Northern boreal forests are characterized by accumulation of accumulation of peat (e.g., known as paludification). The functioning of northern boreal forest species and their capacity to adapt to environmental changes appear to depend on soil conditions. Climate warming is expected to have particularly pronounced effects on paludified boreal ecosystems and can alter current forest species composition and adaptation by changing soil conditions such as moisture, temperature regimes, and soil respiration. In this paper, we review and synthesize results from various reported studies (i.e., 88 research articles cited hereafter) to assess the effects of climatic warming on soil conditions of paludified forests in North America. Predictions that global warming may increase the decomposition rate must be considered in combination with its impact on soil moisture, which appears to be a limiting factor. Local adaptation or acclimation to current climatic conditions is occurring in boreal forests, which is likely to be important for continued ecosystem stability in the context of climate change. The most commonly cited response of boreal forest species to global warming is a northward migration that tracks the climate and soil conditions (e.g., temperature and moisture) to which they are adapted. Yet, some constraints may influence this kind of adaptation, such as water availability, changes in fire regimes, decomposer adaptations, and the dynamic of peat accumulation. In this paper, as a study case, we examined an example of potential effects of climatic warming on future paludification changes in the eastern lowland region of Canada through three different combined hypothetical scenarios based on temperature and precipitation (e.g., unchanged, increase, or decrease). An increase scenario in precipitation will likely favor peat accumulation in boreal forest stands prone to paludification and facilitate forested peatland expansion into upland forest, while decreased or unchanged precipitation combined with an increase in temperature will probably favor succession of forested peatlands to upland boreal forests. Each of the three scenarios were discussed in this study, and consequent silvicultural treatment options were suggested for each scenario to cope with anticipated soil and species changes in the boreal forests. We concluded that, despite the fact boreal soils will not constrain adaptation of boreal forests, some consequences of climatic warming may reduce the ability of certain species to respond to natural disturbances such as pest and disease outbreaks, and extreme weather events. View Full-Text

The accumulation of organic material on top of the mineral soil over time (a process called paludification) is common in Northern Boreal coniferous forests. This natural process leads to a marked decrease in forest productivity overtime. Topography both at the surface of the forest floor (i.e., above ground) and the subsurface (i.e., top of mineral soil which is underground) is known to play a critical role in the paludification process. Until recently, the availability of more accurate topographic information regarding the surface and subsurface was a limiting factor for land management and modeling of spatial organic layer thickness (OLT) variability, a proxy for paludification. However so far, no research has assessed which of these two topographic variables has the greatest influence on paludification. This study aims to assess which topographic variable (surface or subsurface) better explains paludification, using high-resolution remote sensing technology (i.e., Light Detection and Ranging: LiDAR and Ground Penetrating Radar: GPR). To this end, field soil measurements were made in over 1614 sites distributed throughout the reference Valrennes Experimental site in Canadian northern coniferous forests. Then, a machine learning model (i.e., Random Forest, RF) was implemented to rank a set of selected predictor topographic variables (i.e., slope, aspect, mean curvature, plan curvature, profile curvature, and topographic wetness index) using the Mean Decrease Gini (MDG) index as an indicator of importance. Results showed that overall 83% of the overall variance was explained by the RF selected model, while the derived subsurface topography predictors had the lowest MDGs for predicting paludification. On the other hand, the surface slope predictor had the highest MDGs and better explained paludification. This finding would be particularly useful for implanting sustainable management strategies based on the surface variables of paludified northern boreal forests. This study has also highlighted the potential of LiDAR data to provide surface topographic spatial detail information for planning and optimizing forest management activities in paludified boreal forests. This is even of great importance when we know that LiDAR variables are easier to obtain compared to GPR derived variables (subsurface topographic variables).

