|
Chilean Journal of Agricultural Research, Vol. 70, No. 2, April-June, 2010, pp. 251-258 Research Effects of nitrate and labile carbon on denitrification of southern temperate forest soils Efectos del nitrato y carbono lábil en la desnitrificación en suelos de bosques templados australes Cecilia A Pérez1, Martín R Carmona2, José M Fariña1, Juan J Armesto3 1 Pontificia Universidad Católica de Chile, Centre for Advanced Studies in Ecology & Biodiversity (CASEB), Alameda 340, Santiago, Chile Correspondence Address: Cecilia A Pérez, Pontiificia Universidad Catblica de Chile, Centre for Advanced Studies in Ecology & Biodiversity (CASEB), Alameda 340, Santiago, Chile, cperez@bio.puc.cl Date of Submission: 02-Mar-2009 Code Number: cj10027 Abstract The pressure for anthropogenic land use changes and logging of temperate forests in southern Chile is rapidly increasing, with its potentially high impacts on the capacity of soils to retain important limiting elements. We tested the hypotheses that logging increases the denitrification rates and nitrate and C limitation of denitrifiers activity would be higher in soils of unlogged, old-growth forests than in soils of logged forests. Potential denitrification rates were estimated by the acetylene inhibition assay in intact soil cores in laboratory short-term aerobic incubations using the following treatments: 0.7 mmol NO 3 -N addition, the same nitrate addition plus 23.3 mmol C-glucose, and controls (no additions) with and without 10% v/v acetylene. Forest logging did not significantly change soil nitrate content and C lability (e.g. soil C/N ratio). A nested two-factor ANOVA for repeated measures showed that denitrification was enhanced by nitrate plus labile C additions in both forests, suggesting that in both logged and unlogged forests labile C and nitrate limit denitrifiers activity. Increases were up to one order of magnitude when glucose was added to nitrate treated soils; from 373 ± 113 to 3 353 ± 451μg N 2 0-N m -2 d -1 in the unlogged, oldgrowth forest and from 1 369 ± 941 to 12 192 ± 7 474 μg N 2 0-N m 2 d -1 in the logged forest. We conclude that, denitrification would be enhanced in logged forests in the longer term due to a greater nitrate and labile C availability of both in disturbed soils.Keywords: acetylene inhibition assay, lowland evergreen forests, selective logging, nitrogen availability Resumen La presión por el cambio en el uso del suelo y la tala de los bosques templados de Chile está aumentando rápidamente, con altos impactos potenciales sobre la capacidad de retención en el suelo de elementos limitantes del crecimiento y regeneración del bosque. Evaluamos las hipótesis que la tala selectiva del bosque incrementaría las tasas de desnitrificación y que la limitación de N y C para la actividad de los desnitrificadores sería más elevada en los bosques testigos que en los manejados. Las tasas de desnitrificación potencial se estimaron mediante el método de inhibición con acetileno en suelos intactos, a través de incubaciones de laboratorio a corto plazo, con los siguientes tratamientos: adición de 0.7 mmol N-NO3, nitrato más la adición de 23.3 mmol C-glucosa, y controles con y sin 10% v/v acetileno. La tala selectiva no cambió significativamente la labilidad del C de la materia orgánica del suelo (e.g. C/N) ni el contenido de nitrato en el suelo. Un análisis de varianza anidado de dos factores para medidas repetidas muestra que las tasas de desnitrificación fueron estimuladas con la adición de nitrato más C lábil en ambos bosques, lo cual sugiere que en ambos tipos de bosque tanto el nitrato como el C lábil son limitantes de la actividad de los desnitrificadores. Los incrementos fueron de más de un orden de magnitud cuando la glucosa se agregó en los suelos tratados con nitrato: desde 373 ± 113 a 3353 ± 451 µg N-N2O m-2 d-1 en los bosques testigos, y desde 1369 ± 941 a 12192 ± 7474 µg N- N2O m-2 d-1 en el bosque bajo tala selectiva. Concluimos que las tasas de desnitrificación pueden estimularse en el largo plazo por la tala selectiva intensiva del bosque, debido a un incremento en la disponibilidad de nitrato y C lábil en el suelo perturbado. Palabras clave: ensayo de inhibición con acetileno, bosques siempre verdes de zonas bajas, tala selectiva, disponibilidad de N
