Search

Non-native plant invasions are often catalyzed by wildfires and may alter the structure, function, and composition of ecosystems. Of particular concern are non-native annual grasses that promote increased fire frequency and compete with native vegetation for early season moisture. While these effects are well studied in non-forested ecosystems, relatively few studies document invasions of non-native annual grasses in forested ecosystems. We document non-native plant invasion following repeated high-severity wildfire in mixed-conifer/hardwood forests dominated by Douglas-fir (Pseudotsuga menziesii) in the Klamath Mountains of northern California. Despite great concern regarding the loss of conifer dominance and transitions to non-forested states related to recent and projected increases in wildfire activity, there is little known about invasions of non-native plant species following fire in this region. Non-native plant species capable of long-distance dispersal had an average total cover of 26% and were ubiquitous across a systematic, gridded sample of twenty 800 m2 plots. Non-native forbs included multiple species of common post-disturbance invaders (e.g. Cirsium vulgare, Senecio sylvaticus, Lactuca serriola). Non-native annual grasses comprised more of the vegetative cover than non-native forbs (17% vs. 9%), and were dominated by two species, Aira caryophyllea and Vulpia myuros, which reached as high as 80% cover. Our findings indicate that non-native annual grasses are a novel stressor that may accelerate projected loss of conifer dominance in the Klamath Mountains by facilitating fire spread and increasing competition for early season soil moisture. Additional monitoring will be crucial to understanding the impacts of non-native plants as fire activity in this region continues to increase.

We compared early seral development between stands subject to single and repeated high-severity wildfire in low-elevation, mixed conifer – hardwood forests in the Klamath Mountains, California, USA. We used a before–after, control–impact (BACI) approach to assess changes in the density of conifer regeneration and the cover of multiple components of vegetation structure (conifers, hardwoods, shrubs, forbs, and graminoids) and compare pathways of seral development between plots that burned once and plots that burned twice. Fifty-three field plots were established 6 years following a high-severity fire in 2004. Nineteen of these plots experienced a second high-severity wildfire 11 years later (2015), and all plots were remeasured in 2016–2017. Conifer regeneration was abundant following the first fire but was greatly reduced by the second fire. Plots that did not reburn increased in conifer, hardwood, and shrub cover, whereas plots that reburned increased in forb cover and decreased in shrub, hardwood, and conifer cover. Despite conifer loss, we found little evidence of shifts to nonforested states following repeated fire due to resilience of resprouting hardwoods. Our results indicate that repeated high-severity fire has the potential to protract early seral development and catalyze transitions from mixed conifer – hardwood forest to hardwood-dominated early seral conditions.

Tree mortality is of considerable concern, but the magnitude and extent of forest canopy decline are relatively unknown in landscapes with high levels of topographic complexity, spatial heterogeneity, and species diversity. We assessed 30 years of canopy decline, including a 5-year period characterized by extreme drought, in one of North America’s most diverse landscapes in the Klamath Mountains of northern California, USA. (1) Characterize tree mortality by species, (2) Quantify temporal and spatial patterns of remotely-sensed canopy decline in relation to climate, (3) Compare canopy decline among vegetation types, topographic settings, and stand structural classes during drought. We characterized tree mortality by species with field data and quantified the role of climate on canopy decline with a 30-year Landsat time series. We assessed and compared the role of topography and stand structure on canopy decline during drought. Most tree mortality and canopy decline occurred at higher elevations in Shasta red fir (Abies magnifica var. shastensis) and subalpine forests. Annual area of canopy decline was positively correlated with summer temperature and minimum vapor pressure deficit but not precipitation. The area of canopy decline was three times greater during the drought. The magnitude of decline was greatest at higher elevations, on more exposed, southwestern slopes, and in stands with old-growth structure. Stands in valleys and low slopes experienced relatively little decline. Our study demonstrates the vulnerability of high elevation, old-growth forests to increasing temperature and suggests the potential for refugia from drought in diverse, heterogeneous landscapes.