The accumulation of organic material on top of the mineral soil over time (a process called paludification) is common in Northern Boreal coniferous forests. This natural process leads to a marked decrease in forest productivity overtime. Topography both at the surface of the forest floor (i.e., above ground) and the subsurface (i.e., top of mineral soil which is underground) is known to play a critical role in the paludification process. Until recently, the availability of more accurate topographic information regarding the surface and subsurface was a limiting factor for land management and modeling of spatial organic layer thickness (OLT) variability, a proxy for paludification. However so far, no research has assessed which of these two topographic variables has the greatest influence on paludification. This study aims to assess which topographic variable (surface or subsurface) better explains paludification, using high-resolution remote sensing technology (i.e., Light Detection and Ranging: LiDAR and Ground Penetrating Radar: GPR). To this end, field soil measurements were made in over 1614 sites distributed throughout the reference Valrennes Experimental site in Canadian northern coniferous forests. Then, a machine learning model (i.e., Random Forest, RF) was implemented to rank a set of selected predictor topographic variables (i.e., slope, aspect, mean curvature, plan curvature, profile curvature, and topographic wetness index) using the Mean Decrease Gini (MDG) index as an indicator of importance. Results showed that overall 83% of the overall variance was explained by the RF selected model, while the derived subsurface topography predictors had the lowest MDGs for predicting paludification. On the other hand, the surface slope predictor had the highest MDGs and better explained paludification. This finding would be particularly useful for implanting sustainable management strategies based on the surface variables of paludified northern boreal forests. This study has also highlighted the potential of LiDAR data to provide surface topographic spatial detail information for planning and optimizing forest management activities in paludified boreal forests. This is even of great importance when we know that LiDAR variables are easier to obtain compared to GPR derived variables (subsurface topographic variables).

Soil data and soil mapping are indispensable tools in sustainable forest management. In northern boreal ecosystems, paludification is defined as the accumulation of partially decomposed organic matter over saturated mineral soils, a process that reduces tree regeneration and forest growth. Given this negative effect on forest productivity, spatial prediction of paludification in black spruce stands is important in forest management. This paper provides a description of the soil database to predict organic layer thickness (OLT) as a proxy of paludification in northeastern Canada. The database contains 13,944 OLT measurements (in cm) and their respective GPS coordinates. We collected OLT measurements from georeferenced ground plots and transects from several previous projects. Despite the variety of sources, the sampling design for each dataset was similar, consisting of manual measurements of OLT with a hand probe. OLT measurements were variable across the study area, with a mean ± standard deviation of 21 ± 24?cm (ranging from a minimum of 0?cm to a maximum of 150?cm), and the distribution tended toward positive skewing, with a large number of low OLT values and fewer high OLT values. The dataset has been used to perform OLT mapping at 30-m resolution and predict the risk of paludification in northeastern Canada (Mansuy et al., 2018) [1]. The spatially explicit and continuous database is also available to support national and international efforts in digital soil mapping.

In northern boreal ecosystems, paludification is defined as the accumulation of partially decomposed organic matter over saturated mineral soils, a process that reduces tree regeneration and forest growth. Given this negative effect on forest productivity, spatial prediction of paludification in black spruce stands is important in forest management. Here, we used the Random Forest approach to predict organic layer thickness (OLT) as a proxy of paludification in northeastern Canada, where forests tend to paludify naturally. The RF approach involved regression and classification models using a suite of 20 environmental predictors derived from multiple sources. The performance of each model was evaluated using cross-validation and an independent dataset based on conventional ecological survey maps from a provincial forest inventory. Importance measures of the predictors indicated that slope, topographic position index, spectral bands 4 and 5 from Landsat, latitude, and PALSAR_HH were the most important variables explaining the spatial distribution of OLT for both models. Cross-validated relative root mean square error (± standard deviation) for the regression model was estimated at 20.66%?±?0.576, with R2 of 0.41?±?0.020, whereas the average out-of-bag error for the classification model was estimated at 44.75%. However, both models performed better in predicting high risk of paludification (OLT values >40?cm). With predicted OLT values averaging 44.07?±?16.80?cm (range 4.25–104.58?cm), the spatial patterns were in accordance with the results of previous studies at the national and landscape scale. These results highlight that ecological types such as black spruce–sphagnum on thin-to-thick organic deposit, with ombrotrophic drainage, are particularly prone to paludification (OLT depth?>?40?cm) throughout the study area. Limitations of the models and applications for decision-making in forest management are discussed.