Introduction Soils of unpolluted, old-growth temperate forests in Chile and Argentina present a high ratio of net nitrification in relation to net mineralized N; either potential (Perez et al., 1998; Satti et al., 2003) or in situ (Perez et al., 1998; Decker and Boerner, 2003). However nitrate pools in forest soils and nitrate losses to stream waters are extremely low because of a high N retention in soil organic matter (Perakis and Hedin, 2001). The almost complete transformation of nitrate to ammonium by dissimilatory nitrate reduction has been postulated as a mechanism of N retention within an old-growth Andean forest ecosystem in this region (Huygens et al., 2007). However, there is little information about the magnitude of denitrification, i.e. microbial reduction of reactive N oxides to nitrous oxide and elemental N in these forests (Perez et al., 2003). During denitrification C acts as an electron donor and nitrate as an electron acceptor, so denitrifiers are active in C-rich and water saturated soils, where nitrate serves as electron acceptor instead of oxygen. In the temperate region of the Northern Hemisphere, it is well documented for conifer, hardwood or riparian forests that the addition of nitrate and labile C to soils increases the rates of denitrification (Robertson et al., 1987; Ashby et al., 1998; Henrich and Haselwandter, 1997; Regina et al., 1998; Jordan et al., 1998; Mohn et al., 2000; Laverman et al., 2001; Wallenstein et al. 2006). However, such studies have not been conducted in comparatively less polluted southern temperate forests where nitrate retention in soils is remarkably high. Two anthropogenic factors that can drastically disrupt N retention mechanisms in unpolluted southern temperate forests: 1) the global increase in atmospheric reactive N from neighbouring agricultural activities that increase the emissions of ammonia (Godoy et al., 2005) and, 2) the pressure of land use change from forests to agriculture and exotic tree plantations (Armesto et al., 2009). Well-documented effects of such alterations are reported for air-polluted northern temperate forests, especially in areas that have a long history of logging and agricultural land use. The evidence shows that forest logging increases N availability through higher rates of mineralization (Reynolds et al., 2000; Thibodeau et al., 2000; Hope et al., 2003; Lindo and Visser, 2003; Inagaki et al., 2008) and decomposition (Prescott, 1997; Brais et al., 2002), leading to increased availability of labile C (Chatterjee et al., 2008) and higher denitrification (Robertson and Tiedje, 1988). However, the effects of altering the inputs of reactive N and land use change in N transformations are less understood in southern temperate forests, where N is strongly limiting for tree growth (Perakis and Hedin, 2001; Vann et al., 2002; Satti et al., 2003; Diehl et al., 2008), and soils are dominated by recalcitrant forms of C, as indicated by the relatively high soil C/N ratios (Perez et al., 2009). Such analysis is relevant because one of the important microbial transformations in the context of global change is denitrification, as it may counteract eutrophication in aquatic ecosystems and also may be a significant source of N 2 0, an important greenhouse gas. In the present study we tested the hypotheses that increases of soil N availability and labile C (i.e. lower soil C/N ratios) following logging of southern temperate forests should enhance denitrification rates in soils, and secondly, under nitrate plus glucose addition denitrification should be more enhanced in unlogged forests, because of the postulated stronger limitation of these elements. The main objective of this work was to evaluate the effects of soil available N as nitrate and C as glucose for denitrification rates using short-term aerobic laboratory incubations of soils of logged and unlogged old growth forests from southern Chile. Although lab incubations are only indicative of processes occurring in the field, we expect that our data offers insights into the effects of logging on N transformations and allows us to make more precise predictions about N dynamics in southern temperate forests. Material and Methods Study area Soils were collected from evergreen rain forests in central Chiloe Island (42 o 37′S, 73 o 46′W), Chile. In this region lowland, old-growth rainforests are dominated by broad-leaved evergreen tree species, mainly "tepa" Laureliopsis philippiana (G. Looser) Schodde and different Myrtaceae species (Gutierrez et al., 2009). The study sites are located on the foothills of the Coastal Range, between 100-200 m.a.s.l. Soils are well