Litter decomposition, microbial biomass and activity of ... - Springer Link

Nutrient Cycling in Agroecosystems 69: 257–267, 2004. © 2004 Kluwer Academic Publishers. Printed in the Netherlands.

257

Litter decomposition, microbial biomass and activity of soil organisms in three agroforestry sites in central Amazonia Dariusz Kurzatkowski1,2, Christopher Martius2,*, Hubert Höfer3, Marcos Garcia4, Bernhard Förster5, Ludwig Beck3 and Paul Vlek1,2 1Institute

for Agronomy in the Tropics, University of Göttingen, Göttingen, Germany; 2Zentrum für Entwicklungsforschung (ZEF), Walter-Flex-Strasse 3, D-53113 Bonn, Germany; 3Staatliches Museum für Naturkunde (SMNK) Karlsruhe, Germany; 4Embrapa Amazônia Ocidental, Manaus, Brazil; 5ECT Ökotoxikologie GmbH, Flörsheim, Germany; *Author for correspondence (email: [email protected]) Received 11 July 2003; accepted in revised form 16 April 2004

Key words: Agroforestry systems, Amazonia, Litter decomposition, Macrofauna, Microbial respiration

Abstract Soil organisms play a central role in the decomposition of organic matter. The activity of soil organisms was comparatively examined in three experimental sites in central Amazonia 共Brazil兲: a peach palm monoculture 共Bactris gasipaes兲 a, rubber tree plantation 共Hevea sp.兲, and an agroforestry system 共four tree species planted in rows, the space between covered by upcoming secondary vegetation兲. The overall decomposition rates in the systems and the role of different groups of soil organisms 共macrofauna, mesofauna, microflora兲 were studied with leaf litter 共Vismia guianensis) enclosed in litter bags. Microbial respiration and biomass 共SIR method兲 in litter and soil were measured 共IRGA兲. Microbial respiration in all sites decreased in the gradient litter ⬎ topsoil 共0–5 cm兲 ⬎ soil at 5–15 cm. The highest decomposition rate was always observed in the litter bags of coarse mesh size, pointing to the crucial role of the macrofauna in maintaining a high decomposition rate of the organic material in all systems. The Hevea 共k ⫽ 3.4兲 and the Bactris plantation 共k⫽3.1兲 both showed the highest decomposition rates, followed by the polyculture system 共k⫽1.9兲. The Bactris plantation also had the highest level of microbial respiration and biomass in litter and soil. We discuss these findings in the light of data on rainfall, pH and canopy closure. They suggest that microclimate is a more important factor than biomass in determining litter decomposition rates and activity of soil organisms at these sites.

Introduction The decomposition of plant detritus is an important process in ecosystems, as it allows the nutrient and carbon cycle to be closed 共Lavelle et al. 1993兲. As decomposition and soil organic matter formation are main functions of soil organisms, these ultimately have a crucial role in maintaining soil fertility and plant growth. Soil fauna can improve the growth conditions of plants 共Beare et al. 1992; Reddy 1992; Tian et al. 1995兲. Understanding the decomposition processes fully would allow to develop management op-

tions for agroecosystems that make use of the natural ecological processes to be devised, thus minimizing external inputs. As organic matter is key to many important soil functions, we also expect that managing the soil fauna will help in halting the alarming degradation of soils in the tropics 共Martius et al. 2001兲. Developing this concept fully is of particular importance in the rainforests of the Amazon region, which are notoriously deficient in mineral nutrients 共Stark and Jordan 1978; Salick et al. 1983兲. There is a general lack of taxonomic and ecological knowledge on soil organisms 共Dunger and Fiedler