Boreal black spruce (Picea mariana) forests are prone to developing thick organic layers (paludification). Black spruce is adapted to this environment by the continuous development of adventitious roots, masking the root collar and making it difficult to age trees. Ring counts above the root collar underestimate age of trees, but the magnitude of age underestimation of trees in relation to organic layer thickness (OLT) is unknown. This age underestimation is required to produce appropriate age-correction tools to be used in land resource management. The goal of this study was to assess aging errors that are done with standard ring counts of trees growing in sites with different degrees of paludification (OLT; 0–25 cm, 26–65 cm, >65 cm). Age of 81 trees sampled at three geographical locations was determined by ring counts at ground level and at 1 m height, and real age of trees was determined by cross-dating growth rings down to the root collar (root/shoot interface). Ring counts at 1 m height underestimated age of trees by a mean of 22 years (range 13–49) and 52 years (range 14–112) in null to low vs. moderately to highly paludified stands, respectively. The percentage of aging-error explained by our linear model was relatively high (R2adj = 0.71) and showed that OLT class and age at 0-m could be used to predict total aging-error while neither DBH nor geographic location could. The resulting model has important implications for forest management to accurately estimate productivity of these forests.

In Quebec, reforestation by means of coniferous plantation began in the early 1980 and peaked in 1990. However, one wonders if these efforts bore fruits given the high degree of vegetation competition encountered in some plantations. The study objectives were to evaluate and compare changes in forest cover of coniferous plantations established since 1985 on clay soils in the boreal forest. The evaluation was conducted using vegetation indices extracted from images provided by Landsat TM and ETM + for the periods 2000, 2005 and 2010. Digital ortho-photos and field data were used to assess the effectiveness of these indices. The Tasseled Cap Greenness Index proved to be the most efficient to discriminate the type of cover with a Kappa index of 50%. The results show that over 50% of the plantations were classified as mixed (deciduous – softwood) and 25% were classified as deciduous in 2000, while in 2010, 51 % of plantations were mixed and 11 % deciduous. Using a generalized linear mixed model with Laplace approximation, we have established the relationship between changes in cover, age and type of the plantation. The likelihood of an initial deciduous cover to evolve toward a mixed or coniferous cover was nearly 68 %. The plantation type, its age and cover and their interactions all had significant effects on the probability of a cover to change over time. This project contributed to the assessment of coniferous plantations established between 1980 and 1995 over an area of 18797 ha and confirms that remote sensing tools are effective for tracking large areas while providing quality information for the planning of silvicultural treatments.

