue· in the present. However, successional concepts cannot cessional forests). However, the habitat requirements be accepted naively. The validity of this concept, and its for population persistence are, like disturbance, com­ resulting popularity (or lack thereof), is directly related plex and span many scales in both space and time. In to the complexity of the system studied and the scale space, wildlife habitat spans spatial domains measured at which the system is evaluated. Simple systems have in meters (e.g., specific resources for denning), to ki­ fewer succession pathways and fewer species; hence, lometers (e.g., sufficient resources to support a local the vegetative trajectories are more predictable. For viable population), to hundreds of kilometers (e.g., a example, in the western forests of the United States, mosaic of resources and connectivity sufficient to sup­ tree communities are often simple and contain large port long-term metapopulation [Levins 1969, 1970] areas of intact natural vegetative communities. In persistence and abundance). Additionally, at all spatial these systems, succession-based classification systems sca1es, juxtaposition and spatial patterns of habitats (e.g., habitat types; Pfister and Arno 1980) and con­ are important (e.g., Iverson ~tal. 1987). Temporally, cepts (e.g., potential natural vegetation [PNV]; Kuchler the definition of habitat is even more complex. Again, 1964) are popular. In highly modified landscapes con­ scales vary from almost instantaneous (e.g., the timing cepts, however, concepts like PNV become abstract as of ice breakup in the Arctic) to millennia! (e.g. , the pro­ none of the vegetation within a study area may exist cesses of erosion and deposition that create soils, caves, in its putative potential condition (Zerbe 1998). For E F FECTS O F DISTURBANCE AND SU C CESSI O N O N WILD LIF E H ABITAT 145 these reasons, looked at objectively, succession is often process, we discuss and label this process as part of a a problematic concept, but one that has proved useful separate body of literature. Similarly, we are prone to in many communities and without which our ability view what we do as being more of a disturbance than to study systems and build predictive models would be what other organisms do. For example, beavers (Cas­ severely limited. However, regardless of its historical tor canadensis) instigate a variety of disturbances (tree validity, there is good reason to doubt its relevance as a felling, house and dam construction), oftentimes re­ tool to predict future habitat conditions; the likelihood moving all accessible trees from the area adjacent to of rapid directional climate change and the increasing their pond (Martell et al. 2006) and thereby creating presence of exotic species compromise our ability to a shifting pattern of disturbance that can, over many use past patterns to predict the future. Another con­ generations, affect large areas (Naiman et al. 1986). cept, community assembly theory, which views com- . However, studies of beaver activity are seldom framed munities as being the result of a continuous process of or titled as being disturbance studies. species invasion and extinction (Lodge 1993), may be By choosing to concentrate our studies on mid-scale more germane. disturbances, such as fires, and the subsequent changes in habitat, we have, unfortunately, also chosen events that occur at temporal scales that are inconvenient­ Disturba nce and Its Re lationship to postdisturbance successional trajectories frequently W ildli fe Commun ities require many years to evolve. To study these processes, Defin ing Disturbance it is therefore necessary either to set up very long-term The spatial scaling of habitat, including the habitat studies or trade space for time, with its associated as­ patch structure generated by disturbance, is ultimately sumptions and weaknesses. As noted earlier, trading defined by the grain at which specific organisms per­ space for time requires the assumption of a high de­ ceive the world (Kotliar and Wiens 1990) and opera­ gree of similarity and transferability across disturbance tionally defined at some arbitrary time scale. Thus, events and subsequent recoveries. Given these con­ disturbances also are subject to these same scalings: straints, there are two types of disturbances that occur an event that constitutes a significant disturbance for at scales and with frequencies that allow both study a woodland salamander (e.g. , Ensetina sp.) may be of and reasonable transferability: anthropogenic changes little importance to a wolverine (Gulo gulo). Thus, such as logging and land conversion, and fires; a great when we consider disturbance ecology, we should ide­ deal of what we identify as the study of disturbance ally begin with operational environment of the affected therefore concerns these types of disturbances. Likely organisms, clearly identify the time frame of interest, for these reasons, and based on practical considerations and define disturbances accordingly. This, however, is such as the ease of aging trees and hence inferring past seldom done. When we consider disturbance, we often disturbance events, this body of literature has a strong have a particular scale in mind, generally an intermedi­ bias toward the study of forested systems. ate scale. When we label events as disturbances, we en­ In addition to often being appropriate in scale, an­ vision events such as hurricanes or forest fires and par­ thropogenic disturbance is highly researchable because ticularly anthropogenic activities such as logging, land we seek to control many aspects of our disturbances. clearing, and infrastructure development: scientific For example, in forestry we apply standardized silvicul­ articles that claim to study disturbance overwhelm­ tural treatments: a clear-cut treatment, in addition to ingly study these types of events. However, this scale is removing all trees, generally will also include removal not the organism's scale; rather, it is our scale. In part, of residual debris and soil scarification, leaving a fairly this is simply a matter of semantics: how we define homogenous postdisturbance environment that is disturbance rather than what we study. For example, replicated across multiple treatment areas. Treatment we study the effects of tree-fall and subsequent forest blocks also tend to be of similar size-both econom­ gap dynamics on wildlife; Forsman eta!. (2010) found ics and policy dictate this. On Forest Service lands in eleven studies of gap effects on forest birds. However, the United States, for example, clear-cut treatments while forest gap' formation is an obvious disturbance have for many years been limited in size and spacing 146 RESEARCH AND CONSERVATION to ten to twenty ha blocks interspersed with uncut ar­ disinterest-is the norm not just for studies of topical eas (Franklin and Forman 1987). They also occur at natural disasters but is pervasive across disturbance relatively fixed rates across time and space because types. As noted by Fontaine and Kennedy (2012), in a we desire fairly constant flows of products supplied recent meta-analysis of fire effects on small mammals to a fixed array of mills. Thus, a