There is growing concern that populations of fire-adapted species around the world are at risk of extirpation due to 20th century fire exclusion (i.e. senescence risk). Concurrently, increasing fire activity in many regions is prompting concern that repeated, short-interval fire poses a risk to young stands before trees reach reproductive maturity (i.e. immaturity risk). Conceptual models assessing vulnerability to both risks exist, yet there is a lack of empirically-based studies that assess the effects of anthropogenically-altered fire regimes across the entire range of a fire-adapted species. We used geospatial data on fire occurrence (1900–2015) and a systematic sample of recently re-measured (∼2005–2015) forest inventory plots across the range of a serotinous conifer, knobcone pine (Pinus attenuata), in California and southwestern Oregon, U.S.A. to assess the relative threat of immaturity and senescence risk. We base our assessment on: (1) recent trends in area burned, (2) geographic variability and changes in stand-scale population structure, and (3) patterns of local colonization, extirpation, and population change. The area burned across the species range was lowest during the early and mid-20th century and increased in the early 1980s. Approximately 39% of the range burned between 1984 and 2015 and fire rotation was 81 years for the 25-year period ending in 2015. During this period, approximately 6% of the range of knobcone pine experienced repeated fire, and only 4% of the reburned area (∼0.25% of the entire range) experienced multiple fires at an interval less than 10 years. Most populations of knobcone pine currently exist in a mature or decadent state and we found evidence of recent senescence-related extirpation across approximately 15% of its range. However, we also observed a cumulative expansion of knobcone pine during the last decade, and the rate of colonization (less than half of which was associated with fire) was almost double that of extirpation. Despite high rates of colonization and expansion of knobcone pine in response to recent fires, vulnerability to extirpation varies geographically and remains high in some parts of the species range where wildfire activity has been low. This study provides a rare empirically-based look at the contribution of extirpation and colonization to the range-wide dynamics of a serotinous tree. Our findings underlie the importance of recognizing the different temporal and spatial scales at which contemporary disturbance regimes threaten fire-adapted species, and highlight the potential for alternative pathways of persistence and expansion of serotinous species.

Whitebark pine (Pinus albicaulis. Engelm.) is vulnerable to a number of threats including an introduced pathogen (Cronartium ribicola J.C. Fisch.), epidemic levels of native mountain pine beetle (Dendroctonus ponderosae Hopkins), fire suppression, and climate change. To describe the structure of whitebark pine populations in two national parks in the southern Cascades (Crater Lake, Oregon, USA (CRLA) and Lassen Volcanic, California, USA (LAVO) National Parks), we surveyed trees in 30 &times, 50 &times, 50 m plots in both parks. We used these plots to describe the extent of white pine blister rust (the disease caused by Cronartium ribicola), mountain pine beetle occurrence, and to elucidate factors influencing the presence of pests and pathogens, cone production, and canopy kill. In each plot, we recorded data related to tree health, including symptoms of blister rust and mountain pine beetle, and reproductive vigor (cone production). In both parks, encroachment from other species, particularly mountain hemlock (Tsuga mertensiana (Bong.) Carri&egrave, re), was negatively associated with cone production. In CRLA, water stress was a good predictor of blister rust infection and cone production. For CRLA and LAVO, the presence of mountain pine beetle and blister rust was associated with higher canopy kill for whitebark pine. Lastly, we found evidence for a pest-pathogen interaction, mountain pine beetle attack was greater for trees that showed symptoms of blister rust infection in CRLA. Our results indicate that whitebark pine populations in the southern Cascade Range are experiencing moderate levels of blister rust infection compared with other sites across the species range, and that competition from shade-tolerant species may result in an additional threat to whitebark pine in both parks. We present our findings in the context of park management and situate them in range-wide and regional conservation strategies aimed at the protection and restoration of a declining species.

Pathogens and insect pests have become increasingly important drivers of tree mortality in forested ecosystems. Unfortunately, understanding the relative contributions of multiple mortality agents to the population decline of trees is difficult, because it requires frequent measures of tree survival, growth, and recruitment, as well as the incidence of mortality agents. We present a population model of whitebark pine (Pinus albicaulis), a high-elevation tree undergoing rapid decline in western North America. The loss of whitebark pine is thought to be primarily due to an invasive pathogen (white pine blister rust; Cronartium ribicola) and a native insect (mountain pine beetle; Dendroctonus ponderosae). We utilized seven plots in Crater Lake National Park (Oregon, USA) where 1220 trees were surveyed for health and the presence of blister rust and beetle activity annually from 2003–2014, except 2008. We constructed size-based projection matrices for nine years and calculated the deterministic growth rate (λ) using an average matrix and the stochastic growth rate (λs) by simulation for whitebark pine in our study population. We then assessed the roles of blister rust and beetles by calculating λ and λs using matrices in which we removed trees with blister rust and, separately, trees with beetles. We also conducted life-table response experiments (LTRE) to determine which demographic changes contributed most to differences in λ between ambient conditions and the two other scenarios. The model suggests that whitebark pine in our plots are currently declining 1.1% per year (λ = 0.9888, λs = 0.9899). Removing blister rust from the models resulted in almost no increase in growth (λ = 0.9916, λs = 0.9930), while removing beetles resulted in a larger increase in growth (λ = 1.0028, λs = 1.0045). The LTRE demonstrated that reductions in stasis of the three largest size classes due to beetles contributed most to the smaller λ in the ambient condition. Our work demonstrates a method for assessing the relative effects of different mortality agents on declining tree populations, and it shows that the effects of insects and pathogens can be markedly different from one another. In our study, beetle activity significantly reduced tree population growth while a pathogen had minimal effect, thus management actions to stabilize our study population will likely need to include reducing beetle activity.