Northern Canadian boreal forest is characterised by accumulation of a thick organic soil layer (paludification). Two types of paludification are recognised on the basis of topography and time since the last fire, viz., permanent paludification that dominates in natural depressions within the landscape, and reversible paludification that occurs on flat or sloping terrain over time following fire or mechanical site preparation. Accurate information about the occurrence of permanent or reversible paludification is required for land resource management. Such information is useful for the identification of locations of existing paludified areas where investment after harvesting should help to achieve greater productivity. This study investigated the potential for using a semi-automated method that was based on geomorphological analysis to map and differentiate between the two paludification types at the landscape scale within the Canadian Clay Belt region. For the purposes of this study, slope, topographic position index (TPI), and topographic wetness index (TWI) were generated from a LiDAR digital terrain model. TPI and TWI are, respectively, predictors of surface morphology (i.e., depressions vs flat areas) and moisture conditions (i.e., wet vs dry), and were used to explain paludification occurrence. A semi-automated classification method based on TPI and slope was firstly used to create six initial topographic position classes: deep-depressions, lower-slope depressions, flat surfaces, mid-slopes, upper-slopes, and hilltops. Each of these six classes was then combined with TWI classes (representing moisture conditions: wet, moderately wet, and dry) and this combination assisted in assigning each resulting class to one of the two paludification types. Slope and TWI values were used in sub-dividing the lower slope depression class, based on slope, into significantly different sub-classes, namely open and closed depressions (Tukey's HSD, P < 0.001). The distribution of field data (e.g., tree basal area, organic layer and fibric horizon thicknesses) within each position class provided additional information for corroborating the assignment of each class to a defined paludification type. The proposed semi-automated classification provided a relatively simple and practical tool for distinguishing and mapping permanent and reversible paludification types with an overall accuracy of 74%. The tool would be particularly useful for implementing strategies of sustainable management in remote boreal areas where field survey information is limited.

La forêt boréale canadienne a une importance écologique et économique considérable. Toutefois, les forêts d'épinettes noires situées dans la ceinture d'argile, une région boréale de l'est de l'Amérique du Nord, sont sujettes à la paludification. Ce phénomène est un processus naturel par lequel une couche organique s'accumule sur le sol forestier conduisant à une diminution importante de la productivité de ces forêts. Théoriquement, il existe deux types de paludification, à savoir la paludification permanente et réversible. La paludification permanente se produit dans des endroits où les conditions d'humidité du sol sont élevées (ex., reliefs plats, dépressions topographiques); alors que la paludification réversible intervient dans des sites à pente faible ou moyennement forte au fil du temps en réponse à une perturbation telle qu'un feu peu sévère. L'épaisseur de la couche organique (ECO) et la topographie constituent des paramètres clefs de la présence de la paludification dans cette région et y affectent négativement la productivité. La recherche proposée dans cette thèse consiste à approfondir la compréhension et la détection du phénomène de paludification dans les forêts d'épinette noire dans une perspective de maintien ou d'accroissement de la productivité des arbres. Cette thèse a pour objectif principal de déterminer et de sélectionner les variables permanentes des site permettant d'expliquer les écarts de productivité de la pessière à épinette noire observés dans la ceinture d'argile à l'aide des méthodes à haute résolution et des données recueillies sur le terrain. L'expression de ces variables a été ensuite utilisée pour prédire la productivité actuelle et potentielle des sites soumis à la paludification. Les objectifs spécifiques de cette thèse étaient de (1) détecter et identifier d'une manière continue l'interface sol minéral/couche organique afin de cartographier la topographie du sol minéral à l'échelle des sites paludifiés; (2) étudier d'une façon quantitative les relations entre l'ECO et la topographie (au niveau du sol minéral et de la surface) afin de caractériser ces relations à l'échelle du paysage, notamment la distribution et la variabilité spatiale de l'ECO; (3) identifier les variables topographiques permettant de distinguer et de cartographier la paludification réversible et la paludification permanente à l'échelle du paysage; et (4) évaluer l'effet de l'ECO et des variables topographiques, exprimées à différentes résolutions spatiales, sur la productivité forestière des forêts paludifiées de la ceinture d'argile. Le but ultime est d'améliorer notre compréhension de la façon dont ces variables ainsi que leurs résolutions influencent la productivité des arbres dans les forêts d'épinette noire.

Les résultats du premier chapitre de la thèse ont démontré que la méthode géophysique géoradar, ayant une bonne corrélation de ses résultats avec les données du terrain (r = 0,93; P < 0,001), a permis d'obtenir une cartographie précise, continue et fiable de l'interface couche organique/sol minéral dans des sites faiblement à modérément paludifiés. Cependant, en dépit de son incapacité à cartographier l'interface couche organique/sol minéral dans les sites hautement paludifiés, le recours au géoradar s'est révélé pertinent dans la mise en évidence de l'interface horizon fibrique/couche organique et de sa continuité spatiale. Cela rend le géoradar particulièrement intéressant dans la détection des niveaux d'entourbement constituant ainsi une méthode de détection indirecte prometteuse pour l'aménagement des forêts paludifiées.

