This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy Forest Ecology and Management 261 (2011) 1090–1098 Contents lists available at ScienceDirect Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco Soil CO2 efflux in an Afromontane forest of Ethiopia as driven by seasonality and tree species Yonas Yohannes a,b,∗ , Olga Shibistova a,c , Asferachew Abate a , Masresha Fetene d , Georg Guggenberger a a Institute of Soil Science, Leibniz Universität Hannover, Herrenhäuser Str. 2, 30419 Hannover, Germany Forestry Research Center, Ethiopian Institute of Agricultural Research, P.O. Box 41957, Addis Ababa, Ethiopia VN Sukachev Institute of Forest, SB-RAS, Akademgorodok, 660036 Krasnoyarsk, Russian Federation d Department of Biology, Addis Ababa University, P.O. Box 1176, Addis Ababa, Ethiopia b c a r t i c l e i n f o Article history: Received 27 June 2010 Received in revised form 24 December 2010 Accepted 25 December 2010 Available online 15 January 2011 Keywords: Croton macrostachys Ethiopia Podocarpus falcatus Prunus africana Soil CO2 efflux Soil moisture Soil respiration Soil temperature Tropical forest a b s t r a c t Variability of soil CO2 efflux strongly depends on soil temperature, soil moisture and plant phenology. Separating the effects of these factors is critical to understand the belowground carbon dynamics of forest ecosystem. In Ethiopia with its unreliable seasonal rainfall, variability of soil CO2 efflux may be particularly associated with seasonal variation. In this study, soil respiration was measured in nine plots under the canopies of three indigenous trees (Croton macrostachys, Podocarpus falcatus and Prunus africana) growing in an Afromontane forest of south-eastern Ethiopia. Our objectives were to investigate seasonal and diurnal variation in soil CO2 flux rate as a function of soil temperature and soil moisture, and to investigate the impact of tree species composition on soil respiration. Results showed that soil respiration displayed strong seasonal patterns, being lower during dry periods and higher during wet periods. The dependence of soil respiration on soil moisture under the three tree species explained about 50% of the seasonal variability. The relation followed a Gaussian function, and indicated a decrease in soil respiration at soil volumetric water contents exceeding a threshold of about 30%. Under more moist conditions soil respiration is tentatively limited by low oxygen supply. On a diurnal basis temperature dependency was observed, but not during dry periods when plant and soil microbial activities were restrained by moisture deficiency. Tree species influenced soil respiration, and there was a significant interaction effect of tree species and soil moisture on soil CO2 efflux variability. During wet (and cloudy) period, when shade tolerant late successional P. falcatus is having a physiological advantage, soil respiration under this tree species exceeded that under the other two species. In contrast, soil CO2 efflux rates under light demanding pioneer C. macrostachys appeared to be least sensitive to dry (but sunny) conditions. This is probably related to the relatively higher carbon assimilation rates and associated root respiration. We conclude that besides the anticipated changes in precipitation pattern in Ethiopia any anthropogenic disturbance fostering the pioneer species may alter the future ecosystem carbon balance by its impact on soil respiration. © 2010 Elsevier B.V. All rights reserved. 1. Introduction Despite of the general concern that tropical forests, and particularly intact African forests may play an important role as a carbon sink in the global terrestrial carbon cycle (Ciais et al., 2009; Lewis et al., 2009; Stephens et al., 2007), the size of the carbon stocks and the carbon fluxes in these ecosystems still remain highly uncertain, especially with respect to potential effects of climate change and anthropogenic disturbances. ∗ Corresponding author. Present address: Institute of Soil Science, Leibniz Universität Hannover, Herrenhäuser Str. 2, 30419 Hannover, Germany. Tel.: +49 51176219015; fax: +49 5117625559. E-mail addresses: yonyoh4@gmail.com, yonas@ifbk.uni-hannover.de (Y. Yohannes). 0378-1127/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2010.12.032 Soil CO2 efflux is one of the largest components in the terrestrial carbon budget (Raich and Schlesinger, 1992), with the global magnitude in the order of 98 Pg C per year (Bond-Lamberty and Thomson, 2010). Contributing more than 50% to the ecosystem respiration across variety of biomes (Janssens et al., 2001; Schlesinger, 1997; Shibistova et al., 2002), soil respiration determines the ecosystem carbon balance and thus an ecosystem sink or source activity (Valentini et al., 2000). Yet to date there has only been a limited amount of data available on soil efflux and its abiotic and biotic determinants for Eastern Africa (Werner et al., 2007) and, likewise, comparative studies across major ecosystems are scarce (Bahn et al., 2010). Among the climatic conditions controlling soil respiration in different ecosystems on hourly, weekly and seasonally scale, soil temperature and soil moisture are considered to be the two most Author's personal copy 1091 Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 18 2008 2009 2008 2009 Air temperature, oC 17 200 150 16 100 15 50 14 13 Precipitation, mm influential parameters (Davidson et al., 1998; Reichstein et al., 2003; Rustad et al., 2000). While the relationship between soil CO2 efflux and soil temperature often is described by a simple exponential function (Davidson et al., 2000; Lloyd and Taylor, 1994; Qi et al., 2002), there is no common agreement on the type of relation between soil moisture and soil respiration. For example, working in an old-growth neotropical forest in Costa Rica, Schwendenmann et al. (2003) found a parabolic relation, while Kosugi et al. (2007), working in a Southeast Asian lowland rain forest, identified a linear dependence of soil CO2 efflux on soil moisture. Epron et al. (2004) used an exponential regression to fit a relation between soil water content and the seasonal variability of soil respiration. Strong effect of soil moisture on temporal pattern of soil respiration has been shown in some tropical forests and plantations (e.g., Epron et al., 2004; Hashimoto et al., 2004; Kosugi et al., 2007; Ohashi et al., 2008), but just a weak (Salimon et al., 2004) or even no correlation (Adachi et al., 2006) between soil temperature and soil respiration in primary and secondary forests has been reported. In a tropical plantation of New French Guiana, Bréchet et al. (2009) showed no significant relationship between soil respiration and both, soil temperature and soil moisture on a seasonal scale. Spatial variability of soil CO2 efflux is primarily affected by biotic factors. So Bréchet et al. (2009) stated that variability of soil respiration was mainly explained by leaf litterfall. Oscar (2007) emphasised the role of different tree species and noted that their different fine root production explained much of the spatial variability in soil CO2 efflux. Leaf and