Summary Fire-prone woodlands and savannas world-wide face management challenges resulting from fire exclusion and subsequent encroachment of fire-sensitive trees. In the Pacific Northwest (USA), Quercus garryana oak woodlands and savannas are threatened by encroachment from the native conifer Pseudotsuga menziesii in the absence of fire. In the Bald Hills of Redwood National Park (California, USA), prescribed fire and conifer removal have been used to restore encroached woodlands. We examined the effects of encroachment and restoration on understorey vegetation, comparing four treatments: prescribed fire, prescribed fire and conifer removal, conifer removal, and encroached (control). Treatments including prescribed fire had the greatest native species richness. These two treatments also had the greatest non-native species richness, at both the site level and the treatment level. Woodlands treated with conifer removal and no prescribed fire were intermediate in species richness and diversity compared to burned treatments and encroached woodlands. Encroached woodlands had diminished richness and diversity compared to all restoration treatments. Non-metric multidimensional scaling (NMS) ordination demonstrated that conifer basal area, conifer litter and fine wood were associated with low species richness and diversity and that elevation and thatch were associated with higher species richness and diversity. Indicator species analysis identified that most native species and non-native species were associated with burned woodlands that were never encroached. Synthesis and applications. Our results suggest that both prescribed fire and conifer removal have benefits for understorey plant communities, increasing species richness, diversity and cover in oak woodlands and shifting understorey communities from forest-associated species to more woodland-associated species. Restoration of remnant Quercus garryana oak woodlands is complicated by the persistence and abundance of non-native herbaceous plants.

PREMISE OF THE STUDY Serpentine soils have attracted the attention of evolutionary biologists for decades because of their high number of rare and endemic taxa, though less is known about the ecological factors that govern the diversity and composition of serpentine communities. Theory suggests that vegetation on these low-productivity soils will be relatively resilient to fire, the most common natural disturbance in serpentine systems. METHODS We studied the recovery of vegetation in Darlingtonia fens, a unique habitat dominated by herbaceous perennials, from a major fire that burned ∼202,000 ha in California and Oregon's Klamath Mountains in 2002. We established permanent plots in eight unburned and eight burned fens in 2003 and recorded percent cover of vascular plant species. We re-sampled plots each year through 2007. KEY RESULTS Burned fens had less plant cover than unburned fens for 2 yr after the fire. Average species density was ∼10% lower in burned fens 1 yr after the fire but ∼4-8% higher for the next 4 yr. Burned fens exhibited greater evenness but not until 4 yr after the fire. Differences in community composition were detected between the two fen types, but species ranks were similar, and species neither were added to nor removed from the burned assemblages. CONCLUSIONS Burning of Darlingtonia fens has detectable, albeit modest, effects on serpentine communities. Because fens have little or no canopy cover, fire has little influence on light availability in this system. This relatively small resource change, combined with high soil moisture and well-developed underground organs of fen plants, produces a highly resilient assemblage.

