Assessing and testing prediction uncertainty for single tree-based models: A case study applied to northern hardwood stands in southern Québec, Canada

Estimating prediction uncertainty for a single tree-based model is hindered by the complex structure of these models. In this paper, we addressed this issue with a case study applied to northern hardwood stands in Québec, Canada. SaMARE is a stochastic single tree-based model that was designed for these types of stands. Using a Monte Carlo approach, the model can provide a mean predicted value and its confidence limits for some plot-level attributes.The mean predicted values were compared to observed values in terms of bias and accuracy. In addition to these common statistics, we compared nominal coverage of Monte Carlo-simulated confidence intervals with real (observed) coverage to verify the adequacy of the simulated uncertainty. A comparison was made using several plot-level attributes, which exhibited an increasing discriminative complexity. This complexity ranges from coarse attributes, such as all-species basal area, up to more complex ones, such as basal area for stems of a particular species and with sawlog potential.The results showed that in terms of absolute value, biases were small, but could be relatively high with respect to the average observed value when the discriminative complexity of the attribute increased. The comparison between nominal and real coverage of confidence intervals gave satisfactory results for all-species plot-level attributes. However, for some species-specific attributes, the Monte Carlo-simulated confidence intervals overestimated the real coverage.     

Preferential bromide and pesticide movement to tile drains under different cropping practices

Subsurface drainage systems are useful tools to study chemical leaching in soils. Our objective was to compare the breakthrough behavior of bromide, atrazine (2-chloro-4-ethylamino-6-isopropylamino-s-triazine) and metolachlor [2-chloro-N-(2-ethyl-6-methylphenyl)-N-(2-methoxy-1-methylethyl) acetamid] to tile drains under two fall tillage practices (conventional tillage [CT] with a moldboard plow, and reduced tillage [RT] with a chisel plow) in field plots cultivated with corn (Zea mays L.). Leachate volume were greater in RT than in CT, with no statistical differences. Soil analysis showed that bromide migrated deeper in the soil profile than both herbicides, with little tillage effect. All chemicals were detected in drainage water at the same time and followed an event-driven behavior. Tillage had no effect on atrazine and metolachlor found in drainage water, while bromide concentration peaks were higher in RT than in CT in 1999. Concentration peaks were recorded earlier for atrazine and metolachlor than for bromide. Plots of cumulative relative chemical mass (cumulative mass divided by total mass measured in drainage) as a function of cumulative drainage were mostly linear for bromide, while they were S-shaped for both herbicides. Drainage that corresponded to 50% of relative cumulative mass ranged from 40 to 55% for bromide and from 5 to 28% for both herbicides. Rapid chemical movement to tile drains suggested that preferential flow was important in both CT and RT, and that these tillage practices had little influence on this phenomena.     

Aerial measurement of radioxenon concentration off the west coast of Vancouver Island following the Fukushima reactor accident

In response to the Fukushima nuclear reactor accident, on March 20th, 2011, Natural Resources Canada conducted aerial radiation surveys over water just off the west coast of Vancouver Island. Dose-rate levels were found to be consistent with background radiation, however a clear signal due to ¹³³Xe was observed. Methods to extract ¹³³Xe count rates from the measured spectra, and to determine the corresponding ¹³³Xe activity concentration, were developed. The measurements indicate that ¹³³Xe concentrations on average lie in the range of 30-70 Bq/m³.     

The allometry of plant spacing that regulates food intake rate in mammalian herbivores

The distance that mammalian herbivores can travel without interrupting food processing corresponds to a distance threshold (d*) in plant spacing where change occurs in the mechanisms regulating the functional response. The instantaneous rate of food consumption is controlled by food encounter rate when plant spacing exceeds d*. Below this threshold, food processing in the mouth controls instantaneous intake rate. The distance threshold provides a mechanistic definition of the scale of heterogeneity of the food resource. Recent work indicates that d* should scale positively with the body mass of mammalian herbivores. Here I evaluated the empirical evidence for this positive scaling by investigating (1) herbivores consuming only alfalfa (Medicago sativa), (2) grazers, and (3) herbivores consuming any kind of vegetation. Overall, I found greater evidence for a negative than for a positive scaling of d*. Out of the three groups, only herbivores consuming alfalfa could yield a positive covariation between d* and body mass. However, even this positive scaling became negative when herbivores consumed bites of alfalfa that were only a fraction of their maximum size. Finally, d* also scaled negatively among herbivores foraging in similar food patches. Overall, differences in foraging decisions among mammalian herbivores seem more likely to have been shaped by a negative than a positive scaling of d*.     

Incorporating putatively neutral and adaptive genomic data into marine conservation planning

The availability of genomic data for an increasing number of species makes it possible to incorporate evolutionary processes into conservation plans. Recent studies show how genetic data can inform spatial conservation prioritization (SCP), but they focus on metrics of diversity and distinctness derived primarily from neutral genetic data sets. Identifying adaptive genetic markers can provide important information regarding the capacity for populations to adapt to environmental change. Yet, the effect of including metrics based on adaptive genomic data into SCP in comparison to more widely used neutral genetic metrics has not been explored. We used existing genomic data on a commercially exploited species, the giant California sea cucumber (Parastichopus californicus), to perform SCP for the coastal region of British Columbia (BC), Canada. Using a RAD-seq data set for 717 P. californicus individuals across 24 sampling locations, we identified putatively adaptive (i.e., candidate) single nucleotide polymorphisms (SNPs) based on genotype–environment associations with seafloor temperature. We calculated various metrics for both neutral and candidate SNPs and compared SCP outcomes with independent metrics and combinations of metrics. Priority areas varied depending on whether neutral or candidate SNPs were used and on the specific metric used. For example, targeting sites with a high frequency of warm-temperature-associated alleles to support persistence under future warming prioritized areas in the southern coastal region. In contrast, targeting sites with high expected heterozygosity at candidate loci to support persistence under future environmental uncertainty prioritized areas in the north. When combining metrics, all scenarios generated intermediate solutions, protecting sites that span latitudinal and thermal gradients. Our results demonstrate that distinguishing between neutral and adaptive markers can affect conservation solutions and emphasize the importance of defining objectives when choosing among various genomic metrics for SCP.     

