Search

Spruce budworm (Choristoneura fumiferana; SBW) outbreaks in the northeastern USA and Canada are recurring phenomena leading to large-scale mortality of spruce (Picea sp.) and balsam fir (Abies balsamea (L.) Mill.) forests as susceptibility to SBW is primarily determined by the availability of host species and their maturity. Our study examined several satellite remote sensing (Sentinel-1 C-band synthetic aperture radar (SAR), PALSAR L-band SAR, and Sentinel-2 multispectral) and site variables over space and time to develop a method to produce large-scale SBW stand impact types and susceptibility maps in Maine, USA. We used two machine-learning algorithms (Random Forest, RF; Multi-Layer Perceptron, MLP) to map SBW host species where RF produced better results than MLP. Our best model with site (elevation and aspect) and Sentinel-2 data attained an overall accuracy (OA) of 83.4%. However, the addition of SAR variables did not improve the model further. Combining host species data with age data retrieved from Land Change Monitoring, Assessment, and Projection (LCMAP) products, we demonstrated that SBW susceptibility map (based on stand impact types) could be produced with an OA of 88.3%. The fine spatial resolution (20 m) maps derived from our study provide reliable products for landscape-level SBW interventions in the region. [ABSTRACT FROM AUTHOR]

Leaf traits are expressed either as area-based content (g m−2) or mass-based concentration (mg/g or %) and subsequently scaled to canopy level using leaf area index (LAI) and foliage biomass, respectively. Despite their wide use, it remains unclear whether the two trait expressions and scaling options (by LAI or foliage biomass) yield the same prediction accuracy using multispectral remote sensing data and additional site variables. In this study, we used LAI (m2 m−2) and foliage biomass (kg) to upscale area-based content [leaf mass per area (LMA), equivalent water thickness (EWT), nitrogen (N area) and potassium (K area)] and mass-based concentration [leaf dry matter content (LDMC), leaf water content (LWC), N mass , and K mass)], respectively in two contrasting study sites (coastal vs inland) for an ecologically important yet distinct regional species (red spruce , Picea rubens Sarg.) in Maine, USA. We used partial least squares regression (PLSR) and random forest (RF) to examine the effect of these metrics on model performance and prediction accuracy at both leaf- and canopy-scales using Sentinel-2 satellite bands, vegetation indices, as well as several topographical variables and depth to water table (DWT). The best-performing models (based on cross-validated normalized root mean square error: nRMSE) were used to generate high-resolution canopy trait maps. Results demonstrated that prediction accuracies of area-based leaf traits were generally higher (R2 = 0.41–0.65, nRMSE = 0.14–0.20) than the mass-based concentration (R2 = 0.42–0.65, nRMSE = 0.15–0.25) for both sites and modelling techniques. We also observed that canopy traits upscaled by LAI yield better prediction accuracies (R2 = 0.44–0.70, nRMSE = 0.17–0.20) compared to a foliage biomass upscaling approach (R2 = 0.41–0.57, nRMSE = 0.19–0.24). Among site variables, DWT was a significant variable in modeling canopy traits and showed robust behavior across the contrasting study sites. Overall, our study highlights the importance of trait expression and how it affects foliar trait retrieval accuracy at both leaf- and canopy-scale using satellite multispectral data and site variables. These findings have implications on leaf trait modelling especially in the context of monitoring dynamics in forest health and productivity from space. [ABSTRACT FROM AUTHOR]

North American ash species (Fraxinus spp.) are under dire threat from the invasive pest, emerald ash borer (Agrilus planipennis, EAB), and mapping ash trees is of utmost significance for conservation and prevention efforts. We developed remote sensing techniques to identify ash trees at the individual tree level in a mixed-species forest in Maine, USA, using hyperspectral bands combined with derived spectral vegetation indices (SVIs), texture metrics using Random Forest (RF) and Support Vector Machine (SVM) algorithms. The pixel-based SVM-Optimized model proved to be the most accurate classification method with the lowest number of input variables including six Minimum Noise Fraction (MNF)-reduced bands, four SVIs and two texture variables, achieving 78.2% overall accuracy and 81.0% ash Producer's Accuracy. The technique presented in this study could be used to map ash trees throughout Maine and other states with similar forest types for future preservation efforts. [ABSTRACT FROM AUTHOR]