total aboveground litter (leaf, bark and woody debris) have also been reported to have an effect on soil respiration (Epron et al., 2004). In this context, long term measurements of soil CO2 efflux with concurrent climatic record are needed to understand the influence of abiotic and biotic drivers on the magnitude of soil CO2 efflux. Here, we report on seasonal and diurnal pattern of soil respiration measured underneath a natural mixed evergreen-deciduous Afromontane forest, growing in Ethiopia, East Africa. The Munessa Shashemene forest is one of the largest Afromontane forests in the country. The soil organic carbon pool (0–60 cm depth) of the forest is appreciably high (134 Mg ha−1 ) and is within the range of similar ecosystems (Lemenih and Itanna, 2004). We carried out our studies under three different tree species: Croton macrostachys Hochst, Podocarpus falcatus (Thumb.) R.BR.ex and Prunus africana (Hook. F.) Kalkman. These tree species differ in a variety of traits, such as crown architecture (Bekele-Tesemma et al., 1993), depth and distribution of roots (Fritzsche et al., 2006), and CO2 assimilation rate and water relation (Fetene and Beck, 2004; Lüttge et al., 2003). The tree species also represent different functional types according to Whitmore (1989). Croton macrostachys is a pioneer deciduous tree, while P. falcatus and P. africana are late successional evergreen trees, the first being coniferous and the second broadleaf. Podocarpus falcatus and P. africana are among tree species that are locally threatened by extinction as a result of illegal cutting and encroachment. Such activities created gaps that favor the abundance of the pioneer tree species C. macrostachys and diverse shrubs at the expense of climax species. This anthropogenically driven change in relative abundance of tree species may bring modifications in the stand-level carbon balance. The objective of the present study was, first, to analyze seasonal and diurnal variation in soil CO2 flux rate as a function of soil temperature and soil moisture. Since Ethiopia is characterized by a strong seasonality of precipitation, we assume soil moisture as the major abiotic driving factor. As a second objective, we were addressing the impact of different functional-type forest tree species on soil CO2 efflux. Here, we hypothesised that soil CO2 efflux under deciduous pioneer C. macrostachys shows a larger seasonal variability than that under the late successional tree species. x Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 Month Fig. 1. Monthly total precipitation (bar graph) and mean monthly air temperature (line graph) for the years 2008 and 2009 measured at the Kuke field station being adjacent to the experimental plots (data from Strobl et al., unpublished). x shows missed data in November 2009 due to rain gauge failure. 2. Materials and methods 2.1. Study site The Munessa-Shashemene forest is located in Oromia Regional State on the eastern escarpment of the southern Main Ethiopian Rift Valley, about 240 km south of Addis Ababa. The forest area, 23,000 ha in size, is divided into three blocks; namely Gambo, Sole and Degaga. For the present study, the experimental plots were established in the Degaga block (07◦ 25′ 51′′ N and 038◦ 51′ 52′′ E). The elevation is at 2266–2279 m above sea level. Mean annual temperature is 15 ◦ C and mean annual precipitation is about 1200 (data from Ethiopian Meteorological Agency at Degaga town, 07◦ 26′ 00′′ N and 038◦ 50′ 26′′ E). The study site is located at the central area of the country, with a minor rainy season occurring from March to May and a major rainy season from July to November (Griffits, 1972). Own meteorological records since 2001 show that 80% of the annual precipitation fell in the major rainy season from July to November, and no clear indication of a minor rainy season could be made (Strobl et al., unpublished). Further, the monthly precipitation pattern outside the major rain season varied considerably, e.g. in the years 2008 and 2009 (Fig. 1). Total annual precipitation in both years was similar, with 1133 mm in 2008 and 1036 mm in 2009. Mean annual air temperature corresponded well to the average value, with 14.7 ◦ C in 2008 and 15.3 ◦ C in 2009. Soils of the study area are rich in clay evolved from volcanic parent material (Fritzsche et al., 2007). At the experimental plots they were classified as Mollic Nitisols according to the WRB system (FAO, 1998). Vegetation of the natural forest is dominated by the canopy species P. falcatus and C. macrostachys. Other plant species with relatively less abundance include P. africana, Syzygium guineense (Wild.) DC., Celtis africana Burm. f. and Pouteria adolfi-friederici (Engl.). The forest is strongly degraded by grazing and illegal logging activities leading to its transformation from a primary to secondary forest. 2.2. Experimental setup and soil CO2 efflux measurement Since a major goal of our study was to assess the temporal and spatial variability of the soil CO2 efflux with respect to the interrelationship of tree species with abiotic factors, the study was carried out on an individual tree basis. In June 2008, for each of the three tree species, C. macrostachys, P falcatus and P. africana, three individual juvenile trees of the third height class (Tesfaye et al., 2010) were selected. In total, nine experimental plots were Author's personal copy 1092 Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 Table 1 Basic characteristic of the soils under the studied trees. Tree/soil depth C. macrostachys 0–10 cm 10–25 cm P. falcatus 0–10 cm 10–25 cm P. africana 0–10 cm 10–25 cm a b c d e f Sanda (g kg−1 ) Silta (g kg−1 ) Claya (g kg−1 ) Cb (g kg−1 ) Nb (g kg−1 ) pHc CECd (mmol(+) kg−1 ) BSe (%) Pavail f (mg kg−1 ) 211 ± 6 219 ± 2 379 ± 20 301 ± 24 410 ± 26 480 ± 25 138 ± 20 64 ± 8 11.7 ± 0.6 5.9 ± 0.5 6.7 ± 0.04 6.5 ± 0.44 732 ± 24 498 ± 63 100 100 26 ± 8.5 9.5 ± 5.1 180 ± 3 214 ± 4 409 ± 20 304 ± 20 411 ± 17 482 ± 24 116 ± 11 66 ± 13 11.5 ± 0.5 6.5 ± 0.8 6.5 ± 0.15 6.2 ± 0.20 565 ± 55 376 ± 42 100 100 15 ± 2.4 8.6 ± 2.4 195 ± 5 222 ± 8 405 ± 37 319 ± 27 400 ± 32 459 ± 36 122 ± 9 80 ± 3 12.2 ± 0.3 8.2 ± 0.3 6.6 ± 0.13 6.5 ± 0.16 640 ± 45 497 ± 49 100 100 16 ± 3.0 8.6 ± 0.3 Measured by the pipette method (Gee and Bauder, 1986). Total carbon and nitrogen measured by dry combustion (Elementar Vario EL, Hanau, Germany). Analyzed potentiometrically in 1 M KCl [1:2.5 (m/v)]. Cation exchange capacity determined with the BaCl2 compulsive exchange method (Gillman and Sumpter, 1986). Base saturation. Available phosphorus measured after extraction with Bray I (Bray and Kurtz, 1945). Values are mean and standard deviation of three replicated samples. established, with three replicates for each tree species within a distance of about 100 m. For characterization of the soils underneath the canopies, soil samples were collected from each plot before the start of soil CO2 efflux measurement campaign. Core soil samples (cylindrical steel core with 4.0 cm diameter) were obtained separately in triplicate from 0–10 and 10–25 cm depths under the canopy of each experimental tree. Soil