Predicting tree species presence and basal area in Utah: A comparison of stochastic gradient boosting,generalized additive models,and tree-based methods

Many efforts are underway to produce broad-scale forest attribute maps by modelling forest class and structure variables collected in forest inventories as functions of satellite-based and biophysical information. Typically, variants of classification and regression trees implemented in Rulequest's^© See5 and Cubist (for binary and continuous responses, respectively) are the tools of choice in many of these applications. These tools are widely used in large remote sensing applications, but are not easily interpretable, do not have ties with survey estimation methods, and use proprietary unpublished algorithms. Consequently, three alternative modelling techniques were compared for mapping presence and basal area of 13 species located in the mountain ranges of Utah, USA. The modelling techniques compared included the widely used See5/Cubist, generalized additive models (GAMs), and stochastic gradient boosting (SGB). Model performance was evaluated using independent test data sets. Evaluation criteria for mapping species presence included specificity, sensitivity, Kappa, and area under the curve (AUC). Evaluation criteria for the continuous basal area variables included correlation and relative mean squared error. For predicting species presence (setting thresholds to maximize Kappa), SGB had higher values for the majority of the species for specificity and Kappa, while GAMs had higher values for the majority of the species for sensitivity. In evaluating resultant AUC values, GAM and/or SGB models had significantly better results than the See5 models where significant differences could be detected between models. For nine out of 13 species, basal area prediction results for all modelling techniques were poor (correlations less than 0.5 and relative mean squared errors greater than 0.8), but SGB provided the most stable predictions in these instances. SGB and Cubist performed equally well for modelling basal area for three species with moderate prediction success, while all three modelling tools produced comparably good predictions (correlation of 0.68 and relative mean squared error of 0.56) for one species.     

A stratified model for the assessment of meteorologically adjusted trends of surface ozone

We consider the problem of assessing long-term trends of ozone concentrations measured on a single site located in an urban area. Among the many methods proposed in the literature to eliminate the confounding effect of changing weather conditions, we employ a stratification of daily maxima based on regression trees. Within each stratum conditional independence and Weilbull distribution are assumed for maxima. Long-term trend is defined non-parametrically by the sequence of yearly medians. Models are estimated following the Bayesian approach. The alternative assumptions of common and stratum specific trends are compared and a model with common trend for all strata is selected for the analyzed real dataset. The conditional independence assumption is checked by the comparison with a model including an autoregressive component.     

镉对5种常绿树木若干生理生化特性的影响

研究了5种常绿树木对镉污染胁迫的反应,实验结果表明,在高浓度镉胁迫下,5种树木叶片的叶绿素含量、细胞质膜透性、过氧化氢酶活性及镉富集量等生理生化特性均产生明显变化,但变化不尽一致。其中,黄杨、海桐,杉木抗镉污染能力优于香樟和冬青。      

Habitat-tree protection concepts over 200 years

The protection and sustainable management of habitat trees is an integral part of modern forest nature conservation concepts such as retention forestry. Bats, cavity-nesting birds, arboreal marsupials, and many different saproxylic species depend on habitat trees and their great variety of microhabitats and old-growth characteristics. With a focus on insights from temperate forests, we traced the development of habitat-tree protection over 200 years. The idea was first conceptualized by foresters and natural scientists in the early 19th century. At that time, utilitarian conservation aimed to protect cavity trees that provided roosts and nesting holes for insectivorous bats and birds. By the second half of the 19th century, habitat-tree protection was well known to foresters and was occasionally implemented. Knowledge of the protection of large old trees, a special kind of habitat tree, for sociocultural and aesthetic reasons developed similarly. But, many foresters of that time and in the following decades fundamentally rejected protection of habitat trees for economic reasons. Beginning in the 1970s, forest conservation and integrative forest management became increasingly important issues worldwide. Since then, the protection of habitat trees has been implemented on a large scale. Long-term views on the development of conservation concepts are important to inform the implementation of conservation today. In particular, historical analyses of conservation concepts allow the testing of long-term conservation outcomes and make it possible to study the resilience of conservation approaches to changing social or ecological conditions. We encourage all conservation ecologists to assess the practical and conceptual impact of the initial ideas that led to modern conservation concepts in terms of long-term biodiversity conservation.     