258 1997兲. Until now the microbiological activity and biomass in tropical soils have not been sufficiently studied 共Yang and Isram 1991; Feigl et al. 1995; Wood 1995兲. Soil organisms play a very important role in improving the structure, organic matter content and distribution of nutrient elements in the soil. On the basis of organism size, soil animals are classified into three groups: macrofauna, mesofauna and microfauna 共Swift et al. 1979兲. The macrofauna 共large soil animals like earthworms, termites, diplopods etc.兲 is responsible for the primary decomposition, the first mechanical breakdown of freshly fallen plant material 共Fragoso and Lavelle 1992兲. The role of the mesofauna 共mainly enchytraeids, springtails, mites兲 is often seen in the ‘grazing’ effect by which they increase the productivity of microorganisms. The role of soil microfauna and flora in nutrients cycling in the ecosystem has been recognized for many years. In the present study, we use litter enclosed in nylon bags of different mesh sizes 共litter bags兲 to study the relative importance of these three groups of soil organisms in litter decomposition in these central Amazonian agroecosystems on which little information exists. Decomposition is here defined as the loss of litter from the litter bags. This study is part of the Brazilian-German research program ‘Studies on Human Impact on Floodplains and Forests in the Tropics’ 共SHIFT兲. It belongs more specifically into the framework of a research program on the regeneration of degraded land 共here a former monoculture rubber plantation lying fallow兲 with perennial cultures, as alternatives to the traditional slash-and-burn practice. In this program, a research project on the role of soil organisms in nutrient cycles in natural rainforest and managed agroforestry sites was carried out 共Höfer et al. 2000; Martius et al. 2004a, b, c兲. The study reported here supplemented this project by extending the investigations to monoand polyculture systems that had not been assessed before, namely a peach palm monoculture, a rubber monoculture and another agroforestry system. The aim was to determine the contribution of soil organismic activity to litter decomposition and to estimate the quantity of microbial biomass in the litter and soil on the three agroforestry sites. We used decomposition rates of leaf litter and microbial respiration rate/ microbial biomass as indicators of the contribution of soil organisms to the decomposition process, and thus, as a proxy to soil fertility.

Material and methods The study sites were located in the EmbrapaAmazônia Ocidental research area 30 km north of the city of Manaus, Amazonas state, Brazil 共3°8⬘ S, 59°52⬘ W; 44 m altitude兲. Soils are Xanthic Ferralsol 共FAO/UNESCO classification兲 with 60% clay content 共Höfer et al. 2000兲. The experimental site had been cleared of primary rainforest in 1987 so that a rubber plantation 共Hevea sp.兲 could be established. This plantation was abandoned. The secondary forest that developed at the site was manually cleared in 1992 and the vegetation was burnt. Different experimental plots 共each measuring 48 ⫻ 32 m兲 were established, from which, for this study, two monocultures 共peach palm, Bactris gasipaes; rubber trees, Hevea sp.兲 and one polyculture plantation were selected, each in three replicates. The polyculture consisted of four tree species, Theobroma grandiflorum 共cupuaçu兲, Bactris gasipaes 共peach palm), Bertholletia excelsa 共Brazil nut兲 and Bixa orellana 共Orleans tree兲. Peach palms, rubber trees, and the trees in the mixed culture system were grown in rows with 2, 6 and 4 m spacing, respectively. In the rubber tree plots, the legume Pueraria phaseoloides (Fabaceae) was grown between the trees. The polyculture site has a rather open canopy and no shrub understory, and the Bactris plantation has the most closed canopy of all the sites to be compared. Thus, this study allows the effects of crop diversity 共two monocultures vs. one mixed culture system兲 to be compared with the effects of increased shading. Climate The research was carried out from May 15 to November 15, 1998. In this year, the total annual precipitation was 2545 mm 共data from meteorological station Embrapa, Manaus兲. The driest month was August 共less than 100 mm兲 and the month with greatest rainfall was April 共377 mm兲. The daily average air temperature in 1998 was 28 °C and did not deviate from the mean temperature of the preceding years. The largest temperature fluctuations within a month, were measured in August 共22.1–33.2 °C兲 and the lowest and highest evaporation occurred in April and October, respectively 共Figure 1兲.

259

Figure 1. Average monthly rainfall, mean temperature and evaporation in the study area 共Embrapa Amazônia Ocidental in central Amazonia, near the city of Manaus, Brazil兲.

Litter bags Polyester gauze bags 共25 ⫻ 25 cm兲 were prepared from three mesh sizes 共coarse 1000 ␮m, medium 250 ␮m and fine 20 ␮m兲. Seven to eight grams of air-dried leaves of Vismia guianensis (Guttiferae), collected in a secondary forest near the study sites, were put into each bag. The exact weight of the leaf material in each numbered bag was recorded. The bags were closed by sewing; in the bags of the smaller mesh sizes the seams were additionally sealed with silicon rubber. The 135 litter bags 共45 of each mesh size兲 were exposed at random places on the soil surface in the nine study sites. In the Hevea plantation, the bags were put below the layer of Pueraria plants that covered the soil. At the retrieval dates 共after 27, 88, 115 and 140 days兲 nine litter bags from each study plot were removed, enclosed in plastic bags, and transported to the laboratory. After the bags had been checked for holes made by animals and for penetration by roots or fungal hyphae, the leaves were dried at 65 °C and weighed. To account for the weight of inorganic particles 共e.g., soil兲 adhering to the leaves after field exposure, part of the retrieved material from each bag was burned in a muffle furnace to determine the ash weight. As this ash weight also contains the ash from the minerals that are part of the leaf tissue, another correction factor was used to ac-

