Land management explains major trends in forest structure and composition over the last millennium in California's Klamath Mountains.

SignificanceWe provide the first assessment of aboveground live tree biomass in a mixed conifer forest over the late Holocene. The biomass record, coupled with local Native oral history and fire scar records, shows that Native burning practices, along with a natural lightning-based fire regime, promoted long-term stability of the forest structure and composition for at least 1 millennium in a California forest. This record demonstrates that climate alone cannot account for observed forest conditions. Instead, forests were also shaped by a regime of frequent fire, including intentional ignitions by Native people. This work suggests a large-scale intervention could be required to achieve the historical conditions that supported forest resiliency and reflected Indigenous influence.

Forest restoration and fuels reduction work: Different pathways for achieving success in the Sierra Nevada.

Fire suppression and past selective logging of large trees have fundamentally changed frequent-fire-adapted forests in California. The culmination of these changes produced forests that are vulnerable to catastrophic change by wildfire, drought, and bark beetles, with climate change exacerbating this vulnerability. Management options available to address this problem include mechanical treatments (Mech), prescribed fire (Fire), or combinations of these treatments (Mech + Fire). We quantify changes in forest structure and composition, fuel accumulation, modeled fire behavior, intertree competition, and economics from a 20-year forest restoration study in the northern Sierra Nevada. All three active treatments (Fire, Mech, Mech + Fire) produced forest conditions that were much more resistant to wildfire than the untreated control. The treatments that included prescribed fire (Fire, Mech + Fire) produced the lowest surface and duff fuel loads and the lowest modeled wildfire hazards. Mech produced low fire hazards beginning 7 years after the initial treatment and Mech + Fire had lower tree growth than controls. The only treatment that produced intertree competition somewhat similar to historical California mixed-conifer forests was Mech + Fire, indicating that stands under this treatment would likely be more resilient to enhanced forest stressors. While Fire reduced modeled wildfire hazard and reintroduced a fundamental ecosystem process, it was done at a net cost to the landowner. Using Mech that included mastication and restoration thinning resulted in positive revenues and was also relatively strong as an investment in reducing modeled wildfire hazard. The Mech + Fire treatment represents a compromise between the desire to sustain financial feasibility and the desire to reintroduce fire. One key component to long-term forest conservation will be continued treatments to maintain or improve the conditions from forest restoration. Many Indigenous people speak of "active stewardship" as one of the key principles in land management and this aligns well with the need for increased restoration in western US forests. If we do not use the knowledge from 20+ years of forest research and the much longer tradition of Indigenous cultural practices and knowledge, frequent-fire forests will continue to be degraded and lost.

Tallo: A global tree allometry and crown architecture database.

Data capturing multiple axes of tree size and shape, such as a tree's stem diameter, height and crown size, underpin a wide range of ecological research-from developing and testing theory on forest structure and dynamics, to estimating forest carbon stocks and their uncertainties, and integrating remote sensing imagery into forest monitoring programmes. However, these data can be surprisingly hard to come by, particularly for certain regions of the world and for specific taxonomic groups, posing a real barrier to progress in these fields. To overcome this challenge, we developed the Tallo database, a collection of 498,838 georeferenced and taxonomically standardized records of individual trees for which stem diameter, height and/or crown radius have been measured. These data were collected at 61,856 globally distributed sites, spanning all major forested and non-forested biomes. The majority of trees in the database are identified to species (88%), and collectively Tallo includes data for 5163 species distributed across 1453 genera and 187 plant families. The database is publicly archived under a CC-BY 4.0 licence and can be access from: https://doi.org/10.5281/zenodo.6637599. To demonstrate its value, here we present three case studies that highlight how the Tallo database can be used to address a range of theoretical and applied questions in ecology-from testing the predictions of metabolic scaling theory, to exploring the limits of tree allometric plasticity along environmental gradients and modelling global variation in maximum attainable tree height. In doing so, we provide a key resource for field ecologists, remote sensing researchers and the modelling community working together to better understand the role that trees play in regulating the terrestrial carbon cycle.

Accurate tracking of forest activity key to multi-jurisdictional management goals: A case study in California.