system of clear-cuts and birds, the meta-analysis was limited to short-term is a much more regularized disturbance pattern than (four years or less) responses because data "at longer would occur naturally, and for scientific study, these time scales were too sparse to permit quantitative as­ regularizing factors allow us to achieve a level of rep­ sessment" (Fontaine and Kennedy 2012, abstract). lication that is generally absent in natural events. Of natural disturbances, fires are probably the most ideal ANTHROPOGENIC DISTURBANCE for study. In fire-prone areas, fires occur frequently Within the constraints noted by Fontaine and Kennedy enough to be grouped based on covariates such as in­ (2012; taxa limited to small mammals and birds and ef­ tensity, aspect, and pre-fire vegetation. Thus, fires form fects to the period immediately following disturbance), another major class of disturbance studies. In addition there are still hundreds ofpublished papers document­ to these two heavily studied classes of disturbance, a ing the effects of anthropogenic disturbance on wildlife. third major group of published studies concerns distur­ Luckily, there has been a series of recent meta-analyses bance events that we perceive as being "natural disas­ that provide both syntheses of these papers' findings ters": hurricanes, floods, tsunamis, and so forth. and extensive bibliographic references. Kalies et a!. (2010) performed a meta-analysis of twenty-two papers on the effects of thinning, prescribed burns, and some Wildlife Studies Associated wildfire on birds and rodents in the American South­ with Disturbance west. Not all species or taxa were equally represented; Formal studies of the effects of disturbances on wild­ birds were most commonly studied. Spatial scalJs were life are almost exclusively limited to the immediate small (less than four hundred ha) and temporally close aftermath of the event. This phenomenon is perhaps to the treatment (twenty years or less posttreatment). most apparent in the study of natural disasters. For ex­ Ground-foraging birds and rodents showed neutral ample, a Coogle Scholar search on "Hurricane Hugo" density responses to the treatments, whereas aerial, revealed a total of thirteen journal publications on tree, and bole-foraging birds had positive or neutral the hurricane's effects on animals, of which ten (77%) responses to both small-diameter removal and burn­ were published within the first five years. However, few ing treatments but negative responses to overstory re­ studies are currently underway; I was not able to locate moval and wildfire. Hartway and Mills (2012) looked at a single journal article with animal responses to Hurri­ a variety of anthropogenic treatments to increase nest cane Hugo as the primary subject published after 2002. success, including prescribed burns (nineteen stud­ This pattern is not limited to the study of wildlife: of ies, fifty-two species). They also evaluated the effects seventy-two identi£ed journal publications on all sub­ of livestock exclusion, prescribed burning, removal of jects concerning the effects of Hurricane Hugo, sixty­ predators, and removal of cowbirds (Molothrus ater). seven (83%) were published in the first five years, and Of the four treatments, prescribed burning was not the none were published after 2008. Given that changes most effective but did significantly increase nest suc­ in vegetation wrought by Hugo will affect habitat in cess for most species, though there was a large amount various ways for hundreds of years, there is no biologi­ of variation between species groups. Specifically, preco­ cal rationale for this spate of studies of the exceedingly cial bird species were negatively affected by prescribed ephemeral habitat conditions associated with its imme­ burning. Vanderwel et a!. (2007) performed a meta­ diate aftermath coupled with low levels of interest asso­ analysis of the effects of partial cutting on forest birds. ciated with the still rapidly evolving habitat conditions All forty-two studies examined the effects of cutting that currently exist. Further, this pattern- frequent ten years or less after harvest. Of the thirty-four bird studies immediately after the disturbance followed by species studied, fourteen declined with increasing bar- I EFFECTS OF DISTURBANCE AND SUCCESSION ON WILDLIFE H ABITAT 147 vest intensity and six increased. Not surprisingly, deep erality of the pattern observed by Hossack et al. (2013) forest birds such as ovenbirds (Seiurus aurocapilla) and is unknown. brown creepers (Certhia americana) were the species most sensitive to harvesting. General Strengths and Weaknesses of Disturbance-Related Meta-analyses FIRE Similar to anthropogenic disturbances, a number of Throughout these meta-analyses, a general pattern recent meta-analyses document the effects of fire on emerges: bird response to disturbance is much more wildlife. Not all of these meta-analyses separate natural heavily studied than is mammal response-confin­ fire from anthropogenic treatments such as thinning ing their scope to the effects of fire alone, Leidolf and or prescribed burns. For example, Zwolak (2009) per­ Bissonette (2009) identify 512 studies on birds-and formed a meta-analysis of fire clear-cutting and partial mammal study is almost exclusively limited to small cutting on small mammal populations. Of the fifty-six mammals. The general paucity of habitat disturbance studies used in the meta-analysis, most (fifty-four) doc­ studies for mammals other than small mammals is umented short-term changes. Only eleven documented likely due to several factors. Perhaps the most critical longer-term (ten to twenty years after the disturbance) factor is that for larger mammals, home range sizes effects (Zwolak 2009, supplemental, appendix A). and/or densities are not compatible with abundance Zwolak found that deer mice (Peromyscus maniculatus) estimation within disturbance classes. Another is that increased and red-backed voles (Myodes gapperi) de­ they are far more difficult to locate and count than are creased after disturbance, with the effect size increas­ small mammals and particularly birds. Even for small ing with rising disturbance intensity. These patterns re­ mammals, there is no approach that provides as effi­ laxed as time since disturbance increased. Fontaine and cient an estimate of occurrence and relative abundance Kennedy (2012) performed a meta-analysis of forty­ as a bird point-count. Thus, specific disturbance studies one papers, documenting the short-term effects of fire of many mammalian species of interest such as large and fire surrogates on small mammals and birds. They, carnivores are impeded both by the utilization of large like Zwolak (2009), found that postdisturbance effects areas that likely exceed areas of disturbance and diffi­ on the abundance of species increased with increasing culties in obtaining sufficient sample sizes. The effects fire severity. A total of 119 bird species and 17 small of disturbance on other groups such as herptofauna are mammal species were analyzed by Fontaine and Ken­ far less studied. nedy (2012). While generalizations across this broad of In addition to being limited in types