Forest tree species distribution for Europe 2000-2020: mapping potential and realized distributions using spatiotemporal Machine Learning

Paper describes a data-driven framework based on spatio-temporal ensemble machine learning to produce distribution maps for 16 forest tree species (Abies alba Mill., Castanea sativa Mill. , Corylus avellana L., Fagus sylvatica L., Olea europaea L., Picea abies L. H. Karst., Pinus halepensis Mill., Pinus nigra J. F. Arnold, Pinus pinea L., Pinus sylvestris L., Prunus avium L., Quercus cerris L., Quercus ilex L., Quercus robur L., Quercus suber L. and Salix caprea L.) at high spatial resolution (30 m). Tree occurrence data for a total of 3 million of points was used to train different Machine Learning (ML) algorithms: random forest, gradient-boosted trees, generalized linear models, k-nearest neighbors, CART and an artificial neural network. A stack of 585 coarse and high resolution covariates representing spectral reflectance (Landsat bands, spectral indices; time-series of seasonal composites), different biophysical conditions (i.e. temperature, precipitation, elevation, lithology) and biotic competition (other species distribution maps) was used as predictors for realized distributions, while potential distribution was modelled with environmental predictors only. Logloss and computing time were used to select the three best algorithms to train an ensemble model based on stacking with a logistic regressor as a meta-learner for each species. High resolution (30 m) probability and model uncertainty maps of realized distribution were produced for each species using a time window of 4 years for a total of 6 distribution maps per species for the studied period, while for potential distributions only one map per species was produced. Results of spatial cross validation show that Olea europaea and Quercus suber achieved the best performances in both potential and realized distribution, while Pinus sylvestris and Salix caprea achieved the worst. Further analysis shows that fine-resolution models consistently outperformed coarse resolution models (250 m) for realized distribution (average decrease in logloss: +53%). Realized distribution models achieved higher predictive performances than potential distribution ones. Importance of predictor variables differed across species and models, with the green band for summer and the NDWI and NDVI for fall for realized distribution and the diffuse irradiation and precipitation of the driest quarter being the most important and frequent for potential distribution. The ensemble model outperformed or performed as good as the best individual model in all potential species distributions, while for ten species it performed worse than the best individual model in modeling realized distributions. The framework shows how combining continuous and consistent EO time series data with state of the art ML can be used to derive dynamic distribution maps. The produced time-series occurrence predictions can be used to quantify temporal trends and detect potential forest degradation.

Is Cork Growth a Reliable Proxy for Stem Diameter Growth in Cork Oak (Quercus suber L.)? Implications for Forest Management under Climate Change in Mediterranean Regions

Cork-ring widths have been extensively used in dendroecological studies assessing the relationship between cork growth patterns and climate (precipitation and temperature). Generally, cork growth is assumed as a proxy for stem diameter growth to address cork oak (Quercus suber L.) growth sensitivity to climate and cork yield modeling. Cork growth represents a large part of stem radial increment in this species due to the enhanced activity of phellogen when compared to the cambium activity; thus, similar inter-annual variations of cork-ring widths and tree diameter growth might be expected. However, so far, the influence of rainfall and temperature on stem diameter growth has scarcely been addressed; moreover, it is still not clear whether tree size relates, and in what way, to the variations in radial growth of cork and stem diameter and whether these reflect (proportional) quantitative variations in stem basal area growth. In this study, we computed the annual growth of cork and of stem diameter at breast-height in data series of 47 trees, from 2000 to 2012, corresponding to a full cork production cycle. Results showed a tight link between cork-ring width and stem diameter growth indices. However, while cork growth strongly correlated with climate conditions in autumn–winter prior to the growing season, stem diameter growth correlated with climate conditions of the current growing season, and, more importantly, it was tree size-related. The extrapolation from cork-ring increments to stem basal area growth is likely to progressively underestimate tree growth and biomass increment in larger cork oaks and to further bias it due to climate change effects in the Mediterranean region.

A Cork Cell Wall Approach To Swelling And Boiling

The bark of cork oak (Quercus suber L.) is mostly used for cork stopper production, whereas bark is undergoing a series of industrial procedures, boiling usually leading to changes in the characteristics of the tissue. Trees are traditionally grown under natural conditions; however, irrigation is now being used in plantations. These permanent water availability affects cork-oak development, while its effects on industrial procedures is unknown. This study provides a first insight into the behaviour of the cell walls of cork during the process of swelling and boiling when trees have been grown under irrigation, subject to a specific water regime. Cork tissue was analysed using environmental and scanning electron microscopy under three regimes: raw conditions; following immersion in water; and after boiling. Additionally, the radial expansion of samples was determined. The results showed greater cell-wall expansion in cork from the irrigated site than cork from the traditional rainfed plot, when hydrated for 24h. After boiling, the cell walls of the rainfed site were thinner than in the raw stage, in contrast to the irrigated cork. This study suggests that irrigation during cork-oak growth produces a higher capacity for adsorption, increasing cell-wall thickness from the raw stage to the boiling stage.