count for the minerals in the tissue. This factor was determined using the initially collected leaf material: ‘True’ remaining weight ⫽ 共dry weight of sample – ash in sample兲 ⫹ natural ash content in the leaf. From the resulting weights, the weight loss was calculated 共initial weight – ‘true’ remaining weight兲, and the decomposition was expressed in percent of the initial mass. Negative exponential regressions were fitted through the data using the ‘two-parameter single exponential decay’ fitting procedure of the Sigmaplot software, and the decomposition coefficient k was determined from the regression y ⫽ a · e–kx, with x being time 共days after exposure兲 and y the remaining weight in percent. A, the y axis coefficient, corresponds to initial weight and was set to 100% 共regressions were forced through A⫽100%兲. Microbial respiration and biomass in the soil The CO2 respired from soil samples was used as an index of microbial activity. Measurements were made on two dates: 1st June and 28th October. The soil samples were taken at random at each experimental site using a soil core sampler 共diameter 6.5 cm, length 15 cm兲; eight soil cores from each plot were taken. Each sample was cut into two layers, 0–5 cm and 5–15 cm, not including the litter layer. In the laboratory the soil was sieved through a 4-mm sieve and the

260 organic material, e.g., plant roots or animals 共ants兲, was removed. The soil samples were then stored in a refrigerator at 6 to 8 °C for a maximum of 2 weeks. For the determination of soil moisture and waterholding capacity, 10 g from each soil sample was dried for 24 h at 105 °C. The percent water holding capacity was calculated according to Schlichting and Blume 共1966兲 as 100% dry mass. Soil respiration was measured as carbon dioxide production over time with an Infra Red Gas Analyser 共IRGA兲, in a continuous flow system at a rate of 300 mL fresh air per minute at 22 °C 共Heinemeyer et al. 1989; Höfer et al. 2000兲. A complete measurement lasted 24 hours and consisted of two stages: 共a兲 measurement of basal respiration to assess the metabolic activity of microorganisms, and 共b兲 determination of Substrate Induction Respiration 共SIR兲, by adding a glucose–talcum mixture at a concentration rate of 80 mg g–1 litter dry weight and performing the measurement with optimal soil moisture. The SIR values were used to calculate microbial biomass in the soil 共Anderson and Domsch 1978兲. The respiration rate was calculated from the formula: Respiration 共CO2兲 ⫽ 共Flow兲 ⫻ 共Difference between CO2 of the sample and of ambient air兲/共dry weight of the sample兲. The microbial biomass was calculated by using the Anderson and Domsch regression method and is valid for the temperature of 22 °C, using: x ⫽ 40.09y ⫹ 0.37X where X is 关␮g biomass C · g soil–1兴 and y is 关␮l CO2 · g soil–1兴 h–1. For the determination of microbial respiration and biomass in leaf litter, three litter samples 共about 100– 200 g organic material兲 were randomly taken from each study site. In the laboratory the litter was cut into pieces of approximately 2 cm. The moisture of the litter was standardised with de-ionized water 共3 ml water per 1 g dry weight litter兲 and incubated for one week at 25 °C in an incubator. The respiration was then measured 共5 g amended with glucose 共SIR兲 at a concentration rate of 80 mg/g litter dry weight兲 with the IRGA and the microbial biomass of the litter was calculated as described for soil. For the determination of pH, 25 ml 1 M KCl and 0.01 M CaCl2 were added to 10 g of field-moist sieved soil samples. The mixture was stirred and allowed to rest. The pH was taken in the clear solution. Statistical calculations were made using the program Sigma Stat 2.0.1. One-way ANOVA was used for comparison of the results.