of species ana­ a group are difficult, in many ways their results mirror lyzed, meta-analyses are weakened by the need to re­ those of Vanderwel et al. (2007) and Zwolak (2009): duce studies to groups that can be evaluated based on the largest positive effect was on deer mice in severe a common metric (effect size) and common covariates burns whereas forest birds such as ovenbirds and her­ such as time since disturbance. While such grouping mit thrush ( Catharus guttatus) were negatively affected. is necessary, all disturbances are not created equal; Interestingly, brown creepers were positively affected postdisturbance abundances may be correlated with by moderate intensity disturbances. While the period abundances within the general landscape (e.g., Bretons immediately following disturbance may exhibit the et al. 2005), a covariate that is generally missing from most profound shifts in structure and composition with disturbance studies. Similarly, the effects of past distur­ a relaxation toward the predisturbance state over time bances may have altered systems leading to different re­ (e.g., Zwolak 2009), this is not always the case. Hos­ sponses to the studied disturbance when viewed across sack et al. (2013), studying salamander response to fire many systems. Fox (2011), for example, notes that dry, in Montana, found little change in abundance during disturbance-prone grasslands generically have fewer the period immediately after wildfire but significant granivores than other grasslands; these systems would declines seven to twenty-one years after disturbance. therefore be expected to have different responses to any Because we lack very many long-term studies, the gen- specific fire because they are already adapted to the oc- 148 RESEARCH AND CONSERVATION currence of frequent fires. Assuming that a species can the first five years after the burn. However, unlike other persist through a disturbance and within the immedi­ taxa, some bird species are specifically adapted to these ate postdisturbance landscape, likely the postdistur­ immediate postfire conditions, making their study in bance vegetative trajectory is more important than the these environments especially useful. Notable among time since disturbance. Monamy and Fox (2000), for the bark insectivores is the black-backed woodpecker example, found that recolonization trajectories for an (Picoides arcticus). This species colonizes burned areas old forest specialist differed between two fires based within one year after a fire, occupies burned areas for on time since disturbance but that the two h·ajectories three to five years, and peaks in density around three were concordant when vegetative trajectory rather years after fire (Caton 1996; Saab et al. 2007). Outside than time was evaluated. This pattern appears logical of this window, when postfire environments are opti­ and is likely general. If this is the case, then standard mal, black-backed woodpeckers are difficult to locate approaches to successional studies where time since and are distributed diffusely feeding, opportunistically, disturbance is used to characterize organism response on recently killed trees (Tremblay et al. 2010, Dudley are likely produce weaker results than studies that mea­ et al. 2012). sure vegetative conditions directly. Succession Bird Studies Where Eva lu ation Close to the Defining Succession Disturbance Event is Essential Succession is an idea that emerged in the late 1800s Clearly, more common and replicable sorts of dis­ and was championed by Clements (1916). He en­ turbance, like fire, will be more likely to engender visioned the process as an orderly series of plant as­ meaningful adaptations. However, even for common semblages from the point of initiation through to the disturbance types, such as fire in dry environments, im­ "climax," a stable state. Just as we define distJrbance mediate postfire conditions will be relatively rare on as events that we find disturbing given our viewpoint, the landscape. To take advantage of these conditions, similarly, we tend to consider succession in terms of population responses must be quick and the organisms processes that scale reasonably to our lifetimes. For need either to be common enough to be on or adjacent these reasons, the genesis of succession as a concept to the fire when it occurs or to be mobile enough to began with the observations of the dynamics of forests find these disturbances quickly so as to take advantage (Clements 1916) and is most commonly still evoked in of the ephemeral postdisturbance environment. It is this context. A great deal of thinking about this con­ not surprising that, for example, deer mice show large cept has occurred since Clements wrote "Plant Succes­ increases in postfire environments. They are general­ sion: An Analysis of the Development of Vegetation" ists and therefore present in the area, can reproduce in 1916, but the underlying ideas have remained re­ rapidly, and normally favor bare areas over dense grass markably stable. Primarily, the concept has been em­ (Pearson et al. 2001) and should therefore prosper in bellished, incorporating ideas such as "secondary" suc­ postfire environments. Similarly, avian species, being cession, in which initiation contains elements of the highly mobile, can locate desired environments includ­ predisturbance ecosystem, and various "disclimaxes," ing recent disturbances and can evolve to take advan­ where an area is prevented from evolving into its pre­ tage of these conditions. For these reasons, the study sumed true climax by processes such as frequent dis­ of avifauna in postfire environments is both active turbance or grazing (see Meeker and Merkel1984 for and biologically appropriate. Saab and Powell (2005), a discussion). As mentioned previously, succession has in a review of literature on over two hundred species retained popularity in many fields, particularly where of birds and their associations with fire, showed that ecosystems are simple and primarily composed of na­ aerial, ground, and bark insectivores clearly favored tive species. Even in these simple systems, however, recently burned habitats whereas foliage gleaners pre­ there is a great deal of variability in these trajectories ferred unburned habitats. As with other studies, virtu- due to the effects of pathogens and increasingly due to ally all of the 'associated papers evaluated effects within novel weather patterns owing to climate change. For EFFECTS OF DISTURBANCE AND SUCCESSION ON WILDLIFE HABITAT 149 example, Rehfeldt et al. (2009) showed that histori­ phenomena across broad scales of space and time are cally novel weather patterns, but patterns that are pre­ necessarily Gleasonian, whereas fields that study small, dicted to become increasingly common in the future, local phenomena tend to be Clemensian (Horta! et al. immediately preceded major aspen (Po pulus tremuloi­ 2012). The effects of rapid directional climate change des) die-offs in Colorado. on this debate are to shrink the spatiotemporal domai n One critical aspect of successional thinking is that, across which constancy is a reasonable conceit. We are although the initiation conditions might vary as would, relatively