Le deuxième objectif a été abordé dans deux différents chapitres (II et III) à l'aide d'une approche quantitative de modélisation de l'ECO par arbre de régression. Différentes variables topographiques (élévation, pente, exposition, indice topographique d'humidité (TWI), courbure totale, courbure transversale, et courbure horizontale) ont été utilisées dans les modèles sélectionnés des deux chapitres. D'une façon générale, nous avons démontré que les topographies de surface et du sol minéral influencent l'accumulation de la couche organique à l'échelle du paysage dans la ceinture d'argile. Les résultats du deuxième chapitre ont permis de délimiter les principaux patrons de l'ECO et d'élucider trois relations spatiales entre l'ECO et les variables explicatives: (i) les zones avec une couche organique épaisse (62 cm) avaient des pentes douces (≤ 1,8%); (ii) les zones avec pentes plus raides (> 3,2 %) ont été associées à une couche organique peu profonde (27 cm); et (iii) les résultats les plus significatifs ont été obtenus avec des résolutions 10 et 20 m en comparaison au 1 et 5 m. Le troisième chapitre a permis de mettre en évidence les différentes relations entre la topographie du sol minéral et l'ECO à l'échelle du paysage. La construction d'un modèle numérique d'élévation au niveau du sol minéral à l'échelle du paysage est un élément central de notre démarche. Les modelés développés nous permettent d'affirmer que : (i) la pente du sol minéral, la composition du sol minéral (argile, till et régolithe), le TWI et l'exposition sont les quatre principales variables influençant l'accumulation de la couche organique; (ii) les valeurs seuils de pente du sol minéral > 3,5% et ≤ 2% permettent respectivement de distinguer les zones les plus prometteuses et les plus vulnérables pour l'aménagement forestier; (iii) les zones avec une exposition nord et est étaient associées à une couche organique plus profonde par rapport à celles exposées vers le sud et l'ouest; et (iv) la distinction entre les zones paludifiées et non paludifiées sur la base d'une valeur seuil de la pente du sol minéral de l'ordre de 3,5% constitue un des apports majeurs de cette étude.

Afin de répondre au troisième objectif, une approche semi-automatique de subdivision sous SIG du territoire à l'étude en des entités du paysage distinctes a été réalisée en combinant des données topographiques, notamment l'indice topographique de position (TPI), l'indice topographique d'humidité (TWI) ainsi que la pente de surface. Cette approche s'est révélée efficace, car elle a permis de délimiter des entités possédant des caractéristiques géomorphologiques semblables, notamment en terme de susceptibilité à l'accumulation de la couche organique, et par conséquent ont été assignées à l'un ou l'autre type de paludification, soit réversible ou permanente. Un apport majeur de cette approche semi-automatisée est la mise en évidence de deux sous-entités statistiquement différentes (le test HSD de Tukey, P < 0,001), à savoir des dépressions ouvertes préférentiellement drainées (paludification réversible) et des dépressions fermées potentiellement engorgées (paludification permanente) du fait de leurs positions topographiques et conditions d'humidité. Cela rend l'outil développé particulièrement utile pour la mise en oeuvre des stratégies d'aménagement durable dans les forêts paludifiées.

