Search

Data on net ecosystem exchange (NEE) dynamics and carbon balance of a dry, extensively managed sandy grassland, as measured in Hungary in the years 2003 and 2004 are reported. The grassland was a weak source of carbon in 2003 (80 g C m � 2 ), owing to the exceptionally hot and dry conditions while it was a strong sink in 2004 (� 188 g C m � 2 ), when the amount of precipitation was considerably exceeding the 10 years average. Gross primary production (GPP) values in 2003 and 2004 were 584 and 1112 g C m � 2 , respectively, while ecosystem respiration (Reco) values were 663 and 924 g C m � 2 for these 2 years. GPP declined more than Reco due to drought and heat wave effects in 2003 than in 2004. The ratio between net sink and net source days were 0.55 and 1.11 for 2003 and 2004, respectively. The average of daily NEE sums during source periods did not differ between the 2 years (0.796 and 0.777 g C m � 2 day � 1 ), while for the sink periods the average of daily NEE sums were strongly different (� 0.836 and � 1.677 g C m � 2 day � 1 , for 2003 and 2004). The main difference between the years was found in late winter source activity associated with low temperatures (2003), the degree of summer drought and the absence (2003) or presence (2004) of autumn regrowth. As sink activity potentially may occur in the period April–June, the amount of winter–early spring precipitation proved to be decisive to the carbon balance of the grassland and was much less in 2003 than in 2004. Significant source activity was found during droughts in each year contributing up to 50% to the total source activity. While in the favourable periods the assimilation and respiration components were correlated, significant ecosystem respiration not coupled to current photosynthesis was responsible for large part of the source activity of the grassland. Good correlations were found between satellite derived normalized difference vegetation index (NDVI) and broadband NDVI (NDVIb) values in both years. The relation between GPP and the NDVIb index was significantly different between the main growth periods (April–June) of the 2 years, while it was statistically not significant during the autumn regrowth period (2004). # 2006 Elsevier B.V. All rights reserved.