Soil quality indicators for intensive vegetable production systems in Java, Indonesia

Soil quality indicators for intensive vegetable production systems in Java, Indonesia

Ecological Indicators 18 (2012) 218–226 Contents lists available at SciVerse ScienceDirect Ecological Indicators journal homepage: www.elsevier.com/...

438KB Sizes 0 Downloads 61 Views

Ecological Indicators 18 (2012) 218–226

Contents lists available at SciVerse ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

Soil quality indicators for intensive vegetable production systems in Java, Indonesia Bram Moeskops a,∗ , David Buchan a , Sukristiyonubowo b , Stefaan De Neve a , Bart De Gusseme c , Ladiyani Retno Widowati b , Diah Setyorini b , Steven Sleutel a a

Department of Soil Management, Ghent University, Coupure Links 653, 9000 Gent, Belgium Indonesian Soil Research Institute, Bogor 16123, Indonesia c Department of Biochemical and Microbial Technology, Ghent University, 9000 Gent, Belgium b

a r t i c l e

i n f o

Article history: Received 17 December 2010 Received in revised form 8 November 2011 Accepted 12 November 2011 Keywords: Organic farming Soil quality Ergosterol PLFA Enzyme activity Indonesia

a b s t r a c t We explored the value of ergosterol, phospholipid fatty acid (PLFA) profiles and neutral lipid fatty acid (NLFA) 16:1␻5c as soil quality indicators for the intensive systems of vegetable production in the humid tropical climate of West Java, by comparing organic and conventional management. Additionally, we measured dehydrogenase and ␤-glucosidase activity. A secondary forest was included to obtain reference values under undisturbed conditions. Organic and conventional agriculture, and secondary forest each differed in the composition of the microbial community measured by PLFA profiles. Ergosterol appeared not to be universally applicable as a fungal biomarker and in this respect seemed to be inferior compared to PLFA fungal markers. NLFA 16:1␻5c may provide additional information as an indicator of arbuscular mycorrhizal fungi, but its high variability complicated the interpretation of data. The ratio of cy17:0 to 16:1␻7c was effectively applied as an indicator of physiological stress experienced by the bacterial community. Conventional vegetable production resulted in higher cy17:0/16:1␻7c ratios. Finally, a soil quality index, developed by stepwise canonical discriminant analysis and based on the absolute amount of PLFA 16:0, the relative amount of PLFAs 10Me16:0 and 10Me18:0, and dehydrogenase activity, was successfully validated. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction Numerous reports warn for the overuse of fertilizer and pesticides in vegetable production in Southeast Asia (see Moeskops et al., 2010 for a discussion of Mazlan and Mumford, 2005; Phupaibul et al., 2002; Poudel et al., 1998; Rerkasem, 2005; Schroer et al., 2005). The intensive character of vegetable production in Southeast Asia and Indonesia is further enhanced by continuous cultivation, made possible by the constantly high temperatures which prevail year-round. In the highlands of West Java, fallow periods are restricted to a few weeks only and cultivation of 4–6 crops per year on the same field is not uncommon. When monitoring the on-going land degradation in the tropics, sensitive soil quality indicators are to be identified. Because

∗ Corresponding author. Tel.: +32 9 264 60 61; fax: +32 9 264 62 47. E-mail addresses: [email protected] (B. Moeskops), [email protected] (D. Buchan), [email protected] (Sukristiyonubowo), [email protected] (S. De Neve), [email protected] (B. De Gusseme), [email protected] (L.R. Widowati), [email protected] (D. Setyorini), [email protected] (S. Sleutel). 1470-160X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecolind.2011.11.011

of the role of the microbial community in maintaining soil fertility (Giller et al., 1997) and the suppression of diseases (Janvier et al., 2007), and due to its rapid response to changes in management, microbial indicators seem to be particularly useful for the assessment of soil quality. Phospholipid and neutral fatty acids (PLFAs and NLFAs) and ergosterol are membrane bound cell components that are increasingly used to study the composition of the microbial community in soil. Ergosterol is the predominant sterol in fungal cell membranes. It is specific to higher fungal phyla, but does not occur in arbuscular mycorrhizal fungi (AMF) (Joergensen and Wichern, 2008; West et al., 1987). Due to their specificity and rapid degradation after cell death, PLFAs are reliable measures of the viable cell biomass of specific groups of microorganisms (Zelles, 1999). While PCR-based methods offer more insight into changes in specific microbial populations, PLFA analysis is the most powerful approach to demonstrate change in total microbial community structure (Ramsey et al., 2006). In this study, we will pay particular attention to the ratio of PLFA cy17:0 to PLFA 16:1␻7c. As (Gramnegative) bacteria enter the stationary growth phase, monoenoic ␻7 PLFAs are transformed into cyclopropyl fatty acids and hence the ratio of cy17:0 to 16:1␻7c has been proposed as an indicator of stress in the bacterial community (Bossio and Scow, 1998; Petersen and Klug, 1994). Bossio and Scow (1998) and Petersen and Klug

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

(1994) mention anaerobic conditions, low pH, high temperature and starvation as possible stress factors. Although the PLFA 16:1␻5c is often used as a biomarker for AMF, it also occurs in Gram-negative bacteria (Zelles, 1997). Therefore Olsson (1999) proposed the ratio between NLFA and PLFA 16:1␻5c to distinguish between AMF and Gram-negative bacteria as this ratio is high in AMF (1–200) and low in bacteria (<1). Energy in AMF is mainly stored in neutral lipids, of which NLFA 16:1␻5c comprises more than 60% (Olsson, 1999), while bacteria do not store energy in the form of lipids. “Bacterial” NLFAs found in soil are actually phospholipids of which the phosphate group has been cleaved as the first step in the decomposition process (Bååth, 2003). Analysis of NLFA 16:1␻5c has, to the best of our knowledge, not yet been undertaken in tropical soils. In this study we explore the indicator value of PLFA profiles, NLFA 16:1␻5c and ergosterol for the intensive systems of vegetable production in the humid tropical climate of West Java. We further compare these results with dehydrogenase and ␤-glucosidase activity, which are established indicators for microbial soil quality. Our analysis is based on a comparison of organic and conventional farms. Because of the excessive use of fertilizers and pesticides by conventional vegetable farmers in West Java (Moeskops et al., 2010), differences in soil quality with organic farming methods that rely exclusively on organic inputs are probably large. Large differences would be helpful for the identification of potential soil quality indicators.