sure that climates will change significantly to a certain extent, the trajectory toward climax con­ over the next several centuries and that climate change ditions, the climax conditions are invariant. As Horn is already affecting ecosystems, but our understandings (1975) noted, an invariant steady state is a property of both the details of local climate change and, even of stochastic systems with constant transition proba­ more critically, the eHects of these changes on habitat bilities (Markov chains). Usher (1979) observed that and the wildlife that these habitats support are vague if succession is described as a Markov process, then (Walther et al. 2002). This uncertainty is obviously in­ transition probabilities are likely not constant and will creased by the continuous entry of additional exotic be affected by many fac tors such as population pro­ species whose interactions with current ecosystems cesses. Cadenasso et al. (2002), for example, showed are largely unknown. There is a school of thought that that levels of mammal herbivory play a key role in believes that these specifics, when taken together, are early successional trajectories of old fields. With many important enough to render the entire concept of suc­ exotic species entering ecosystems, the probability of cession suspect (Ricklefs 2008). unforeseen biological feedback obviously increases. For example, the successional trajectories of western Wildlife Research on the Effects North American grasslands to disturbance have been of Succession radically altered by the introduction of a plethora of ex­ otic plant species. In western Montana, spotted knap­ The overwhelming number of disturbance papers that weed (Centaurea maculosa) has occupied vast areas to measure immediate postdisturbance responses and the detriment of both native bunch grasses and native the associated lack of longer-term studies dictates that forbs (Callaway et al. 1999). In an effort to control the direct study of the site-level effects of succession knapweed, exotic gall flies were introduced. They failed are sparse to nonexistent. One of the few cases where to control the knapweed but instead became super­ long-term studies on primary successional dynamics abundant, leading to a variety of secondary trajectories are ongoing is within the area affected by the pyroclas­ involving increased densities of deer mice and associ­ tic explosion of Mount Saint Helens (e.g., Spear eta!. ated hantavirus (Pearson and Callaway 2006). These 2012). Because large-scale population-level studies of grass systems continue to be inundated by new exotics the effects of succession on wildlife are difficult and (e.g., leafy spurge [Euphorbia esula ]) and with climate seldom done, most of our understandings of forest change, other species, such as yellow star-thistle (Cen­ succession on wildlife are associated with habitat use taurea solstitialis ), common in adjacent states, may gain data. For example, we know that some species have foothold. strong associations with certain habitats and that these For a system to move in an orderly manner from dis­ habitats are related to disturbance. The Florida sand turbance toward a steady-state climax, both the species skink (Ples tiodon reynoldsi), for example, is obligately involved and the climatic background need to remain associated with Florida scrub habitat (Schrey et al. constant. This strong requirement for constancy was 2011), a cover type that requires high-intensity fire on first challenged by Gleason (1927), who argued that a a five-to-eighty-year time scale (Laessle 1958). Know­ plethora of factors lead to unique postdisturbance tra­ ing this habitat association, we can infer the relation­ jectories. In the eighty-three years that have elapsed ship between this species and succession. Given this since Gleason's paper was published, the use and accep­ constraint- that we understand habitat relationships tance of successional concepts has split along disciplin­ and know the disturbance regime associated with a ary lines: disciplines like phytogeography that evaluate particular vegetative community-there is a vast lit- 150 RESEARCH AND CONSERVATION erature on the effects of succession on wildlife (see reasonable to ask whether it is a necessary approach. Hunter and Schmiegelow 2011 for a recent review). We could, rather than seek to quantify the specific Interestingly, a great deal of this work focuses on the effects of specific disturbances, endeavor to design two ends of the successional spectrum: early succes­ studies that would allow comparison and generality sional conditions associated with the immediate af­ between disturbances. In the case of hurricanes on the termath of disturbances and the study of a variety of eastern coast of the United States, for example, rather organisms that are dependent on old forests. This may than studying the effects of Hugo, one could ask how be because organisms of interest are associated with much the immediate postdisturbance conditions re ­ these two seres. For ungulates, likely the most heavily lated to Hugo resemble those of other recent strong studied mammal group, the general understanding is "Cape Verde-type" hurricanes that made landfall on that early sera! environments are preferred due to both the southeastern coast of the United States (e.g., Fran the encouragement of forage and increased forage nu­ (1996]; Floyd (1998]; Isabel (2003]). One paper, Rit­ tritional content (e.g., Hobbs and Spowart 1984), and tenhouse et al. (2010), attempted to place the effects a great deal of focus is therefore on the generation and of hurricanes in a broader context, including both maintenance of these conditions. Similarly, because immediate poststorm effects and overall changes in very old forests are in decline worldwide, dependent species composition for birds. Rittenhouse used the species often follow this trend and may become endan­ North American Breeding Bird Survey (BBS; Link and gered. Species such as the spotted owl (Strix occiden­ Sauer 1998) to infer species abundance patterns for talis) and red-cockaded woodpecker (Picoides borealis) the period 1967-2005. This period included a number have become iconic in this regard (Bart and Forsman of hurricane landfalls, notably Hurricane Hugo (Sep­ 1992; Ligon et al. 1986). These older forests are not tember 1989). Study areas were chosen based on the only in decline but often are associated with higher intersection of hurricane tracks and a time series of levels of biological diversity when compared to earlier annual or biennial satellite imagery. The main effect of seres (e.g., Diaz et al . 