samples collected under the canopies of same tree species were bulked. According to Table 1, the soil under C. macrostachys tends to have a higher pH, appears to be richer in organic carbon and nitrogen, and has a larger cation exchange capacity and more available phosphorus than that under the other tree species. The better soil nutritional status under C. macrostachys is likely due to the larger nutrients content in its leaves as compared to the other two tree species (Zech, unpublished). With litter fall, the nutrients are returned to the topsoil. To assess the seasonal and diurnal variation of soil CO2 efflux, five permanent soil collars (20 cm in diameter and 5 cm long) made from PVC were inserted in the soil at randomly selected positions underneath the canopy of each individual tree. To minimize any influence of mechanical disturbance of soil surface on diffusion rates, and to avoid cutting of fine roots, the soil collars were inserted into the soil not more than 2 cm. The collars were sealed at the outside with fine sand. Herbaceous understory vegetation was avoided during collar set-up. However, when, in the following two years period, any vegetation grew inside the collars, it was clipped back. Once inserted, soil collars were left in place throughout the measurement period. Soil CO2 efflux from the forest floor was measured using an Infrared Gas Analyzer (Li-8100, LI-COR, Lincoln, NE, USA) supplied by a LI-8100-103 Soil Survey Chamber. For each measurement, the soil respiration chamber was placed on each collar, and CO2 flux rate was automatically calculated from exponential regression of increasing CO2 concentration over the 2–3 min following chamber equilibration. For each of the tree species mean CO2 efflux rates were calculated from the 15 chamber measurements obtained during individual sampling events. The measurements began on July 11, 2008 and have been carried out on a weekly basis between 12:00 h and 15:00 h until July 24, 2010 with the exception of October–November, 2009 due to an instrument failure. No measurements were also taken during or immediately following rainfall. Diurnal soil CO2 efflux measurements were performed over a 24 h period at a 4 h interval. Three days representing different soil moisture categories were selected: 03 July 2008 (wet season); 06 December 2008 (transition from wet to dry season) and 15 March 2009 (dry season). Soil temperature (◦ C) at a depth of 0.1 m was monitored within 10–20 cm distance of each collar simultaneously with soil CO2 efflux measurements using a thermocouple probe (Li-8100-201) connected to the Li-8100. The volumetric soil water content at 0.06 m depth was measured adjacent to each PVC collar with a theta probe (ML2, Delta-T Device Ltd., Cambridge, UK) at three replicates around the collars (the data is available from May 30, 2009). At a representative place between the plots air temperature underneath the forest canopy, canopy precipitation and soil moisture data were obtained at a weather station. Temperature was measured continuously using 8-bit temperature sensor (Onset Computer Corporation, Bourne, Massachusetts), while precipitation was measured on weekly basis using five randomly distributed polyethylene funnels with a 120 mm upper diameter. Volumetric soil water content was recorded using frequency-domain reflectometer (ECHO probe, Decagon Devices Inc., Pullman, WA) probe installed at 0.1 m depth. 2.3. Data analysis We separated dry and wet periods following the approaches of Gibbs and Maher (1967), where the distribution of precipitation events over a long-term record is divided into sections for each ten percent of the distribution. Such rainfall deciles were calculated from ten years historical rainfall data (1998–2007) obtained from a nearby metrological station (Degaga town 07◦ 26′ 00′′ N and 038◦ 50′ 26′′ E). By definition, the fifth decile is the median, and it is near normal classification of wet and dry periods. According to Gibbs and Maher (1967), we used the fifth decile range as a cut-off for dry and wet period classification. Based on that months with more than 55.2 mm rainfall were defined as wet periods and the drier periods are months that received less than 55.2 mm rainfall. This categorized December 2008 to May 2009, December 2009 and January 2010, and March 2010 as dry months while the other months belonged to the wet seasons. To account for changes in the environmental variables, soil moisture and soil temperature and their interactions with the tree species over the course of the study periods were analyzed with regression approaches to analysis of variance (ANOVA). A general linear model was fitted with soil CO2 efflux rate as response, and tree species and season (wet vs. dry) as categorical explanatory variables. Soil moisture and soil temperature were fitted into the model as numerical covariates. The model terms were fitted sequentially and ANOVA (type I sums of squares) was computed leading to F-tests for the main effects and interactions. We fitted the covariates sequentially in the order soil moisture then soil temperature. Author's personal copy 1093 Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 Homoscedastic residuals were obtained with untransformed values. Statistical differences were significant if the probability of type I error was less than 0.05. On the other hand, to analyze the interannual soil CO2 variability among and between the tree species, we averaged soil CO2 efflux data over the wet and dry periods for both years of observations. Then the data set was analyzed using a three-way ANOVA, with factor effects (tree species, years of observation (first year vs. second year), and season (wet vs. dry)). When ANOVA results indicated significant difference, Tukey’s HSD test was performed. The data for all measurements with both, soil moisture and soil temperature values available from 30 May 2009 to the end of the experiment, were analyzed using nonlinear regression. The objective of the analysis was to describe the dependence of soil respiration on soil moisture alone and jointly with soil temperature under the three tree species separately. Models were fit using non linear least squares method in R 2-11.0 (R Development Core Team, 2009). The performances of the equation were evaluated by goodness-of-fit measures (Root Mean Square Error (RMSE), and r2 ). Since in preliminary graphical analysis, the non-linear relation between soil respiration and soil moisture was prominent, while a potential dependency of soil respiration on soil temperature was only weak, we initially modelled the dependency of respiration on soil moisture by different nonlinear models. The Gaussian model was fitting best for all three species both in terms of maximal r2 values and fitting the downturn of soil respiration rates for high soil moisture values (Eq. (1)). After fitting the first part, different models for including the effect of soil temperature were added: a linear regression (Eq. (2)), a linear regression including an interaction term for temperature and soil moisture (Eq. (3)). After that the reduction of the residual sums of squares (at costs of additional parameters) were tested using F-tests: 2 (1) 2 (2) 2 (3) SR = a ∗ exp(−0.5 ∗ ((SM − x0 )/b) ) SR = a ∗ exp(−0.5 ∗ ((SM − x0 )/b) ) + c ∗ ST SR = a ∗ exp(−0.5 ∗ ((SM − x0 )/b) ) + c ∗ ST + d ∗ ST ∗ SM where SR is soil respiration (␮mol m−2 s−1 ), SM is soil moisture (%), ST is soil temperature (◦ C) and a, b, c, d, and x0 are fitted parameters. To describe the relationships between soil respiration rate (SR) and soil temperature (ST) at a diurnal scale, a single exponential function (Eq. (4)) was used: SR = a ∗ expb∗ST (4) Annual soil CO2 efflux was estimated by extrapolating each weekly measurement to a 7-day period. Where there were missing data due to rainfall incidence or failure of instrument, soil CO2 efflux was estimated based on the Gaussian relation of soil moisture and soil temperature with soil CO2 efflux for the different tree species shown in Table 3. Since the linear relation between the volumetric soil water content measured at the weather station and underneath the individual tree species was close (r2 = 0.94 for C. macrostachys; r2 = 0.93 for P. falcatus; r2 = 0.93 for P. africana), the former was used to simulate the latter in cases of gaps. All graphing and statistical analysis was performed using R 211.0 or SigmaPlot version 11 (Systat Software Inc., San Jose, CA, USA). 3. Results 3.1. Seasonal variability of soil CO2 efflux Soil CO2 efflux rates varied between 2 and 7 ␮mol m−2 s−1 (Fig. 2). In general, the respiration rates followed changes in the precipitation and the resulting changes in the volumetric soil water Table 2 Analysis of variance for the net effect of explanatory variable tree species, season, and covariates (soil moisture and soil temperature), and their interaction. Effect df SS (% SS)a F p-Value Tree Season SM ST Tree:season Tree*SM Tree*ST Season:SM Season:ST SM*ST Residual 2 1 1 1 1 2 2 1 1 1 89 12.970 59.767 33.494 6.531 1.844 5.488 0.332 0.332 12.395 2.017 14.985 8.35 38.49 21.57 4.21 1.18 3.53 0.21 7.98 3.49 1.29 38.516 354.982 198.937 38.792 5.476 16.298 0.984 73.622 32.387 11.981 <0.001 <0.001 <0.001 <0.001 0.005 <0.001 0.3776 <0.001 <0.001 <0.001 a % SS indicate increases in multiple r2 (explained variance) due to the addition of this term. df, degree of freedom; SS, sum of square; SM, soil moisture; ST, soil temperature. content, which varied considerably between 8% and 39%, being on average 16% and 28% for dry and wet months, respectively. So during rainy periods from July to November 2008, from July to November 2009 and from May to July 2010 (end of observation period) soil CO2 efflux rates mostly exceeded 4 ␮mol m−2 s−1 . Drying out of the soil with the onset of dry periods in December 2008 and December 2009 resulted in a continuous decline in the soil respiration rates. Remarkable is the pronounced response of the soil CO2 efflux on individual rain events during or at the end of a dry period, e.g. in January or April 2009. Table 2 displays the net effect of tree species, season, the covariates soil moisture and soil temperature, and their interaction on the dependent variable soil CO2 efflux. The ANOVA shows that seasonal changes exerted the strongest influence on soil CO2 efflux in the Munessa forest, followed by the covariate soil moisture. Seasonality accounted for 38% of variability, with F (1, 89) = 354.98, p < 0.001, indicating significant difference in the response variable soil CO2 efflux in wet and dry periods. Soil moisture explained 22% of the overall variances with F (1, 89) = 198.93, p < 0.001. There was also a significant interaction effect of soil moisture and seasonal variability on soil CO2 efflux with F (1, 89) = 73.62, p < 0.001. Using the data set available from 30 May 2009, we assessed the dependence of soil CO2 efflux rate on soil moisture individually under the three different tree species. Under all three tree species soil respiration rates steadily increased with increasing volumetric soil water content threshold level (Fig. 3). After exceeding this threshold, soil respiration decreased with increasing volumetric soil water content. The Gaussian Equation (1) shows that the effect of soil moisture had a significant (p < 0.05) relationship with soil CO2 efflux that explained about 50% for C. macrostachys, 56% for P. falcatus, and 58% for P. africana (Table 3). In contrast to soil moisture, soil temperature at 0.1 m soil depth did not vary much, ranging from about 15 ◦ C to 20 ◦ C. The higher soil temperatures were recorded during the dry periods (Fig. 2). The influence of soil temperature on the overall variation of soil CO2 efflux was smaller as compared with that of soil moisture and tree species. Considering soil temperature as covariate and incorporated in the general linear model, it explains only 4% of the variability of soil CO2 efflux (Table 2). This is also mirrored in Fig. 4, showing a high degree of scatter between soil temperature and soil respiration. We also used the same data set to show the joint effects of soil moisture and soil temperature on soil CO2 efflux rate. The inclusion of the soil temperature in the data analysis (Eq. (2)) increased the r2 values only weakly and slightly reduced the RMSE (Table 3). The intercepts of the models with the linear trend of soil temperature terms were slightly lower than those with just soil moisture (Table 3). However, the pLinear test, which compares the addi- Author's personal copy 1094 Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 24 a 22 40 35 20 30 25 18 20 16 15 14 10 12 5 Precipitation b 9 7 200 150 5 100 3 50 Precipitation [mm] 204 177 149 92 121 64 37 9 324 352 298 241 268 213 185 157 129 73 101 45 17 327 355 299 271 215 1 243 Soil respiration [µmol m-2 s-1] Soil moisture [%] Soil temperature [°C] 45 C. macrostachys P. falcatus P. africana Soil moisture 0 |-------------2008------------|-------------------------------2009---------------------------|-------------------2010--------------| DOY Fig. 2. Seasonal courses of daily mean volumetric soil water content and soil temperature under the canopy (a), weekly canopy precipitation and soil CO2 efflux rate measured under three tree species (b). Each data point for soil respiration and soil temperature is a mean of fifteen measurements and for precipitation is a mean of five measurements. Error bars indicate standard deviation. Data gaps are due to rain events or instrument failure. Periods with light grey background indicate dry periods as calculated by the approaches of Gibbs and Maher (1967). tional effects of the linear regression (soil temperature) relative to the previous Gaussian model (soil moisture term in Eq. (1)), was not significant (Table 3). Inclusion of the joint effects of the linear trend depended on soil temperature and the interaction of soil P. falcatus P. africana 10 -2 -1 Soil respiration [µmol m s ] C. macrostachys moisture and soil temperature (Eq. (3)) brought small differences in r2 and the RMSE values, indicating that the interaction term was more influential than soil temperature alone for C. macrostachys (p < 0.001) and P. africana (p = 0.002). 