2. Materials and methods 2.1. Experimental set-up At three locations in the West Java highlands, soils from an organic (OF) and a conventional vegetable farm (CF), within 1 km from each other, were selected. Two of these locations were situated in the Cisarua district (Bogor regency), further referred to as Cisarua1 and Cisarua2, the third one in the Ciwidey district (Bandung regency). The climate of West Java is still fully humid equatorial according to the Köppen–Geiger classification, but approaching the monsoonal equatorial climate in the east (Kottek et al., 2006). This means the climate is characterized by two seasons: a rainy season from October to April with about 80% of the annual precipitation and a dry season from May to September. More details about climate and altitude of the research sites can be found in Moeskops et al. (2010). Soils at all sites belonged to the Andisol order. Texture and management of the selected fields are summarized in Table 1. The OF in Cisarua1 was established in 1999. At the OF in Cisarua2 a distinction was made between a site that had been organically cultivated for 24 years and plots that had been converted from conventional management only three years before sampling. Also at the OF in Ciwidey two distinct sites were considered. At the first site, vegetable production was started in 1992 with organic principles adopted in 2002. The second site was overgrown with brushwood until the beginning of 2008 when it was cleared for organic vegetable production. This second site will be referred to as ‘OF-cleared site’. Finally, a secondary forest on a loam soil was selected in Ciwidey, within the 1 km radius from the OF, that provided natural reference values. At the two organic farms in Cisarua vegetables are cultivated on small beds of 10 m2 . Following the principle of intercropping, the same crop is never planted on adjacent beds. At the organic farm of Ciwidey crops are grown in groups of 3–15 beds of 8 m2 each. On all three organic farms, a second vegetable is generally intercropped between the rows of the main crop. Conventional vegetable production is also small-scale. The area of a single field, with one main crop and sometimes an intercrop, ranges between 500 and 5000 m2 . Organic and conventional fields of the

219

same location did not always have the same crops. Although different crops may have different effects on the microbial community, such crop dependent effects were not apparent from earlier data (presented in Moeskops et al., 2010). Furthermore, a wide range of vegetables is cultivated in a rapid and continuous succession at both the organic and conventional farms, which reduces the possibility of microbial communities being adapted to any specific crop. Whereas the organic farms applied a uniform fertilization rate for all crops, the conventional farmers applied variable rates of fertilizer (and pesticides) according to the crop grown. Hence, the rates given for the conventional farms only apply to the crops grown at the moment of sampling. The conventional farmers in Cisarua1 and Cisarua2 purchased dried poultry litter and mixed this before application with excreta from their own goats. The conventional farmer in Ciwidey only applied chemical fertilizer. Organic fertilizers applied at the organic farms were more variable in composition, but always consisted of a composted mixture of crop residues and animal manures (chicken and goat in Cisarua1, chicken in Ciasura2, cattle and chicken in Ciwidey). The organic farms in Cisarua applied smaller amounts of compost to each newly transplanted crop, while at the organic farm in Ciwidey higher compost doses were applied, but less frequently so. 2.2. Soil sampling Because the research sites were differently organized, the soil sampling strategy was designed to be site-specific. At the organic farms in Cisarua1 and Cisarua2 (both for 24-year OF and 3-year OF) three separate beds of 10 m2 , spread evenly over the respective site, were selected as replicates. At the organic farm in Ciwidey two times three adjacent replicate beds of 8 m2 were selected: three beds at the older organic site and three at the OF-cleared site. At the conventional fields, three plots of 10 m2 were selected, spaced approximately 5–10 m apart. In all replicates 15 samples were taken from the 0–15 cm soil layer and bulked into one composite sample per plot. The three replicates within each site can be considered as pseudo-replicates and not as completely independent replicates. The reason for not selecting different fields was the difficulty in finding organic fields and farmers willing to cooperate with the research. While the use of replicates within fields is not optimal, it is used and acknowledged also in other soil science papers (e.g. Sleutel et al., 2009). All sites were sampled in September 2008, around harvest of the crops. Soil water conditions during sampling were comparable at all sites. 2.3. Soil analyses 2.3.1. General soil properties Determination of general soil properties was carried out on airdried and sieved (2 mm) soil. pH–KCl was measured in 1 N KCl extracts (soil:KCl ratio of 1:2.5). Total C and N contents were measured with a Variomax CNS elemental analyzer (Elementar GmbH, Hanau, Germany) applying the Dumas method. Since pH–KCl values were acidic (less than 6.5), free carbonates were assumed not to be present and total carbon contents were considered equivalent to organic carbon contents. Texture was determined by the combined sieve and pipette method according to Gee and Bauder (1986). 2.3.2. Fatty acid analysis Soil samples for fatty acid analysis were freeze-dried and sieved (2 mm) after sampling and subsequently stored at −18 ◦ C until extraction. NLFAs and PLFAs were extracted using a modified Bligh and Dyer technique (Bligh and Dyer, 1959). Four gram freeze-dried soil was weighed in glass tubes. Lipids in the soil samples were extracted twice by adding 3.6 ml phosphate

220

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

Table 1 Management data and USDA soil texture of selected fields. Location

Management (texture) Organic (sandy clay loam)

Cisarua1 Conventional (sandy clay loam)

Solanum lycopersicum L. (tomato) Capsicum frutescens L. (chilli) Brassica oleracea L. (broccoli)

Organic –24 years (clay loam)

Capsicum frutescens L. (chilli) Daucus carota L. (carrot) Brassica oleracea L. (broccoli, cauliflower) Lactuca sativa L. (lettuce) Amaranthus hybridus L. (smooth amaranth) Arachis hypogaea L. (peanut) Phaseolus vulgaris L. (French bean) Crotalaria juncea L. (green manure)

Organic –3 years (clay loam)

Capsicum frutescens L. (chilli) Solanum lycopersicum L. (tomato) Allium fistulosum L. (scallion) Vigna angularis (Willd.) Ohwi & H. Ohashi (azuki bean) Lactuca sativa L. (lettuce) Brassica oleracea L. (broccoli, kai-lan) Ocimum basilicum L. (basil)

Conventional (clay loam)

Brassica oleracea L. (cabbage) Capsicum frutescens L. (chilli) Allium fistulosum L. (scallion)

Organic (sandy loam)

Solanum lycopersicum L. (tomato) Lactuca sativa L. (lettuce) Brassica rapa L. (bok choy)

Cisarua2

Ciwidey

a

Crop Solanum lycopersicum L. (tomato) Brassica oleracea L. (kai-lan, broccoli) Brassica rapa L. (bok choy, choy sum) Brassica juncea (L.) Czern. (leaf mustard) Amaranthus hybridus L. (smooth amaranth)

Organic –cleared site (silt loam) Conventional (sandy loam)

Fertilization

a

Dolomite enriched compost: 6.0 Mg DM = 69 kg N

Manure: 31 Mg DM = 659 kg N (NH4)2SO4: 5.0 kg N, 5.8 kg S phosphate: 21 kg P2O5 KCl: 14 kg K2O

Compost: 11 Mg DM = 80 kg N

Manure: 27 Mg DM = 574 kg N urea: 197 kg N NPK: 23 kg N, 23 kg P2O5, 23 kg K2O

Lime enriched compost: 125 Mg DM = 1.24 Mg N per year

Zea mays L. (baby corn) Zea mays L. (sweet corn)

Pesticides

Extract from tobacco leaves (Nicotiana tabacum L.)