Results Decomposition in litter bags The largest weight loss of Vismia leaves was observed from the coarse mesh bags 共1000 ␮m兲共Figure 2兲. At the end of the experiment 共140 days兲, the Vismia leaves in the coarse mesh bags showed 71% loss of the original weight in the Bactris plantation, 69% loss in Hevea, and 46% loss in the polyculture system. The variation between individual coarse mesh bags was much greater than between the individual bags of medium and fine mesh within the plots. In the bags of medium and fine mesh size the greatest decomposition was found in the Bactris plantation. Here, after 140 days, 17% of the leaf material had decomposed in the medium mesh bags, and 14% in the fine mesh bags. Weight loss from the coarse mesh bags was statistically significant from that from the bags of medium and fine mesh 共p ⬍ 0.05兲, whereas differences between the medium and the fine mesh size were not statistically significant. From the remaining weights of the exposed Vismia leaves, the daily and annual decomposition rates k were calculated using non-linear regression 共exponential decay; Table 1兲. The highest annual decomposition rates were found in the coarse mesh size bags in the plantations of Hevea 共k ⫽ 3.4兲 and Bactris 共k ⫽ 3.2兲. The rate was very low in the polyculture system 共k ⫽ 1.9兲. The decomposition rates in the medium and fine mesh bags are almost one order of magnitude lower than those in the coarse mesh bags. Soil respiration The basal soil respiration in the lower soil stratum 共5–15 cm兲 was always approximately half of that in the upper 5 cm 共statistically significant at p ⫽ 0.026 in June and p ⫽ 0.043 in October兲. In all cases, the basal respiration appeared to be highest in the Bactris plantation 共Table 2兲. However, these differences were statistically not significant. Overall, the basal respiration in June and October did not differ statistically. Microbial biomass in the soil Microbial biomass in the soil 共Table 3兲 showed similar trends as the basal respiration: it was smaller in the lower soil layer than in the topsoil 共statistically significant with p ⫽ 0.004 in June and p ⫽ 0.031 in

261

Figure 2. Decomposition of Vismia leaves in the litter bags 共coarse mesh: 1000 ␮m, medium-size mesh: 250 ␮m, fine mesh: 20 ␮m; see text兲. Each data point is based on 3 replicates. Error bars indicate standard deviation.

262 Table 1. Daily and annual decomposition rates 共k兲 of Vismia leaves in litterbags. R ⫽ correlation coefficient, p ⫽ probability. Retrieval Bactris

Mesh width

Coarse Medium Fine Polyculture Coarse Medium Fine Hevea Coarse Medium Fine

k 共day兲

k 共year兲

R

p

0.009 0.001 0.001 0.006 0.001 0.001 0.009 0.001 0.001

3.13 0.49 0.39 1.95 0.38 0.33 3.40 0.37 0.34

0.41 0.68 0.69 0.34 0.75 0.76 0.46 0.82 0.66

0.0013 0.0001 ⬍ 0.0001 0.0011 ⬍ 0.0001 ⬍ 0.0001 0.0015 ⬍ 0.0001 ⬍ 0.0001

October兲, the only exception being in the Bactris plantation 共although the differences between sites are not significant兲. Microbial biomass in the soil was generally 共but not significantly兲 higher in October. Microbial respiration and biomass in the litter The CO2 respiration from litter was up to one order of magnitude greater than the respiration in the topsoil, and two orders of magnitude greater than the respiration at 5–15 cm soil depth 共Table 4兲. As in the soil, microbial respiration was always 共but without statistical significance兲 highest in Bactris. In Bactris, respiration was higher in October, but in the other two sites, it was lower in this month. However, the differences between months were not significant. Concerning the microbial biomass 共which was greatest in Bactris兲, the site differences were not significant, and the differences between June and October were also not significant 共Table 5兲. Soil pH The pH was measured in each soil sample from the respiration measurements. All pH values were relatively low 共Table 6兲. Particularly in the polyculture system all pH values were below pH 4. The highest pH 共4.5兲 was found in the topsoil under Bactris. The pH readings in the topsoil 共0–5 cm兲 were higher than at 5–15 cm.