8 6 4 2 0 1 5 9 13 17 21 25 29 33 37 41 1 5 9 13 17 21 25 29 33 37 41 1 5 9 13 17 21 25 29 33 37 41 Soil moisture [%] Fig. 3. Relationship between soil CO2 efflux rate (SR) and volumetric soil water content (SM) at 0.06 m soil depth. The parameters of the Gaussian function (Eq. (1)) are shown in Table 3. Author's personal copy 1095 Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 Table 3 Comparative values of regression coefficients, r2 and RMSE using the Gaussian and a linear regression, a linear regression including an interaction term for soil temperature and soil moisture. Parameter a b c d xo pC pLinear pLT T:M pInteraction r2 RMSE C. macrostachys P. falcatus Eq. (1) Eq. (2) 5.29 18.14 – – 27.27 <0.0001 – – – 0.4975 0.931 4.84 11.71 0.02 – 27.45 <0.0001 0.3245 – – 0.4975 0.900 Eq. (3) 2.34 9.88 0.03 0.006 18.32 <0.0001 – <0.0001 <0.0001 0.5317 0.900 P. africana Eq. (1) Eq. (2) Eq. (3) Eq. (1) Eq. (2) 6.42 17.27 – – 30.60 <0.0001 – – – 0.556 1.137 7.32 18.81 −0.05 – 30.65 <0.0001 0.1683 – – 0.558 1.136 5.82 17.90 −0.137 0.006 24.33 <0.0001 – 0.056 0.050 0.562 1.132 5.67 18.26 – – 30.28 <0.0001 – – – 0.574 0.866 4.91 16.48 0.05 – 30.44 <0.0001 0.0811 – – 0.577 0.864 Eq. (3) 3.23 14.31 0.012 0.0042 24.29 <0.0001 0.002 0.002 0.585 0.856 Eq. (1): SR = a*exp(−0.5*((SM − x0 )/b)2 ), Eq. (2): SR = a*exp(−0.5*((SM − x0 )/b)2 ) + c*ST and Eq. (3): SR = a*exp(−0.5*((SM − x0 )/b)2 ) + c*ST + d*ST*SM, where a, b, c, d and x0 are the fitted parameters; r2 is the correlation coefficient, SM is volumetric soil moisture content in %, SR is soil respiration in ␮mol m−2 s−1 and ST is soil temperature in ◦ C, pC (p value of complete model), pLinear (the linear regression of ST relative to the previous model including the Gaussian model for SM). pLT T:M (p-value for adding a linear trend and a linear interaction for ST and SM to the Gaussian model fitted with SM). pInteraction (p-value for adding just a linear interaction term ST:SM to the Gaussian model with linear regression for SM). of soil CO2 efflux rates was observed for all the plots in the dry period. 3.2. Diurnal variability of soil CO2 efflux The volumetric soil water contents in the days representing wet, transition and dry conditions were 26%, 19% and 11%, respectively. The diurnal pattern of the soil temperature was similar for the different seasons, with minima in early morning and maxima in afternoon (not shown). The lowest mean daily soil temperature was recorded under all trees in December, and there were no significant differences in soil temperature for P. falcatus and P. africana plots between dry and wet seasons (Fig. 5). Under C. macrostachys, the soil temperatures were higher as compared to other plots due to the less dense canopy and showed maxima during the dry and sunny March term. The season was exerting a strong influence on the soil CO2 efflux rates. Largest values were obtained in the July measurement, representing the wet period and optimal soil moisture, while in March the smallest diurnal CO2 efflux rates were obtained. The impact of the season on temperature dependency of the diurnal CO2 efflux is also reflected by the Q10 values calculated from the exponential function in Fig. 5 (i.e., exp10b ). Q10 values of 3.2 for C. macrostachys, 4.5 for P. falcatus, and 2.7 for P. africana during the wet season indicated a higher temperature sensitivity of soil respiration than respective Q10 values of 2.0. 1.6, and 2.0 in the transition season. In contrast to wet and transition seasons where the pattern of the soil CO2 efflux rates followed the diurnal soil temperature fluctuations, no diurnal temperature dependency It has been shown that the tree species was the third significant factor affecting soil CO2 efflux variability F (2, 89) = 38.51, p < 0.001 (Table 2). Tree species contributed with 8% to the overall variance. Further, the significant interaction effects of tree species and soil moisture (p < 0.001) indicates that the change in soil CO2 efflux rate as a function of soil moisture was not uniform under the different tree species. This is also obvious from the Gaussian functions (Table 3). The threshold level when higher soil moisture leads to a decline of the soil CO2 efflux decreased in the order P. falcatus (volumetric soil water content 31%), P. africana (30%), C. macrostachys (27%). Likewise, absolute largest soil respiration rates at optimum soil moisture were observed under P. falcatus (Fig. 3). In contrast, it appears that the decrease in soil CO2 efflux with decreasing soil moisture was least for C. macrostachys followed by P. africana, while P. falcatus exhibited the steepest decline in soil respiration with drying out of the soil. This result is mirrored by the three-way ANOVA showing that soil respiration did vary significantly (p < 0.05) between the tree species during wet periods (Fig. 6). In both years under study, the mean soil respiration rate during the wet period P. falcatus P. africana -1 Soil respiration [µmol m s ] C. macrostachys 3.3. Tree species variability in soil CO2 efflux -2 8 6 4 2 0 13 15 17 19 21 13 15 17 19 21 13 15 17 Soil temperature [°C] Fig. 4. Relationship between soil CO2 efflux rate (SR) and soil temperature (ST) at 0.1 m soil depth. 19 21 Author's personal copy 1096 Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 C. macrostachys P. africana P. falcatus Soil respiration [µmol m-2 s-1] 9 ; r = 0.50 SR = 0.58*exp ; r = 0.57 SR = 2.00*exp July SR = 0.69*exp December SR = 1.66*exp March 8 7 ; r = 0.78 SR = 0.86*exp ; r = 0.27 ; r = 0.76 SR = 1.31*exp ; r = 0.89 6 5 4 3 2 1 13 15 17 19 13 21 15 17 19 21 13 15 17 19 21 Soil temperature [°C] Fig. 5. Diurnal dependence of soil CO2 efflux rate (SR) on soil temperature (ST) at 0.1 m soil depth. Each data point is the mean of fifteen measurements made per sampling hour. The equations represent July and December measurements, respectively. was highest under P. falcatus. In contrast, there were no significant differences (p > 0.05) in the soil CO2 efflux rate under the three tree species during the dry season. 4. Discussion 4.1. Variability of soil respiration related to soil moisture Soil respiration [µmol m-2 s-1] Soil respiration under the canopy of all three tree species showed pronounced seasonal variation. It is evident that these seasonal changes are generally driven by the precipitation pattern and its response variable volumetric soil water content. The strong positive relationship between soil moisture and soil respiration is in agreement with previous studies in tropical forest ecosystems where a majority of biological processes coincides with moisture dynamics (e.g., Epron et al., 2004, 2006; Hashimoto et al., 2004; Nsabimana et al., 2009; Salimon et al., 2004; Werner et al., 2007). In general, the highest soil respiration rate during the wet periods may have resulted from the high physiological activity of both plants and microorganisms (Lee et al., 2002) in not limiting soil moisture conditions. Besides this general effect of soil moisture on soil respiration rates, rewetting of the soil after dry periods causes a short-lived but strong increase in soil CO2 efflux. This observation is known as “Birch effect” and is the result