-1

-1

Profenofos (194 mg l ), mancozeb (1778 mg l ) and deltamethrin applied once a week

No application of pesticides

-1

Emamectin benzoate (25 mg l ), propineb -1 (2100 mg l ) and mancozeb applied once a week Extract from wild plants: Toona sureni (B.l) Merr., Acmella paniculata (Wall. ex DC) R.K. Jansen, Mucuna pruriens (L.) Urban, Datura metel L., Tithonia diversifolia (Hemsl.) A. Gray No application of pesticides

Urea: 177 kg N

Mancozeb applied two times during growth cycle

Fertilization rates are given per ha and per growth cycle (85 days) unless otherwise stated. DM = dry matter.

buffer (pH 7.0), 4 ml chloroform and 8 ml methanol. Suspensions were shaken for 1 h and afterwards centrifuged for 10 min (1258 × g). The supernatants of both extraction cycles were collected in separatory funnels and 8 ml phosphate buffer and 8 ml chloroform were added to enhance phase separation. The next day, the lipid layers were transferred to new tubes, dried under N2 and re-dissolved in chloroform. The lipid extracts were separated into neutral, glyco- and phospholipids by chloroform, acetone and methanol respectively using SPE silica columns (Chromabond, Macherey-Nagel GmbH, Düren, Germany). Chloroform (NLFAs) and methanol (PLFAs) fractions were dried under N2 . The dried lipids were then re-dissolved in 1 ml methanol:toluene (1:1 v/v) and 1 ml 0.2 M methanolic KOH. Samples were incubated at 35 ◦ C for 15 min to allow transesterification to methyl esters. After cooling to room temperature, 2 ml hexane:chloroform (4:1 v/v), 1 ml 1 M acetic acid and 2 ml water were added. After vortexing, the samples were centrifuged for 5 min (805 × g). The hexane layers, containing the methylated fatty acids, were transferred to pointed tubes. The aqueous phases were washed twice with hexane:chloroform. The combined hexane phases were dried under N2 . The fatty acid methyl esters were finally re-dissolved in 0.3 ml of hexane containing methyl nonadecanoate fatty acid (19:0) as an internal standard. PLFAs and NLFAs were determined by GC–MS on a Thermo Focus GC coupled to a Thermo DSQ quadrupole MS (Thermo Fisher Scientific Inc., Waltham, USA) in electron ionization mode. Samples were chromatographically separated with a Restek capillary column Rt-2560 (100 m × 0.25 mm i.d., 0.2 ␮m film thickness; Restek, Bellefonte, USA). Following Bossio and Scow (1998) and Kozdrój and van Elsas (2001), the sums of marker PLFA concentrations for selected microbial groups were calculated. For Gram-positive bacteria the sum of i15:0, a15:0, i16:0, i17:0 and a17:0 was used. The PLFAs 16:1␻7c, 18:1␻7c and cy17:0 were considered to be typical for Gram-negative bacteria. The sum of 10Me16:0 and 10Me18:0 was

regarded as an indicator for the actinomycetes. The total bacterial community was assumed to be represented by the sum of the marker PLFAs for Gram-positive and Gram-negative bacteria, and PLFAs 15:0 and 17:0. PLFAs cy19:0 and 18:2␻6,9c co-eluted, preventing chromatographic separation and accurate quantification of 18:2␻6,9c, a common fungal biomarker. Instead of 18:2␻6,9c, 18:1␻9c was used as a fungal biomarker (Joergensen and Wichern, 2008; Kozdrój and van Elsas, 2001). The ratio between NLFA and PLFA 16:1␻5c was calculated to distinguish between AMF and Gram-negative bacteria, but we did not take the threshold value proposed by Olsson (1999), namely 1, as an absolute limit. According to Bååth (2003), PLFA 16:1␻5c is indicative of AMF if the NLFA/PLFA ratio of 16:1␻5c is higher than the NLFA/PLFA ratios of bacterial fatty acids with similar PLFA concentrations as PLFA 16:1␻5c. In our study bacterial PLFAs i17:0 and a17:0 had similar concentrations as PLFA 16:1␻5c. The ratio of PLFAs cy17:0 to 16:1␻7c served as an index for physiological stress in the bacterial community (Bossio and Scow, 1998; Petersen and Klug, 1994). The Shannon diversity index, as a measure of general diversity (Shannon and Weaver, 1949), was obtained considering only the data of these PLFAs that contributed more than 1% to the total PLFA pool. 2.3.3. Ergosterol Extraction and quantification of ergosterol was based on the method developed by Gong et al. (2001). In a glass vial, 2 g freezedried and sieved (2 mm) soil was mixed with 4 g glass beads (2 g of 290–420 ␮m and 2 g of 850–1230 ␮m). After the addition of 6 ml methanol, the vial was vortexed and subsequently shaken intensively for 1 h on a linear shaker. The soil mixture was then allowed to precipitate for 15 min, and a 1.5 ml aliquot of the supernatant was transferred into an Eppendorf microfuge tube. After centrifugation for 10 min at 10,000 × g the supernatant was loaded for analysis on a Dionex HPLC (P580 pump, TCC-100 column oven;

