by Audrey Verreault | 30 January 2019
Published in Canadian Journal of Forest Research 45(5): 553-565. https://doi.org/10.1139/cjfr-2014-0125
As fire is a major disturbance in boreal forests, it is now recognized that it has to be taken into account in forest management planning. Moreover, as the time of exposure to fire is related to stand productivity, combining information on productivity and fire should help in assessing the potential to sustainably manage forests. We present a method to assess potential vulnerability to the risk of fire and illustrate it in the boreal coniferous forest of Quebec. This method takes into account some sources of uncertainty related to the estimation of productivity and fire risk. Spatialization of stand productivity from growth and yield curves allowed us to compute the area comprised of productive stands of each district with or without considering fire risk. Results showed that productive area is generally decreasing with decreasing degree-days, increasing elevation, or in relation to surficial geology. Furthermore, districts with moderate to good productivity were found to be vulnerable to fire when burn rates were greater than 0.333%·year–1. Our innovative approach allowed us to assess the vulnerability of the districts to fire and could be helpful in many regions in the context of a projected increase in future area burned under climate change.
by Claire Morin | 30 January 2019
Published in Canadian Journal of Forest Research 45: 579-593. https://doi.org/10.1139/cjfr-2014-0273
Interest in northern forests is increasing worldwide for both timber production and climate change mitigation. Studies exploring forest productivity at an early age after fire and its determining factors are greatly needed. We studied forest productivity, defined as the combined quality of stocking and growth, of 116 10- to 30-year-old postfire sites. The sites were spread over a 90 000 km2 area north of the Quebec commercial forestry limit and were dominated by Picea mariana (Mill.) B.S.P. and Pinus banksiana Lamb. Seventy-two percent of our sites were classified as unproductive, mainly because of poor growth. Because growth was mostly determined by climatic factors, afforestation alone may not be sufficient to increase stand productivity in our study area. In addition, our results suggest that P. banksiana on dry sites may be less resilient to fire than previously thought, presumably because of poor site quality and climate. Overall, this is one of the first studies to explore productivity issues at an early age in natural northern forests, and the analysis scheme that defines forest productivity as the result of growth and stocking could provide a useful tool to identify similar issues elsewhere.
by Claire Morin | 30 January 2019
Published in Canadian Journal of Forest Research 45: 541-552. https://doi.org/10.1139/cjfr-2014-0124
The capacity of a forest stand to produce timber is related to the interactions that exist between its regeneration capacity, physical site characteristics (climate, surficial deposit, drainage), and disturbances. Minimally, to be sustainably managed, a forest needs to be sufficiently productive and able to regenerate after a disturbance so that its productive capacity is maintained or enhanced. To this effect, we evaluated timber productivity over a large area (175 000 km2) covering the latitudinal extent of closed-canopy black spruce (Picea mariana (Mill.) B.S.P) forest. Site index and relative density index were used to identify stands that cannot reach a minimum volume of trees of minimum size over one rotation. A nonparametric method was used to estimate their values for all stands within the study area. This imputation used either physical site attributes alone to assess potential productivity independent of stand history or physical and vegetation site attributes to assess current productivity. The proportion of productive stands was then estimated at the scale of landscapes ranging from 39 to 2491 km2. Physical site factors alone explain 84% of the variability in the percentage of potentially productive stands (78% for currently productive stands); their combination resulted in an abrupt transition in productivity over the study area. However, burn rate alone also explains 63% of variation in the proportion of currently productive stands and 41% of the relative difference between percentages of potentially or currently productive stands. These results have implications for strategic forest management planning at land classification stage, as timber production area is assumed to remain stable through time, whereas it is apparently related to the disturbance rate.
by Marie-Claude Boileau | 30 January 2019
Published in ISCHP 2015: 5th International Scientific Conference on Hardwood Processing. 15-17 septembre 2015. Québec, QC. 107-114
The establishment of precise relationships between the characteristics of standing trees and log grade volumes would improve forest management decisions in hardwood stands. In this study, a two-part conditional model (to account for the excess of zero responses within certain log grades) was used as a framework to establish this relationship for some important species in hardwood and mixed stands. Models were developed for sugar maple (Acer saccharum Marsh.), yellow birch (Betula alleghaniensis Britt.), paper birch (Betula papyrifera Marsh.) and American beech (Fagus grandifolia Ehrh.). In the model assessment process, we also tested whether the tree grade classification based on stem quality evaluation could be simplified without losing too much information about log product assortment. Higher variations among species were observed in the predictions of log grade occurrences, while the estimations of conditional volume followed similar and increasing trends with tree size. The Akaike information criteria (AIC) showed that the specifically designed tree grade classification best predicted both log grade occurrence and log grade volume for sugar maple and yellow birch, but not for paper birch and American beech. For these two species, grouping the highest quality classes of the tree grade classification system was found to be the best predictor of log grade occurrence and volume. This result can probably be explained by insufficient sampling, especially for large trees of higher grades. Our results support previous findings that tree grade classification would improve the prediction of log product assortments in northern hardwood stands.
by Claire Morin | 30 January 2019
Published in Forest Ecology and Management 375: 172-181. https://doi.org/10.1016/j.foreco.2016.05.037
Fire is considered the major disturbance in boreal forests. Nonetheless, in several areas logging has become the primary driver of forest dynamics. In many areas of the boreal forest, stands may undergo paludification (i.e. the accumulation of thick, poorly decomposed organic layers over the mineral soil) in the prolonged absence of fire, which reduces forest productivity. Whereas high-severity fires (HSF) may restore forest productivity by burning the soil organic layer (SOL), low-severity fires (LSF) mainly burn the soil surface and do not significantly reduce SOL thickness. In the Clay Belt region of eastern Canada, an area prone to paludification, forest stands have historically been harvested by clearcutting (CC), but concerns about the protection of soils and tree regeneration lead to the replacement of CC by careful logging (CL). Whereas CC disturbs the SOL and is thought to favor tree growth, CL has little impact on the SOL. Furthermore, it has been suggested that prescribed burning after clearcut (CCPB) could also be used to control paludification. Using a retrospective approach, this study sought to understand how CC, CL, and CCPB compare to LSF and HSF with respect to soil properties, SOL thickness, vegetation ground cover, tree nutrition, and stand height in paludified black spruce stands of the Clay Belt region. HSF led to significantly taller trees than CL and LSF, but did not differ from CC and CCPB. Foliar N was significantly higher in HSF and CCPB sites relative to CL and LSF, with an intermediate value in CC sites. Ground cover of Rhododendron groenlandicum was significantly lower in HSF and CC sites relative to LSF, with intermediate values in CL and CCPB sites. Sphagnum spp. ground cover was significantly lower in HSF and CCPB sites relative to CL, with intermediate values in CC and LSF sites. High-severity fire sites had a significantly thinner SOL than the four other disturbances. Finally, regression tree analysis showed that SOL thickness represented the best predictor of tree height, whereas segmented regression showed that tree height was negatively correlated to SOL thickness and revealed a cut-off point circa 23 cm, which suggests that tree growth is impeded beyond this threshold. These results support the idea that management strategies intending to regenerate paludified forests should primarily aim at reducing organic layer thickness, either through mechanical disturbance or combustion.
