Introduction ACTA BIOLOGICA SLOVENICA LJUBLJANA 2004 Vol. 47, Št. 1: 13-20 Sprejeto (accepted): 2003-11-21 Seasonal and daily pattern, temporal and spatial variability of ecosystem CO2-exchange in a temperate Pannonian loess grassland Szilvia FOTI 1, Janos BALOGH2, Szilard CZOBEL1, Zoltan NAGY1, Sandor BARTHA3 & Zoltan TUBA1,2 1Department of Botany and Plant Physiology, Faculty of Agricultural and Environmental Sciences, Szent Istvan University, H-2103 Gi:idi.ill6 ( e-mail: fotiszi@freemail.hu) 2 Departmental Research Group of Hungarian Academy of Sciences, Szent Istvan University, H-2103 Gi:idi.i116 3 Institute of Ecology and Botany, Hungarian Academy of Sciences, H-2163 Vacrat6t Abstract. In the photosynthetically most active spring, summer and autumn vege- tation period the investigated grassland did maintain a relatively strong daytime carbon gain. During winter the grassland displayed a slight daytime carbon loss. These data suggest that the grassland was a weak sink for carbon in the investigated period. COz- exchange variability during the day seemed to be independent from that of the daily photosynthetic radiation. Thus other factors like soil respiration, soil moisture content and temperature and their interactions could be responsible for the high daily variabil- ity of grassland COz-exchange. The considerable temporal (daily and seasonal) vari- ability of the grassland COz-exchange can be considered as a characteristic feature of the grassland COz-exchange. In the investigated loess grassland vegetation the variabil- ity of COz-exchange showed clear dependence on measuring area, which is obvious in the CV of NEE. We hypothetised that the spatial scale with the lowest variability is the characteristic area of the grassland ecosystem's COz-exchange (COz-exchange physi- ological unit). In general decreased variability indicates a more regulated state. Keywords: carbon-balance, temperate grassland, chamber technique, net ecosystem CO2-exchange, carbon gain Abbreviations: Net Ecosystem Exchange (NEE), Photosynthetic Photon Flux Density (PPFD), air temperature (Ta), COz-concentration (Ca), Coefficient of Variation (CV), Net COz-uptake (A), leaf area index (LAI) Detailed leaf-, individual- and macroscale CO2-exchange studies have long been conducted on grasslands (BREYMEYER ET AL. 1996). On the other hand there are relatively few information avail- able on the micro-scale level, more specifically in temperate grasslands including loess grasslands. 14 Acta Biologica Slovenica, 47 (1), 2004 Out of the physiological processes, the carbon cycle based on photosynthesis-respiration bal- ances of ecosystems is of primary importance. Grasslands are characterised by the fact that large part of their organic matter (C-content) is contained in the belowground living plant parts and in the soil (RICE ANO GARCIA 1994) and this feature is important also in terms of their C-cycle. The temporal pattem of the C-balance of a grassland is also influenced by soil respiration and microbial activity (DbRR ANO MDNNICH 1987). Investigation of the CO2-exchange rates of these grasslands is necessary to estimate their significance in the global carbon cycle and global climate change. Temporal dynamics of the photosynthetical activity and its relation to climate have been inten- sively studied up to the present date. There are less information available on the relationships between the climate and phenology (leaf structure phenology) and the temporal variability of the COrexchange. Phenology should definitely be considered in a study conducted on a yearly tempo- ral scale. Net COrexchange is primarily affected by senescence in the autumn period and not by the abrupt weather changes (HAM ANO KNAPP 1998). Moreover, patches of a grass stand are not alike. Considerable structural variability can be per- ceived even in intensively managed grasslands with profound consequences on their primary func- tioning. This also implies the necessity of carrying out scale-dependent stand level physiological studies when spatial variability is concemed. Closed chamber techniques are suitable tools for study- ing