An essential component of sustainable forest management is accurate monitoring of forest activities. Although monitoring efforts have generally increased for many forests throughout the world, in practice, effective monitoring is complex. Determining the magnitude and location of progress towards sustainability targets can be challenging due to diverse forest operations across multiple jurisdictions, the lack of data standardization, and discrepancies between field inspections and remotely-sensed records. In this work, we used California as a multijurisdictional case study to explore these problems and develop an approach that broadly informs forest monitoring strategies. The State of California recently entered into a shared stewardship agreement with the US Forest Service (USFS) and set a goal to jointly treat one million acres of forest and rangeland annually by 2025. Currently, however, federal and state forest management datasets are disjoint. This work addresses three barriers stymying the use of federal and state archival records to assess management goals. These barriers are: 1) current databases from different jurisdictions have not been combined due to their distinct data collection processes and internal structures; 2) datasets have not been comprehensively analyzed, despite the need to understand the extent of previous treatments as well as the rate of current activity; and 3) the spatial accuracy of archival datasets has not been evaluated against remotely-sensed data. To reduce these barriers, we first aggregated existing archival forest management records between 1984 and 2019 from the USFS' Forest Activity Tracking System (FACTS) and the California Department of Forestry and Fire Protection (CAL FIRE) using a qualitative scalar of treatment intensity. Combined FACTS and CAL FIRE completed footprint acres - defined as unique areas of land where a treatment was completed at any time since 1984 - have decreased since a peak in 2008. At most, 300,000 footprint acres are completed each year, 30% of the million-acre goal. Prescribed fires - defined as direct burning operations - have risen over time, according to the FACTS hazardous fuels dataset but prescribed fire records in CAL FIRE's dataset have rapidly increased since 2016. We also refined the spatial and temporal detail of the aggregated management record using the Continuous Change Detection and Classification algorithm on satellite remote sensing data to produce a state-wide time series map of harvest disturbances. A comparison of the algorithm's refined data to the archival record potentially suggests over-reporting in both FACTS and CAL FIRE's archival datasets. Our integrated dataset provides a better assessment of current treatments and the path towards the 1-million-acre a year goal. The refined dataset leverages the strengths of complementary, albeit imperfect, monitoring strategies from archives and remotely-sensed detection.

Evaluating a new method for reconstructing forest conditions from General Land Office survey records.

Historical forest conditions are often used to inform contemporary management goals because historical forests are considered to be resilient to ecological disturbances. The General Land Office (GLO) surveys of the late 19th and early 20th centuries provide regionally quasi-contiguous data sets of historical forests across much of the Western United States. Multiple methods exist for estimating tree density from point-based sampling such as the GLO surveys, including distance-based and area-based approaches. Area-based approaches have been applied in California mixed-conifer forests but their estimates have not been validated. To assess the accuracy and precision of plotless density estimators with potential for application to GLO data in this region, we imposed a GLO sampling scheme on six mapped forest stands of known densities (159-784 trees/ha) in the Sierra Nevada in California, USA, and Baja California Norte, Mexico. We compared three distance-based plotless density estimators (Cottam, Pollard, and Morisita) as well as two Voronoi area (VA) estimators, the Delincé and mean harmonic Voronoi density (MHVD), to the true densities. We simulated sampling schemes of increasing intensity to assess sampling error. The relative error (RE) of density estimates for the GLO sampling scheme ranged from 0.36 to 4.78. The least biased estimate of tree density in every stand was obtained with the Morisita estimator and the most biased was obtained with the MHVD estimator. The MHVD estimates of tree density were 1.2-3.8 times larger than the true densities and performed best in stands subject to fire exclusion for 100 yr. The Delincé approach obtained accurate estimates of density, implying that the Voronoi approach is theoretically sound but that its application in the MHVD was flawed. The misapplication was attributed to two causes: (1) the use of a crown scaling factor that does not correct for the number of trees sampled and (2) the persistent underestimate of the true VA due to a weak relationship between tree size and VA. The magnitude of differences between true densities and MHVD estimates suggest caution in using results based on the MHVD to inform management and restoration practices in the conifer forests of the American West.

A Word to the Wise: Advice for Scientists Engaged in Collaborative Adaptive Management.

Collaborative adaptive management is a process for making decisions about the environment in the face of uncertainty and conflict. Scientists have a central role to play in these decisions. However, while scientists are well trained to reduce uncertainty by discovering new knowledge, most lack experience with the means to mitigate conflict in contested situations. To address this gap, we drew from our efforts coordinating a large collaborative adaptive management effort, the Sierra Nevada Adaptive Management Project, to offer advice to our fellow environmental scientists. Key challenges posed by collaborative adaptive management include the confusion caused by multiple institutional cultures, the need to provide information at management-relevant scales, frequent turnover in participants, fluctuations in enthusiasm among key constituencies, and diverse definitions of success among partners. Effective strategies included a dedication to consistency, a commitment to transparency, the willingness to communicate frequently via multiple forums, and the capacity for flexibility. Collaborative adaptive management represents a promising, new model for scientific engagement with the public. Learning the lessons of effective collaboration in environmental management is an essential task to achieve the shared goal of a sustainable future.