Overlooked biodiversity loss in tropical smallholder agriculture

Smallholder agriculture is the main driver of deforestation in the western Amazon, where terrestrial biodiversity reaches its global maximum. Understanding the biodiversity value of the resulting mosaics of cultivated and secondary forest is therefore crucial for conservation planning. However, Amazonian communities are organized across multiple forest types that support distinct species assemblages, and little is known about smallholder impacts across the range of forest types that are essential for sustaining biodiversity. We addressed this issue with a large-scale field inventory of birds (point counts) and trees (transects) in primary forest and smallholder agriculture in northern Peru across 3 forest types that are key for Amazonian biodiversity. For birds smallholder agriculture supported species richness comparable to primary forest within each forest type, but biotic homogenization across forest types resulted in substantial losses of biodiversity overall. These overall losses are invisible to studies that focus solely on upland (terra firma) forest. For trees biodiversity losses in upland forests dominated the signal across all habitats combined and homogenization across habitats did not exacerbate biodiversity loss. Proximity to forest strongly predicted the persistence of forest-associated bird and tree species in the smallholder mosaic, and because intact forest is ubiquitous in our study area, our results probably represent a best-case scenario for biodiversity in Amazonian agriculture. Land-use planning inside and outside protected areas should recognize that tropical smallholder agriculture has pervasive biodiversity impacts that are not apparent in typical studies that cover a single forest type. The full range of forest types must be surveyed to accurately assess biodiversity losses, and primary forests must be protected to prevent landscape-scale biodiversity loss.     

Evaluating the Role of Plantations as Carbon Sinks: An Example of an Integrative Approach from the Humid Tropics

/ Despite their fast growth, tropical plantations are a small sink of atmospheric carbon because they occupy only a small area in relation to other land uses worldwide. Proper design and management of plantations can increase biomass accumulation rates, making them more effective C sinks. However, fast-growing plantations can extract large amounts of nutrients from the soil, and site fertility declines may limit sustained plantation forestry after a few rotations. We measured aboveground biomass accumulation, carbon sequestration, and soil chemistry in three young plantations of 12 indigenous tree species in pure and mixed designs in the humid lowlands of Costa Rica. Annual biomass increments for the three mixed plantations ranged from 10-13 Mg/ha. The mixtures of four species gave higher biomass per hectare than that obtained by the sum of one fourth hectare of each species in pure plots. At this early age of the plantations, estimated annual C sequestration values were comparable to other reports from young plantations of exotic species commonly grown in the tropics. Four years after planting, decreases in soil nutrients were apparent in pure plots of some of the fastest growing species, while beneficial effects on soils were noted under other species. The mixed plots showed intermediate values for the nutrients examined and, sometimes, improved soil conditions. A mixture of fast and slower growing species yields products at different times, with the slower growing species constituting a longer term sink for fixed carbon. Examination of the role of tropical plantations as C sinks necessitates integrative approaches that consider rates of C sequestration, potential deleterious effects on ecosystem nutrients, and economic, social, and environmental constraints.KEY WORDS: Native trees; Aboveground biomass; Stem increments; Rotation length; Soil nutrients; Economics     

Technogenic stress and its effect on deciduous trees (an example from parks in Yaroslavl)

A comparative analysis of several morphological, physiological, and biochemical characteristics of leaves has been performed in birch (Betula pendula Roth), linden (Tilia cordata Mill.), and poplar (Populus nigra L.) trees growing in industrial areas of Yaroslavl with different levels of anthropogenic pollution. The results show that high concentrations of pollutants in the atmosphere suppress the growth of leaves, and the number and severity of necrotic lesions in them increase. Water content in leaves becomes lower, whereas ash content increases. The contents of photosynthetic pigments change as well: the leaves contain smaller amounts of chlorophylls a and b, whereas the accumulation of carotenoids increases. Judging from changes in the test parameters, P. nigra is more resistant to pollution than B. pendula and T. nordata.     

Risk assessment in support to land-use planning in Europe: Towards more consistent decisions?