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

Dionex Corp., Sunnyvale, USA) equipped with a C18 reversedphase column (Allsphere ODS-2 5 ␮m, 250 mm × 4.6 mm; Grace, Deerfield, USA). Ergosterol could be measured after a retention time of 8.46 min at 282 nm using a UVD340S detector (Dionex Corp.). Methanol was used as the mobile phase at a flow rate of 1.5 ml min−1 . The column temperature was kept at 30 ◦ C. An additional spike experiment resulted in an average recovery of 98.6% for the soils in this study. As a consequence, no corrections for incomplete extraction were necessary. 2.3.4. Enzyme activities Dick (1994) pointed out that soil enzyme activities are very suitable as indicators of soil quality, based on their relationship to soil biology and soil functioning, rapid response to changes in soil management and their ease of measurement. Two enzyme activities were determined in this study. Dehydrogenase activity was selected as it represents a general measure of microbial activity (Alef and Nannipieri, 1995). ␤-Glucosidase is a hydrolytic enzyme and was selected, because an earlier study on organic and conventional vegetable fields in West Java (Moeskops et al., 2010), revealed that ␤-glucosidase activity had a relatively low variability compared to other hydrolases and was significantly higher under organic than under conventional vegetable production. The activity of ␤-glucosidase was measured according to a procedure modified from Eivazi and Tabatabai (1988; cited in Alef and Nannipieri, 1995) in which p-nitrophenyl-␤-d-glucoside is degraded to p-nitrophenol (PNP) during a 1 h incubation of fresh soil in Modified Universal Buffer pH 6.0 at 37 ◦ C. The produced PNP was extracted by Tris buffer pH 12, but also CaCl2 was added to the suspension to prevent dispersion of clay minerals. Dehydrogenase activity was determined according to Casida et al. (1964) as the reduction rate of triphenyltetrazolium chloride to triphenyl formazan (TPF) during a 24 h incubation in the dark of fresh soil in Tris buffer pH 7.8 at 37 ◦ C. TPF was extracted from the soil using methanol. Both enzyme activities were measured in triplicate with one blank on fresh soil stored at 4 ◦ C. Concentrations of PNP and TPF were determined with a Hitachi 150-20 spectrophotometer (Hitachi Ltd., Tokyo, Japan) at 400 nm and 485 nm respectively. More detailed procedures of the enzyme activity measurements are given in Moeskops et al. (2010). 2.4. Data processing To compare the relative composition of the microbial community in the different soil samples, PLFA concentrations were converted to percentages of the total PLFA concentration of the respective soil sample. Fisher’s canonical discriminant analysis (CDA) was applied to this percentage distribution with Tibco Spotfire S+ (version 8.1, TIBCO Software Inc., Palo Alto, USA) using correlation coefficients. Fisher’s CDA transforms data in order to discriminate between predefined groups (Huberty, 1994). In our analysis three groups were considered: CF, OF and secondary forest. After removal of all PLFAs that contributed less than 1% to the total pool of PLFAs, 20 PLFAs were retained for CDA. Statistical comparison between treatments was carried out for each location separately (at the 0.05 level of significance) using SPSS (version 15.0, SPSS Inc., Chicago, USA). First the homogeneity of variances was tested with Levene’s test. t-Tests and ANOVA/Tukey’s post hoc test were used to compare homoscedastic populations. Welch’s analysis-of-variance and Games–Howell’s multiple comparisons were applied for heteroscedastic populations. Pearson’s correlations coefficients mentioned in the text were also calculated with SPSS. Finally, we developed and validated a soil quality index based on biochemical and chemical soil parameters. The index was

221

developed by stepwise CDA in SPSS on OF and CF data presented by Moeskops et al. (2010), who found very large differences in soil quality between CF and OF in West Java. Moeskops et al. (2010) compared data from OF and CF collected in 2007. In this paper we used a similar approach, but the conventional farms sampled were different farms than these from Moeskops et al. (2010). The organic farms were the same, but we sampled now different fields on these farms as compared to Moeskops et al. (2010). Therefore, the validation of the index using the data in the current paper is justified. Stepwise CDA is a technique that allows to select the variables with the highest power to discriminate between predefined groups or treatments from a more extended data set (Puglisi et al., 2005, 2006). At each step of this method, the variable that minimizes the overall Wilks’ Lambda is entered into the model. Maximum significance of F to enter was set to 0.1, minimum significance of F to remove was 0.25. Data were scaled in order to base the CDA on correlation coefficients. In total 16 parameters and ratios between parameters were considered in calculating the index: soil organic C (SOC) and total N (TN) content, pH–KCl, the concentration of PLFA 16:0 (nmol g−1 dry soil), dehydrogenase and ␤-glucosidase activity, the proportions of the sums of marker PLFAs of Grampositive, Gram-negative and total bacteria, actinomycetes, fungi and AMF (16:1␻5c) to the total PLFA pool, the fungi to bacteria ratio, the Gram-positive to Gram-negative bacteria ratio, cy17:0/16:1␻7c and the Shannon diversity index of PLFAs. PLFA 16:0, the most ubiquitous PLFA, was considered as a measure for microbial biomass. The fungi to bacteria ratio and the Gram-positive to Gram-negative bacteria ratio were calculated by dividing the respective sums of marker PLFAs. 3. Results 3.1. Chemical soil properties In Cisarua1, pH and SOC and TN content did not differ between OF and CF (Table 2). In Cisarua2 on the other hand, TN content was significantly higher under long-term than under 3-year OF and significantly higher under 3-year OF than under CF. SOC content was significantly higher under long-term OF than under CF. In Ciwidey, SOC content was significantly higher under OF than under CF, and the OF-cleared site had a significantly higher SOC content than the older organic site. The OF-cleared site had a significantly higher TN content than the older organic site and CF. The OF-cleared site and CF had a comparable pH that was significantly lower than that of the older organic site. 3.2. Fatty acids As was also reported by Bååth (2003), NLFA 16:1␻5c measurements varied considerably in our study and as a result so did NLFA/PLFA ratios (Table 3). In Cisarua1, the NLFA/PLFA ratio was significantly higher under CF compared to OF. In Ciwidey, the NLFA/PLFA ratio was remarkably high at the OF-cleared site (more than five times higher than at the older organic site). In Ciwidey, all ratios were higher than 1 and thus certainly for this location there was little doubt about the AMF origin of PLFA 16:1␻5c. In Fig. 1 the logarithm of the NLFA/PLFA ratios of the bacterial fatty acids i17:0 and a17:0 is plotted against the logarithm of the amount of PLFA. A significant linear decrease could be observed in the log ratio with increasing log PLFA amounts both for Cisarua1 and Cisarua2. This means that the main reason for different NLFA/PLFA ratios was variable amounts of PLFA with a constant background amount of NLFA (Bååth, 2003). Both in Cisarua1 and Cisarua2, the fatty acid 16:1␻5c had higher NLFA/PLFA ratios than would be expected for