Climate change is impacting forest ecosystems, which support key ecosystem services and the general well-being of natural resource-dependent communities in Northeastern, USA. Understanding the determinants of climate change risk perceptions among forest resource stakeholders is critical to eliciting broad support for adaptation. We examined social-psychological drivers of climate change risk perceptions using hierarchical regression based on an online survey of 211 forest stakeholders, representing a wide range of subsectors, in Maine, USA. Using the climate change risk perceptions model (CCRPM), we explained 70% of the variance in risk perception. Political orientation, belief in climate change, social norms, affect, and experience with weather-related impacts were all significant predictors of perceived risk. Mediation results demonstrate that experience with weather-related impacts influences risk perceptions indirectly via attribution and holistic affect. This study advances our understanding of the social-psychological determinants of climate change risk perceptions, with implications for communication and outreach strategies. [ABSTRACT FROM AUTHOR]

We conducted a spatially explicit vulnerability assessment of the forest industry in Maine, USA, to climate change in an effort to (1) advance a spatial framework for assessing forest industry vulnerability and (2) increase our understanding of Maine's specific vulnerabilities to climate change in order to guide decision-making. We applied a bottom-up indicator approach to evaluate exposure, sensitivity, and adaptive capacity to climate change using both biophysical and social indicators, largely driven by participatory processes. Our approach enabled us to synthesize and aggregate indicators of regional importance to evaluate vulnerability, allowing us to simultaneously examine combinations of potential changes. We found that each Maine county had its own unique combination of exposure, sensitivity, and adaptive capacity indicators, with overall vulnerability highest in the rural northern and western parts of the state, where forest industry activities are most prevalent. However, results also indicate that although increased stress from climate-related changes can negatively affect Maine's forest via high exposure, reduced sensitivities and increased adaptive capacity have the potential to largely decrease overall vulnerability in many parts of the state. [ABSTRACT FROM AUTHOR]

Climate change will have a significant impact on the forest industry and will require strategies that promote sustainable forest management. Understanding perceptions of climate change impacts is critical to supporting the use of adaptation strategies, informing future research, and supporting decision-making. We describe a multi-method approach using nominal group technique and semi-structured interviews to identify and understand experts' concerns in regards to future climate change impacts on the forest industry in Maine, USA. A review of the existing scientific literature helped inform the development of the interview and nominal group techniques. Experts prioritized the greatest and most likely climate change impacts on the forest industry as: forest health threats imposed by insects and pathogens, extreme precipitation events, shifts in forest composition, invasive species, and changes in forest productivity. Interviews and current scientific understanding largely support the prioritization of the impacts, but also elucidate uncertainties in regards to climate change perceptions (e.g., timing and magnitude of future impacts), highlight the need for continued research that specifically addresses how climate change will affect the forest industry, and demonstrate that climate change presents not only a perceived threat but also an opportunity in the form of increased forest productivity and economic growth. [ABSTRACT FROM AUTHOR]