drained, with a high content of organic matter in the A,, horizon (ca. 50%). Soil texture belongs to silty-loam. Soil type belongs to Cambisols. Within an area of ca. 2 km 2 a small watershed was selected which was logged 11 yr ago with selective logging, i.e., removal of 50% of all the basal area of the stand. An adjacent unlogged old growth forest of ca. 300 yr old was sampled as control. Detailed descriptions of the structure and dynamics of this old-growth forest are given by Gutierrez et al. (2009). Selective logging required a silvicultural plan approved by CONAF (Chilean Forest Service). Selective cut of L. philippiana individuals, >50 cm diameter at breast height, from an area of about 100 ha left behind 35-40% of the original canopy cover. Timber of this species is used for industrial production of wooden panels. It is worth mentioning that selective logging was applied only to one watershed in the study area, and therefore a suitable replicate of this treatment was unavailable for the same soil and topography. In order to capture as much spatial variability as possible, we sampled across the complete logged area, avoiding the edge effects. Experimental design In each forest three springs were selected as reference for sampling point location. The reference springs were about 150-200 m from each other. From each, a sample point was located 12 m away following the contour line, perpendicular to the springs, and a second point was sampled in the opposite direction. In total there were six sample points per forest (control and logged). Seasonally, from April 2005 until January 2008 (11 sampling dates), pointing each point we obtained samples of surface mineral soil (A h , 0-10 cm). Soil samples were taken with a shovel and sieved on site using a 2 mm mesh size for chemical characterization. One additional soil sample per point was taken with a 100 cm 3 steel cylinder for determination of bulk density. These samples were dried at 70ºC for at least 2 d and then weighted and referred to sample volume. Soil extraction of ammonium and nitrate proceeded with a 0.021 mol L -1 of aluminum potassium sulfate dodecahydrate (Merck, Darmstadt, Germany, KAl(SO 4 ) 2 FNx01 12 H 2 O) solution (1:4 m/v) and determined by means of fractionated steam distillation (Perez et al., 1998). In order to estimate soil C lability, we used the C/N ratio of soil organic matter assuming that lower soil C/N ratio indicated higher C lability. Soil samples were ground for the determination of total N and C by means of flash combustion with and element analyzer (NA 2500 Carlo Erba Element Analyzer, Lakewood, New Jersey, USA). Soil reaction was determined in a 1:4 soil:water suspension with a glass electrode (Horiba, Kyoto, Japan). This proportion was used considering the high amount of organic matter in theses soils. Parallel to these analyses, four soil cores of the surface horizon (Ah) were taken with a 100 cm 3 steel cylinder in each sample point for the determination of denitrification rates by the acetylene inhibition assay (Groffman et al., 1999). Net nitrification and nitrate concentration in these soils are high (Pιrez et al., 2009) and therefore nitrate reductase inhibition by acetylene is minimal. Soil cores were carefully placed inside 500 mL hermetic glass jars and stored for up to 6 h before incubation at room temperature (16 to 22ºC). These jars are sealed with a canning lid fitted with a rubber stopper. Before experimental treatment, soil cores were vented to equilibrate with ambient atmosphere. From the four soil cores, one was added 4 mL 0.7 mmol NO 3 -N from a potassium nitrate (KNO 3 , Fluka, Buchs, Switzerland) solution (AN), and a second sample was added the same 4 mL of the nitrate solution plus 4 mL of a 23.3 mmol C-glucose (D-C 6 H 12 O 6 , Vetec, Rio de Janeiro, Brazil) solution (ANG). Two soil cores were left untreated as controls. As the addition of either 4 or 8 mL of water did not significantly change water content of soils in both forests, control soil cores with distilled water were not considered. From the two controls cores, one was incubated without acetylene (C), which allows estimating the reduction capacity of N 2 O to N 2 . The second control with acetylene (A) allows the accumulation of N 2 O to measure it by gas chromatography. A 10% v/v acetylene atmosphere was created in the nutrient addition soil cores and in the (A) control soil core. Gas samples were taken with plastic syringes and needles stuck