222

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

Table 2 General soil properties and ergosterol contents. Location

Management

Cisarua1

Organic Conventional

Cisarua2

Ciwidey

pH–KCl

SOC (%)

Total N (%)

C/N ratio

Ergosterol (␮g ergosterol g−1 dry soil)

5.02 (0.11) 5.05 (0.09)

4.10 (0.16) 4.21 (0.17)

0.42 (0.03) 0.40 (0.01)

9.8 (0.3) 10.6 (0.2)

1.90 (0.19)b 0.99 (0.20)a

Organic – 24 years Organic – 3 years Conventional

5.59 (0.01) 5.53 (0.05) 4.85 (0.21)

3.90 (0.19)b 3.25 (0.17)ab 2.96 (0.22)a

0.38 (0.02)c 0.32 (0.02)b 0.25 (0.01)a

10.2 (0.1) 10.4 (0.1) 12.0 (0.8)

1.60 (0.18) 1.84 (0.24) 1.36 (0.23)

Organic Organic – clrd. site Conventional Secondary forest

5.92 (0.08)b 5.06 (0.03)a 5.25 (0.06)a 4.92 (0.34)

4.80 (0.14)b 6.80 (0.22)c 3.59 (0.14)a 7.58 (1.66)

0.43 (0.01)a 0.65 (0.03)b 0.37 (0.02)a 0.53 (0.13)

11.3 (0.1)b 10.5 (0.2)ab 9.8 (0.3)a 14.5 (1.1)

1.08 (0.19)a 1.49 (0.22)a 1.86 (0.32)a 3.99 (0.37)b

Values in parentheses indicate standard errors. Data in italics were analyzed by Welch/Games–Howell. Secondary forest was excluded from statistical analysis on general soil properties because of too high variability. Significant differences are indicated by different letters per location (P < 0.05); no letters if no significant differences were found.

bacterial fatty acids (Fig. 1), indicating that also at those locations PLFA 16:1␻5c was indicative of AMF. In the remaining part of this paper PLFA 16:1␻5c will therefore be considered as a marker PLFA for AMF.

Fig. 1. NLFA/PLFA ratios plotted against the amount of PLFA. Regression for i17:0 and a17:0 includes all field replicates. For 16:1␻5c site averages and standard errors are shown; a. Cisarua1, b. Cisarua2.

In Cisarua1 no significant differences were found in absolute marker PLFA concentrations between OF and CF, although concentrations were always higher under OF (Table 3). In Cisarua2, all microbial groups considered were significantly more abundant under long-term OF than under CF. No significant differences were found between 3-year OF and CF. In Ciwidey, all microbial groups were more abundant under secondary forest than in agricultural soil, significantly so for fungi, Gram-positive, Gram-negative and total bacteria. With regard to actinomycetes and AMF, Welch’s test revealed a significant effect of soil management, but the power of the Games–Howell’s multiple comparisons was not strong enough to detect significant pairwise differences. The OF-cleared site had significantly higher marker PLFA concentrations of Gram-positive and total bacteria than the older organic site and CF. The first dimension of Fisher’s CDA discriminated between forest and agriculture, while the second dimension separated organic management from conventional management (Fig. 2). The first dimension strongly and positively correlated with PLFA 18:1␻7c, but negatively with cy17:0 (Table 4). This would mean that relatively less bacteria are in the stationary growth phase in the forest soil compared to cultivated soils. The first dimension was further strongly and negatively correlated with PLFA 20:5, a marker PLFA for protozoa (Kozdrój and van Elsas, 2001), suggesting that there were proportionally less protozoa under secondary forest than in agricultural soil. Burke et al. (2003) also found that protozoa are relatively more important members of the microbial community

Fig. 2. Scatter plot of Fisher’s CDA on PLFAs.

0.645 (0.015) 0.786 (0.015) 0.828 (0.047) 0.635 (0.008) 2.65 (0.01)ab 2.67 (0.01)b 2.68 (0.02)b 2.60 (0.01)a

First dimension

Second dimension

PLFA

Corr. coeff.

PLFA

Corr. coeff.

17:0 cy17:0 18:0 18:1␻7c 24:0 20:5␻3,6,9,12,15

−0.587 −0.591 −0.519 0.780 −0.476 −0.521

18:1␻5c

0.475

in agricultural sites than in forest habitats. The second dimension of the CDA was most strongly correlated with PLFA 18:1␻5c. Monounsaturated PLFAs, such as 18:1␻5c, have been considered as indicators for high substrate availability (Bossio and Scow, 1998; Moore-Kucera and Dick, 2008). OF exhibited lower cy17:0 to 16:1␻7c ratios than CF at all sites (Table 3). In Cisarua2, this difference was significant between longterm OF and CF, while 3-year OF took an intermediate position. In Ciwidey, Welch’s test revealed a significant effect of soil management, but the power of the Games–Howell’s multiple comparisons was not strong enough to detect significant pairwise differences. The low cy17:0/16:1␻7c ratio found under secondary forest corroborated the results of Fisher’s CDA. OF and CF had similar Shannon indices at all three locations (Table 3). In Ciwidey, secondary forest had a significantly lower PLFA diversity than CF and the OF-cleared site. Values in parentheses indicate standard errors. Data in italics were analyzed by Welch/Games–Howell. Significant differences are indicated by different letters per location (P < 0.05); no letters if no significant differences were found.

3.01 (0.21)a 3.65 (0.34)a 2.59 (0.28)a 9.86 (0.34)b 3.65 (0.28) 3.34 (0.19) 1.93 (0.15) 8.65 (0.75) 32.66 (1.65)a 42.03 (1.31)b 29.29 (1.69)a 104.25 (2.97)c 4.96 (0.20) 7.39 (0.55) 3.80 (0.25) 20.41 (1.98) 18.81 (1.24)a 27.54 (0.67)b 18.00 (1.82)a 64.88 (1.81)c Organic Organic – clrd. site Conventional Secondary forest Ciwidey

12.58 (0.86)a 12.99 (0.83)a 9.25 (0.75)a 36.43 (1.15)b

30.62 (1.38)b 23.72 (1.82)a 22.68 (0.17)a Organic – 24 years Organic – 3 years Conventional Cisarua2

223

Table 4 Pearson correlation coefficients between mol% of PLFAs and CDA dimensions with P ≤ 0.001.

