hide
Free keywords:
-
Abstract:
The carbon cycle strongly interacts with the
nitrogen cycle. Several observations show that the effects
of global change on primary production and carbon storage
in plant biomass and soils are partially controlled by N
availability. Nevertheless, only a small number of terrestrial
biosphere models represent explicitly the nitrogen
cycle, despite its importance on the carbon cycle and on
climate. These models are difficult to evaluate at large
spatiotemporal scales because of the scarcity of data at the
global scale over a long time period. In this study, we
benchmark the capacity of the O–CN global terrestrial
biosphere model to reproduce temporal changes in leaf area
index (LAI) at the global scale observed by NOAA_AVHRR
satellites over the period 1982–2002. Using a satellite
LAI product based on the normalized difference vegetation index of global inventory monitoring and modelling studies
dataset, we estimate the long-term trend of LAI and we
compare it with the results from the terrestrial biosphere
models, either with (O–CN) or without (O–C) a dynamic
nitrogen cycle coupled to the carbon–water-energy cycles.
In boreal and temperate regions, including a dynamic N
cycle (O–CN) improved the fit between observed and
modeled temporal changes in LAI. In contrast, in the tropics,
simulated LAI from the model without the dynamic N
cycle (O–C) better matched observed changes in LAI over
time. Despite differential regional trends, the satellite
estimate suggests an increase in the global average LAI
during 1982–2002 by 0.0020 m2 m-2 y-1. Both versions
of the model substantially overestimated the rate of change
in LAI over time (0.0065 m2 m-2 y-1 for O–C and
0.0057 m2 m-2 y-1 for O–CN), suggesting that some
additional limitation mechanisms are missing in the model.
We also estimated the relative importance of climate, CO2
and N deposition as potential drivers of the temporal
changes in LAI.We found that recent climate change better
explained temporal changes in LAI when the dynamic N
cycle was included in the model (higher ranked fit for
O–CN vs. O–C). Using the O–C configuration to estimate
the direct effect of climate on LAI, we quantified the
importance of climate-N cycle feedbacks in explaining the
LAI response. We found that the warming-induced release
of N from soil organic matter decomposition explains
17.5 % of the global trend in LAI over time, however,
reaching up to 40.9 % explained variance in the boreal
zone, which is a more important contribution than
increasing anthropogenic nitrogen deposition. Our analysis
supports a strong connection between warming, N cycling,
and vegetation productivity. These findings underscore the
importance of including N cycling in global-scale models
of vegetation response to environmental change.
