1. Unit of Biosystem Physics 1. O BJECTIVES To analyze impacts of grazing on carbon dioxide (CO 2 ) fluxes (F) exchanged by a meadow. 2. E XPERIMENTAL SITE Situation: Belgium, Dorinne Terrestrial Observatory (l50° 18’ 44’’ N; 4° 58’ 07’’ E; 248 m asl.). Climate: temperate oceanic. Mean annual temperature: 10°C. Annual precipitation: 800 mm. Type: permanent grassland. Surface: 4.2 ha. Slope: moderate (1 to 2 %). Ruminant livestock system: intensive (≈ 2 LU ha -1 ). Breed of cattle: Belgian Blue. 3. M ETHODS 3.1 Long term and short term effects of grazing Long term effects: result from biomass consumption by the cattle and from the cattle effluents that modify assimilation and respiration fluxes. This could only be quantified by comparing fluxes before and after grazing periods. Short term effects: due to cattle respiration that is a part of Total Ecosystem Respiration (TER) and should be measured during its presence in the field.. 3.2 Data sets Data set I: long term effects  2 complete years of eddy covariance measurements made at DTO: 12 May 2010 – 12 May 2012 (data from vegetative periods). Data set II: short term effects  livestock confinement experiments (summer 2012). Impacts of grazing on carbon dioxide fluxes of an intensively grazed grassland in Belgium Jérôme Elisabeth 1, Beckers Yves 2, Bodson Bernard 3, Dumortier Pierre 1, Moureaux Christine 3, Aubinet Marc 1 1 University of Liege, Gembloux Agro-Bio Tech, Unit of Biosystem Physics, 8 Avenue de la Faculté, B-5030 Gembloux, Belgium 2 University of Liege, Gembloux Agro-Bio Tech, Animal Science Unit, 2 Passage des Déportés, B-5030 Gembloux, Belgium 3 University of Liege, Gembloux Agro-Bio Tech, Crops Science Unit, 2 Passage des Déportés, B-5030 Gembloux, Belgium. This research was funded by The « Direction Generale opérationnelle de l’Agriculture, des Ressources naturelles et de l’Environnement - Région Wallonne » Project n° D31-1278, January 2012 - December 2013 Contact Person: Jérôme Elisabeth - University of Liege – Gembloux Agro-Bio Tech (GxABT) - Unit of Biosystem Physics, 8 Avenue de la Faculté - 5030 Gembloux - Belgium Tel : +32 (0)81 62 24 90 - Fax : +32 (0)81 62 24 39 e-mail : Elisabeth.Jerome@ulg.ac.be 4. R ESULTS 4.1 Long term effects 5. C ONCLUSIONS We found a long term and a short term impact of grazing on CO 2 fluxes exchanged by a meadow. 3.3 Long term effects Analyzing and comparing fluxes between grazing and non grazing periods  data set I divided into different intervals corresponding to grazing or non-grazing periods. 1. Response of cumulated gap filled F CO2 calculated for each period to grazing intensity  could be blurred by the flux variability due to responses to meteorological drivers. 2. Response of differences between parameters of interest of the last and the first 5-day window of each grazing or non-grazing period to grazing intensity. Parameters of interest were obtained by fitting a 5-day window F CO2 - PPFD relationship on daytime eddy covariance measurements: 3.4 Short term effects Figure 1: Schematic representation of Dorinne Terrestrial Observatory. Localization of the micro- meteorological station and eddy- covariance set-up. Black area represents the confinement zone used to analyze short term impacts of grazing on carbon dioxide fluxes. GPP max (µmol m -2 s -1 ) = gross primary productivity at light saturation; R d (µmol m -2 s -1 ) = dark respiration; α (µmol µmol -1 ) = apparent quantum efficiency. R d,10 (µmol m -2 s -1 ) = dark respiration normalized at 10°C; T ref = reference temperature set to 10°C; T S (°C) = average soil temperature at 2 cm depth for the considered 5- day window; E 0 (K) = respiration sensitivity to temperature. E 0 values were deduced from the annual response of nighttime half-hourly filtered CO 2 fluxes to T S (353 K and 359 K for the 5 days windows in Year 1 and Year 2, respectively). Figure 3: a) Nighttime CO 2 flux evolution and b) daytime CO 2 flux response to radiation of two successive days with or without cattle for confinement experiment II. Average stocking rate for day with cattle was 27 livestock unit ha -1. Data set II was filtered for u * and stationarity and environmental conditions were equivalent between the two successive days. Errors bars are the random error of measurement. Table 1: Results of confinement experiments. R l is the livestock respiration. Figure 2: a) Cumulated carbon dioxide flux (F CO2 ), and b), c), responses of difference of the last and the first 5 days window day fitting regression parameters for specific periods to grazing intensity. a) b) c) a) b) 4.2 Short term effects Figure 2a: cumulated F CO2 ↑ with grazing intensity but the trend was not significant  response blurred by responses to other driving variables (radiation, soil temperature, drought, etc.). Figure 2b: ∆GPP max significantly ↓ with grazing intensity: above-ground biomass ↓ due to defoliation by grazing  ↓ plant assimilation. Figure 2c: No significant impact of grazing intensity on R d,10 : net impacts of grazing on R d,10 less important than those of climate and seasonality  due to the combination of contradictory effects: ↓ autotrophic plant respiration and ↑ heterotrophic respiration.  Discrimination of indirect impacts of grazing on CO 2 fluxes from flux response to climate only possible after gathering and treating two years of measurements taken under various climatic conditions. Figure 3, Table 1: Nighttime and daytime fluxes always higher with cattle on the plot than without. Confinement experiment: 2 days 1. First day: cattle (27 LU ha -1 ) confined in the main wind direction area of the eddy covariance set-up (1.76 ha); 2. Second day: removed from the plot. 3 independent estimations of livestock respiration (R l ): 1. Nighttime eddy covariance measurements; 2. Daytime eddy covariance measurements;  Comparison of filtered half-hourly data made at 24h interval, in presence of cattle or not (considering equivalent environmental conditions: air temperature within 3°C - wind speed within 3 m s -1 - radiation within 75 µmol m -2 s -1 - wind direction within confinement area). 3. Carbon intake measurements. Table 1: -R l estimations around 2 kg C LU -1 d -1 and not significantly different between the two experiments and data sets.  confidence in estimates based on F CO2 measurements. -R l estimations > HM measurements confirms partially R l estimations > F CO2 measurements. Short term impacts Increase of CO 2 fluxes in presence of cattle. Could be quantified and observed only thanks to confinement experiment. Confinement experiment: only way to study short term impacts of grazing. Long term impacts Decrease of photosynthetic capacity due to cattle consumption of vegetation. Discrimination from flux response to climate only after treating two years of measurements.