Linking Cork to Cork Oak Landscapes: Mapping the Value Chain of Cork Production in Portugal

Traditional farming landscapes in South and Central Portugal, known as montados, are affected by global socio-economic and biophysical pressures, putting the sustainability of the systems in jeopardy. Cork oak trees (Quercus suber L.) are characteristic features of these complex agro-silvo-pastoral agroforestry systems, delivering a globally important product, cork. The increasingly distant, global scale of decision making and trade can consequently be observed on the local, landscape, scale. In this study, we use a value chain approach to test the concept that landscape products can ensure sustainable management of the landscape of origin. We interviewed agents—cork producers, intermediaries, industrial transformers, and winemakers—about the challenges they perceived in the business and how these were connected to the landscape of origin. We illustrate the network of agents and sub-actors involved in the sector and highlight the most prominent concerns. We conclude that this approach can reveal the major points for determining the future of the montado, and we suggest that collaboration amongst value chain agents can be a pathway to landscape sustainability.

Anticandidosic Activity and Acute Toxicity of Quercus suber L. Bark Extracts

The cork oak (Quercus suber L.), endemic essence of the Mediterranean Basin, is commonly used in traditional pharmacopoeia. The main objective of this work is to enhance the valorization of this plant species through the study of the anticandidosic activity of cork oak bark methalonic extracts in order to develop an efficient natural formulation for Candidiasis treatment.The anticandidosic activity of methanolic extracts of Q. suber bark stemming from decoction, maceration and Soxhlet methods of extraction in was tested on five different Candida albicans strains. Our results showed that all the tested extracts displayed an inhibitive activity, which varies according to the obtained extract and the tested strain. The best anticandidosic potential was observed with extracts obtained with Soxhlet method. The study of the acute toxicity showed that the lethal dose is 1150 mg/kg in mice, which remained moderately toxic according to Hodge and Sterner classification scale. Thus, this extract can be used in phytotherapy without danger in doses that are below 300 mg/kg of corporal weight. Based on these results, we can conclude that Cork oak bark extracts can be used to treat Candida albicans infections.

Identification and in-silico characterization of taxadien-5α-ol-O-acetyltransferase (TDAT) gene in Corylus avellana L.

Paclitaxel® (PC) is one of the most effective and profitable anti-cancer drugs. The most promising sources of this compound are natural materials such as tissue cultures of Taxus species and, more recently, hazelnut (Corylus avellana L.). A large part of the PC biosynthetic pathway in the yew tree and a few steps in the hazelnut have been identified. Since understanding the biosynthetic pathway of plant-based medicinal metabolites is an effective step toward their development and engineering, this paper aimed to identify taxadiene-5α-ol-O-acetyltransferase (TDAT) in hazelnut. TDAT is one of the key genes involved in the third step of the PC biosynthetic pathway. In this study, the TDAT gene was isolated using the nested-PCR method and then characterized. The cotyledon-derived cell mass induced with 150 μM of methyl jasmonate (MeJA) was utilized to isolate RNA and synthesize the first-strand cDNA. The full-length cDNA of TDAT is 1423 bp long and contains a 1302 bp ORF encoding 433 amino acids. The phylogenetic analysis of this gene revealed high homology with its ortholog genes in Quercus suber and Juglans regia. Bioinformatics analyses were used to predict the secondary and tertiary structures of the protein. Due to the lack of signal peptide, protein structure prediction suggested that this protein may operate at the cytoplasm. The homologous superfamily of the T5AT protein, encoded by TDAT, has two domains. The highest and lowest hydrophobicity of amino acids were found in proline 142 and lysine 56, respectively. T5AT protein fragment had 24 hydrophobic regions. The tertiary structure of this protein was designed using Modeler software (V.9.20), and its structure was verified based on the results of the Verify3D (89.46%) and ERRAT (90.3061) programs. The T5AT enzyme belongs to the superfamily of the transferase, and the amino acids histidine 164, cysteine 165, leucine 166, histidine 167, and Aspartic acid 168 resided at its active site. More characteristics of TDAT, which would aid PC engineering programs and maximize its production in hazelnut, were discussed.