of burst mineralization of labile soil organic matter that has been accumulated during the dry period and is available to microor- First year wet periods First year dry periods Second year wet periods Second year dry periods 8 5 c a b a b 4 d c 6 b a b b b 2 0 C. macrostachys P. falcatus P. africana Fig. 6. Comparison of soil CO2 efflux between wet and dry periods for both years of observation (July 11, 2008 to July 10, 2009 and July 11, 2009 to July 17, 2010). Bar graphs following a different letter are significantly different (Tukey’s HSD, p < 0.05). Error bars indicate standard deviation. ganisms after re-wetting of the soil (Birch, 1964; Jarvis et al., 2007). At our study, volumetric soil water content was positively related to soil CO2 efflux rates only when it was below around 30%, at higher values there appeared a negative relation. Since volumetric soil water contents of up to 42% were measured, soil respiration rates were below the maximum values under these wet conditions. Excess soil moisture may negatively affect CO2 efflux rates by reducing soil aeration and thus CO2 diffusivity (Janssens and Pilegaard, 2003). Oxygen deficit as result of too high soil moisture decreases activity of plant roots (Adachi et al., 2006) and the heterotrophic decomposition of soil organic matter (Linn and Doran, 1984). This may be particularly the case in the clayey soils under study. An alternative explanation is based on the photosynthetic activity of the trees. The periods with the most amount of precipitation are associated with conditions of low photosynthetic active radiation in the Munessa forest with c. 1–4 mol m−2 leaf area d−1 (Seyoum et al., in preparation). This translates to small carbon assimilation rates and, consequently, to small rates of autotrophic soil respiration. 4.2. Variability of soil respiration related to soil temperature We examined possible seasonal effects of soil temperature on soil respiration and found little relation for the investigated forest ecosystem. The apparent weak contribution of soil temperature unlike to soil moisture is partly due to the relatively small temperature fluctuations in this ecosystem being not sufficient enough to drive seasonal variations in soil respiration. Further, soil temperature tended to be higher during the dry periods when soil CO2 efflux was restrained by low soil moisture. Similar observations have been found in other tropical forest ecosystem where soil temperature is relatively constant within the year and poorly correlated with soil respiration (e.g., Davidson et al., 2000; Hashimoto et al., 2004; Kiese and Butterbach-Bahl, 2002; Nsabimana et al., 2009). In other biomes such as boreal and temperate forests, there is a wide range of seasonal temperatures that leads to seasonal variation in soil respiration rate (Davidson et al., 1998; Malhi et al., 1999; Shibistova et al., 2002). Over diurnal pattern, when soil moisture is assumed to be almost constant, soil temperature is considered to be a major control of soil CO2 efflux (e.g., Chang et al., 2008). Also we found a positive relation between soil temperature and soil respiration, as long as soil moisture is not limiting as is the case at the July and December measurements. The Q10 values suggest that the more favorable is the soil moisture the more prominent is the tem- Author's personal copy Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 perature dependency of the diurnal soil CO2 efflux. This may be attributed to increasing soil organic matter mineralization rates with increasing temperature. But according to Tang et al. (2005), largest soil respiration rates in the afternoon rather suggest that biotic processes, coupled to photosynthesis, carbon allocation to roots and autotrophic may control diurnal CO2 efflux. 4.3. Effects of trees on soil CO2 efflux Our analysis demonstrated that different tree species affected soil respiration. At optimum soil moisture, the soil CO2 efflux under P. falcatus exceeded that under the other two tree species (Fig. 3). Several biotic factors such as assimilation rate, root density, phenological differences, soil microbial activity and diversity (Epron et al., 2006; Raich and Tufekcioglu, 2000; Vanhala, 2002) might explain this observation. During the wet season with dominantly clouded sky, the shade tolerant P. falcatus is having a physiological advantage as compared with the light demanding C. macrostachys in terms of carbon assimilation (Seyoum et al., in preparation). So it seems reasonable that the higher soil respiration rate during the wet period under the canopy of P. falcatus than under the other two tree species is associated with the cumulative metabolic activities of the plant. The situation changes when the dry season is progressing. With decreasing soil moisture, soil respiration decreased under all three tree species, however, the change in magnitude of soil CO2 efflux rate under P. falcatus was larger than under C. macrostachys and P. africana. The least soil moisture sensitivity was observed for C. macrostachys. Probably, this reflects the different sensitivity of the three tree species on soil moisture. Seyoum et al. (in preparation) analyzed particularly larger values of carbon assimilation and transpiration for C. macrostachys than for the other two tree species during the dry season. This pronounced photosynthetic activity of C. macrostachys during the dry period probably results in a relative large root respiration, thus causing a less pronounced decline in CO2 efflux during conditions of low soil moisture than for the other two tree species. 4.4. Annual soil CO2 efflux Cumulative annual soil CO2 efflux was highest under P. falcatus with 144 mol m−2 in the first year and 162 mol m−2 in the second year of observation. Respective values for C. macrostachys were 131 mol m−2 and 135 mol m−2 , and for P. africana 140 mol m−2 and 156 mol m−2 . With that the cumulative annual soil CO2 efflux falls within the range of other published estimates of secondary tropical forests in eastern Amazon (150 mol CO2 m−2 year−1 ; Davidson et al., 2000), tropical monsoon forests in Thailand (213 mol CO2 m−2 year−1 ; Hashimoto et al., 2004), monospecifc forest plantation in Rwanda (112 mol CO2 m−2 year−1 , Nsabimana et al., 2009), and moist tropical lowland forest in Panama (128 mol CO2 m−2 year−1 , Sayer et al., 2007). In most studies soil collars were installed randomly over the whole experimental plots. In the present study collars were installed randomly but underneath the canopy of our experimental trees. Wiseman and Seiler (2004) showed that measurement position had significant effect on soil respiration, and soil CO2 efflux rates were consistently near the trees. This suggested that the estimated annual soil respiration probably reflects a comparatively higher CO2 efflux rate than at a completely randomized design at the forest ecosystem level. As shown in Table 2, the season is having the strongest influence on the soil respiration. The mean soil CO2 efflux rate was higher during the wet period than during the dry period, independently of the study tree and the year of observation (Fig. 6). When the two study years are compared, there were no differences in the soil CO2 efflux rates neither for the wet period nor for the dry period 1097 (except the second year wet period observation under P. africana). However, during the first year of observation six months were classified as dry periods, whereas at the second year only three months fell into this category because of a more homogenous distribution of the precipitation. Hence, the higher soil CO2 efflux rates during the second year of observation could be related to fewer periods where soil respiration was restrained due to low soil moisture. This shows that the precipitation pattern is having a strong impact on the soil cumulative soil respiration in the Afromontane forest. 