Silica uptake and release in live and decaying biomass in a northern hardwood forest.

In terrestrial ecosystems, a large portion (20-80%) of the dissolved Si (DSi) in soil solution has passed through vegetation. While the importance of this "terrestrial Si filter" is generally accepted, few data exist on the pools and fluxes of Si in forest vegetation and the rate of release of Si from decomposing plant tissues. We quantified the pools and fluxes of Si through vegetation and coarse woody debris (CWD) in a northern hardwood forest ecosystem (Watershed 6, W6) at the Hubbard Brook Experimental Forest (HBEF) in New Hampshire, USA. Previous work suggested that the decomposition of CWD may have significantly contributed to an excess of DSi reported in stream-waters following experimental deforestation of Watershed 2 (W2) at the HBEF. We found that woody biomass (wood + bark) and foliage account for approximately 65% and 31%, respectively, of the total Si in biomass at the HBEF. During the decay of American beech (Fagus grandifolia) boles, Si loss tracked the whole-bole mass loss, while yellow birch (Betula alleghaniensis) and sugar maple (Acer saccharum) decomposition resulted in a preferential Si retention of up to 30% after 16 yr. A power-law model for the changes in wood and bark Si concentrations during decomposition, in combination with an exponential model for whole-bole mass loss, successfully reproduced Si dynamics in decaying boles. Our data suggest that a minimum of 50% of the DSi annually produced in the soil of a biogeochemical reference watershed (W6) derives from biogenic Si (BSi) dissolution. The major source is fresh litter, whereas only ~2% comes from the decay of CWD. Decay of tree boles could only account for 9% of the excess DSi release observed following the experimental deforestation of W2. Therefore, elevated DSi concentrations after forest disturbance are largely derived from other sources (e.g., dissolution of BSi from forest floor soils and/or mineral weathering).

Decreased water flowing from a forest amended with calcium silicate.

Acid deposition during the 20th century caused widespread depletion of available soil calcium (Ca) throughout much of the industrialized world. To better understand how forest ecosystems respond to changes in a component of acidification stress, an 11.8-ha watershed was amended with wollastonite, a calcium silicate mineral, to restore available soil Ca to preindustrial levels through natural weathering. An unexpected outcome of the Ca amendment was a change in watershed hydrology; annual evapotranspiration increased by 25%, 18%, and 19%, respectively, for the 3 y following treatment before returning to pretreatment levels. During this period, the watershed retained Ca from the wollastonite, indicating a watershed-scale fertilization effect on transpiration. That response is unique in being a measured manipulation of watershed runoff attributable to fertilization, a response of similar magnitude to effects of deforestation. Our results suggest that past and future changes in available soil Ca concentrations have important and previously unrecognized implications for the water cycle.

The promise and peril of intensive-site-based ecological research: insights from the Hubbard Brook ecosystem study.

Ecological research is increasingly concentrated at particular locations or sites. This trend reflects a variety of advantages of intensive, site-based research, but also raises important questions about the nature of such spatially delimited research: how well does site based research represent broader areas, and does it constrain scientific discovery? We provide an overview of these issues with a particular focus on one prominent intensive research site: the Hubbard Brook Experimental Forest (HBEF), New Hampshire, USA. Among the key features of intensive sites are: long-term, archived data sets that provide a context for new discoveries and the elucidation of ecological mechanisms; the capacity to constrain inputs and parameters, and to validate models of complex ecological processes; and the intellectual cross-fertilization among disciplines in ecological and environmental sciences. The feasibility of scaling up ecological observations from intensive sites depends upon both the phenomenon of interest and the characteristics of the site. An evaluation of deviation metrics for the HBEF illustrates that, in some respects, including sensitivity and recovery of streams and trees from acid deposition, this site is representative of the Northern Forest region, of which HBEF is a part. However, the mountainous terrain and lack of significant agricultural legacy make the HBEF among the least disturbed sites in the Northern Forest region. Its relatively cool, wet climate contributes to high stream flow compared to other sites. These similarities and differences between the HBEF and the region can profoundly influence ecological patterns and processes and potentially limit the generality of observations at this and other intensive sites. Indeed, the difficulty of scaling up may be greatest for ecological phenomena that are sensitive to historical disturbance and that exhibit the greatest spatiotemporal variation, such as denitrification in soils and the dynamics of bird communities. Our research shows that end member sites for some processes often provide important insights into the behavior of inherently heterogeneous ecological processes. In the current era of rapid environmental and biological change, key ecological responses at intensive sites will reflect both specific local drivers and regional trends.