Key message: Simulated spruce budworm (SBW; Choristoneura fumiferana (Clem.)) defoliation generally becomes ubiquitous in 3 years after its initiation in agreement with historical observations despite varying environmental and stand conditions over large ranges. Current-year defoliation has almost no correlation with defoliation more than 1 year ago at the same location, which may be related to the role of SBW dispersal in sustaining defoliation across space and time. Mitigation practices like insecticide spraying may be more efficient if applied early to initial spots (epicenters) of defoliation, while management probably should focus on improving forests' resilience to withstand repeated defoliation by altering species composition. Context: SBW defoliation during its periodic and extensive outbreaks greatly affects forest productivity at large spatial and temporal scales. A generalized modeling framework that simultaneously accounts for both highly variable spatial and temporal dynamics of SBW outbreaks has not been developed. Aims: To develop a flexible parametric spatiotemporal model to explicitly predict defoliation in continuous space and time in order to evaluate the dynamics of SBW outbreaks across a complex forested landscape. Methods: A novel model was developed on extensive defoliation data covering approximately 50,000 km2 and 10 years of the last SBW outbreak during the 1970s and 1980s in Maine, USA. Simulations of various outbreak scenarios were performed using this model. Results: The developed model provided a sufficient fit of the data (R2 of 0.63 and mean bias of +0.3%) and was relatively consistent with expectations. Simulations show that defoliation generally becomes ubiquitous in 3 years despite varying environmental and stand conditions. Current-year defoliation has almost no correlation with defoliation more than 1 year ago at the same location, which may be related to the role of SBW dispersal in sustaining defoliation across space and time. Conclusion: Mitigation practices like insecticide spraying may be more efficient if applied early to initial spots (epicenters) of defoliation, while management probably should focus on improving forests' resilience to withstand repeated defoliation by altering species composition. Our model provides quantitative information flexible in spatial and temporal scales yet directly usable in existing forest growth and yield modeling frameworks and management decision support systems. This generalized spatiotemporal model is readily extendable for evaluating spatial and temporal dynamics of other forms of defoliation across complex forest landscapes. [ABSTRACT FROM AUTHOR]

A few severe drought events occurred in the Northeast (NE) USA in recent decades and caused significant economic losses, but the temporal pattern of drought incidents and their impacts on agricultural systems have not been well assessed. Here, we analyzed historical changes and patterns of drought using a drought index (standardized precipitation-evapotranspiration index (SPEI)), and assessed drought impacts on remotely sensed vegetation indices (enhanced vegetation index (EVI) and normalized difference vegetation index (NDVI)) and production (yield) of the wild blueberry fields in Maine, USA. We also analyzed the impact of short- and long-term water conditions of the growing season on the wild blueberry vegetation condition and production. No significant changes in the SPEI were found in the past 71 years, despite a significant warming pattern. There was also a significant relationship between the relatively long-term SPEI and the vegetation indices (EVI and NDVI), but not the short-term SPEI (one year). This suggests that the crop vigor of wild blueberries is probably determined by water conditions over a relatively long term. There were also significant relationships between 1-year water conditions (SPEI) and yield for a non-irrigated field, and between 4-year-average SPEI and the yield of all fields in Maine. The vegetation indices (EVI and NDVI) are not good predictors of wild blueberry yield, possibly because wild blueberry yield does not only depend on crop vigor, but also on other important variables such as pollination. We also compared an irrigated and a non-irrigated wild blueberry field at the same location (Deblois, Maine) where we found that irrigation decoupled the relationship between the SPEI and NDVI or EVI. [ABSTRACT FROM AUTHOR]

A fine-resolution region-wide map of forest site productivity is an essential need for effective large-scale forestry planning and management. In this study, we incorporated Sentinel-2 satellite data into an increment-based measure of forest productivity (biomass growth index (BGI)) derived from climate, lithology, soils, and topographic metrics to map improved BGI (iBGI) in parts of North American Acadian regions. Initially, several Sentinel-2 variables including nine single spectral bands and 12 spectral vegetation indices (SVIs) were used in combination with forest management variables to predict tree volume/ha and height using Random Forest. The results showed a 10–12 % increase in out of bag (OOB) r2 when Sentinel-2 variables were included in the prediction of both volume and height together with BGI. Later, selected Sentinel-2 variables were used for biomass growth prediction in Maine, USA and New Brunswick, Canada using data from 7738 provincial permanent sample plots. The Sentinel-2 red-edge position (S2REP) index was identified as the most important variable over others to have known influence on site productivity. While a slight improvement in the iBGI accuracy occurred compared to the base BGI model (~2%), substantial changes to coefficients of other variables were evident and some site variables became less important when S2REP was included. [ABSTRACT FROM AUTHOR]