5. Conclusions The soil CO2 fluxes from soil to atmosphere in the Munessa forest is primarily controlled by soil moisture, and with that by the precipitation pattern. Therefore, the length of dry and wet seasons is of utmost importance for the soil CO2 efflux and the whole carbon balance of this ecosystem. In Ethiopia, predicted future changes in precipitation pattern are likely to have considerable direct effects on soil CO2 efflux over most of the year. However, this study also shows that biological variability is an important factor, because different functional types of trees are responding differently on precipitation pattern with respect to soil CO2 efflux. Since anthropogenic impact to the forest structure, e.g. by changes in the intensity of grazing-induced disturbance, favors the pioneer C. macrostachys at the cost of the late successional tree species, this adds another important driving factor in the seasonal and annual variability of soil respiration at local scales. Interestingly the late successional P. falcatus shows a larger variability is soil CO2 efflux rates than the pioneer C. macrostachys that colonizes disturbed niches. Acknowledgments We would like to thank Deutsche Forschungsgesellschaft (DFG) for financial support of the study within the project package PAK 188. We thank Deksiso Bulcha, Getu Tadesse, Temesgen Yohannes, Abule Loya, and Awol Assefa for their assistance and support in collecting data in the field. We also thank Roger-Michael Klatt, Ulrike Pieper, Pieter Wiese and Holger Ciglasch for their laboratory assistance in soil analysis. Likewise we are grateful to Frank Schaarschmidt and Hany El Kateb for their advice in statistical analysis. References Adachi, M., Bekku, Y.S., Rashidah, W., Okuda, T., Koizumi, H., 2006. Differences in soil respiration between different tropical ecosystems. Appl. Soil Ecol. 34, 258–265. Bahn, M., Reichstein, M., Davidson, E.A., Grunzweig, J., Jung, M., Carbone, M.S., Epron, D., Misson, L., Nouvellon, Y., Rouspard, O., Savage, K., Trumbore, S.E., Gimeno, C., Yuste, J.C., Tang, J., Vargas, R., Janssens, I.A., 2010. Soil respiration at mean annual temperature predicts annual total across vegetation types and biomes. Biogeoscience 7, 2147–2157. Bekele-Tesemma, A., Birnie, A., Tengnaes, B., 1993. Useful Trees and Shrubs for Ethiopia, Identification, Propagation and Management for Agricultural and Pastoral Communities. Regional Soil Conservation Unit/SIDA, RSCU, Nairobi. Birch, H.F., 1964. Mineralisation of plant nitrogen following alternate wet and dry conditions. Plant Soil 20, 43–49. Bond-Lamberty, B., Thomson, A., 2010. Temperature-associated increases in the global soil respiration record. Nature 464, 579–582. Bray, R.H., Kurtz, L.T., 1945. Determination of total organic and available forms of phosphorus in soils. Soil Sci. 59, 39–45. Bréchet, L., Ponton, S., Roy, J., Freycon, V., Coûteaux, M., Bonal, D., Epron, D., 2009. Do tree species characteristics influence soil respiration in tropical forests? A test based on 16 tree species planted in monospecific plots. Plant Soil 319, 235–246. Chang, S.C., Tseng, K.F., Wang, C.P., Wu, J.T., 2008. Soil respiration in subtropical montane forest in Taiwan. Agric. For. Meteorol. 148, 788–798. Ciais, P., Piao, S.-L., Cadule, P., Friedlingstein, P., Chédin, A., 2009. Variability and recent trends in the African terrestrial carbon balance. Biogeoscience 6, 1935–1948. Davidson, E.A., Belk, E., Boone, R.D., 1998. Soil water content and temperature as independent or confounded factors controlling soil respiration in a temperate mixed hardwood forest. Global Change Biol. 4, 217–227. Author's personal copy 1098 Y. Yohannes et al. / Forest Ecology and Management 261 (2011) 1090–1098 Davidson, E.A., Verchot, L.V., Cattánio, J.H., Ackerman, I., Carvalho, J.E.M., 2000. Effects of soil water content on soil respiration in forests and cattle pastures of eastern Amazonia. Biogeochemistry 48, 53–69. Epron, D., Nouvellon, Y., Roupsard, O., Mouvondy, W., Mabiala, A., Saint-Andre, L., Joffre, R., Jourdan, C., Bonnefond, J., Berbigier, P., Hamel, O., 2004. Spatial and temporal variations of soil respiration in Eucalyptus plantation in Congo. For. Ecol. Manage. 202, 149–160. Epron, D., Bosc, A., Bonal, D., Freycon, V., 2006. Spatial variation of soil respiration across a topographic gradient in a tropical rain forest in French Guiana. J. Trop. Ecol. 22, 565–574. FAO, ISRIC, ISSS, 1998. World Reference Base for Soil Resources, vol. 84. Food and Agriculture Organization of the United Nations, Rome, Italy. Fetene, M., Beck, E., 2004. Water relations of indigenous versus exotic tree species, growing at the same site in a tropical montane forest southern Ethiopia. Trees 18, 428–435. Fritzsche, F., Abate, A., Fetene, M., Beck, E., Stephan, W., Guggenberger, G., 2006. Soil–plant hydrology of indigenous and exotic trees in an Ethiopian montane forest. Tree Physiol. 26, 1043–1054. Fritzsche, F., Zech, W., Guggenberger, G., 2007. Soils of the Main Ethiopian Rift Valley escarpment: a transect study. Catena 70, 209–219. Gee, G.W., Bauder, J.W., 1986. Particle-size analysis. In: Klute, A. (Ed.), Methods of Soil Analysis Part 1, second ed., Agronomy Series No. 9. American Society of Agronomy, Madison, Wisconsin, pp. 383–411. Gibbs, W.J., Maher, J.V., 1967. Rainfall Deciles as Drought Indicators. Bureau of Metrology, Bulletin No. 48. Common Wealth of Australia, Melbourne, p. 37. Gillman, G.P., Sumpter, E.A., 1986. Modification to the compulsive exchange method for measuring exchange characteristics of soils. Aust. J. Soil Res. 24, 61– 66. Griffits, J.F., 1972. Ethiopian highlands. In: Griffits, J.F. (Ed.), Climate of Africa. World Survey of Climatology, vol. 10, pp. 369–387. Hashimoto, S., Tanaka, N., Suzuki, M., Inoue, A., Takizawa, H., Kosaka, I., Tanaka, K., Tantasirin, C., Tangtham, N., 2004. Soil respiration and soil CO2 concentration in a tropical forest, Thailand. J. For. Res. 9, 75–79. Janssens, I.A., Pilegaard, K., 2003. Large seasonal changes in Q10 of soil respiration in a beech forest. Global Change Biol. 9, 911–918. Janssens, I.A., Lankreijer, H., Matteucci, G., Kowalski, A.S., Buchmann, N., Epron, D., Pilegaard, K., Kutsch, W., Longdoz, B., Gruenwald, T., Montagnani, L., Dore, S., Rebmann, C., Moors, E.J., Grelle, A., Rannik, U.E., Morgenstern, K., Oltsev, S., Clement, R., Gudmundsson, J., Minerbi, S., Berbiger, P., Ibrom, A., Moncrieff, P.J., Aubinet, M., Bernhoffer, C., Jensen, N.O., Vesala, T., Granier, A., Schulze, E.