2005). The emphasis on early hurricanes was seen in the year immediately following sera! and old growth forests is also due to a perceived the hurricane, where modest decreases in abundance overabundance of mid-aged forests owing to a variety of and increases in species richness were similar across human activities. Under natural disturbance regimes, focal areas. Interestingly, although hurricanes caused forest age structures frequently follow negative expo­ a threefold increase in the extent of disturbed forest nential or Weibull distributions (Van Wagner 1978), within the focal areas, this factor was associated with or distributions containing significant areas of both changes in community similarity for only three of the old forest and areas that have recently been disturbed. thirteen avian groups examined and was not associated Conversely, human-caused disturbance generally leads with patterns of avian abundance or species richness. to truncated age distributions and can simultaneously Rittenhouse et al. (2010) attribute the lack of rela­ decrease early sera! conditions through activities such tionships between disturbance levels and population as fire suppression (e.g., Betts et al. 2010). responses to landscape scale heterogeneity; refuges were present in areas where intact forests were the pre­ disturbance land cover. This understanding is similar Disturbance Viewed in a Landscape to that of Brotons et al. (2005), who found that local Context: Studies of Both Disturbance abundances in areas exterior to disturbances affected and Succession postdisturbance abundances within the disturbed In the case of black-backed woodpeckers, postfire con­ area. Thus, to understand postdisturbance trajectories, ditions are clearly essential habitat, but for many other evaluating the larger landscape may be essential. While species these environments are rare and noncritical. Rittenhouse et al. (2010) were able to analyze data at However that may be, based orr this review, it is clear spatial and temporal scales sufficient to infer general­ that targeted studies of disturbance occur almost ex­ ized relationships, studying disturbances at this scale clusively during the immediate aftermath of the event. is difficult and was only possible in this case because While it is clear that this has been our approach, it is of the presence of the BBS, which is relatively unique EFFECTS OF DISTURBANCE AND SUCCESSION ON WILDLIFE HABITAT 151 both in its longevity and in its spatially comprehensive­ clear. In some cases, low-severity fire appears to have ness. However, even though Rittenhouse eta!. (2010) little effect (Roberts et al. 2011). Clark eta!. (2013) were able to apply forty years of BBS data, because hur­ associated negative effects with fire, but sites they ricanes occurred sequentially within the time period, studied had been burned and salvage logged, and pre­ statistical statements were limited to a period within fire timber harvest had occurred. Due to sample size ten years after disturbance. constraints, Clark et al . (2013) could not separate an­ A rare example of a meta-analysis of both distur­ thropogenic and natural disturbance factors. Further, bance and succession is found in Schieck and Song while Franklin et al. (2000) found short-term negative (2006). They performed a meta-analysis of studies on effects, the heterogeneous fire-generated landscape boreal forest birds for a period of 0 - 125 years after of northwestern California featured high reproduc­ disturbance. Interestingly, they found that while bird tive and survival rates (Franklin eta!. 2000)-the fire communities changed after a disturbance, composition mosaic in general appeared to produce excellent owl changes were largely associated with numbers of oc­ habitat. In short, the relationship between spotted owls currences rather than actual changes in species com­ and disturbance is complex and our understandings position. All species with measurable counts (greater are still very nebulous. Using derived information to than five) were present in all forest types. The ability of conserve spotted owls into the future is also compli­ organisms to persist in all forest types and conditions cated by the twin factors of climate change and exotic obviates the need to colonize areas except for after the species introduction. Owl demographics are sensitive most extreme events. It also suggests that immediate both to regional climate and local weather (Glenn et al . postfire abundance studies may not provide informa­ 2010). Further, fire frequency is anticipated to increase tive data on the landscape-level responses of popula­ in the Pacific Northwest with climate change (Mc­ tions to disturbance and succession. Kenzie et al. 2004 ). Lastly, the invasion of the :barred owl (Strix varia) is having a profound negative effect on spotted owl demographics across its range (Dug­ Species Relationships to Disturbance and ger et al. 2011), fundamentally changing the expected Succession: The Case of the Spotted Owl population responses to specific vegetative patterns. In Given the current approach of conducting largely inde­ short, the spotted owl provides a case study in all of pendent and descriptive studies of postdisturbance or­ the complexities highlighted in this paper. The effects ganism response, it is reasonable to ask how well these of disturbance are complex, depending both on time studies composite into generalized understandings fo r since disturbance and the scale at which the question any organism. The spotted owl, being one of the most is asked. Effects of any specific type of disturbance are heavily studied species, provides a good example of the difficult to quantify because, at the scale of spotted owl difficulties associated with determining the relation­ home ranges, a variety of disturbance types have often ships between an organism and disturbance while ap­ occurred: fire, logging, and postfire salvage. Lastly, our plying this methodology. As mentioned previously, the understandings concerning the effects of postdistur­ spotted owl has become an iconic organism represent­ bance succession on the quality of spotted owl habitat ing an old forest obligate (Bart and Forsman 1992), and have been fundamentally altered by the invasion of an for this reason, the long-term viability of owls in fi re­ exotic competitor. prone forests has been questioned (Spies et al. 2006). However, spotted owls exist in a wide variety of for­ Approaches to the Problem ests, including some that are highly modified (Thome eta!. 1999). Short-term relationships to anthropogenic It is clear that our traditional approaches, involving disturbance have been highly studied, with uniformly primarily small-scale, short-duration descriptive stud­ negative relationships (e.g., Fors~an et al. 1984; Frank­ ies within disturbed areas (and sometimes including lin eta!. 2000), except where the dominant prey are undisturbed control areas), while producing much woodrats (Neotoma fusc ipes; Thome et a!. 1999). Fire knowledge of local importance, do not provide a co­ is much less heavily studied, and the effects are less herent approach to increasing our knowledge of the 152 RESEARCH AND CONSERVATION more general effects of disturbance and succession on collect appropriate data to assess biological processes wildlife. It is equally clear that the traditional approach across broader spatiotemporal domains. Because of the of trading space for time to study succession is becom­ high level of scientific design associated with targeted ing less and less tenable due to the increasing pool of monitoring, intermediate short-term products can exotic organisms and the directional nature of climate both increase the immediate utility of the monitoring change. What is less clear is how to move