AUTHORS: Nathan Gill*, Clark University; Dan Jarvis, Vermont Technical College; Tom Veblen, University of Colorado; John Rogan, Clark University; Dominik Kulakowski, Clark University
ABSTRACT: Understory vegetation and ground cover drive many important ecosystem processes, including tree seedling regeneration. The exact effect of ground cover on tree seedling establishment, survival, and growth is dependent on biophysical context. In subalpine forests, this context is largely determined by disturbances such as beetle outbreak, blow down, and fire. Compounded disturbances that overlap in short succession can alter stand properties and trajectories in ways that are not predictable from the additive impact of individual disturbances. The aim of this study is to examine how compounded Dendroctonus rufipennis (Spruce Beetle) outbreak followed by fire and compounded wind blowdown followed by fire each influence the relationship between post-fire ground cover and conifer regeneration. We measured categorical ground cover percentages and conifer regeneration densities in permanent plots from 2003-2014 after stand-replacing fires of 2002 burned stands that had been blown down in 1997, affected by SB outbreak in the 1940s, or neither. We created mixed-effect models to measure the relationships between stand attributes and post-fire ground cover as well as between post-fire ground cover and conifer regeneration densities. Ground cover patterns were sensitive to compounded disturbances, and the relationships between conifer regeneration and ground cover were fundamentally different when fire was preceded by another disturbance. Conifer regeneration densities increased with increased litter coverage in stands that only burned, but decreased with increased litter coverage in stands that were blown down and then burned. Similarly, herbaceous cover changed from facilitative to competitive when fire was preceded by SB outbreak. These compound effects via ground cover are at play across stands of varying composition and structural stage and potentially across broad spatiotemporal scales.
AUTHORS: Jeremy Larroque, Département de Sciences Biologiques, Université de Montréal; Simon Legault, Département de Sciences Biologiques, Université de Montréal; Rob Johns, Canadian Forest Service, Natural Resources Canada; Lisa Lumley, Royal Alberta Museum; Canadian Forest Service, Natural Resources Canada; Michel Cusson, Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Centre; Patrick M. A. James, Département de Sciences Biologiques, Université de Montréal
ABSTRACT: High periodic variations in population size are common across natural populations. In some cases, preventive management strategies could reduce economic losses associated with population outbreaks, requiring a clear understanding of the spatial population dynamics of irruptive species. However, factors governing their spatial dynamics are still not fully understood. It is generally considered that dispersal (“epicenter hypothesis”) and spatial correlation in environmental stochasticity (“oscillatory hypothesis” or Moran effect) can synchronize populations over wide areas. Our objective was to identify the relative support of these two mechanisms in the outbreaks of an economically important irruptive forest insect, the spruce budworm (Choristoneura fumiferana) in the province of Quebec (Canada). AMOVA, cluster analysis, isolation by distance and sPCA were used to characterize spatial and temporal genomic variation using 1370 SBW larvae sampled over four years (2012-2015) and genotyped at more than 190,000 SNP loci. We found evidence of weak spatial genetic structure at the scale of Quebec. The little structure that did exist decreased between 2012 and 2015. We also found genetic evidence of a long-distance dispersal event over > 140 km. Results thus suggest that dispersal is the key mechanism involved in driving population synchrony at this stage of the outbreak. Early intervention management strategies that aim to control source populations have the potential to be effective through limiting dispersal. However, the timing of such interventions relative to the outbreak cycles and local dynamics will greatly influence their probability of success.
AUTHORS: Brian R. Sturtevant*, USDA Forest Service; Matthew Garcia, University of Wisconsin; Jacques Régnière, Yan Boulanger, Barry J. Cooke – Natural Resources Canada; Joseph J. Charney, Gary L. Achtemeier – USDA Forest Service; Johanne Delisle, Marc Rhainds, Rémi Saint-Amant – Natural Resources Canada
ABSTRACT: The spatiotemporal dynamics of eastern spruce budworm outbreaks in North American boreal and sub-boreal forests may be sensitive to long-distance dispersal patterns that are assisted by meteorological processes. We provide an overview of temperature-constrained functional relationships between insect mass, wing area, and wingbeat frequencies, and employ these relationships to simulate realistic flight altitude distributions. An agent-based model of budworm flight, conditioned on mesoscale numerical weather simulations of temperature and wind fields, produces vertical flight density distributions that arise in combination with atmospheric boundary layer thermal profiles. We confronted our simulated moth density distributions with observed vertical profiles of moth density inferred from weather surveillance radar during budworm 2013 and 2017 outbreak events across the Gulf of St. Lawrence River in southeastern Canada. These data were used to iteratively refine the most uncertain flight parameters. Consequent dispersal patterns from broader-scaled budworm flight simulations were consistent with observations of long-distance migration events at sites in southern Quebec and New Brunswick (Canada) and in northern Maine (USA). We were able to validate outbreak source populations using a distinct indicator in differential loadings of parasitic mites. The biologically-based and meteorologically-conditioned budworm flight model holds promise as an operational tool for assessing budworm outbreak events, with the potential to inform both land management agencies and a concerned public as the eastern North American outbreak shifts into more human-dominated regions of the southeastern boreal and Acadian forests.