small-scale spatial variability and dynamics of CO2 gas-exchange (ANGELL ET AL. 2001). The temperate Salvio-Festucetum rupicolae loess grassland (steppe) has also been investigated in structural, dynarnical and conservation terms for decades resulting in a considerable amount of information in these subject (Z6LY0MI ANO FEKETE 1994, V1RAGH ANO FEKETE 1984). At the same tirne, studies conducted on the spatial organization of these grasslands have revealed that the spatial scale of the most important processes (species exchanges, coexistence pattems, and diversity) is in the order of a few dm2 to m2 (MUCINA ANO BARTHA 1999). The aims of the present work are: i) to explore the daily and seasonal courses of CO2 gas- exchange at a fixed spatial scale, ii) to describe the temporal (daily and seasonal) and spatial vari- ability (heterogeneity) of the photosynthetical activity and iii) to study the spatial scale dependence of CO2 gas-exchange in a Pannonian loess grassland (steppe) ecosystem. Materials and Methods Study site The measurements were conducted in the years 2000-2001 on the loess grassland situated at the village Kerepes (G6d6116-Monor-Irsa Hills, 170 m a.s.l., 25 km south-east from Budapest). The cli- mate is a temperate continental with hot dry summers and cold winters; mean annual precipitation 550-600 mm or less; annual mean temperature of 11 °C; and large annual amplitude of temperature changes (22 °C). The investigated loess grassland The vegetation is a xeric temperate loess steppe (Salvio-Festucetum rupicolae pannonicum Z6lyomi). The community is dominated by Festuca rupicola, Chrysopogon gryllus, Stipa dasyphyl- la, Cytisus austriacus, and Carex humilis. S. F6ti, J. Balogh, S. Cz6bel, Z. Nagy, S. Bartha & z. Tuba: Seasonal and daily pattern .. . 15 The parent rock is sandy loess with thick humus- and nutrient-rich A layer. The original grassland is made up of more than 90 species. It is a perennial and overwintering, vertically well-structured (60- 80 cm height) grassland, with many broad-leaved dicotyledonous species. The loess grassland can be considered as the representative of the European temperate xeric grassland (steppe) vegetation. Measurements of the net ecosystem CO2-exchange rates Net ecosystem CO2-exchange rates (parallel with transpiration, air temperature, relative humid- ity, vapour pressure and stomatal conductance) were measured by using a portable closed-loop IRGA (LI-COR 6200, operated in absolute mode) sampling the air in a cylinder-shaped plexi-chamber of 60 cm diameter and 70 cm height, with three replicates in five plots. Mixing of the air in the cham- ber was achieved by operation of mixing fans . Ambient conditions (Ta and Ca) at the beginning of each measurement have been re-established by lifting the chamber while the fans were running. The duration of a measurement was 10 to 25 seconds, therefore the changes in the Ta and Ca in the cham- ber were small. PPFD values and canopy surface temperatures were recorded using ceptometers (Decagon) and an infra red thermometer (Raytek MX4). CO2 gas-exchange values from five plots were used to calculate average net photosynthesis values and daytime C-balances. Coefficient of variation was also calculated for each average. Daily courses of gas-exchange have been measured from sunrise to sunset (1.5-2 hours intervals) seasonally, (04/10/2000, 20/03/2001 , 23/05/2001, 03/07/2001) to consider phenological effects, too. Set-up for measuring spatial scale-dependence of the ecosystem COz-exchange The set-up for estimating spatial dependence of CO2 gas-exchange on measuring area consisted of six chambers with different diameters. Ground areas of the six gas exchange chambers follow a logarithmic scale with the diameter of the chambers doubling from 7 .5 cm to 240 cm. The height of each chamber is 70 cm. The cylinder-jacket of the chambers has been arched from UV-B resistant water clean plexiglass. The air motion within the chambers is supplied by outer fan except for the two largest chambers, where the ventilation systems are within the chamber. The chambers are suit- ed for measurements in closed