Recognising the importance of establishing appropriate separation distances between hazardous installations and vulnerable residential areas for mitigating the effects of industrial accidents, the European legislation for the control of major accident hazards - the so-called Seveso II Directive - calls for procedures ensuring that technical advice is taken systematically into account for land-use planning (LUP) purposes. Due to historical, administrative, cultural and other reasons, these European Union’s Member States which have consolidated procedures for addressing this issue, have employed different approaches, methods and criteria, with a potential for great divergence in the resulting land-use planning decisions. In order to address this situation and to increase consistency and ‘defendability’ of land-use planning decisions in the EU, a European Working Group has been established and is operating under the coordination of the European Commission’s Joint Research Centre (JRC). This Group, consisting of experts from the EU Member States, the industry and the academia, is trying to understand the different approaches and their implications to LUP decision-making, to develop guidelines in support to these decisions and to examine data sources and tools for consistent application of risk assessment in support to LUP. This paper presents the activities of the Group, reviews the situation with respect to LUP in Europe and discusses whether a direction towards more consistent LUP decisions is being followed in Europe.     

Evaluation of Fenton and ozone-based advanced oxidation processes as mature landfill leachate pre-treatments

Fenton treatment (Fe²⁺/H₂O₂) and different ozone-based Advanced Oxidation Processes (AOPs) (O₃, O₃/OH⁻ and O₃/H₂O₂) were evaluated as pre-treatment of a mature landfill leachate, in order to improve the biodegradability of its recalcitrant organic matter for subsequent biological treatment. With a two-fold diluted leachate, at optimised experimental conditions (initial pH 3, H₂O₂ to Fe²⁺ molar ratio of 3, Fe²⁺ dosage of 4 mmol L⁻¹, and reaction time of 40 min) Fenton treatment removed about 46% of chemical oxygen demand (COD) and increased the five-day biochemical oxygen demand (BOD₅) to COD ratio (BOD₅/COD) from 0.01 to 0.15. The highest removal efficiency and biodegradability was achieved by ozone at higher pH values, solely or combined with H₂O₂. These results confirm the enhanced production of hydroxyl radical under such conditions. After the application for 60 min of ozone at 5.6 g O₃ h⁻¹, initial pH 7, and 400 mg L⁻¹ of hydrogen peroxide, COD removal efficiency was 72% and BOD₅/COD increased from 0.01 to 0.24. An estimation of the operating costs of the AOPs processes investigated revealed that Fe²⁺/H₂O₂ was the most economical system (8.2 € m⁻³ g⁻¹ of COD removed) to treat the landfill leachate. This economic study, however, should be treated with caution since it does not consider the initial investment, prices at plant scale, maintenance and labour costs.     

Comparison and ranking of different modelling techniques for prediction of site index in Mediterranean mountain forests

Forestry science has a long tradition of studying the relationship between stand productivity and abiotic and biotic site characteristics, such as climate, topography, soil and vegetation. Many of the early site quality modelling studies related site index to environmental variables using basic statistical methods such as linear regression. Because most ecological variables show a typical non-linear course and a non-constant variance distribution, a large fraction of the variation remained unexplained by these linear models. More recently, the development of more advanced non-parametric and machine learning methods provided opportunities to overcome these limitations. Nevertheless, these methods also have drawbacks. Due to their increasing complexity they are not only more difficult to implement and interpret, but also more vulnerable to overfitting. Especially in a context of regionalisation, this may prove to be problematic. Although many non-parametric and machine learning methods are increasingly used in applications related to forest site quality assessment, their predictive performance has only been assessed for a limited number of methods and ecosystems.In this study, five different modelling techniques are compared and evaluated, i.e. multiple linear regression (MLR), classification and regression trees (CART), boosted regression trees (BRT), generalized additive models (GAM), and artificial neural networks (ANN). Each method is used to model site index of homogeneous stands of three important tree species of the Taurus Mountains (Turkey): Pinus brutia, Pinus nigra and Cedrus libani. Site index is related to soil, vegetation and topographical variables, which are available for 167 sample plots covering all important environmental gradients in the research area. The five techniques are compared in a multi-criteria decision analysis in which different model performance measures, ecological interpretability and user-friendliness are considered as criteria.When combining these criteria, in most cases GAM is found to outperform all other techniques for modelling site index for the three species. BRT is a good alternative in case the ecological interpretability of the technique is of higher importance. When user-friendliness is more important MLR and CART are the preferred alternatives. Despite its good predictive performance, ANN is penalized for its complex, non-transparent models and big training effort.