Pour atteindre le quatrième objectif, deux modèles ont été explorés pour la modélisation de la productivité (exprimée par l'indice de qualité de station (IQS) dans notre cas) en utilisant une approche par arbre de régression. Le premier modèle contient les variables topographiques et l'ECO, alors que le deuxième modèle inclut seulement les variables topographiques issues des données LiDAR. Les résultats de cette modélisation ont démontré que l'ECO, l'exposition et la pente sont les trois variables les plus importantes pour expliquer la productivité forestière à l'échelle du paysage; et pour déterminer des seuils d'ECO et des variables topographiques qui permettent de caractériser, à la fois, des zones productives et improductives. En effet, les zones avec une productivité élevée étaient associées à une couche organique peu profonde (< 35 cm) et à des pentes orientées sud-ouest et dont la valeur est supérieure à 2,2%, favorisant ainsi une plus forte croissance des arbres; en revanche, les zones avec une faible productivité avaient une couche organique très profonde (> 85 cm), favorisant l'invasion de mousses et de sphaignes. Du point de vue de l'échelle (résolutions), le premier modèle semble relativement indépendant de l'échelle, alors que la réponse du deuxième modèle augmentait significativement avec la taille du pixel. Ces résultats pourraient donc être appliqués à des échelles opérationnelles et là où des informations sur l'ECO sont disponibles afin de prédire la productivité. Des cartes thématiques prédictives de la distribution spatiale de la productivité ont été réalisées avec nos deux modèles.

Les résultats de cette thèse ont permis d'approfondir les connaissances sur les variables permettant d'expliquer les écarts de productivité dans la pessière à épinette noire ainsi que sur l'importance des variables topographiques dans la modélisation de l'ECO et la productivité à l'échelle du paysage. De plus, cette étude a permis de caractériser la distribution spatiale des deux types de paludification (permanente et réversible) à l'échelle du paysage. Quoique cette étude s'intéresse plus particulièrement à la pessière à épinette noire, les connaissances acquises pourraient être applicables à d'autres territoires paludifiés de la forêt boréale.

Mineral soil topography is difficult to describe in boreal regions because of the thick overlying organic layer despite its presumed importance in determining where and at what rate an organic layer will accumulate (paludification). The overall purpose of this study was to examine the relationship between mineral soil topography and OLT at the landscape scale. More specifically, these relationships can be used to map the distribution and spatial variability of paludification across the landscape, thereby exploring the potential to discriminate between the two commonly known paludification types (permanent and reversible). Seven topographic variables (elevation, slope, aspect, mean curvature, plan curvature, profile curvature and topographic wetness index) were generated from a digital elevation model that we developed for the mineral soil surface (MS-DEM). OLT data were collected from field measurements across the landscape by manual probing and values varied from 5 to 150 cm. The MS-DEM was generated by subtracting OLT field values from the corresponding LiDAR-derived elevation values. Most correlations between OLT and individual predictor variables were weak and illustrated that OLT and its landscape-scale distribution cannot be explained by simple bivariate relationships. Consequently, two regression tree-based models were developed using: (1) only the seven mineral soil topographic variables, and (2) all predictor variables (mineral soil topography and surficial deposits). Mineral soil slope was the most important variable for both models and corresponded to the first level of splitting the dataset into homogenous landscape units in terms of organic layer thickness. Surficial deposit, topographic wetness index (TWI) and aspect were also related to OLT and proved to be contributing to the development of the two models. Model 1 explained 0.34 of the OLT variability and offer simple models with few landscape units that are easy to interpret. Model 1 splitting rules allowed the combination of different maps (slope, TWI and aspect) for producing a landscape units map, on which OLT was determined and related to increasing paludification categories. A good overall accuracy of 74% was achieved for this map. Model 2 was the best model in terms of estimate quality (R2adj = 0.52). Both models were successful in discriminating highly paludified landscape units. Except for one landscape unit that was assigned to permanent paludification type, both models were unable to further subdivide more landscape units into reversible and permanent paludification, suggesting that both of these types interact within the same landscape unit. This study demonstrated that the combination of topographic information from remotely sensed LiDAR data and field OLT measurement data has the potential to be useful for defining both promising and vulnerable areas for forest management.