Ephemeral disturbances have long-lasting impacts on forest invasion dynamics.

Ephemeral disturbances are common in many systems. Often, these brief events are assumed to be a nuisance with little long-term ecological impact. We quantified the impact of the ephemeral forest disturbance caused by gypsy moth canopy defoliation on exotic plant invasion in eight hardwood forests in the Delaware Water Gap National Recreation Area, USA. Six years following the gypsy moth disturbance, we were able to predict 59% of the increase in Alliaria petiolata abundance and 42% of the increase in Microstegium vimineum abundance with models incorporating an interaction between disturbance severity and propagule pressure. In addition, we found that the disturbance timing had substantial impacts on the invasion dynamics of Alliaria petiolata. Our results suggest that ephemeral disturbances can have important and long-lasting impacts on plant communities, and highlight the need to account for variations in disturbance characteristics and the role of propagule pressure in determining the disturbance-invasion relationship.

Beyond reducing fire hazard: fuel treatment impacts on overstory tree survival.

Fuel treatment implementation in dry forest types throughout the western United States is likely to increase in pace and scale in response to increasing incidence of large wildfires. While it is clear that properly implemented fuel treatments are effective at reducing hazardous fire potential, there are ancillary ecological effects that can impact forest resilience either positively or negatively depending on the specific elements examined, as well as treatment type, timing, and intensity. In this study, we use overstory tree growth responses, measured seven years after the most common fuel treatments, to estimate forest health. Across the five species analyzed, observed mortality and future vulnerability were consistently low in the mechanical- only treatment. Fire-only was similar to the control for all species except Douglas-fir, while mechanical-plus-fire had high observed mortality and future vulnerability for white fir and sugar pine. Given that overstory trees largely dictate the function of forests and services they provide (e.g., wildlife habitat, carbon sequestration, soil stability) these results have implications for understanding longer-term impacts of common fuel treatments on forest resilience.

The importance of quantifying propagule pressure to understand invasion: an examination of riparian forest invasibility.

The widely held belief that riparian communities are highly invasible to exotic plants is based primarily on comparisons of the extent of invasion in riparian and upland communities. However, because differences in the extent of invasion may simply result from variation in propagule supply among recipient environments, true comparisons of invasibility require that both invasion success and propagule pressure are quantified. In this study, we quantified propagule pressure in order to compare the invasibility of riparian and upland forests and assess the accuracy of using a community's level of invasion as a surrogate for its invasibility. We found the extent of invasion to be a poor proxy for invasibility. The higher level of invasion in the studied riparian forests resulted from greater propagule availability rather than higher invasibility. Furthermore, failure to account for propagule pressure may confound our understanding of general invasion theories. Ecological theory suggests that species-rich communities should be less invasible. However, we found significant relationships between species diversity and invasion extent, but no diversity-invasibility relationship was detected for any species. Our results demonstrate that using a community's level of invasion as a surrogate for its invasibility can confound our understanding of invasibility and its determinants.

Nonlinear shifts in infectious rust disease due to climate change.

Range shifts of infectious plant disease are expected under climate change. As plant diseases move, emergent abiotic-biotic interactions are predicted to modify their distributions, leading to unexpected changes in disease risk. Evidence of these complex range shifts due to climate change, however, remains largely speculative. Here, we combine a long-term study of the infectious tree disease, white pine blister rust, with a six-year field assessment of drought-disease interactions in the southern Sierra Nevada. We find that climate change between 1996 and 2016 moved the climate optimum of the disease into higher elevations. The nonlinear climate change-disease relationship contributed to an estimated 5.5 (4.4-6.6) percentage points (p.p.) decline in disease prevalence in arid regions and an estimated 6.8 (5.8-7.9) p.p. increase in colder regions. Though climate change likely expanded the suitable area for blister rust by 777.9 (1.0-1392.9) km2 into previously inhospitable regions, the combination of host-pathogen and drought-disease interactions contributed to a substantial decrease (32.79%) in mean disease prevalence between surveys. Specifically, declining alternate host abundance suppressed infection probabilities at high elevations, even as climatic conditions became more suitable. Further, drought-disease interactions varied in strength and direction across an aridity gradient-likely decreasing infection risk at low elevations while simultaneously increasing infection risk at high elevations. These results highlight the critical role of aridity in modifying host-pathogen-drought interactions. Variation in aridity across topographic gradients can strongly mediate plant disease range shifts in response to climate change.