system. The measurements have been carried out on 13th June 2001, in nine patches with three replications at each chamber size. Results and Discussion Daily courses of net CO2-exchange in the four seasons Daily courses of net ecosystem CO2-exchange in the four seasons (including soil and root respi- ration in addition to photosynthesis), air temperature and PPFD average values from the five meas- ured patches are presented (Fig. 1.). The aim was to describe the seasonal features of NEE, as based on these measurements. Autumn (04/10/2000) Air temperature was higher than expected at this date and ranged between 16.2-32.9 °C. Cloudiness caused 20-30 o/o CV considering PPFD. There was a strong correlation between NEE and PPFD until midday. After midday, stomata1 limitation of photosynthesis due to water shortage caused this correlation to become weaker at high Ta values. NEE reached its maximum (5 .37 µmol CO2m 2s·1) in the early moming hours, this value was far below those measured in spring or summer, which is 16 Acta Biologica Slovenica, 47 (1), 2004 the consequence of the decreased photosynthetically active LAI due to the autumn senescence of many species. -"! : 1 ; 10 o E ' ~ " o g, C\l ~ {i ~ -10 N Q -15 C.) ... ~ 1!!i " ~ 10 o i ' :g, o 1 5i ~ •/ 8 -15 / 4/10/2000 •· ·•· 23/05/2001 _r-. (' ·"'-: · \ 20/03/2001 • . • ··· • .. ~ 7/07/2001 l :_: ""' ,,. 1500 ~ 25 ~ \._ \ · Q 20 - 1000 \ .i:'6 15 l- e u. o.. " o.. l:, 2000~ ·~ 30 " 1500 ~ 25 - o ~ [ 20 -~ 10000 15 ..... u. o.. " soo o.. ~ "] ' I -20 ~-- _________ .,___ ___________ _, ' 20 Time (hours) Time (hours) Figure l.: Daily course of loess grassland C02 exchange (dots with error bars, solid line) measured on 60 cm diameter stand plots, of the photosynthetic photon flux density (PPFD, triangles, dashed line) and of air temperature (Tair, squares, dotted line) on four seasonally different days. One symbol represents the average value of five stand plots. Late Winter - Early spring (20/03/2001) Soil respiration increased at Ta values above 15°C, balancing the C02-uptake. As a result NEE was negative throughout the day caused also by dawn frosts, small LAI values and below optimal Ta values for photosynthesis while PPFD was high. Spring (23/05/2001) Growth is most intensive in loess grasslands in this period of the year. NEE followed the rapid- ly increasing Ta and PPFD values during the moming with the maximum of 9.51 µmol C02m 2s·1• From midday Ta was above 30°C and NEE declined with low soil water content (12% by volume) and high vapor pressure deficit of the air. Summer (03/07/2001) NEE followed the rapidly increasing Ta and PPFD values during the moming with the maximum of 10.68 µmol C02m 2s· 1• Ta range was narrower than that measured in May. Increasing cloud cover from early aftemoon caused decreasing PPFD and hence NEE values, with the break of the meas- urements due to aftemoon rains. The observed large seasonal fluctuation of the daily C02-exchange rates also indicates that this pattem is subjected to large variation season by season and year by year due to the fluctuation of the climatic factors. Among others this inter-seasonal variability underlines the necessity of the contin- uous long-term measurements. S. F6ti, J. Balogh, S. Cz6bel, Z. Nagv, S. Bartha & Z. Tuba: Seasonal and daily pattern. .. 17 Temporal variability of grassland COz-exchange Daily course of variability of NEE in representative, stili rather variable, randomly selected patches of the loess grassland was investigated. PPFD JO r 1soo ii . t o~ i 15 )( : ~. " • E ON • • 2, ZJO . ~ • 1000 (.) .r> •i· .., ~- Cl ~ lJ.. o o. > ~ ,t '! t • -~ o o. (.) 100 " \l • o ~ - % roo .. ·, . . , . ' . ' . . : , • . . ~ 23/05/2001 03/07/2001 roo 2roo n=40 n=40 ~ o 400 $' i 2000 ~ t ., " o C, * t . ~: C: " >JO 10)() ii 15 )( • • " o E o ZJO t i • » • 2, (.) . o . " 1000 Cl lJ.. o ~ li" o. > . 1 • • t o. (.) 100 : "'° . . . e" . () } . 