2.85 (0.04) 14.80 (2.11) 3.68 (0.73) 3.20 (0.70)

0.556 (0.038)a 0.678 (0.020)ab 0.728 (0.030)b 2.67 (0.01) 2.68 (0.02) 2.66 (0.01) 1.06 (0.19) 1.71 (0.22) 1.03 (0.28) 4.72 (0.11)b 3.88 (0.37)ab 3.36 (0.13)a 4.46 (0.29)b 3.35 (0.31)a 3.04 (0.09)a 7.61 (0.02)b 6.36 (0.28)a 6.19 (0.22)a 16.09 (0.88)b 12.79 (0.94)ab 11.91 (0.47)a

48.92 (2.15)b 38.34 (2.92)a 36.26 (0.53)a

cy17:0/16:1␻7c H

0.653 (0.014) 0.723 (0.029) 0.89 (0.13)a 1.50 (0.12)b

NLFA/PLFA 16:1␻5c Fungi

3.08 (0.25) 2.36 (0.19)

AMF

2.81 (0.12) 2.36 (0.12) 36.96 (1.61)b 31.36 (1.04)b 6.04 (0.24) 4.96 (0.31) 10.77 (0.51)b 8.56 (0.53)a

Total bacteria Actinomycetes Gram-negative Gram-positive

Organic Conventional Cisarua1

24.60 (1.35) 21.40 (0.50)

Management Location

Table 3 Concentrations of marker PLFAs (nmol g−1 dry soil), NLFA/PLFA ratios of 16:1␻5c, Shannon diversity indices (H) and cy17:0/16:1␻7c ratios.

2.65 (0.01) 2.66 (0.01)

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

3.3. Ergosterol In Cisarua1 and Cisarua2 ergosterol contents were higher under OF than under CF, but this difference was significant only in Cisarua1 (Table 2). In Ciwidey, ergosterol content was significantly higher under secondary forest than under agriculture, but no significant differences were found between the three agricultural sites. 3.4. Enzyme activities In Cisarua1 and Cisarua2, dehydrogenase and ␤-glucosidase activities were higher under OF than under CF (Fig. 3). In Cisarua1 this difference was significant for ␤-glucosidase. In Cisarua2, both ␤-glucosidase activity and dehydrogenase activity were significantly higher under long-term OF than under CF. Short-term OF only had a significantly higher dehydrogenase activity than CF. In Ciwidey, lowest activities were found under CF and at the older organic site. 3.5. Soil quality index Stepwise CDA using the data of Moeskops et al. (2010) resulted in the following soil quality index (parameters listed in order of entrance into the model): 2.58 × PLFA 16 : 0 + 0.828 × proportion of actinomycetes PLFAs to total PLFA pool + 1.28 × dehydrogenase activity As regards the data of Moeskops et al. (2010), soil quality scores clearly separated OF from CF. Soil quality scores obtained from that study were significantly higher under OF than under CF (P < 0.01). No significant difference was found between short-term and longterm OF in Cisarua2 (P > 0.05) (ANOVA models as described in Moeskops et al., 2010). In the current study, the soil quality index was significantly higher under OF compared to CF in Cisarua1 (Table 5). In Cisarua2,

224

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

Table 5 Soil quality index scores (SQI). Location

Management

SQI

Cisarua1

Organic Conventional

6.01 (1.08)b 2.43 (0.39)a

Cisarua2

Organic – 24 years Organic – 3 years Conventional

8.84 (0.38) 7.51 (1.32) 5.78 (0.15)

Ciwidey

Organic Organic – cleared site Conventional

3.31 (0.52)a 8.62 (1.40)b 0.63 (0.62)a

Values in parentheses indicate standard errors. Significant differences are indicated per location by different letters (P < 0.05); no letters if no significant differences were found.

no significant differences were found, but OF had higher indices than CF. In Ciwidey, the OF-cleared site had a significantly higher soil quality index than the older OF site and CF. 4. Discussion 4.1. Fungal biomarkers Ergosterol content ranged between 0.99 and 1.90 ␮g g−1 dry soil under agriculture, and reached 3.99 ␮g g−1 dry soil under secondary forest. These concentrations are higher than those reported for tropical Andisols under agriculture in Nicaragua: 0.08–0.49 ␮g g−1 dry soil for CF and 0.14–0.69 ␮g g−1 dry soil for OF (Castillo and Joergensen, 2001; Joergensen and Castillo, 2001). Ergosterol contents reported by Turgay and Nonaka (2002) for Japanese Andisols under vegetable production (0.79–1.55 ␮g g−1 dry soil) and young forest (1.90–2.39 ␮g g−1 dry soil) were comparable to those found in our study. As already explained, chromatographic separation of PLFA 18:2␻6,9c from PLFA cy19:0 was not successful, except for 10 samples. Pearson’s correlation

coefficient between PLFA 18:1␻9c and PLFA 18:2␻6,9c calculated for those 10 samples was 0.747 (P < 0.05). Together with the sizeable amount of reports found in the literature (e.g. Bååth, 2003; Joergensen and Wichern, 2008; Kozdrój and van Elsas, 2001), this confirms PLFA 18:1␻9c as a suitable fungal biomarker. Surprisingly, results for PLFA 18:1␻9c and ergosterol did not correlate (r = 0.304, P = 0.464). The only significant difference in ergosterol content between OF and CF was found in Cisarua1, while no significant difference could be found for 18:1␻9c at that location. In Cisarua2, on the other hand, significant differences were found for 18:1␻9c, but not for ergosterol. The lack of correlation between PLFA 18:1␻9c and ergosterol thus contrasts with the results of Klamer and Bååth (2004) who found a good correlation between ergosterol and PLFA 18:2␻6,9c. Högberg (2006) suggested that ergosterol is only a reliable biomarker for fungi in relatively undisturbed soils, which the vegetable cultivated soils of West Java are obviously not. Helfrich et al. (2008) and Zhao et al. (2005) reported that ergosterol could not capture the considerable decrease in fungal biomass following fungicide application. Mille-Lindblom et al. (2004) showed that the decomposition of ergosterol, in soil without living fungi, is a rather slow process with a half-life of ca. 3–5 months. We therefore conclude that ergosterol cannot be considered a reliable indicator of fungal biomass for the intensive vegetable production systems in West Java. The indicator value of ergosterol is further impaired by the fact that coefficients of variation in this study were higher for ergosterol than for PLFA 18:1␻9c, even though both compounds were extracted from the same freeze-dried, sieved and homogenized samples. However, ergosterol may still be useful as an indicator when larger differences in fungal biomass are expected such as those between secondary forest and agriculture. 4.2. Fatty acid ratios Bossio and Scow (1998) and Petersen and Klug (1994) mention among others decreasing pH and starvation as possible factors that

Fig. 3. a. ␤-Glucosidase activity and b. dehydrogenase activity. Error bars indicate standard errors. Dehydrogenase activity in Ciwidey was analyzed by Welch/Games–Howell. Significant differences are indicated by different letters per location (P < 0.05); no letters if no significant differences were found.