Simulation of the effects of forest harvesting under changing climate to inform long-term sustainable forest management using a biogeochemical model.

Process ecosystem models are useful tools to provide insight on complex, dynamic ecological systems, and their response to disturbances. The biogeochemical model PnET-BGC was modified and tested using field observations from an experimentally whole-tree harvested northern hardwood watershed (W5) at the Hubbard Brook Experimental Forest (HBEF), New Hampshire, USA. In this study, the confirmed model was used as a heuristic tool to investigate long-term changes in hydrology, biomass accumulation, and soil solution and stream water chemistry for three different watershed cutting intensities (40%, 60%, 80%) and three rotation lengths (30, 60, 90 years) under both constant (current climate) and changing (MIROC5-RCP4.5) future climate scenarios and atmospheric CO2 through the year 2200. For the no future cutting scenario, total ecosystem stored carbon (i.e., sum of aboveground biomass, woody debris and soil) reached a maximum value of 207 t C ha-1 under constant climate but increased to 452 t C ha-1 under changing climate in 2200 due to a CO2 fertilization effect. Harvesting of trees decreased total ecosystem stored carbon between 7 and 36% for constant climate and 7-60% under changing climate, respectively, with greater reductions for shorter logging rotation lengths and greater watershed cutting intensities. Harvesting under climate change resulted in noticeable losses of soil organic matter (12-56%) coinciding with loss of soil nutrients primarily due to higher rates of soil mineralization associated with increases in temperature, compared with constant climate conditions (3-22%). Cumulative stream leaching of nitrate under climate change (181-513 kg N ha-1) exceeded constant climate values (139-391 kg N ha-1) for the various cutting regimes. Under both climate conditions the model projected greater sensitivity to varying the length of cutting period than cutting intensities. Hypothetical model simulations highlight future challenges in maintaining long-term productivity of managed forests under changing climate due to a potential for a deterioration of soil fertility.

How do riparian woody seedlings survive seasonal drought?

In semi-arid regions, a major population limitation for riparian trees is seedling desiccation during the dry season that follows annual spring floods. We investigated the stress response of first-year pioneer riparian seedlings to experimental water table declines (0, 1 and 3 cm day(-1)), focusing on the three dominant cottonwood and willows (family Salicaceae) in California's San Joaquin Basin. We analyzed growth and belowground allocation response to water stress, and used logistic regression to determine if these traits had an influence on individual survival. The models indicate that high root growth (>3 mm day(-1)) and low shoot:root ratios (<1.5 g g(-1)) strongly predicted survival, but there was no evidence that plants increased belowground allocation in response to drawdown. Leaf δ(13)C values shifted most for the best-surviving species (net change of +3.5 per mil from -30.0 ± 0.3 control values for Goodding's willow, Salix gooddingii), implying an important role of increased water-use efficiency for surviving water stress. Both S. gooddingii and sandbar willow (S. exigua) reduced leaf size from controls, whereas Fremont cottonwood (Populus fremontii) sustained a 29% reduction in specific leaf area (from 13.4 to 9.6 m(2) kg(-1)). The functional responses exhibited by Goodding's willow, the more drought-tolerant species, may play a role in its greater relative abundance in dry regions such as the San Joaquin Basin. This study highlights the potential for a shift in riparian forest composition. Under a future drier climate regime or under reduced regulated river flows, our results suggest that willow establishment will be favored over cottonwood.

Characterization of the woody biomass feedstock potential resulting from California's drought.

