Vegetation productivity

Indicator Assessment
Prod-ID: IND-480-en
Also known as: LSI 009
Created 25 Apr 2019 Last modified 07 Feb 2020
18 min read

This item is open for comments. Login with your Eionet account in order to see and add comments. See comments section below

Vegetation productivity indicates the spatial distribution and change of the vegetation cover - a key characteristic of ecosystem condition.  On average, vegetation productivity increased most in South Eastern Europe, over croplands and wetlands in the Steppic region and grasslands and sparsely vegetated lands and in the Black Sea and Anatolian regions. Vegetation productivity declined most over croplands and grasslands in the Atlantic region as well as over wetlands in the Alpine region. Increasing precipitation especially in the South Eastern regions had the strongest influence on productivity increase. Decreasing number of frost days were strongly correlated to increasing productivity in the Pannonian region and to decreasing productivity in the Atlantic region.  Land use change was a stronger driver of productivity changes than climatic variations.  Converting lands to agriculture and agriculture internal conversions resulted in highest productivity increase whereas increasing urban sprawl resulted in most productivity decrease. 

Key messages

Vegetation productivity indicates the spatial distribution and change of the vegetation cover - a key characteristic of ecosystem condition. 

On average, vegetation productivity increased most in South Eastern Europe, over croplands and wetlands in the Steppic region and grasslands and sparsely vegetated lands and in the Black Sea and Anatolian regions.

Vegetation productivity declined most over croplands and grasslands in the Atlantic region as well as over wetlands in the Alpine region.

Increasing precipitation especially in the South Eastern regions had the strongest influence on productivity increase. Decreasing number of frost days were strongly correlated to increasing productivity in the Pannonian region and to decreasing productivity in the Atlantic region. 

Land use change was a stronger driver of productivity changes than climatic variations. 

Converting lands to agriculture and agriculture internal conversions resulted in highest productivity increase whereas increasing urban sprawl resulted in most productivity decrease. 

Vegetation productivity change in Europe during 2000-2016.

Linear trend in yearly vegetation productivity


Note:
Linear trend of yearly growing season vegetation productivity. Vegetation productivity was derived within the growing season from the PPI vegetation index. The layers shows the R value of the fitted linear regression model. Temporal extent 2000-2016. Spatial resolution 500m. The vegetation index used in the indicator is the Plant Phenology Index (PPI, Jin and Eklundh, 2014). PPI is based on the MODIS Nadir BRDF-Adjusted Reflectance product (MODIS MCD43 NBAR. The product provides reflectance data for the MODIS “land” bands (1 - 7) adjusted using a bi-directional reflectance distribution function. This function models values as if they were collected from a nadir-view to remove so called cross-track illumination effects. The product is distributed with 500 m pixel size (MODIS Sinusoidal Grid) with 8-days compositing period. The web map service was derived from a floating point time series.

Data sources:

Data provenance info is missing.

In order to concentrate on strong productivity change signals, a linear trend was fitted over the yearly productivity values and significant (p<0.1) slopes were selected. For significant pixels productivity change was calculated from the fitted regression line with „Yfit,start“ and „Yfit,end“ as the start and end points, respectively. Changes between „Yfit,start“ and „Yfit,end“ were expressed in percentage of the „Yfit,start“ value in order to ensure comparability of change values.

Vegetation productivity changes were significantly different between ecosystems and biographical regions as well as between their combinations (two-way unbalanced ANOVA, p < 0.0001; F=8.821). Decreasing and increasing productivity showed strongest significant differences between biographical regions (F=30.020) than between ecosystems and between their combinations (F=6.331 respectively F=2.408).

Both climatic variations and land use changes had a significant effect on vegetation productivity dynamics (analysis of covariance (ANCOVA), F=124.5, p<0.001) in Europe. Land use change had stronger effect on productivity changes than climatic trends (Type III Sun-of-Squares, F=147.1, p<0.001). To a lesser extent precipitation variations and changes in the number of frost days highly significant influence on how productivity change (F=145.4 and F=28.1, respectively). Standardized model parameters indicated that converting other land covers to agriculture (LCF5) resulted in highest productivity increase and that increasing urbanisation resulted in most productivity decrease. Climatic and land use change effects on vegetation productivity are analysed in subsequent sections.

EXPLORE PRODUCTIVITY TREND STATISTICS HERE.

On average, vegetation productivity increased most in South Eastern Europe. Highest increase was seen in Steppic croplands and grasslands (over 90% increase). Vegetation productivity increased over 80% in grasslands of the Black Sea and Anatolian regions, in sparsely vegetated lands of the Black Sea region and in wetlands of the Steppic regions. In the northern latitudes productivity increased with 70% in Arctic woodlands and forests. The observed increasing trends in the Arctic region is probably due to the shorter periods of snow cover rather than more productivity. The Atlantic biogeographical regions was the only region where on average vegetation productivity decreased especially over grasslands (11% decrease) and croplands (10% decrease). The Alpine and Boreal regions showed low productivity increase (9% and 15%, respectively) compared to other regions, although productivity of alpine wetlands decreased with as much as 11%.