into the rubber stopper, at 0, 2 and 6 h of incubation and stored in 3 mL Venojets (Terumo, Leuven, Belgium). Samples were frozen until analyzed. The N 2 O concentration in the gas samples was determined by gas chromatography (Shimadzu GC 8A, Kyoto, Japan), equipped with a Porapack column Q 80/100 and electron capture detector. Carrier gas used was a mixture of methane and argon. Calibration curves were prepared with a 1 mg L-1 nitrous oxide balance in N of Scotty analyzed gases. Denitrification rates were estimated from N 2 O-N concentration difference between 6 and 2 h of incubation and referred to an area basis. All analyses were conducted at the Biogeochemistry Laboratory in the Pontificia Universidad Catσlica de Chile. Statistical analyses In order to test for the independency among sample points, the spatial autocorrelation of denitrification rates was estimated using Mantel′s test (Manly, 1997). This test evaluates the correlation among the matrix of topographic distances among sample points and the matrix of differences in denitrification rates. The degree of significance of the correlation was tested using 999 permutations of a random matrix of differences in denitrification rates and comparing the observed correlation to the resulting null distribution(Manly, 1997). Mantel′s test was performed with the program PopTools (Hood, 2000). A non-significant correlation indicates statistical independence among sample points. The effects of sampling date (months), logging treatment (forest type) and nutrient additions (nitrate and nitrate + glucose) on denitrification rates were assessed by a nested ANOVA considering repeated measures (Zar, 1996). One-way ANOVA and a posteriori Tukey′s tests were used to identify differences soil characteristics between logged and unlogged forests. When the assumption for variance homogeneity was not fulfilled, data were either log transformed or ranked. Forward stepwise multiple regressions were used to detect the soil variable responsible for denitrification rates: e.g.; soil pH, bulk density, water content, C/N ratio, nitrate and ammonium content. Statistical analyses were performed with the program Statistica 5.1 (Statsoft, 1997). P values were significant at 0.05 level. Results and Discussion The correlation coefficient of Mantel′s test was nonsignificant (r = 0.021, P = 0.731), indicating a spatial independence among soil sample points in both logged and unlogged forests. Surface soils of logged and oldgrowth forests did not significantly differ in most chemical properties and bulk density, but ammonium and water content were higher in the logged forest [Table - 1]. Denitrification rates in unlogged and logged forests were significantly correlated with soil nitrate content (r2 = 0.594, F2.9 = 6.59, P = 0.017). A significant effect of logging (F1.38 = 6.037, P = 0.019) and nutrient addition (F3.38 = 27.26, P < 0.001) on denitrification rates was observed and no effect of sampling date was detected (F10.380 = 0.061, P = 1). A-posteriori Tukey′s tests indicate that: i) denitrification rates did not significantly differ between control cores with and without acetylene, and ii) the addition of nitrate and nitrate plus glucose significantly increased the rates of denitrification in logged and unlogged forest soils, showing an overall trend to be higher under selective logging [Figure - 1]. For all the interactive effects, it is worth mentioning that the 3-way interaction of silvicultural treatment FNx01 nutrient addition FNx01 sampling date was not significant (F30.380 = 0.511, P = 0.986) indicating that both logged and unlogged forests responded in the same way to nutrient addition in the different sampling periods. Averaging the values for all sampling dates, the increase in denitrification following nitrate plus glucose addition was one order of magnitude [Table - 2] in both logged and unlogged forests. Annual rates of denitrification estimated for these temperate forests during the study period were 1.3 ± 0.4kg N ha -1 in the unlogged old-growth forest and 5.4 ± 3.8 kg N ha -1 in the logged forest. Peaks of denitrification were recorded during austral spring in old growth forest (October 2005 and 2007) and during austral summer (January 2007) in logged forest. The lack of significant differences between control cores, i.e. with and without acetylene, suggests that in the first place, nitrate reductase was not inhibited by acetylene secondly, N 2 O is the main end