Patterns of species occurrence and abundance are infl uenced by abiotic factors and biotic interactions, but these factors are diffi cult to disentangle without experimental manipulations. In this study, we used observational and experimental approaches to investigate the role of temperature and interspecifi c competition in controlling the structure of groundforaging ant communities in forests of the Siskiyou Mountains of southwestern Oregon. To assess the potential role of competition, we fi rst used null model analyses to ask whether species partition temporal and/or spatial environments. To understand how thermal tolerances infl uence the structure of communities, we conducted a laboratory experiment to estimate the maximum thermal tolerance of workers and a fi eld experiment in which we added shaded microhabitats and monitored the response of foragers. Finally, to evaluate the roles of temperature and interspecifi c competition in the fi eld, we simultaneously manipulated shading and the presence of a dominant competitor (Formica moki). Th e foraging activity of species broadly overlapped during the diurnal range of temperatures. Species co-occurrence patterns varied across the diurnal temperature range: species were spatially segregated at bait stations at low temperatures, but co-occurred randomly at high temperatures. Th e decreased abundance of the co-occurring thermophilic Temnothorax nevadensis in shaded plots was a direct eff ect of shading and not an indirect eff ect of competitive interactions. Th ermal tolerance predicted the response of ant species to the shading experiment: species with the lowest tolerances to high temperatures showed the greatest increase in abundance in the shaded plots. Moreover, species with more similar thermal tolerance values segregated more frequently on baits than did species that diff ered in their thermal tolerances. Collectively, our results suggest that thermal tolerances of ants may mediate competitive eff ects in habitats that experience strong diurnal temperature fl uctuations. A major goal of ecology is to identify the processes that con trol patterns of community structure. At large spatial scales, historical processes determine properties of the regional species pool (Ricklefs 2004), which in turn infl uence the composition of local communities (Webb et al. 2002, Wiens and Donoghue 2004). At smaller spatial scales, both abiotic fi ltering (pH, temperature, water availability, etc.), and biotic interactions (interspecifi c competition, parasites, herbivory, etc.) may operate to determine which species occur, and cooccur, together within the local community (Weiher and Keddy 1999, Cavender-Bares et al. 2009).

The large-scale conversion of old forests to tree plantations has made it increasingly important to understand how understory vegetation responds to such landscape changes. For instance, in some forest types a reduction in understory richness and cover is thought to result from the development of canopy closure in plantations, although there is a paucity of empirical data demonstrating this relationship. We used a 420-year forest chronosequence as a case study to assess the relationship between stand age, tree canopy cover and understory vascular plant richness and composition in the Siskiyou Mountains of Oregon. The chronosequence consisted of six young managed (age 7–44) and nine older unmanaged (age 90–427) stands. All stands were similar in underlying geology, slope, elevation, and aspect. We found a non-linear relationship between stand age and richness, in which richness was highest in the youngest stands, reached a low in mid-aged stands (∼55 years), then increased in the oldest stands. We also found that percent tree canopy cover was correlated with total understory cover, richness, diversity, and species composition. In general, young stands were characterized by high shrub and graminoid cover and old stands were characterized by an abundant herb layer. Our work suggests that a major component of our study landscape is currently entering the forest stage (canopy closure) characterized by low levels of vascular plant species richness and cover. We use our results to discuss the potential effects of future forest management on understory plants.

Aim A major endeavour of community ecology is documenting non-random patterns in the composition and body size of coexisting species, and inferring the processes, or assembly rules, that may have given rise to the observed patterns. Such assembly rules include species sorting resulting from interspecific competition, aggregation at patchily distributed resources, and co-evolutionary dynamics. However, for any given taxon, relatively little is known about how these patterns and processes change through time and vary with habitat type, disturbance history, and spatial scale. Here, we tested for non-random patterns of species co-occurrence and body size in assemblages of ground-foraging ants and asked whether those patterns varied with habitat type, disturbance history, and spatial scale. Location Burned and unburned forests and fens in the Siskiyou Mountains of southern Oregon and northern California, USA. Methods We describe ground-foraging ant assemblages sampled over two years in two discrete habitat types, namely Darlingtonia fens and upland forests. Half of these sites had been subject to a large-scale, discrete disturbance ‐ a major fire ‐ in the year prior to our first sample. We used null model analyses to compare observed species co-occurrence patterns and body-size distributions in these assemblages with randomly generated assemblages unstructured by competition both within (i.e. at a local spatial scale) and among (i.e. at a regional scale) sites. Results At local spatial scales, species co-occurrence patterns and body-size ratios did not differ from randomness. At regional scales, co-occurrence patterns were random or aggregated, and there was evidence for constant body-size ratios of forest ants. Although these patterns varied between habitats and years, they did not differ between burned and unburned sites. Main conclusions Our results suggest that the operation of assembly rules depends on spatial scale and habitat type, but that it was not affected by disturbance history from fire.

