Acclimation of Russian forests to recent changes in climate

Assessments made over the past few decades have suggested that boreal forests may act as a sink for atmospheric carbon dioxide. However, the fate of the newly accumulated carbon in the living forest biomass is not well understood, and the estimates of carbon sinks vary greatly from one assessment to another. Analysis of remote sensing data has indicated that the carbon sinks in the Russian forests are larger than what has been estimated from forest inventories. In this study, we show that over the past four decades, the allometric relationships among various plant parts have changed in the Russian forests. To this end, we employ two approaches: (1) analysis of the database, which contains 3196 sample plots; and (2) application of developed models to forest inventory data. Within the forests as a whole, when assessed at the continental scale, we detect a pronounced increase in the share of green parts (leaves and needles). However, there is a large geographical variation. The shift has been largest within the European Russia, where summer temperatures and precipitation have increased. In the Northern Taiga of Siberia, where the climate has become warmer but drier, the fraction of the green parts has decreased while the fractions of aboveground wood and roots have increased. These changes are consistent with experiments and mathematical models that predict a shift of carbon allocation to transpiring foliage with increasing temperature and lower allocation with increasing soil drought. In light of this, our results are a possible demonstration of the acclimation of trees to ongoing warming and changes in the surface water balance. Independent of the nature of the observed changes in allometric ratios, the increase in the share of green parts may have caused a misinterpretation of the satellite data and a systematic overestimation by remote sensing methods of the carbon sink for living biomass of the Russian forest.     

Estimating regional plant biodiversity with GIS modelling

In this study, we analysed a statewide species database together with a county-level geographic information system (GIS) to build a model based on well-surveyed areas to estimate species richness in less surveyed counties. The Illinois Plant Information Network (ILPIN), a species-based database on all the vascular flora of Illinois, contains county distributions (totalling nearly 90,000) for each taxon and information on the taxonomy, ecology, biology, and ecodistribution. We compiled a statewide database with 112 variables on climate, landuse (current and historic), landscape pattern, soils and human population. We used a subset of this database to build a regression model for assessing native plant species richness for thirty-three botanically well-surveyed counties in Illinois. The best model was then used to predict the richness of the remaining sixty-nine less botanically surveyed counties. The model involved GIS (Arc/Info) and statistics (S-PLUS), including spatial statistics (S+SpatialStats). The resultant model had an R ² of 0.80 and used the following variables: percentage of the county in cropland, the percentage with soils somewhat limiting for agriculture, the percentage of urban land, and the average size of farms. Although this particular model is not transferable to other locations without validation, the methodology shown here should be useful in estimating species richness patterns across regions where botanical sampling is heterogeneous.