AUTHORS: Marion Germain*, University of Quebec in Montreal; Marc-André Villard, University of Quebec in Rimouski; Louis de Grandpré, Service Canadien des Forêts; Patrick James, University of Montreal; Dan Kneeshaw, University of Quebec in Montreal; Udaya Vepakomma, FPInnovations; Jean-François Poulin, WSP
ABSTRACT: Spruce budworm outbreaks are the most significant disturbance in North American boreal forests. Large-scale, spatially synchronous outbreaks occur periodically, causing significant mortality or growth reduction in spruce and fir over large areas. The current outbreak was first detected in 2006 on the North Shore of the St. Lawrence River and has affected >7 million ha so far. Efficient forest protection against defoliation requires early intervention, but early detection of outbreaks remains challenging. Ground-based surveys cannot be applied over large areas and aerial surveys of current defoliation cannot be used to guide early intervention. In this context, we investigated whether bird population densities can be used as indicators of future defoliation to guide early intervention. Specifically, we modelled the relationships between the occurrence of Tennessee Warbler, Cape May Warbler, and Bay-breasted Warbler and cumulative spruce budworm defoliation at different temporal lags. Using data from a large-scale (>174 000 km²) bird survey conducted between 2006 and 2016 (>1500 point counts) and annual defoliation data in Québec’s North Shore region, we explored the numerical response of each focal bird species to defoliation from 3 years before the count to 6 years after while accounting for spatial variation. Preliminary results confirm the numerical increase of each “budworm warbler” species with increasing defoliation. We also expect species-specific patterns in numerical response, with Cape May Warbler increasing earlier than Bay-breasted and Tennessee Warblers due to the different ways that each of these species use different levels of the tree crown. We expect that our landscape-level models of how warblers respond to defoliation will serve as an effective tool for forest protection by helping to identify outbreaks and to guide early intervention strategies.
AUTHORS: Barry J. Cooke*, Canadian Forestry Service; Brian R. Sturtevant, USDA Forest Service; Louis-Etienne Robert, University of Montreal; Daniel Kneeshaw, University of Quebec at Montreal
ABSTRACT: The spruce budworm (Choristoneura fumiferana Clem.) and forest tent caterpillar (Malacosoma disstria Hbn.) are early-season defoliators of spruce/fir trees and aspen/maple trees, respectively. Both exhibit periodic outbreaks – albeit at different time scales – so they are often perceived as stereotypical “cyclic” forest defoliator species. We contrasted the outbreak dynamics of these two defoliators as they related to forest landscape structure via tree-ring studies within a common landscape. The Border Lakes landscape is a large (20,000 km2) ecoregion containing contrasting land management zones with clear differences in forest landscape structure (i.e., concentration and spatial configuration of host species for each defoliator) while minimizing the confounding influence of climate. We found that outbreaks spruce budworm were more strongly periodic, more synchronous, and more severe in regions with higher concentrations of its host trees, with analogous results for forest tent caterpillar related more strongly to forest fragmentation metrics. However, we were surprised to find each species exhibited complex patterns of spatio-temporal autocovariance that led to a significant departure from purely cyclic, synchronous behaviour. Temporally, cycle peaks were distributed tri-modally, not uni-modally, as predator-prey theory would predict. Spatially, successive outbreak cycles tended to occur in disparate parts of the study area, which is not consistent with Moran’s theorem of cycle synchronization. The net emergent effect was a breakdown in cycle synchrony in those parts of the landscape where host trees were sparse. Spruce budworm tended cycle synchronously where forest tent caterpillar did not, and vice versa. Taken together, this suggests that forest landscape structure modulates cycle amplitude and synchrony, regardless of the herbivore-plant species association. It further suggests a homeostatic mechanism whereby severe outbreaks associated with high abundances of host-tree species tend to be followed by less severe outbreaks occurring in more diverse residual forests.
AUTHORS: Chris Jones, Center for Geospatial Analytics, North Carolina State University; Aaron Moody, Geography, UNC-Chapel Hill; Ross Meentemeyer, Center for Geospatial Analytics, North Carolina State University
ABSTRACT: Forest pathogens can have large-scale impacts on forest composition and can interact with natural disturbance regimes to dramatically change forest composition. But the long-term impacts of and interactions with natural disturbance regimes for newly introduced pests and pathogens are not well understood because current disease models don’t account for disease-related mortality or interactions with other disturbances. Many models of forest growth and succession work ignore the spatial distribution of species across a landscape. Therefore, we have combined a dynamic epidemiological model, a fire-behavior model, and a forest landscape simulation model (LANDIS-II) to understand how these disturbances interact and change forest composition over the course of a century using Phytophthora ramorum as our case study invasive pathogen. Three disturbance scenarios (fire, disease, and fire and disease) were used in order to understand the interacting effects of fire and disease on forest composition. The model scenarios were simulated from 1990 to 2090 using projected daily climate data. Data from these scenarios were aggregated to the entire study area and to ecoregion levels for analysis purposes. Biomass for four key species (tanoak, coast live oak, California bay laurel, and California black oak) affected by P. ramorum was analyzed over the 100-year simulation for the 3 disturbance scenarios. Our results suggest that changes in species composition are driven by asymmetries in host competency and species-specific response to fire across the study system. Additionally, initial species composition serves to either magnify or mitigate the effects of P. ramorum in the study system.