Two-dimensional wavelet analysis of spruce budworm host basal area in the Border Lakes landscape

Increases in the extent and severity of spruce budworm (Choristoneura fumiferana Clem.) outbreaks over the last century are thought to be the result of changes in forest structure due to forest management. A corollary of this hypothesis is that manipulations of forest structure and composition can be used to reduce future forest vulnerability. However, to what extent historical forest management has influenced current spatial patterns of spruce budworm host species is unknown. To identify landscape-scale spatial legacies of forest management in patterns of spruce budworm host species (i.e., Abies balsamea and Picea spp.), we analyzed remotely sensed forest data from the Border Lakes landscape of northern Minnesota and northwestern Ontario. Our study area contains three regions with different management histories: (1) fine-scale logging patterns in Minnesota, (2) coarse-scale logging patterns in Ontario, and (3) very limited logging history in the Boundary Waters Canoe Area and adjacent Quetico Provincial Park. We analyzed forest basal-area data using wavelets and null models to identify: (1) at which scales forest basal area is structured, (2) where those scales of pattern are significantly present, and (3) whether regions of local significance correspond to regional boundaries that separate the study area. Results indicate that spatial patterns in host basal area are created by nonstationary processes and that these processes are further constrained by lakes and wetlands. Wavelet analysis combined with significance testing revealed a bimodal distribution of scale-specific wavelet variance and separate zones of host species basal area that partially correspond with regional boundaries, particularly between Minnesota and the Wilderness region. This research represents one of the first comparisons of forest spatial structure in this region across an international border and presents a novel method of two-dimensional wavelet analysis that can be used to identify significant scale-specific structure in spatial data.     

Is Diversification a Suitable Option to Reduce Drought-Induced Risk of Forest Dieback? An Economic Approach Focused on Carbon Accounting

Environmental Modeling & Assessment - Extreme or recurrent drought events are the principal source of stress on forests, impairing their overall health. They result in financial losses for...     

Searching behavior and use of sampling information by free-ranging bison (Bos bison)

I examined the searching behavior of free-ranging plains bison (Bos bison bison) in their natural habitat, and determined whether their assessment of food patch quality was influenced by the short-term sampling information acquired during search. Bison used area-concentrated search during their winter foraging activity. Their movements between areas of suitable food patches were influenced by local environmental conditions, being sometimes less sinuous, and at other times more sinuous, than expected from a correlated random walk model. Bison also systematically avoided digging in areas where plants of low profitability lay under the snow. Where they dug, there was evidence that a bison's perception of food quality varied during a foraging bout, and was therefore influenced by short-term sampling information. After controlling for forage quality, I found that small feeding craters were more likely to be preceded by samples of high quality food patches. My observations suggest that bison take advantage of the structural characteristics of their environment during searching activity, and base foraging decisions on local rather than global availability.     

Effects of climatic data low-pass filtering on the ICBM temperature- and moisture-based soil biological activity factors in a cool and humid climate

The air temperature (T_air), total precipitation (TP) and potential evapotranspiration (PET) are standard input data for soil carbon dynamic models, i.e., for calculating temperature and moisture effects on soil biological activity. The resolution needed depends on objectives, the complexity of models and inbuilt pedotransfer functions. The Introductory Carbon Balance Model (ICBM) soil climate front end model calculates a multiplicative soil-temperature (r_{e_temp}) and -moisture (r_{e_wat}) factor with a daily time-step to estimate soil biological activity, i.e., r_{e_crop} = r_{e_temp} × r_{e_wat}. Our objective was to determine how much r_{e_temp}, r_{e_wat} and r_{e_crop} are affected by low-pass filtering of the climatic input data for a cool, humid temperate region. To achieve this we conducted spectral analysis on T_air, TP, PET and r_{e_crop} in the frequency domain. Thereafter we applied Fourier low-pass filters of 5, 15, 30, 60 and 180 days on T_air, TP, PET and tracked their effects through the soil climate model's state variables and outputs. This was done using a sandy and a heavy clay soil and an 89-year daily time-series from a meteorological station in Quebec (Canada). The Fourier spectra showed that the variance for T_air, PET and r_{e_crop} was dominated by an annual cycle, as could be expected. There was no yearly cycle for TP. The variation in r_{e_temp} explained most of the variance in r_{e_crop}. The soil climate module outputs were not sensitive to low-pass filtering of PET. A daily time-step was needed to avoid overestimating r_{e_crop} for the sandy soil. Using a weekly time-step for TP and T_air allowed us to explain about 80% of the variance in r_{e_crop} for the heavy clay soil. This study also indicates that the standard leaf (and green) area index functions for calculating transpiration should receive more attention, since they have significant effects on the state and output variables.