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

cause the cy17:0/16:1␻7c ratio to increase. No correlation between pH and cy17:0/16:1␻7c was found in this study, but the positive correlation between the second dimension of Fisher’s CDA and PLFA 18:1␻5c pointed to lower substrate availability under conventional vegetable production. Nevertheless, it is not probable that soil microbial communities under conventional management were resource limited as application rates of manure are also high in the conventional systems (also at the CF in Ciwidey, although no manure was applied during the sampled growth cycle). Furthermore, there was no correlation between cy17:0/16:1␻7c and SOC content (r = 0.052, P > 0.05). Higher physiological stress under CF compared to OF is more probably due to the intensive use of pesticides of which the negative impacts on the soil microbial community have already been discussed by Moeskops et al. (2010). Further, the high stress signature of the OF-cleared site in Ciwidey is remarkable. It is possible that the soil microbial community was severely disturbed by the clearance of the brushwood and subsequent cultivation and had not yet reached a new stage of equilibrium. The high NLFA/PLFA ratio of 16:1␻5c at the OF-cleared site was probably also a consequence of the conversion of brushwood into farmland. As indicated by PLFA 16:1␻5c, AMF biomass was reduced to an amount comparable with the older organic site, but storage structures, apparently not fully decomposed after six months, still contained large amounts of the NLFA 16:1␻5c, resulting in high NLFA/PLFA ratios. The NLFA/PLFA ratio of 16:1␻5c was further significantly higher under CF than under OF in Cisarua1. This again points to a relatively high presence of storage structures compared to the amount of hyphae. In a field study by Gryndler et al. (2006), AMF spores were also significantly more abundant, while AMF hyphal length was significantly lower, when manure was combined with mineral fertilizer than when manure alone was applied. The low diversity of the soil microbial community in the secondary forest, may be an indication of resource limitation. Jangid et al. (2008) found that oligotrophic bacteria outcompeted copiotrophic bacteria in forest soil in Georgia (USA) resulting in lower bacterial diversity than in agricultural soils. Manure and compost are indeed much richer nutrient sources than forest litter. Upchurch et al. (2008) further proposed that the higher bacterial diversity observed in managed agricultural soils results from greater (seasonal) variation in the plant community and increased immigration of wind or animal borne bacteria to open cropland.

4.3. Correlations between parameters Absolute concentrations of marker PLFAs for Gram-positive and Gram-negative bacteria, actinomycetes, total bacteria, AMF and fungi were all positively correlated with dehydrogenase activity (P < 0.05). ␤-glucosidase activity was positively correlated with Gram-negative and total bacteria and fungi (P < 0.05). These correlations confirm the functional link between microbial biomass, microbial activity and organic matter turnover (as measured by ␤glucosidase activity). Fungi are considered to be the main actors in the process of organic matter decomposition (Killham, 1994). In our study higher ␤-glucosidase activities corresponded to larger relative amounts of the PLFA 18:1␻9c (0.825, P < 0.01). The cy17:0/16:1␻7c ratio was negatively correlated with absolute concentrations of the AMF marker PLFA 16:1␻5c (−0.771, P < 0.05, excluding secondary forest) indicating that AMF are negatively affected by conditions that are as well stressful for the bacterial community, like conventional management. The negative impact of conventional agricultural management on AMF has been reported by several other studies (e.g. Bending et al., 2004; Kurle and Pfleger, 1994; Mäder et al., 2002). Together with their obligate symbiotic nature, the susceptibility to disturbance makes

225

AMF important potential indicators of soil fertility in sustainable agricultural systems (Bending et al., 2004).

4.4. Comparison of indicators In Cisarua1, most microbial parameters pointed to increased soil quality under OF compared to CF. Enzyme activities, marker PLFA contents and the soil quality index were higher, while the cy17:0/16:1␻7c ratio was lower under OF. However, only ␤glucosidase activity and the soil quality index showed significant differences. In Cisarua2, enzyme activities under 24-year and 3year OF were comparable and higher than under CF (significantly so for dehydrogenase). In contrast to the findings of Moeskops et al. (2010), most marker PLFA contents were significantly lower under 3-year OF than under 24-year OF. This would suggest that the microbial community actually had not yet fully recovered from the conventional farming methods. The intermediate position of the soil microbial community after three years of organic farming was reflected in the cy17:0/16:1␻7c ratio and the soil quality index. No significant differences were found between the older OF site and CF in Ciwidey. A negative effect of agriculture on the microbial community could be noticed in Ciwidey since the total amount of bacteria marker PLFAs was significantly lower at the older OF site than at the OF-cleared site, which was until recently overgrown with brushwood. Also enzyme activities were lower at the older OF site than at the OF-cleared site (but not significantly). Like in Cisarua1 and Cisarua2, the information obtained from the individual microbial indicators was also in Ciwidey adequately summarized in the soil quality index. The soil quality index was much higher at the OF-cleared site than at the older OF site and under CF.

5. Conclusions Because field trials comparing organic and conventional management were not available, we had to rely in this study on fields managed by farmers. As a result application rates of organic and chemical fertilizers, and pesticides could not entirely be controlled. Therefore some caution may be required in interpretation of the results. Nevertheless, some general conclusions can be drawn. Ergosterol appeared not to be universally applicable as an indicator for fungi and in this respect seems to be inferior compared to PLFA markers (18:1␻9c or 18:2␻6,9c). NLFA 16:1␻5c may provide additional information on AMF, but its high variability complicates the interpretation of data. The ratio of cy17:0 to 16:1␻7c was effectively applied as an indicator of physiological stress experienced by the bacterial community. Because dehydrogenase activity was included in the soil quality index and ␤-glucosidase activity was not, dehydrogenase activity is probably more sensitive to differences in management than ␤-glucosidase activity. The soil quality index developed from the data of Moeskops et al. (2010) was successfully validated in this study and summarized the information obtained from the individual parameters and indices well. This means stepwise CDA provides a practicable approach to develop indices. The soil quality index presented in this study seems to be useful to assess soil quality of vegetable production systems in the humid tropics, but we recommend further testing to asses its range of application.