Discussion Decomposition in the litter bags The use of bags with different mesh size allows us to assess the specific role of the different soil fauna groups in the decomposition of the enclosed material. The mesh sizes used here 共20 ␮m, 250 ␮m and 1000 ␮m兲 allow the activities of the microflora alone, of microflora plus mesofauna, and of the whole soil fauna community including macrofauna to be investigated. These mesh sizes have been tested before under similar conditions in Amazonia 共Anderson et al. 1983; Beck et al. 1998兲, and proved to separate clearly between the taxa that are normally attributed to these groups; e.g., microflora 共bacteria, protozoa, and fungi兲, mesofauna 共mostly enchytraeids, springtails and mites兲, and macrofauna 共other arthropods and earthworms兲共Höfer et al. 2000兲. The largest weight loss from leaves in this experiment was always observed in the bags of coarse mesh size. The fact that the decomposition coefficient is almost one order of magnitude larger in coarse than in medium and fine mesh bags 共Table 1兲 shows that the macrofauna predominantly controls the decomposition of leaf litter. The most important decomposer groups in the macrofauna are earthworms and termites, but also Diplopoda and Isopoda 共Höfer et al. 2000兲. According to estimates in Martius 共1994兲, termites alone should consume approximately one third of the annual litter production in Amazonian rain forests. The function of primary decomposers lies in the physical breakdown of plant matter as well as in its chemical degradation, often achieved with the help of symbiotic microorganisms 共cf. Eutick et al. 1978; Mishra and Sen-Sarma 1985; Bignell 1994兲. The assimilation of organic matter by primary decomposers is low; much C is released as CO2 into the atmosphere. This contributes to the reduction of the C/N ratio of the organic residues, which, in turn, facilitates their further processing by the mesofauna and microorganisms. The large variation shown in the coarse mesh bags is probably related to the foraging behaviour and distribution of the macrofauna 共principally termites and earthworms兲 which occurs in patches rather than being equally distributed over the forest floor 共cf. Salick et al. 1983兲.

263 Table 2. Average basal soil respiration 共␮l CO2 g–1 soil h–1兲. Study sites

June

Depth

0–5 cm

Bactris Polyculture Hevea

October 5–15 cm

0–5 cm

5–15 cm

Basal resp.

Std. error

Basal resp.

Std. error

Basal resp.

Std. error

Basal resp.

Std. error

2.08 1.36 1.14

0.81 0.80 0.31

0.58 0.52 0.37

0.27 0.39 0.17

2.34 1.42 1.60

1.09 0.79 0.65

1.15 0.72 0.63

0.68 0.54 0.21

Table 3. Average microbial biomass in the soil 共␮g Cmic g–1 soil兲. Study sites Depth


June

October

0–5 cm

5–5 cm

0–5 cm

5–15 cm

Microbial biomass

Std. error

Microbial biomass

Std. error

Microbial biomass

Std. error

Microbial biomass

Std. error

375 283 290

161 146 67

136 123 148

49 50 36

412 228 353

269 68 158

165 119 159

51 28 38

Table 4. Average values of litter respiration 共␮l CO2 h–1 g–1 litter兲. Study sites

June


October

Litter respiration

Std. error

Litter respiration

Std. error

284 207 218

131 70 53

352 187 226

91 66 117

Table 5. Average microbial biomass in the litter 共␮g Cmic g–1 litter兲. Study sites

June


October

Microbial biomass

Std. error

Microbial biomass

Std. error

2842 2533 2042

1137 1780 1292

3375 1878 1847

961 908 460

Table 6. Median pH values from the study sites 共each value is based on 24 measurements兲. Study sites

June

October

Depth

0–5 cm

5–15 cm

0–5 cm

5–15 cm


4.4 3.7 4.0

4.0 3.8 3.8

4.5 3.8 4.1

4.0 3.9 3.9

Soil respiration Soil respiration is defined as the emission of CO2 from the metabolism of both aerobic and anaerobic

microorganisms in the soil 共bacteria, fungi, algae and protozoa; Anderson 1982兲. Together with data on the activity of the soil fauna, soil respiration and microbial biomass are ecologically relevant parameters for the description of structure and function of the decomposer community of ecosystems. The activity of microorganisms and the microbial biomass in the soil are influenced by many factors, among which soil moisture, temperature and density are the most important 共Anderson 1975; Kröckel and Stolp 1986兲. In laboratory measurements, not all of these factors can be taken into account. Removal, transport, storage and treatment of the test samples change the physical qualities of the soil. In our stud-