8 , • I • ' ' • ' ; 1 = ~ .. ' ] . '" 05 09 13 17 05 09 13 17 Time (hours) Time (hours) Figure 2.: Daily courses of the coefficient of variation of loess grassland C02 exchange measured on 60 cm diameter single stand plots (CV, dots) and ofthe photosynthetic photon flux density (PPFD, diamonds) on four seasonally different various days. (One symbol represents the CV value of three independent meas- urements on the same plots, n=number of measurements.) Spatial variability of the litter decomposition and soil respiration rates are the probable causes of the highly varying NEE at the tirne of the spring measurement. High CV s of NEE in October are most probably caused by highly varying PPFD during the day as opposed to the situation experi- enced in May, when CV(NEE) was much smaller due to the steadily changing light conditions. Temporal variabilities of NEE in the patches and the average variability considering ali the patches are presented in Tab. 1. 18 Acta Biologica Slovenica, 47 (1), 2004 Concerning the daily temporal variability of grassland CO2-exchange, one of the most remark- able observation was that after sunrise the COrexchange variability during the day seems to be rather independent from the considerable changes of the daily photosynthetic radiation. Thus other factors like soil respiration, soil moisture content and temperature and their interactions can be responsible for the high daily variability of grassland COrexchange. The considerable temporal ( daily and seasonal) variability of the grassland COrexchange can be considered as the characteris- tic feature of the grassland COrexchange. This reflects the necessity of the high number and contin- uous measurements during the days and as much possible during the seasons. Tablel: Daily temporal and spatial variability of CO2 exchance rates on five different plots with 60 cm diameter (plot l- plot 5) of temperate loess grassland (04/10/2001). Variability is expressed as % value of variation coefficient (CV%). Periods of measurements Plot 1 Plot 2 Plot 3 Plot 4 Plot 5 Period's averaae 08:07 - 08:24 54,0 229,0 219,0 49,0 98,0 98,1 08:44 - 08:58 18,0 16,0 29,0 21,0 73,0 58,5 09:17 - 09:30 991,0 330,0 66,0 84,0 54,0 92,5 10:34 - 10:49 115,0 108,0 34,0 24,0 48,0 61,9 11 :43 - 11 :55 34,2 74,6 21,3 18,5 14,2 41,0 12:43 -12:56 245,8 187,9 283,9 39,5 185,4 246,9 13:46 - 13:59 23,9 89,8 85,9 56,7 52,8 80,9 14:43 - 14:56 48,7 52,5 114,6 164,0 167,3 212,5 15:15 -15:36 135,0 39,7 15,4 513,3 180,8 568,9 16:27 - 16:43 24,4 22,5 47,6 55,6 2'1-,9 46,0 Spatial variability/spatial heterogeneity of' grassland CO2-exchange The spatial variability of grassland COrexchange rate can be seen as the sign of the spatial het- erogeneity of the ecosystem COrexchange. CV s of NEE (Tab. l.) also demonstrate the spatially dif- ferent behaviour of the five measured patches. Coenological studies have proved that a few dm2 to m2 sampling unit size is suitable for finding the highest variability in species composition and combinations in this grassland (MucINA AND BARTHA 1999, BARTHA ET AL.1997). Consequently, variability of LAI is also high at micro-scale (CV over 30% at 60cm) explaining in part the background of the spatial variability of COrexchange. From this one can conclude that parallel measurements of many unevenly distributed grassland plots with the same diameter are required. Daily maximum values of grassland CO2-exchange and daytime carbon gains Daily maximum values of NEE (Fig. 3.) are in good agreement with both the daytime carbon gain values (528, -864, 2624 and 2171 mgCm-2 on the four representative days, respectively) and with the phenological stages of the vegetation. The maximum of NEE and hence the carbon balance are negative in March, the auturnn NEE values are considerably lower than the ones in the summer. Consequently, in the photosynthetically most active spring, summer and auturnn vegetation period the investigated grassland did presumably maintain a relatively strong daytime carbon gain, while during winter the grassland displayed a slight daytime carbon loss. The above data indicate that in the year of 2000-2001 the investigated overwintering grassland vegetation was very probably acting as a carbon sink. S. F6ti, J. Balogh, S. Cz6bel, Z. Nagy, S. Bartha & Z. Tuba: Seasonal and daily pattem ... 19 14 3000 12 2500 10 2000 ,-; "' , · E 8 0 15 1500 .