Although ants perform many critical functions in forested ecosystems, little is known about how they respond to timber harvesting, especially in temperate systems. We examined ground-foraging ant communities and 11 forest characteristics in temperate conifer forests of southwestern Oregon, USA that ranged in age from 5 to 427 years. Seven forest characteristics were related to stand age and were summarized using principal components analysis (PCA). In this case study, species richness was nearly three times higher in young clearcuts compared to closed-canopy and old stands, and worker number was highest in young stands, lowest in closed-canopy stands, and intermediate in old stands. Using stepwise multiple regression, we found that both species richness and worker number were significantly related to PCA axis 1, which represented the environmental changes that accompany stand development, and canopy variability. Though species richness was high in recently logged stands, our study suggests that, in this system, the ant community does not resemble those found in mature forests until over 100 years following disturbance. Because ants modify their environment and perform ecosystem functions like seed dispersal, the alteration of ant communities may cascade through other parts of ecosystems.

We investigated the effects of the abiotic environment, plant community composition and disturbance by fire on ant assemblages in two distinct habitat types in the Siskiyou Mountains in northern California and southern Oregon, USA. Sampling over 2 years in burned and unburned Darlingtonia fens and their adjacent upland forests, we found that the effects of disturbance by fire depended on habitat type. In forests, fire intensity predicted richness in ant assemblages in both years after the fire, and plant community composition predicted richness 2 years after the fire. No factors were associated with richness in the species-poor fen ant assemblages. Speciesspecific responses to both habitat type and disturbance by fire were idiosyncratic. Assemblage composition depended on habitat type, but not disturbance by fire, and the composition of each assemblage between years was more dissimilar in burned than unburned sites.

We investigated the long-term recovery of Trillium ovatum (Liliaceae) following clear-cut logging by assessing demographic trends across a forest chronosequence that ranged in age from 2 to 424 years. In each of 20 sites across the chronosequence, we used 50 50 m plots to assess the population size, age and stage structure, seed production, and the spatial distribution of seedlings relative to reproductive plants. We found that Trillium populations were significantly reduced in younger sites and a significant positive relationship of population size and stand age, suggesting that population recovery following standreplacement disturbances may take centuries. Slow recovery rates of Trillium can partly be explained by short seed dispersal distances. We found 67% of new Trillium recruits within only one meter of potential parent plants. Despite the general trend of slow recovery, a few recently disturbed populations showed fast population growth. In these cases, high rates of post-disturbance recruitment were related to the number of individuals surviving the disturbance event. We found that local, within-site seed production can most likely mitigate limited dispersal abilities of Trillium. The rate of Trillium recovery following disturbance, therefore, depends on the number of plants that have persisted through the disturbance event and serve as within-site seed sources. Our study suggests that the nature of the stand-replacement disturbance (i.e., magnitude and intensity) may be critical in determining the recovery of understory plants in managed forests. # 2005 Elsevier B.V. All rights reserved.

Cascading ecological effects of anthropogenic habitat fragmentation have been studied primarily in extreme cases (e.g., the isolation of habitat fragments in a novel habitat matrix such as suburban developments, reservoirs, or agricultural fields), with less attention to more subtle and widespread cases, such as habitat fragmentation due to timber harvest. Few studies have used rigorous demographic data to demonstrate the direct and indirect effects of habitat fragmentation. We trapped deer mice (Peromyscus maniculatus) at five sites over two years in southwest Oregon, USA, and used multi-state capture- recapture models to estimate deer mouse survival and movement in clearcuts, forest-frag- ment edges, forest-fragment interiors, and contiguous forests. We also estimated deer mouse densities in fragmented and unfragmented forests and combined deer mouse demographic studies with trillium (Trillium ovatum) seed predation trials to link deer mouse changes to reduced trillium recruitment previously observed at the same study sites. Mouse survival was highest in clearcuts, intermediate in forest fragments, and lowest in unfragmented (control) forests. Mouse movement among clearcuts, forest edges, and forest interiors was common over short time intervals. Collectively, demographic rates led to mouse densities that were 3-4 times higher at forest-fragment sites than at unfragmented sites. Trillium seeds were ;3 times more likely to be depredated in areas of elevated relative mouse abundance than in areas of lower relative abundance. Forest fragmentation has favored mouse populations, resulting in increased seed predation that may decrease recruitment rates and increase local extinction risks for trillium.