forward in and provide feedback on its overall efficacy. Secondly, a manner that leads to more rapid increases in our monitoring efforts should take advantage of new tech­ knowledge of these processes while existing in a world nologies and, specifically, genetic sampling (Schwartz that is in a rapid state of biological flux. The simple et al. 2007). Not only do these methods often provide answer is to expand our studies in both space and time. the least expensive approach to sampling species (e.g., This would allow the formalization of the context in Kendall and McKelvey 2008), but genetic data are which disturbance occurs, including the effects of nearly unique in that, if the samples are retained, these landscape pattern and habitat juxtaposition, and would samples can be reanalyzed at a later date. Even without avoid invoking constancy assumptions to infer future the formal retention of genetic data, the past collection states. There are, however, major problems associated of materials that contain genetic data has allowed ret­ with this approach. The first is cost. Increasing the rospective studies of populations. For example, Miller size and duration of a study quadratically increases the and Waits (2003) were able to retrospectively calculate costs. Further, funding large, long-term projects is diffi­ the effective population size of grizzly bears in Yellow­ cult given the competitive grant paradigm under which stone National Park by analyzing 110 museum speci­ a large proportion of science is funded. Competitive mens collected between 1912 and 1981. Because of this grants generally have both spending and time limits ability to retrospectively analyze data collected in the that preclude their direct utility for long-term fund­ past, genetically based monitoring is less depe!"ldent on ing; long-term studies are often patched together by current technologies than are most other m6nitoring acquiring multiple grants, often from different sources. approaches. With genetic data, new technologies can Additionally, there are many ways that these programs be applied to extant samples, allowing a fully modern can fail and many examples of historical fai lures due analysis of all data regardless of collection time. This to funding drying up, data loss due to inadequate ar­ ability also provides a safeguard against data loss and chiving, failures to maintain quality control across time allows post-hoc testing quality controls: assays can be and space, and poor initial designs leading to the collec­ repeated both to recover data and to ensure that the tion of large quantities of relatively useless data. Lastly, original analyses were correct. there is the strong, but often unacknowledged, power of serendipity associated with many independent stud­ Conclusions ies. The ability to capture serendipity is decreased if scientific resources divert from a large number of There has been little to no attempt in the literature to small, highly independent studies and commit to a view disturbance in terms of the operational environ­ small number of highly directed, long-term projects. ment of the studied species; we define disturbances There are, however, approaches that merge the at anthropocentric scales. While this is biologically fields of science and monitoring and which may prove problematic, it is so ubiquitous that it is impossible to useful in documenting landscape-level changes asso­ write a review article without also adopting this con­ ciated with disturbance and subsequent succession. ceit. However, given this constraint, there is a vast body Targeted monitoring (Yoccoz et a!. 2001; Nichols and of literature associated with both the effects of distur­ Williams 2006) combines principles of scientific de­ bance and succession on wildlife. However, due both sign including clearly articulated goals and specified er­ to practicalities and interest levels, the vast majority ror levels with the large-scale; long-term requirements of specific studies on the effects of disturbance focus for biological monitoring. Increasing the proportion of on the period immediately postdisturbance; few stud­ monitoring that is specifically targeted could allow a ies extend beyond five years postdisturbance. Further, reallocation of extant long-term funding into efforts to most studies are small in scope and are not designed to EFFECTS OF DISTURBANCE AND SUCCESSION ON WILDLIFE HABITAT 153 test concepts; most generalizations are based on com­ sociated wildlife communities were fo llowing expecta­ positing these small-scale primary studies through lit­ tions as well as when and where systems were changing erature reviews and meta-analyses, of which there are in unexpected ways. many. However, as the case of the spotted owl demon­ strates, it is difficult, even with many largely unrelated LITERATURE CITED small-scale studies of specific areas, to generate coher­ Allen, C. D., A. K. Macalady, H. Chenchouni, D. Bacbelet, ent understandings of general relationships between N. McDowell, M. Vennetier, T. Kitzbe.rger, A. Rigling, D. D. an organism and its disturbance environment. Succes­ Breshears, E. H. Hogg, P. Gonzalez, R. Fensham, Z. Zhang, sional studies generally trade space for time, assuming ]. Castro, N. Demidova, J,·H. Lim, G. Allard, S. W. Runnjng, that patterns associated with older habitats generated A. Semerci, and N. Cobb. 2010. "A Global Overview of by past disturbances will indicate future patterns in Drought and Heat-Induced Tree Mortality Reveals Emerg· ing Cli mate Change Risks fo r Forests." Forest Ecology and newly disturbed areas. While this practice is directly re­ Management 259:660- 84. lated to the paucity of longer-term disturbance-related Bart, ]., and E. D. Forsman. 1992. "Dependence of Northern studies, it is also a necessity as many of the successional Spotted Owls Strix occidental is caurina on Old-Growth processes we care about occur across hundreds of years. Forests in the Western USA:' Biological Conservation This approach has historically proven to be extremely 62:95 - 100. Betts, M.G., ]. C. Hagar,]. W. Rivers, ]. D. Alexander, K. useful, but it contains a variety of hidden assumptions Mcgarigal, and B. C. Mccomb. 2010. "Thresholds in Forest that are unlikely to be met as the rate of biological Bird Occurrence. as a Function of the Amount of Early-Sera! change increases worldwide. Broadleaf Forest at Landscape Scales." Ecological Applica· We are currently seeing forest mortality at levels tions 20:2116-30. unprecedented in recent history (Allen et al. 2010). Bormann, F. H., and G. E. Likens. 1979. Pauern and Process in a Forested Ecosystem: Disturbance, Development and the Steady These large-scale disturbances provide the opportu­ State. Springer-Verlag, New York. nity for ecosystems to evolve to meet climatic exigen­ Brotons, L., P. Pons, and S. Herrando. 2005. "Colonization cies and are the engine of predicted broad-scale biome of Dynamic Mediterranean Landscapes: Where Do Birds shifts. Given this, we should not assume that relation­ Come from After Fire?" Journal of Biogeography 32:789-98. ships based in large part on disturbance events in the Cadenasso, M. L., S. T. A. Pickett, and P. ]. Morin. 