264 ies, the IRGA measurements were made in sieved soil at a constant temperature of 22 °C with standardized moisture content. The data therefore allow assessing the microbial potential of the soil, but they do not represent the ‘real’ on-site respiration of the soil in the plantation. The microbial biomass was calculated on the basis of the values obtained with the Substrate Induced Respiration 共SIR兲 method. The microbial biomass is defined as the part of organic substance in the soil that consists of living organisms. In the soil only 2–30% of the microbial mass is living biomass. The SIR method therefore is a measure of the metabolically active microflora 共Anderson and Domsch 1978兲. In this study, the measurement of soil microorganism activity in litter and soil was carried out at two sampling occasions, in June and October, representing late rainy season and dry season, respectively. No significant differences were found between the first and second date. The microbial respiration in the litter layer was about 100 times greater than in the topsoil 共0–5 cm兲. The respiration in the topsoil was about 3 times greater than in the lower soil stratum 共5–15 cm兲. In similar experiments in Central Amazonia respiration values in the litter were 40 ⫻ higher than in the soil 共Förster et al. 1999兲. A probable explanation is the decrease of nutrients and organic matter in the gradient from litter to the deeper soil layer. We tested for a correlation between rainfall 共Figure 1兲 at the soil sampling time and the values of the basal respiration and microbial biomass, but no significant effect was found. We also tested for a correlation between soil pH 共Table 6兲 and basal respiration as well as microbial biomass. The pH controls the nutrient availability in soils; however, pH had no significant effect on the activity or biomass of soil micro-organisms. Site parameters and decomposition The decomposition rates in the coarse mesh bags decreased from the Hevea plantation over the Bactris plantation to the polyculture site. In the medium and fine mesh bags, decay rates were higher in Bactris than in the two other sites, which did not differ significantly 共Table 1兲. Similarly, basal respiration as well as microbial biomass in both soil strata and in the litter layer were largest in the Bactris plantation, but did not differ between Hevea and the polyculture system 共Tables 2 and 3兲.

The trees in the Hevea plantation were spaced at 6 ⫻ 4 m, and the canopy was on average 53% closed

measured with digital photos 共Figure 3兲. However, Pueraria phaseoloides was grown beneath the trees on these plots. This is a leguminous plant used for additional N input and to prevent soil erosion that forms a highly closed vegetation mat covering the soil up to approximately 0.7 m height. Under this mat, we expect soil temperature variation to be reduced and soil moisture to be preserved. The polyculture system consisted of the rows of the planted trees, between which secondary forest was admitted to grow. The canopy cover was on average closed to 17% 共Figure 3兲, and much of the soil was directly exposed to solar radiation, and thus, to high temperature variations. In a parallel study of another polyculture system the minimum temperature at night in the beginning of the dry season 1997 was 23 °C and the maximum during the day was 40 °C 共Martius et al. 2004a兲. In contrast, the canopy closure in the Bactris plantation was 87% on average 共very similar to canopy closure in the primary forest; cf. Figure 3兲. In comparison to the other two systems, the soil under Bactris seemed to be most protected from abiotic stress factors 共like high solar radiation, large temperature variation兲. There was also much leaf litter on the ground 共palm leaves and their heavy ‘stems’兲, and the soil fauna diversity might have been richer here 共cf. Vohland and Schroth 1999兲. These factors might explain the high decomposition rate – in all mesh sizes – and the large microbial biomass observed in Bactris. In the Hevea plantation, although the trees were widely spaced, the macrofauna presumably could thrive under the protection of the Pueraria. This is consistent with the large weight loss in the coarse litter bags observed here. In some bags only leaf stems were found after the first removal date. The litter bag data were compared with data from litter bag experiments carried out parallel to this study in adjacent sites, namely in two polyculture systems 共‘System IV’兲, in one secondary forest site and a primary forest site 共Höfer et al. 2000兲. The polyculture system IV consisted of different tree species than the system studied here 共cf. Höfer et al. 2000兲. Interestingly, the decomposition rates of Vismia leaves in coarse bags in the Hevea and the Bactris plantations were higher than in the primary forest. Also, the coarse bags in the polyculture system II 共present study兲 seem to have performed better than

265

Figure 3. Average canopy closure 共% black pixels in digital camera images; cf. Martius et al. 2004a兲 in all study sites 共Hevea stand; polyculture system; Bactris stand兲 and a primary forest at Embrapa Amazonia Ocidental, Manaus, Amazônia. Box plots are for 10 samples per site. Dotted lines are averages, straight lines within box are medians. Box limits are 25/75% percentiles; whiskers are 10/90% percentiles, and circles 5/95% percentiles. In the top row an example of one typical canopy closure photograph is given for each site.