-D., Lindroth, A., Dolman, A.J., Jarvis, P.D., Cuelemans, R., Valentini, R., 2001. Productivity overshadows temperature in determining soil and ecosystem respiration across European forests. Global Change Biol. 7, 269–278. Jarvis, P., Rey, A., Petsikos, C., Wingate, L., Rayment, M., Pereira, J., Banza, J., David, J., Miglietta, F., Borghetti, M., Manca, G., Valentini, R., 2007. Drying and wetting of Mediterranean soils stimulates decomposition and carbon dioxide mission: the “Birch effect”. Tree Physiol. 27, 929–940. Kiese, R., Butterbach-Bahl, K., 2002. N2 O and CO2 emissions from three different tropical forest sites in the wet tropics of Queensland, Australia. Soil Biol. Biochem. 34, 975–987. Kosugi, Y., Mitani, T., Itoh, M., Noguchi, S., Tani, M., Matsuo, N., Takanashi, S., Ohkubo, S., Nik, A.R., 2007. Spatial and temporal variation in soil respiration in a Southeast Asian tropical rainforest. Agric. For. Meteorol. 147, 35–47. Lee, M., Nakane, K., Nakatsubo, T., Mo, W., Koizumi, H., 2002. Effects of rainfall events on soil CO2 flux in a cool temperate deciduous broad-leaved forest. Ecol. Res. 17, 401–409. Lemenih, M., Itanna, F., 2004. Soil carbon stocks and turnovers in various vegetation types and arable lands along an elevation gradient in southern Ethiopia. Geoderma 123, 177–188. Lewis, S.L., Lopez-Gonzalez, G., Sonke, B., Affum-Baffoe, K., Baker, T.R., Ojo, L.O., Phillips, O.L., Reitsma, J.M., White, L., Comiskey, J.A., Djuikouo, K.M-N., Ewango, C.E.N., Feldpausch, T.R., Hamilton, A.C., Gloor, M., Hart, T., Hladik, A., Lloyd, J., Lovett, J.C., Makana, J.-R., Malhi, Y., Mbago, F.M., Ndangalasi, H.J., Peacock, J., Peh, K.S.H., Sheil, D., Sunderland, T., Swaine, M.D., Taplin, J., Taylor, D., Thomas, S.C., Votere, R., Wöll, H., 2009. Increasing carbon storage in intact African tropical forests. Nature 457, 1003–1006. Linn, D.M., Doran, J.W., 1984. Effect of water filled pore space on CO2 and NO production in tilled and non tilled soils. Soil Sci. Soc. Am. J. 48, 1267– 1272. Lloyd, J., Taylor, J.A., 1994. On the temperature dependence of soil respiration. Funct. Ecol. 8, 315–323. Lüttge, U., Berg, A., Fetene, M., Nauke, P., Peter, D., Beck, E., 2003. Comparative characterization of photosynthetic performance and water relations of native trees and exotic plantation trees in an Ethiopian forest. Trees 17, 40–50. Malhi, Y., Baldocchi, D.D., Jarvis, P.G., 1999. The carbon balance of tropical, temperate and boreal forests. Plant Cell Environ. 22, 715–740. Nsabimana, D., Klemedtson, L., Kaplin, B.A., Wallin, G., 2009. Soil CO2 flux in six monospecific forest plantations in Southern Rwanda. Soil Biol. Biochem. 41, 396–402. Ohashi, M., Kumagai, T., Kume, T., Gyokusen, K., Saitoh, T.M., Suzuki, M., 2008. Characteristics of soil CO2 efflux variability in an aseasonal tropical rainforest in Borneo Island. Biogeochemistry 90, 275–289. Oscar, J.V-B., 2007. Relationships among litterfall, fine-root growth, and soil respiration for five tropical tree species. Can. J. For. Res. 37, 1954–1965. Qi, Y., Xu, M., Wu, J., 2002. Temperature sensitivity of soil respiration and its effects on ecosystem carbon budget: nonlinearity begets surprises. Ecol. Model. 153, 131–142. R Development Core Team, 2009. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Raich, J.W., Tufekcioglu, A., 2000. Vegetation and soil respiration: correlations and controls. Biogeochemistry 48, 71–90. Raich, J.W., Schlesinger, W.H., 1992. The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus 44B, 81–99. Reichstein, M., Rey, A., Freibauer, A., Tenhunen, J., Valentini, R., Banza, J., Casals, P., Cheng, Y., Grünzweig, J.M., Irvine, J., Joffre, R., Law, B.E., Loustau, D., Miglietta, F., Oechel, W., Ourcival, J.M., Peressotti, A., Ponti, F., Qi, Y., Rambalm, S., Rayment, M., Romanya, J., Rossi, F., Tedeschi, V., Tirone, G., Xu, M., Yakir, D., 2003. Modeling temporal and large-scale spatial variability of soil respiration from soil water availability, temperature and vegetation productivity indices. Global Biogeochem. Cycles, doi:10.1029/2003GB002035. Rustad, L.E., Huntington, T.G., Boone, R.D., 2000. Controls on soil respiration: implications for climate change. Biogeochemistry 48, 1–6. Salimon, C.A., Davidson, E.A., Victoria, R.L., Melo, A.W.F., 2004. CO2 flux from soil in pastures and forests in southwestern Amazonia. Global Change Biol. 10, 833–843. Sayer, E.J., Powers, J.S., Tanner, E.V.J., 2007. Increased litterfall in tropical forest boosts the transfer of soil CO2 to the atmosphere. PLoS ONE 2 (12), e1299, doi:10.1371/journal.pone.0001299. Schwendenmann, L., Veldkamp, E., Brenes, T., O’Brien, J.J., Mackensen, J., 2003. Spatial and temporal variation in soil CO2 efflux in an old-growth neotropical rain forest, La Selva, Costa Rica. Biogeochemistry 64, 111–128. Schlesinger, W.H., 1997. Biogeochemistry: An Analysis of Global Change. Academic Press, San Diego, CA. Shibistova, O., Lloyd, J., Zrazhevskaya, G., Arneth, A., Kolle, O., Knohl, A., Astrakhantseva, N., Shijneva, I., Schmerler, J., 2002. Annual ecosystem respiration budget for a Pinus sylvestris stand in central Siberia. Tellus 54B, 568–589. Stephens, B.B., Gurney, K.R., Tans, P.P., Sweeney, C., Peters, W., Bruhwiler, L., Ciais, P., Ramonet, M., Bousquet, P., Nakazawa, T., Aoki, S., Machida, T., Inoue, G., Vinnichenko, N., Lloyd, J., Jordan, A., Heimann, M., Shibistova, O., Langenfelds, R.L., Steele, L.P., Francey, R.J., Denning, A.S., 2007. Weak northern and strong tropical land carbon uptake from vertical profiles of atmospheric CO2 . Nature 316, 1732–1735. Tang, J., Baldocchi, D.D., Xu, L., 2005. Tree photosynthesis modulates soil respiration on a diurnal time scale. Global Change Biol. 11, 1298–1304. Tesfaye, G., Teketay, D., Fetene, M., Beck, E., 2010. Regeneration of seven indigenous tree species in a dry Afromontane forest southern Ethiopia. Flora 205, 135–143. Valentini, R., Matteucci, G., Dolman, A.J., Schulze, E.D., Rebmann, C., Moors, E.J., Granier, A., Gross, P., Jensen, N.O., Pilegaard, K., Lindroth, A., Grelle, A., Bernhofer, C., Grunwald, T., Aubinet, M., Ceulemans, R., Kowalski, A.S., Vesala, T., Rannik, U., Berbigier, P., Loustau, D., Guomundsson, J., Thorgeirsson, H., Ibrom, A., Morgenstern, K., Clement, R., Moncrieff, J., Montagnani, L., Minerbi, S., Jarvis, P.G., 2000. Respiration as the main determinant of carbon balance in European forests. Nature 404, 861–865. Vanhala, P., 2002. Seasonal variation in the soil respiration rates in coniferous forest soils. Soil Biol. Biochem. 34, 1375–2137. Werner, C., Kiese, R., Butterbach-Bahl, K., 2007. Soil-atmosphere exchange of N2 O, CH4 , and CO2 and controlling environmental factors for tropical rain forest sites in western Kenya. J. Geophys. Res. 112, doi:10.05029/02006jd007388. Whitmore, T.C., 1989. Canopy gaps and the two major groups of forest trees. Ecology 70, 536–538. Wiseman, P.E., Seiler, J.R., 2004. Soil CO2 efflux across for age classes of plantation loblolly pine (Pinus taeda L.) on the Virginia Piedmont. For. Ecol. Manage. 192, 297–311.