Understanding biological invasions requires information on the history of spatial spread, as well as measures of landscape and biotic features that control habitat invasibility. Because invasive species often spread quickly over large areas, attaining these two sets of information simultaneously is uncommon. We studied the spread of a fatal nonnative root pathogen, Phytophthora lateralis, across a heterogeneous landscape of its host, Port Orford cedar (Chamaecyparis lawsoniana). Within our 37-km2 study area in southwestern Oregon and northwest California, Port Orford cedar populations are generally restricted to riparian zones along creeks. The pathogen is spread between watersheds in two ways: (1) by spore-infested material being dislodged from vehicles, and (2) by animals or people moving infested mud (i.e., via foot traffic). Using dendrochronological techniques, we determined the date of infection for dead cedars and reconstructed spread history across our study area from 1977 to 1999. Twenty-six of t...

Studies of plants in fragmented habitats have focused on single ecological processes, such as seed production or seed dispersal, that may be altered by increased fragmentation and that operate as the mecha- nism(s) that increase extinction probability. We examined a suite of potential mechanisms to explain demo- graphic shifts toward extinction in populations of Trillium ovatum , a long-lived herbaceous perennial found in the understory of western North American conifer forests. Past work has shown that populations of T. ova- tum found within z 65 m of forest-clearcut edges in southwestern Oregon have had almost no new recruit- ment since the edges were formed. We hypothesized that changes in abiotic conditions and biotic interactions present along edges are responsible for reduced recruitment. In eight populations in eight separate fragments, we evaluated the relationship of distance of the populations to the forest edge with respect to six processes: flowering phenology (timing), seed production, pollination- and resource-limitation of seed set, seed dis- persal, seed predation, and germination. Those factors that showed a significant relationship with edge dis- tance were then compared with recruitment of younger age classes. Two processes were significantly different near edges and were highly correlated with decreased recruitment: decreased seed production due to changes in pollination and increased seed predation by rodents. Our study (in conjunction with previous studies) sug- gests that several ecological processes show no significant relationship with edge distance and can be elimi- nated as possible mechanisms of reduced recruitment: flowering phenology, resource-limitation of seed set, seed dispersal, germination, herbivory, and survivorship of established plants. Edges influence some but not all components of a plant's life history. Thus, determining shifts in only one part of a life history will be inad- equate for testing the prediction of increasing extinction probabilities in fragmented landscapes. Future stud- ies should include enough information to conduct comprehensive analyses, such as matrix projections and sensitivity analyses.

I studied the influence of forest fragmentation on an understory herb, Trillium ovatum, in the Siskiyou Mountains of Oregon, where logging practices over the past 35 yr have created a mosaic of fragments surrounded by clearcuts and tree plantations. The age of trillium plants can be estimated by counting the annual constrictions on their rhizomes. Based on data collected by Whittaker in 1949 (i.e., pre-fragmentation) and a survey I conducted in 1995, I estimated that the process of clearcutting and subsequent conifer planting results in the mortality of almost all trillium (∼97.6%). In general, the remaining plants are not recruiting new individuals, even in sites clearcut 30 yr ago. Thus, trillium is restricted to smaller amounts of remnant, uncut forest. My study also demonstrated that populations in forest remnants that were within ∼65 m of forest-clearcut edges have had almost no recruitment of young plants since the time of the adjacent clearcutting, while forest interior populations contained higher recruitment levels. Projections based on these recruitment estimates indicated that edge populations will decline in size and interior populations will not decline. This study provides the first evidence of demographic changes in plant populations resulting from habitat fragmentation, and it offers evidence for the mechanisms responsible for such demographic changes.

Theoretical study of invasion dynamics has suggested that spatial heterogeneity should strongly influence the rate and extent of spreading organisms. However, empirical support for this prediction is scant, and the importance of understanding heterogeneity for real-world systems has remained ambiguous. This study quantified the influence of host and environmental heterogeneity on the dynamics of a 19-year disease invasion by the exotic and fatal pathogen, Phytophthora lateralis, within a stream population of its host tree, Port Orford cedar (Chamaecyparis lawsoniana). Using dendrochronology, we reconstructed the invasion history along a 1350-m length of infected stream, which serves as the only route of pathogen dispersal. Contrary to theoretical predictions, the temporal progression of the disease invasion was not related to a host's downstream spatial position, but instead was determined by two sources of heterogeneity: host size and proximity to the stream channel. These sources of heterogeneity influenced both the epidemic and endemic dynamics of this pathogen invasion. This analysis provides empirical support for the influence of heterogeneity on the invasion dynamics of a commercially important forest pathogen and highlights the need to incorporate such natural variability into both invasion theory and methods aimed at controlling future spread.