those in system IV in which the decomposition rate is as low as in the secondary forest site 共Figure 4兲. On the other hand, the decomposition rates in the medium and small mesh bags were highest in the primary forest, and lowest in the Hevea plantation and in polyculture system II. This shows that site conditions may have different effects on macrofauna than on mesofauna and microorganisms. The microbial biomass data were also compared with data from adjacent sites 共Table 7兲. The microbial biomass in the Bactris plantation was in the range of the microbial biomass in the primary forest, the secondary forest and the polyculture system IV. The

microbial biomass in the Hevea plantation and in the polyculture system II was lower. This is consistent with the findings from the litter bags where the decomposition in the fine mesh bags 共mainly due to microbial activity兲 was lowest in the Hevea plantation and in the polyculture system II. The interpretation of the decomposition rates from litter bag experiments is constrained by the fact that the Vismia leaves used represent just one litter type. However, if we assume that the differences in the decomposition rates can be attributed to either microclimatic factors or to the site biodiversity, then the obvious conclusion is that a closed canopy 共as in the

266

Figure 4. A comparison of decomposition rates as established with Vismia leaves enclosed in litter bags. First three columns: data from the present study; last four columns: data from adjacent sites: polyculture system ‘IV’ 共four forest tree species within secondary growth stand兲, primary rain forest, and 13-year old secondary forest 共from Höfer et al., in preparation兲. Table 7. Comparison of microbial biomass in the soil depth 0–5 cm of the study sites of the present study 共June sampling兲 and from adjacent sites.

This study

Förster et al. 共1999兲

Study site

Microbial biomass in topsoil 共0–5 cm兲共␮g Cmic g–1 soil兲

Bactris Polyculture Hevea Polyculture IVa Polyculture IVb Secondary forest Primary forest

375 282 290 370 393 411 378

Bactris plantation兲, and even a closed mat of soil-covering annual plants 共as the Pueraria-covered soil in the Hevea plantation兲, both assumed to provide protection from microclimatic stress, seems to enhance the overall decomposition rate 共coarse bags兲 much more than an increase in crop 共tree兲 biodiversity 共as in the polyculture system兲. The fact that overall decomposition rates increase while mesofauna and microorganism activity decrease 共both in litter bags and microbial activity measurements兲 shows that mesofauna and microflora react much more sensibly to the environmental factors than the macrofauna. However, a shift in macrofauna density or diversity or both must have occurred in the Hevea and the Bactris, allowing a decomposition rate to be achieved that is even higher than in the primary forest.

We could not prove the links between site microclimate and decomposition rate/microbial biomass, as microclimate was not measured in this study, but the data strongly indicate that the management of microclimatic conditions might provide a better enhancement of beneficial soil organisms than an increase in tree biodiversity alone. Further experiments, e.g., using a smaller distance between the planted trees, or a combination of polyculture systems with an undercover of annual plants, should be undertaken to assess the role of biodiversity vs. plantation microclimate in managing beneficial soil fauna, and to study the effects of these manipulations on the long-term nutrient status of these agroforestry systems.