product of denitrification, and thirdly, N 2 O accumulation in surface soil samples may be the product of both nitrification and denitrification. Remarkable low values are obtained for bulk density in the A h horizon of logged and unlogged forests, but even lower values (0.07 g cm -1 ) were reported for montane oldgrowth forests of Chiloι Island (Zarin et al., 1998), which is explained by the high content of organic matter in these soils. Extrapolating from our laboratory results to estimate of denitrification rates in the field must be taken with caution, because of important differences in temperature and other physical conditions. However, the extended sampling period of this study, which considered 3 yr of seasonal sampling, provides a likely estimated of temporal variability of denitrifier populations in these southern forests. Annual rates of denitrification reported here were similar to those reported for other temperate forests, reaching 1.65 and 4.73 kg N ha -1 , and following the same tendency, from undisturbed and to logged forests, respectively (Barton et al., 1999). The trend to higher denitrification rates in selectively logged stands in Chiloe confirms other findings that logging promotes an increase of denitrification rates (Dutch and Ineson,1990; meson et al., 1991; Griffiths and Swanson, 2001; Pu et al., 2001) associated with higher nitrate availability in disturbed soils (Robertson and Tiedje, 1988; Groffiuan and Tiedje, 1989; Wallenstein et al., 2006). Similarly, increased emissions of N 2 0-N from soils (i.e. production of N 2 0 from nitrification plus denitrification) have been detected in logged tropical rain forests of Malaysia (Yashiro et al., 2008). From our experimental evidence, the short-term addition of nitrate plus glucose significantly increased denitrification rates in both logged and unlogged forest, suggesting that both factors are limiting microbial activity. Similarly, labile C has enhanced denitrification when added together with nitrate to soils of northern temperate air-polluted and secondary forests (Struwe and Kjoller, 1989; Henrich and Haselwandter, 1991; Ashby et al., 1998). Consequently, nutrient pulses enhance denitrifiers activity in polluted as well in the less polluted southern temperate forests. Our data suggests that gaseous losses of unreactive N may be an important pathway leading to nitrate loss from forest soils and reduced nitrate losses to southern temperate forest streams, and hence they must be considered in modelling and balance approaches to understanding N transformations. We suggest that native denitrifiers in unpolluted old-growth forest soils are ready to act, e.g. genes can be readily expressed, when logging effects on organic matter decomposition increase the availability of N in the soil solution, loosing the documented tightness of the N cycle in these forests. As a consequence, logging of southern temperate lowland forests might not only affect soil sustainability, but also it can be a significant source of an intense greenhouse gas to the atmosphere. Conclusions Carbon lability, as estimated by soil C/N ratios, did not significantly change with logging. In contrast, higher nitrogen availability in logged forest soils was expressed in higher soil ammonium content. The addition of nitrate plus glucose significantly increased denitrification rates in logged and unlogged forests suggesting that both nitrate and labile carbon are limiting denitrifiers activity in forest soils. Our short-term laboratory experiments suggest that the forest under selective logging have the potential for higher losses of N 2 O to the atmosphere. In a longer term, if selective logging significantly increases the content of nitrate and labile carbon in forest soils; i.e. lowering soil C/N ratio, denitrification rates could be greatly stimulated. Acknowledgements Support for this study was provided by FONDECYT 1050830 (2005), FONDECYT FONDAP 1501-0001 and Iniciativa Cientifica Milenio (P05-002). We thank the following people for allowing the access to the study sites: Javier Bruna, Arturo Gallardo, Elemias Gomez. This is a contribution to the research programs of Senda Darwin Biological Station and the Biogeochemistry laboratory at Pontificia Universidad Catolica de Chile.[44] References
Copyright 2010 - Chilean Journal of Agricultural Research The following images related to this document are available:Photo images[cj10027t2.jpg] [cj10027f1.jpg] [cj10027t1.jpg] |
|