Regional tree die-off events generate large quantities of standing dead wood, raising concern over catastrophic wildfire and other hazards. Governmental responses to tree die-off have often focused on incentivizing biomass energy production that utilizes standing dead trees removed for safety concerns. However, the full distribution of potential woody bioenergy feedstock after tree die-off has not been evaluated due to the complexities of surveying and precisely measuring large forested areas. In this paper, we present a novel method for estimating standing dead biomass at a fine spatial resolution that combines aerial survey data with forest structure maps. Using this method, we quantify biomass generated by the unprecedented tree die-off that occurred in California following a 4-year drought and widespread pest outbreaks. The results are used to estimate feasibly recoverable feedstock for energy production. We find that approximately 95.1 million bone-dry tons (BDT) of dead biomass resulted from 2012-2017 mortality, with a lower bound of 26.2 million BDT. In other words, of the aboveground live tree biomass in 2012, ~1.3-4.8% died by 2017. Of the standing dead biomass, 29% meets minimum constraints for potential cost-effective bioenergy feedstock. This proportion drops to as low as 15% in the most affected areas due to terrain slope, wilderness status, and other factors, highlighting the need to complement disposal via biomass energy with other strategies to mitigate the risks of the tree mortality crisis, which is likely to only become more severe over time due to climate change.

Acceleration of exotic plant invasion in a forested ecosystem by a generalist herbivore.

The successful invasion of exotic plants is often attributed to the absence of coevolved enemies in the introduced range (i.e., the enemy release hypothesis). Nevertheless, several components of this hypothesis, including the role of generalist herbivores, remain relatively unexplored. We used repeated censuses of exclosures and paired controls to investigate the role of a generalist herbivore, white-tailed deer (Odocoileus virginianus), in the invasion of 3 exotic plant species (Microstegium vimineum, Alliaria petiolata, and Berberis thunbergii) in eastern hemlock (Tsuga canadensis) forests in New Jersey and Pennsylvania (U.S.A.). This work was conducted in 10 eastern hemlock (T. canadensis) forests that spanned gradients in deer density and in the severity of canopy disturbance caused by an introduced insect pest, the hemlock woolly adelgid (Adelges tsugae). We used maximum likelihood estimation and information theoretics to quantify the strength of evidence for alternative models of the influence of deer density and its interaction with the severity of canopy disturbance on exotic plant abundance. Our results were consistent with the enemy release hypothesis in that exotic plants gained a competitive advantage in the presence of generalist herbivores in the introduced range. The abundance of all 3 exotic plants increased significantly more in the control plots than in the paired exclosures. For all species, the inclusion of canopy disturbance parameters resulted in models with substantially greater support than the deer density only models. Our results suggest that white-tailed deer herbivory can accelerate the invasion of exotic plants and that canopy disturbance can interact with herbivory to magnify the impact. In addition, our results provide compelling evidence of nonlinear relationships between deer density and the impact of herbivory on exotic species abundance. These findings highlight the important role of herbivore density in determining impacts on plant abundance and provide evidence of the operation of multiple mechanisms in exotic plant invasion.

Deer herbivory alters forest response to canopy decline caused by an exotic insect pest.

Hemlock woolly adelgid (HWA; Adelges tsugae) infestations have resulted in the continuing decline of eastern hemlock (Tsuga canadensis) throughout much of the eastern United States. While the initial impacts of HWA infestations have been documented, our understanding of forest response to this disturbance remains incomplete. HWA infestation is not occurring in isolation but within a complex ecological context. The role of potentially important interacting factors, such as elevated levels of white-tailed deer herbivory, is poorly understood. Despite the potential for herbivory to alter forest successional trajectories following a canopy disturbance, little is known about herbivory-disturbance interactions, and herbivory is rarely considered in assessing forest response to a co-occurring disturbance. We used repeated censuses of deer exclosures and paired controls (400 paired plots) to quantify the impact of deer herbivory on tree seedling species abundance in 10 eastern hemlock ravines that span a gradient in HWA-induced canopy decline severity. Use of a maximum likelihood estimation framework and information theoretics allowed us to quantify the strength of evidence for alternative models developed to estimate the impacts of herbivory on tree seedling abundance as a function of varying herbivore density and canopy decline severity. The exclusion of deer herbivory had marked impacts on the abundance of the studied seedling species: Acer rubrum, Acer saccharum, Betula lenta, Nyssa sylvatica, Quercus montana, and Tsuga canadensis. For all six species, the relationship between seedling abundance and deer density was either exponential or saturating. Although the functional form of the response varied among seedling species, the inclusion of both deer density and canopy decline severity measures consistently resulted in models with substantially greater support. Canopy decline resulted in higher proportional herbivory impacts and altered the ranking of herbivory impacts by seedling species. Our results suggest that, by changing species' competitive abilities, white-tailed deer herbivory alters the trajectory of forest response to this exotic insect pest and has the potential to shift future overstory composition.