2002. "Ex· past will necessarily provide adequate guidance when perimental Test of the Role of Mammalian Herbivores on Old Field Succession: Community Structure and Seedling predicting the ecological trajectories associated with Survival." journal of the Torrey Botanical Society 129:228-37. current processes. Cal laway, R. M., T. H. Deluca, and W. M. Belliveau. 1999. Rittenhouse et al. (2010) demonstrated that the "Biological-Control Herbivores May Increase Competitive utility of broad-scale temporally stable monitoring for Ability of the Noxious Weed Centaurea maculosa." Ecology evaluating the effects of disturbance on wildlife com­ 80:1196- 1201. Caton, E. L. 1996. "Effects of Fire and Salvage Logging on the munities and monitoring efforts such as BBS wil l most Cavity Nesting Bird Community in Northwestern Mon· likely become increasingly important as climate change tana." PhD dissertation, University of Montana, Mjssoula, and exotic species produce increasingly novel ecosys­ Montana. tems (Hobbs et al. 2009); we should support these ef­ Clark, D. A., R. G. Anthony, and L. S. Andrews. 2013. "Rela· forts and seek to increase the number of taxa that are tionship between Wildfire, Salvage Logging, and Occupancy monitored. However, these sorts of generalized surveys of Nesting Territories by Northern Spotted Owls." journal of Wildlife Management. doi : 10.1002/jwmg.523. will always be weak for evaluating specifics of distur­ Clements, F. E. 1916. Plant Succession: An Analysis of the De­ bance and succession. Given the pivotal role distur­ velopment of Vegetation. Carnegie Institute of Washington, bance plays in ecological change, more specific focus Washington D. C. is probably warranted. To meet this need, we can de­ Covington, W. W. , and M. M. Moore. 1994. "Southwestern sign targeted monitoring studies specifically designed Ponderosa Forest Structure: Changes Since Euro- American Settlement." Journal of Forestry 92:39-47. to assess the effects of disturbance and succession on Daubenmire, R. 1952. "Forest Vegetation of Northern Idaho wildlife populations. ideally, these would represent a and Adjacent Washington, and Its Bearing on Concepts redirection of existing monitoring efforts, allowing us of Vegetation Classification." Ecological Mon ographs 22: to ascertain whether successional trajectories and as- 301-30. 154 RESEARCH AND CONSERVATION

Dlaz, I. A.,].]. Armesto, S. Reid, K. E. Sieving, and M. F. Horn, H. S. 1975. "Markovian Properties of Forest Succession." Willson. 2005. "Linking Forest Structure and Composition: In Ecology and Evolution of Communities, edited by M. L. Avian Diversity in Successional Forests of Chiloe Island, Cody and]. M. Diamond, 196- 211. Belknap Press of Har­ Chile:' Biological Conservation 123:91-101. vard University Press, Cambridge, Massachusetts. Dudley, ]. G., V. A. Saab, and ]. P. Hollenbeck. 2012. "Foraging­ Horta!, J., P. De Marco Jr., A. M. C. Santos, and J. A. F. Habitat Selection of Black-Backed Woodpeckers in Forest Diniz-Filho. 2012. "Integrating Biogeographical Processes Burns of Southwestern Idaho." The Condor 114:348-57. and Local Community Assembly." Journal of Biogeography Dugger, K. M., R. G. Anthony, and L. S. Andrews. 2011. 39:627-28. "Transient Dynamics of Invasive Competition: Barred Owls, Hossack, B. R., W. H. Lowe, and P. S. Corn. 2013. "Rapid Spotted Owls, Habitat, and the Demons of Competition Increases and Time-Lagged Declines in Amphibian Occu­ Present." Ecological Applications 21:2459- 68. pancy after Wildfire." Conservation Biology 27:219-28. Fontaine, ]. B., and P. L. Kennedy. 2012. "Meta-analysis of Hunter Jr., L. , and F. K. A. Schmiegelow. 2011. Wildlife, Forests Avian and Small-Mammal Response to Fire Severity and and Forestry: Principles of Managing Forests for Biological Fire Surrogate Treatments in U.S. Fire-Prone Forests." Diversity. 2nd edWon. Prentice Hall, Upper Saddle River, Ecological Applications 22:1547-61. New Jersey. Forsman, E. D., E. C. Meslow, and H. M. Wight.1984. "Distri­ Kalies, E. L. , C. L. Chambers, and W. W. Covington 2010. bution and Biology of the Spotted Owl in Oregon." Wi.ldlife "Wildlife Responses to Thinning and Burning Treatments Monographs 87:1- 64. in Southwestern Conifer Forests: A Meta-analysis." Forest Forsman, ). T., P. Reunanen, ]. )okimaki, and M. Miinkkonen. Ecology and Management 259:333- 42. 2010. "The Effects of Small-Scale Disturbance on Forest KendaLl, K. C. , and K. S. McKelvey. 2008. "Hair Collection." In Birds: A Meta-analysis." Canadian journal of Forest Research Noninvasive Survey Methods for North American Carnivores, 40:1833- 42. edited by R. A. Long, P. MacKay,). C. Ray, and W. ]. Zielin­ Fox, B. ]. 2011. "Review of Small Mammal Trophic Structure ski,l41-82. Island Press, Washington D.C. in Drylands: Resource Availability, Use, and Disturbance." Kotliar, N. B. , and ]. A. Wiens. 1990. "Multiple Scales of Patchi­ Journal of Mammalogy 92:1179- 92. ness and Patch Structure: A Hierarchical Framework for Franklin, A. B. , D. R. Anderson, R. ]. Gutierrez, and K. P. the Study of Heterogeneity." Oikos 59:253- 60. Burnham. 2000. "Climate, Habitat Quality, and Fitness in Kuchler, A. W. 1964. "Potential Natural Vegetation of the Northern Spotted Owl Populations in Northwestern Cali­ Conterminous United States." American Geographical Society fornia." Ecological Monographs 70:539- 90. Special Publication 36:1-54. Franklin,). F., and R. T. T. Forman. 1987. "Creating Landscape Laessle, A. M. 1958. "The Origin and Successional Relation­ Patterns by Forest Cutting: Ecological Consequences and ship of Sandhill Vegetation and Sand-Pine Scrub." Ecologica l Principles." Landscape Ecology 1:5-18. Monograph s 28:361 - 87. Gleason, H. A. 1927. "Further Views on the Succession­ Leidolf, A., and}. A. Bissonette. 2009. "The Effects of Fire on Concept." Ecology 8:299-326. Avian Communities: Spatia-Temporal Attributes of the Glenn, E. M., R. G. Anthony, and E. D. Forsman. 2010. "Popu­ Literature, 1912-2003." International Journal of Wildland lation Trends in Northern Spotted Owls: Associations with Fire 18:609- 22. Climate in the Pacific Northwest." Biological Conservation Levins, R. 1969. "The Effects of Random Variation of Different 143:2543-52. Types on Population Growth." Proceedings of the National Iverson, G. C., P. A. Vohs, and T. C. Tacba. 1987. "Habitat Use Academy of Sciences 62:1061 - 65. by Mid-Continent Sandhill Cranes during Spring Migra­ --. 1970. "Extinction." In Some Math ematical Questions tion." journal of Wildlife Management 51:448-58. in Biology, edited by M. Gerstenhaber, 77-107. American Habeck, ). R., and R. W. Mutch. 1973. "Fire-Dependent Forests Mathematical Society, Providence, Rhode Island. in the Northern Rocky Mountains." Quaternary Research Ligon, J. D., P. B. Stacey, R. N. Conner, C. E. Bock, and C. S. 3:408- 24. Adkisson. 1986. "Report of the American Ornithologists' Hartway, C., and L. S. Mills. 2012. "A Meta-analysis of the Union Committee for the Conservation of the Red­ Effects of Common Management Actions on the Nest Cockaded Woodpecker." Th e Auk 103:848- 55. Success of North American Birds." Conservation Biology Link W. A., and]. R. Sauer. 1998. "Estimating Population 26:657-66. Change from Count Data: Application to the North Hobbs, N. T., and R. A. Spowart. 1984. "Effects of Prescribed American Breeding Bird Survey." Ecological Applications Fire on Nutrition of Mountain Sheep and Mule Deer 8:258-68. during Winter and Spring:· Jour~al of Wildlife Management Lodge, D. M. 1993. "Species Invasions and Deletions: Commu­ 48:551- 60. nity Effects and Responses to Climate and Habitat Change." Hobbs, R. )., E. Higgs, and). A. Harris. 2009. "Novel Ecosys­ In Biotic Intera ctions and Global Change, edited by P.M. tems: Implications for Conservation and Restoration." Kareiva, }. G. Kingsolver, and R. B. Huey, 367-87. Sinauer Trends in Eoology and Evolution 24:599- 605. Associates, Sunderland, Massachusetts. EFFECTS OF DISTURBANCE AND SUCCESSION ON WILDLIFE HABITAT 155