Significant productivity changes during 2000-2016 show distinct and local spatial patterns in Europe (Figure 1). The mid European latitudes in the Atlantic and Continental regions show a mixture of low (-50% < x < 50%) increasing or decreasing productivity changes which may be attributed to fluctuations. The Mediterranean, Eastern and Southern Europe and Turkey on the other hand show distinctive and strong (over 50%) significant productivity change patterns. Changes mostly point to increasing productivity with local hot-spots of strong productivity decrease (more than 50% decrease), especially in the Iberian peninsula. In higher latitudes the southern part of Scandinavia, mostly south of Sweden and Finland showed increasing productivity.

Observed increase in vegetation productivity was confirmed also by other studies published in scientific papers. Guay et al. (2014) used global NDVI datasets GIMMSg and GIMMS3g together with NDVI extracted from SeaWiFS, SPOT-VGT and MODIS datasets showing that about 40 % of the studied areas exhibited similar changes in vegetation productivity regardless the used NDVI dataset. Significant increase in vegetation productivity was detected for over 15 % of the study area whereas opposite trend was found for only 3 % of the area. Huang at al. (2017) studied long term productivity shifts for the high northern latitude areas (above +50°). On the other hand, Pan et al. (2018) highlights the possibility that increasing vegetation “browning” (i.e. productivity decline) may be in fact masked by overall vegetation greening (i.e. productivity increase). This is demonstrated by using the GIMMS3g NDVI dataset for the period between 1982 and 2013. More than 60% increase of the browning area accelerated after 1994 is reported by the authors. Increase browning trends are then reported for all latitudes of the northern hemisphere. The authors then conclude that although most of the vegetated areas exhibited overall greening trend, greening-to-browning reversal occurred on all continents and affected much larger area then browning-to-greening reversal.

Climatic drivers of vegetation productivity change in major European ecosystems.

Vegetation productivity was correlated to climatic drivers, such as the number of frost days, precipitation and temperature, in a multivariate linear regression. Using the Corine Land Cover (CLC) 2000-2018 series, only those grid cells were considered in this analysis where the CLC layers indicated no land use change between 2000-2018. Furthermore, only those grid cells were selected for the analysis where the trend in vegetation productivity was significant (p<0.1).

The climatic drivers and the dependent variable were standardised so that the variances of the dependent and independent variables are 1 and their mean is 0. All variables were detrended before the regression runs. Detrending and standardisation were performed to avoid spurious regression results as well as to obtain normalised slopes, which enable the direct comparison of productivity change as a result of variations in the diverse climatic drivers. Significance of the regression coefficient was measured as p<0.1. Due to standardizing all variables results are presented in standardized regression coefficients that show the change in the dependent variable measured in standard deviations.

In absolute terms, rainfall variations were the strongest driver of increasing vegetation productivity in Europe (Figure 3), especially in south Easter regions: affects were strongest in the Steppic and Pannonian regions and strong in the Black Sea and Anatolian regions. While variations in the number of frost days were relatively important in the Pannonian and Atlantic regions, the effect of temperature variations on vegetation productivity was comparably lower all over Europe. Regionally, increasing precipitation resulted in highest increasing vegetation productivity in Pannonian and Steppic croplands, grasslands and sparse vegetation. Decreasing number of frost days resulted in increasing vegetation productivity in most parts of Europe. Strongest frost-productivity association was observed in the Pannonian region with increasing productivity over croplands, grasslands sparsely vegetated lands and woodlands.

EXPLORE CLIMATIC DRIVERS OF VEGETATION PRODUCTIVITY TRENDS HERE.

Regarding the influence of temperature, the found patterns are supported by scientific results for grasslands. A lengthening of the growing season due to increased temperatures is found to possibly promote productivity while higher temperatures in summer might reduce productivity (Angert et al. 2005, Hufkens et al. 2016, Petri et al. 2012, in Petri et al. 2018). The role of ground temperatures in spring for productivity of grasslands is mentioned by Wingler et al. (2017) and spring as well as early summer are identified to be the most important season for grassland productivity in Ireland by Hurtado-Uria et al. (2013, in Wingler et al. 2017). The role of frost during spring for forest productivity is evident from historic events where severe temperature declines lead to only partial recovery of deciduous trees coverage in Europe and North America (Guh et al. 2008, Augspurger 2009, Hufkens 2012, Ningre et al 2009, Keyling et al 2012). Comparably fast de-acclimation in winter and spring (Kalberer et al. 2006, in Vitasse et al. 2014) implies high reactivity to sudden warm swells and comparably high frost vulnerability, peaking at leave emergence (Neuner et al 2007, in Vitasse et al. 2014). Regarding precipitation, the availability of water is recognized to be an important driver of productivity across the globe (Nemani et al. 2003, Garbulsky et al. 2010, Alström et al 2015, Seddon et al. 2016, in Knapp et al. 2016). Temperate grasslands - which exist in environments with a wide range of precipitation ranging from 150 to 1000mm/year - are sensitive to small changes in water availability (Knapp et al. 2016).

Land use drivers of vegetation productivity change in major European ecosystems.

Climatic drivers of productivity dynamics

Frost frequency