267 References Anderson J.P.E. 1975. Einflüsse von Temperatur und Feuchte auf Verdampfung, Abbau und Festlegung von Diallat im Boden. Z. Pflanzenschutz. 7: 141–146. Anderson J.P.E. 1982. Soil respiration. In: Page A.L., Miller R.H. and Keeney D.R. 共eds兲, Methods of Soil Analysis: Part 2, Chemical and Microbiological Properties, 2nd Ed. American Society of Agronomy, Madison, Wisconsin, 831 pp. Anderson J.P.E. and Domsch K.H. 1978. Physiological method for the quantitative measurement of microbial biomass in soils. Soil. Biol. Biochem. 10: 215–221. Anderson J.M., Proctor J. and Vallack H. 1983. Ecological studies in four contrasting lowland rain forest in Gunung Mulu National Park, Sarawak III. Decomposition processes and nutrient losses from leaf litter. J. Ecol. 71: 503–527. Beare M.H., Parmelee R.W., Hendrix P.F., Cheng W., Coleman D.C. and Crossley D.A. 1992. Microbial and faunal interactions and effects on litter nitrogen and decompositions in agroecosystems. Ecol. Monogr. 62: 569–591. Beck L., Gasparotto L., Förster B., Franklin E., Garcia M., Harada A. et al. 1998. The role of soil fauna in litter decomposition in primary forests, secondary forests and a polyculture plantation in Amazonia 共SHIFT Project ENV 52兲: Methodological considerations. Proceedings of the Third SHIFT-Workshop, Manaus, 15–19 March. BMBF, Bonn, Germany, pp. 471–481. Dunger W. and Fiedler H.J. 1997. Methoden der Bodenbiologie. Fischer, Stuttgart, Germany. Feigl B.J., Sparling G.P., Ross D.J. and Cerri C.C. 1995. Soil microbial biomass in Amazonian soils: Evaluation of methods and estimates of pool sizes. Soil Biol. Biochem. 27共11兲: 1467–1472. Förster B., Farias M. and Luizão R. 1999. Microbial respiration and biomass in tropical forest soil and litter. Proceedings of the Third SHIFT-Workshop, Manaus, 15–19 March. BMBF, Bonn, Germany, pp. 483–485. Fragoso C. and Lavelle P. 1992. Earthworm communities of tropical rainforests. Soil Biol. Biochem. 24: 1397–1408. Höfer H., Martius C., Hanagarth W., Garcia M., Franklin E., Römbke J. and Beck L. 2000. Soil fauna and litter decomposition in primary and secondary forest and a mixed culture system in Amazonia. Final report of SHIFT project ENV 52. BMBF, Bonn, Germany. Höfer H., Martius C., Luizão F., Garcia M.B.V. and Beck L. 2004. Macrofauna determines decomposition rates in Amazonian anthropogenic and natural ecosystems. 共in preparation兲 Kröckel L. and Stolp H. 1986. Influence of the water regime on denitrification on aerobic respiration in soil. Biol. Fertil. Soils 2: 15–21.

Lavelle P.E., Blanchart A., Martin A., Spain F., Toutain I, Barois I. and Schaeffer R. 1993. Hierarchical model for decomposition in terrestrial ecosystems: application to soils of the humid tropics. Biotropica 25共2兲: 130–150. Martius C. 1994. Diversity and ecology of termites in Amazonian forests. Pedobiologia 38共5兲: 407–428. Martius C., Höfer H., Garcia M.V.B., Römbke J. and Hanagarth W. 2004a. Microclimate in agroforestry systems in central Amazonia: does canopy closure matter to soil organisms? Agrofor. Syst. 60: 291–304. Martius C., Höfer H., Garcia M.V.B., Römbke J. and Hanagarth W. 2004b. Litter fall, litter stocks and decomposition rates in rain forest and agroforestry sites in central Amazonia. Nutr. Cycling Agroecosyst. 68: 137–154. Martius C., Silva E.G., Garcia M. and Morais J.W. 2004c. Environmental parameters control soil- and litter-inhabiting termites in rain forest and agroforestry systems in central Amazonia. Ecography, in press. Martius C., Tiessen H., and Vlek P.L.G. 2001. The management of organic matter in tropical soils: What are the priorities? Nutr. Cycling Agroecosyst. 61: 1–6. Reddy M.V. 1992. Effects of microarthropod abundance and abiotic variables on mass loss and concentration of nutrients during decomposition of Azadirachta indica leaf litter. Trop. Ecol. 33: 89–96. Salick J.R., Herrera C.F. and Jordan C.F. 1983. Termitaria: nutrient patchiness in nutrient-deficient rain forest. Biotropica 15: 1–7. Stark N. and Jordan C.F. 1978. Nutrient retention by the root mat of an Amazonian rain forest. Ecology 59: 434–437. Swift M.J., Heal M.W. and Anderson J.M. 1979. Decomposition in Terrestrial Ecosystems. Studies in Ecology Vol. 5. Blackwell Scientific Publication, Oxford, UK. Tian G., Brussaard L. and Kang B.T. 1995. Breakdown of plant residues with contrasting chemical compositions under humid tropical conditions: Effect of earthworms and millipedes. Soil Biol. Biochem. 27: 277–280. Vohland K. and Schroth G. 1999. Distribution patterns of the litter macrofauna in agroforestry and monoculture plantations in central Amazonia as affected by plant species and management. Appl. Soil Ecol. 13: 57–68. Wood M. 1995. The role of bacteria and actinomycetes in litter decomposition in the tropics. In: Reddy M.V. 共ed.兲, Soil Organisms and Litter Decomposition in the Tropics. Westview Press, Boulder, Colorado, pp. 15–56. Yang J.C. and Isram H. 1991. Microbial biomass and relative contributions of bacteria and fungi in soil beneath tropical rain forest, Hainan Island, China. Trop. Ecol. 7: 385–393.