Martell, K A. , A. L. Foote, and S. G. Cumming. 2006. "Ripar­ Ryki el Jr., E.]. 1985. "Towards a Definition of Ecological Di s­ ian Disturbance Due to Beavers (Castor canadensis) in turbance." Australian Journal of Ecology 10:361-65. Alberta's Boreal Mixed wood Forests: Implications for Forest Saab, V. A. , and H. D. W. Powell. 2005. "Fi.re and Avian Ecology Management." Ecoscience 13:164-71. in North America: Process Influencing Pattern." Studies in McKenzie, D., Z. Gedalof, D. L. Peterson, and P. Mote. 2004. Avian Biology 30:1-13. "Climatic Change, Wildfire, and Conservation." Conserva­ Saab V. A. , R. E. Russell, and ). G. Dudley. 2007. "Nest Densi­ tion Biology 18:890-902. ties of Cavity-Nesting Birds in Relation to Postfi re Salvage Meeker Jr., D. 0., and D. L. Merkel. 1984. "Climax Theories Logging and Time Since Wildfire." Th e Condor 109:97-108. and a Recommendation for Vegetation Classification: A Schieck, ]., and S. ). Song. 2006. "Changes in Bi rd Communi­ Viewpoint." Journal of Range Management 37:427-30. ties th.roughout Succession Following Fire and Harvest in Miller, C. R., and L. P. Waits. 2003. "The History of Effective Boreal Forests of Western North America: Literature Re­ Population Si ze and Genetic Diversity in the Yellowstone view and Meta-analyses." Canadian Journal of Forest Research Grizzly (Urs us arctos): Implications for Conservation." Pro­ 36:1299-1318. ceedings for the National Academy of Science 100:4334-39. Sch.rey, A. W. , A.M. Fox, H. R. Mushinsky, and E. D. Mccoy. Monamy, V., and B. ). Fox. 2000. "Small Mammal Succession 2011. "Fire Increases Variance in Genetic Characteristics of is Determined by Vegetation Density Rather Than Time Florida Sand Skink (Plestiodon reynoldsi) Local Populations." Elapsed Since Disturbance." Austral Ecology 25:580-87. Molecular Ecology 20:56- 66. Naiman, R. )., ). M. Melillo and J. E. Hobbie. 1986. "Ecosys­ Schwartz, M. K., G. Luikart, and R. S. Waples. 2007. "Genetic tem Alteration of Boreal Forest Streams by Beaver (Castor Monitoring as a Promising Tool for Conservation and Man­ canadensis)." Ecology 67:1254-69. agement." Trends in Ecology and Evolution 22:25-33. Nichols, ]. D., and B. K. Williams. 2006. "Monitoring for Con­ Shafer, A. B. A. , C. I. Cullingham, S. D. Cote, and D. W. Colt­ servation." Trends in Ecology and Evolution 21:668-73. man. 2010. "Of Glaciers and Refugi a: A Decade of Study Nowacki, G. )., and M. D. Abrams_2008. "The Demise of Sheds New Light on the Phytogeography of Northwestern fire and 'Mesophication' of Forests in the Eastern United North America." Molecular Ecology 19:4589- 4621. States." BioScience 58:123-38. Sherry, T. W. , and R. T. Holmes. 1988. "Habitat Selection by Pearson, D. E. , and R. M. Callaway. 2006. "Biological Control Breeding American Redstarts in Response to a Domiy.ant Agents Elevate Hantavirus by Subsidizing Deer Mouse Competitor, the Least Flycatcher." The Auk 105:350-64. Populations." Ecology Letters 9:443-50. Sousa, W. P. 1984. "The Role of Disturbance in Natural Pearson, D. E., Y. K. Ortega, K. S. McKelvey, and L. F. Ruggiero. Communities." Annual Review of Ecology and Systematics 2001. "Small Mammal Community Composition, Relative 15:353-91. Abundance, and Habitat Selection in Native Bunchgrass Spear, S. F., C. M. Crisafull, and A. Storfer. 2012. "Genetic of the Northern Rocky Mountains: Implications for Exotic Structure among Coastal Tailed Frog Populations at Mount Plant Invasions." Northwest Science 75:107-17. St. Helens is Moderated by Post-disturbance Management:' Pfister, R. D. , and S. F. Arno. 1980. "Classifying Forest Habitat Ecological Applications 22:856-69. Types Based on Potential Climax Vegetation." Forest Science Spies, T. A., M. A. Hemstrom, A. Youngblood, and S. Hum­ 26:52-70. mel. 2006. "Conserving Old-Growth Forest Diversity Rehfeldt, G. E., D. E. Ferguson, and N. L. Crookston. 2009. "As­ in Disturbance-Prone Landscapes." Conservation Biology pen, Climate, and Sudden Decline in Western USA." Forest 20:351-62. Ecology and Management 258:2353-64. Thome, D. M. , C. ). Zabel, and L. V. Diller. 1999. "Forest Stand Ricklefs, R. E. 2008. "Disintegration of the Ecological Com­ Characteristics and Reproduction of Northern Spotted munity." The American Naturalisr.172:741-50. Owls in Managed North-Coastal California Forests." Journal Rittenhouse, C. D., A. M. Pidgeon, T. P. Albright, P. D. Culbert, of Wildlife Management 63:44- 59. M. K. Clayton, C. H. Flather, C. Huang, J. G. Masek, and Tremblay,]. A.,]. Ibarzabal, and J.-P. L. Savard. 2010. "Foraging V. C. Radeloff. 2010. "Avifauna Response to Hurricanes: Ecology of Black-Backed Woodpeckers (Picoides arcticus) in Regional Changes in Community Similarity." Global Change Unburned Eastern Boreal Forest Stands." Canadian Journal Biology 16:905-17. of Forest Research 40:991-99. Roberts, S. L. , ). W. van Wagtendonk, A. K. Miles, and D. A. Usher, M. B. 1979. "Markovian Approaches to Ecological Suc­ Kelt . 2011. "Effects of Fire on Spotted Owl Site Occupancy cession." Journal of Animal Ecology 48:413-26. in a Late-Su ccessional Forest:' Biological Conservation Vanderwel, M. C.,]. R. Malcolm, and S. C. Mill s. 2007. 144:610- 19. "A Meta-analysis of Bird Responses to Uniform Partial Russell, R. E., ]. A. Royle, V. A. Saab, ). _F. Lehmkuhl, W. M. Harvesting across North America." Conservation Biology Block, and]. R. Sauer. 2009. "Modeling the Effects of 21:1230- 40. Environmental Disturbance on Wildlife Communities: Van Wagner, C. E. 1978. "Age-Class Distribution and the Forest Avian Responses to Prescribed Fire." Ecological Applications Fire Cycle:' Canadian Journal of Forest Research 8:220- 27. 19:1253-63. Walther, G. -R., E. Post, P. Convey, A. Menzel, C. Parmesan, 156 RESEARCH AND CONSERVATION

T. J. C. Beebee, ).·M. Fromentin, 0. Hoegh·Guldberg, and Zerbe, S. 1998. "Potential Natural Vegetation: Validity and F. Bairlein. 2002. "Ecological Responses to Recent Climate Applicability in Landscape Planning and Nature Conserva· Change." Nature 416:389-95. tion." Applied Vegetation Science 1:165-72. Wiens, J. A. 1989. "Spatial Scaling in Ecology." Functional Ecol· Zwolak, R. 2009. "A Meta-analysis of the Effects of Wildfire, ogy 3:385-97. Clearcutting, and Partial Harvest on the Abundance of Yoccoz, N. G., J. D. Nichols, and T. Boulinier. 2001. "Monitor· North American Small Mammals." Forest Ecology and Man· ing of Biological Diversity in Space and Time." Trends in agement 258:539-45. Ecology and Evolution 16:446-53.