[ i 6 ~ 1000 ~ 4 "' "' ]i -5 500 X 2 ~ " ON o "' (.) o -r- (.) -2 -500 • -4 -1000 04/1012000 20/03/2001 23/0512001 03/07/2001 Figure 3.: Daily maximum values of C02 exchange rates (bars) and the carbon balance values of the inves- tigated days (dots) in the temperate Joess grassland. Diameter of chambers (cm) •~ -• ~ __ l_s __ • • ;) 1~ i ~ . i . \ I:: i / · .X , __ ,/ 1· :: ! / m ~ / " 10 -----------+ o 0.1 1e+1 1e+2 1e+3 1e+4 1e+5 Ground area of chambers (cm2) Figure 4.: Spatial scale dependence of C02 exchange (dots) and its coefficient of variation (CV, diamonds) in temperate loess grassland. (One symbol refers to nine plots measured in three separate replicates on 13/06/2001, the C02 exchange and the ground area values are plotted ona Jogarithmic scale.) Spatial scale-dependence of grassland CO2-exchange Spatial scale-dependence of NEE was investigated with different chamber sizes (different diam- eters, 70cm height). Logarithrnic values are shown in Fig. 4. The regression COz-uptake values vs. chamber size shows good fit (p<<0.01). However CV of NEE shows scale-dependence with mini- mum of variability at 60 cm patch diameter, suggesting this scale to be characteristic unit of this grassland, where the suparindividual regulation is the most pronounced. In the investigated loess grassland vegetation the variability of COz-exchange showed clear spa- tial scale-dependence. The most probable factors which are candidates for causes of variability pat- tem along the investigated space series are: the ratio of covered and uncovered soil surfaces, the spa- 20 Acta Biologica Slovenica, 47 (1), 2004 tial heterogeneity of soil moisture, soil temperature and litter deposition, the changes of species com- position (e.g. dicots/monocots ratio, plant density), the height and the physiognomical and microm- eteorological structure of the canopy in the relation to the changes of the botanical composition. Presumably the spatial scale with the lowest variability can be considered as the characteristic scale of the COz-exchange (CO2-exchange physiological unit) of the grassland ecosystem. But this aspect and relationship between coenological (botanical composition) and physiological scale-dependence should be a matter of future detailed analysis. In general decreased variability indicates a more reg- ulated state. Thus it is probably that the spatial scale with the lowest variability represents the supraindividually most regulated physiological - CO2-exchange- units of the grassland. Ackoowledgements The financial support ofthe Hungarian Scientific Research Foundation (OTKA-32586 project), the MEGARICH 4th and GREENGRASS 5th EU Framework Research Projects is gratefully acknowl- edged. References ANGELL R. F. ET AL 2001: Bowen ratio and closed chamber carbon dioxide flux measurements over sagebrush steppe vegetation. Agricultural and Forest Meteorology 108: 153-161. BARTHA S.ET AL 1997: Spatiotemporal scales of non-equilibrium community dynamics: a methodological challenge. New Zealand Joumal of Ecology 21(2): 199-206. BREYMEYER A. l. ET AL (Eds.) 1996: Globa! Change: Effects on Coniferous Forests and Grasslands. Dč>RR H. & K. O. MiiNNJCH 1987: Annual variation in soil respiration in selected areas of the temperate zone. Tellus 39B: 114-121. HAM J. M. & A. K. KNAPP 1998: Fluxes of CO2, water vapor, and energy from prairie ecosystem during the seasonal transition from carbon sink to carbon source. Agricultural and Forest Meteorology 89: 12-14. MucINA L. & S. BARTHA 1999: Variance in species richness and guild proportionality in two contrasting dry grassland communities. Biologica, Bratislava 54: 67-75. RICE C. V. & F. O. GARCIA 1994: Biologically active pools of soil C and Nin tallgrass prairie. In: DoRAND J. ET AL. (eds.): Defining soil quality for a sustainable environment. Spec. Pubi. No. 35. MSoil Sci. Soc. Am., Madison, WI. VIRAGH K. & G. FEKETE 1984: Degradation stages in a xeroseries: composition, similarity, grouping, coordination. Acta Bot. Hung. 30: 427-459. Z6LYOMJ B. & G. FEKETE 1994: The Pannonian loess steppe: Differentiation in space and tirne. Abstracta Botanica 18: 29-41.