Modeled Changes in Potential Grassland Productivity and in Grass-Fed Ruminant Livestock Density in Europe over 1961–2010

Abstract

About 25% of European livestock intake is based on permanent and sown grasslands. To fulfill rising demand for animal products, an intensification of livestock production may lead to an increased consumption of crop and compound feeds. In order to preserve an economically and environmentally sustainable agriculture, a more forage based livestock alimentation may be an advantage. However, besides management, grassland productivity is highly vulnerable to climate (i.e., temperature, precipitation, CO2 concentration), and spatial information about European grassland productivity in response to climate change is scarce. The process-based vegetation model ORCHIDEE-GM, containing an explicit representation of grassland management (i.e., herbage mowing and grazing), is used here to estimate changes in potential productivity and potential grass-fed ruminant livestock density across European grasslands over the period 1961–2010. Here “potential grass-fed ruminant livestock density” denotes the maximum density of livestock that can be supported by grassland productivity in each 25 km × 25 km grid cell. In reality, livestock density could be higher than potential (e.g., if additional feed is supplied to animals) or lower (e.g., in response to economic factors, pedo-climatic and biotic conditions ignored by the model, or policy decisions that can for instance reduce livestock numbers). When compared to agricultural statistics (Eurostat and FAOstat), ORCHIDEE-GM gave a good reproduction of the regional gradients of annual grassland productivity and ruminant livestock density. The model however tends to systematically overestimate the absolute values of productivity in most regions, suggesting that most grid cells remain below their potential grassland productivity due to possible nutrient and biotic limitations on plant growth. When ORCHIDEE-GM was run for the period 1961–2010 with variable climate and rising CO2, an increase of potential annual production (over 3%) per decade was found: 97% of this increase was attributed to the rise in CO2, -3% to climate trends and 15% to trends in nitrogen fertilization and deposition. When compared with statistical data, ORCHIDEE-GM captures well the observed phase of climate-driven interannual variability in grassland production well, whereas the magnitude of the interannual variability in modeled productivity is larger than the statistical data. Regional grass-fed livestock numbers can be reproduced by ORCHIDEE-GM based on its simple assumptions and parameterization about productivity being the only limiting factor to define the sustainable number of animals per unit area. Causes for regional model-data misfits are discussed, including uncertainties in farming practices (e.g., nitrogen fertilizer application, and mowing and grazing intensity) and in ruminant diet composition, as well as uncertainties in the statistical data and in model parameter values.

Combining livestock production information in a process-based vegetation model to reconstruct the history of grassland management

“…Chang et al (2015b) then added a parameterization of adaptive management through which farmers react to a climatedriven change of previous-year productivity. Though a full N cycle is not included in ORCHIDEE-GM, the positive effect of nitrogen fertilizers on grass photosynthesis rates, and thus on subsequent ecosystem productivity and carbon storage, is parameterized with an empirical function calibrated from literature estimates (version 2.1; Chang et al, 2015b). ORCHIDEE-GM v2.1 was applied over Europe to calculate the spatial pattern, interannual variability (IAV), and the trends of potential productivity, i.e.…”

“…ORCHIDEE-GM v2.1 was applied over Europe to calculate the spatial pattern, interannual variability (IAV), and the trends of potential productivity, i.e. the productivity that maximizes simulated livestock densities assuming an optimal management system in each grid cell (Chang et al, 2015b). This version was further used to simulate NBP and NBP trends over European grasslands during the last 5 decades at a spatial resolution of 25 km and a 30 min time step (Chang et al, 2015a).…”

“…Two sensitive parameters representing photosynthetic capacity (the maximum rate of Rubisco carboxylase activity at a reference temperature of 25 • C; Vc max 25) and the morphological plant traits (the maximum specific leaf area; SLA max ) were reported by Chang et al (2015a) for simulating grassland NPP. The Vc max 25 = 55 µmol m −2 s −1 and SLA max = 0.048 m 2 per g C in ORCHIDEE-GM were previously defined from observations and indirectly evaluated against eddy-flux tower measurements of GPP for temperate C3 grasslands in Europe (Chang et al, , 2015b. The global TRY database gives SLA values for C4 grasses, of 0.0192 m 2 g −1 dry matter (DM) (0.0403 m 2 per g C with a mean leaf carbon content per DM of 47.61 %; Kattge et al, 2011).…”

“…S1 of Chang et al, 2015b), and T grazing,m,k is the number of grazing days in grid cell k at year m. Due to the impact of livestock on grass growth through trampling, defoliation (i.e. biomass intake), etc., and because grassland cannot be continuously grazed during the vegetation period, thresholds of shoot biomass are set for starting, stopping, and resuming grazing (Vuichard et al, 2007a).…”

“…These maps combine historical NPP changes from the process-based global vegetation model ORCHIDEE-GM (Chang et al, , 2015b with gridded grass-biomass use extrapolated from Herrero et al (2013). First, ORCHIDEE-GM is calibrated to simulate the distribution of "potential" (maximal) harvested and grazed biomass from mown and grazed grasslands respectively.…”