Role of funding sources Research Foundation – Flanders (FWO): Bram Moeskops and Bart De Gusseme were PhD fellows at the FWO, Steven Sleutel is post-doctoral research fellow at the FWO.

226

B. Moeskops et al. / Ecological Indicators 18 (2012) 218–226

Faculty Bioscience Engineering (Ghent University): provided a travel grant to Bram Moeskops allowing him to conduct research in Indonesia. Flemish Interuniversity Council (VLIR-UOS): provided a PhD scholarship to L.R. Widowati and provided additional financial support for the research activities in Indonesia. None of the funding sources played a role in the design of the study, the collection and interpretation of the data, the writing of this paper and the decision to submit the paper for publication. Acknowledgements We thank the field staff of the Indonesian Soil Research Institute and the technical staff of the Department of Soil Management (Ghent University) for technical assistance. Special thanks go the Indonesian students Irfan and Deni for their assistance in the lab. References Alef, K., Nannipieri, P. (Eds.), 1995. Methods in Applied Soil Microbiology and Biochemistry. Academic Press, London. Bååth, E., 2003. The use of neutral lipid fatty acids to indicate the physiological conditions of soil fungi. Microb. Ecol. 45, 373–383. Bending, G.D., Turner, M.K., Rayns, F., Marx, M.C., Wood, M., 2004. Microbial and biochemical soil quality indicators and their potential for differentiating areas under contrasting agricultural management regimes. Soil Biol. Biochem. 36, 1785–1792. Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911–917. Bossio, D.A., Scow, K.M., 1998. Impacts of carbon and flooding on soil microbial communities: phospholipid fatty acid profiles and substrate utilization patterns. Microb. Ecol. 35, 265–278. Burke, R.A., Molina, M., Cox, J.E., Osher, L.J., Piccolo, M.C., 2003. Stable carbon isotope ratio and composition of microbial fatty acids in tropical soils. J. Environ. Qual. 32, 198–206. Casida, L.E., Klein Jr., D.A., Santoro, T., 1964. Soil dehydrogenase activity. Soil Sci. 98, 371–376. Castillo, X., Joergensen, R.G., 2001. Impact of ecological and conventional arable management systems on chemical and biological soil quality indices in Nicaragua. Soil Biol. Biochem. 33, 1591–1597. Dick, R.P., 1994. Soil enzyme activities as indicators of soil quality. In: Doran, J.W., Coleman, D.C., Bezdicek, D.F., Stewart, B.A. (Eds.), Defining Soil Quality for a Sustainable Environment. Soil Science Society of America, Madison, pp. 107–124. Gee, G.W., Bauder, J.W., 1986. Particle-size analysis. In: Klute, A. (Ed.), Methods of Soil Analysis, Part 1: Physical and Mineralogical Methods. second ed. Agronomy 9. ASA, SSSA, Madison, pp. 383–411. Giller, K.E., Beare, M.H., Lavelle, P., Izac, A.-M.N., Swift, M.J., 1997. Agricultural intensification, soil biodiversity and agroecosystem function. Appl. Soil Ecol. 6, 3–16. Gong, P., Guan, X., Witter, E., 2001. A rapid method to extract ergosterol from soil by physical disruption. Appl. Soil Ecol. 17, 285–289. ˇ Gryndler, M., Larsen, J., Hrˇselová, H., Rezᡠcová, V., Gryndlerová, H., Kubát, J., 2006. Organic and mineral fertilization, respectively, increase and decrease the development of external mycelium of arbuscular mycorrhizal fungi in a long-term field experiment. Mycorrhiza 16, 159–166. Helfrich, M., Ludwig, B., Potthoff, M., Flessa, H., 2008. Effect of litter quality and soil fungi on macroaggregate dynamics and associated partitioning of litter carbon and nitrogen. Soil Biol. Biochem. 40, 1823–1835. Högberg, M.N., 2006. Discrepancies between ergosterol and the phospholipid fatty acid 18:2␻6,9 as biomarkers for fungi in boreal forest soils. Soil Biol. Biochem. 38, 3431–3435. Huberty, C.J., 1994. Applied Discriminant Analysis. John Wiley & Sons, New York. Jangid, K., Williams, M.A., Franzluebbers, A.J., Sanderlin, J.S., Reeves, J.H., Jenkins, M.B., Endale, D.M., Coleman, D.C., Whitman, W.B., 2008. Relative impacts of landuse, management intensity and fertilization upon soil microbial community structure in agricultural systems. Soil Biol. Biochem. 40, 2843–2853. Janvier, C., Villeneuve, F., Alabouvette, C., Edel-Hermann, V., Mateille, T., Steinberg, C., 2007. Soil health through soil disease suppression: which strategy from descriptors to indicators? Soil Biol. Biochem. 39, 1–23. Joergensen, R.G., Castillo, X., 2001. Interrelationships between microbial and soil properties in young volcanic ash soils of Nicaragua. Soil Biol. Biochem. 33, 1581–1589.

Joergensen, R.G., Wichern, F., 2008. Quantitative assessment of the fungal contribution to microbial tissue in soil. Soil Biol. Biochem. 40, 2977–2991. Killham, K., 1994. Soil Ecology. Cambridge University Press, Cambridge. Klamer, M., Bååth, E., 2004. Estimation of conversion factors for fungal biomass determination in compost using ergosterol and PLFA 18:2␻6,9. Soil Biol. Biochem. 36, 57–65. Kottek, M., Grieser, J., Beck, C., Rudolf, B., Rubel, F., 2006. World map of the Köppen–Geiger climate classification updated. Meteorol. Z. 15, 259–263. Kozdrój, J., van Elsas, J.D., 2001. Structural diversity of microorganisms in chemically perturbed soil assessed by molecular and cytochemical approaches. J. Microbiol. Methods 43, 197–212. Kurle, J.E., Pfleger, F.L., 1994. The effects of cultural practices and pesticides on VAM fungi. In: Pfleger, F.L., Linderman, R.G. (Eds.), Mycorrhizae and Plant Health. APS Press, St. Paul, pp. 101–131. Mäder, P., Fließbach, A., Dubois, D., Gunst, L., Fried, P., Niggli, U., 2002. Soil fertility and biodiversity in organic farming. Science 296, 1694–1697. Mazlan, N., Mumford, J., 2005. Insecticide use in cabbage pest management in the Cameron Highlands, Malaysia. Crop Prot. 24, 31–39. Mille-Lindblom, C., von Wachenfeldt, E., Tranvik, L.J., 2004. Ergosterol as a measure of living fungal biomass: persistence in environmental samples after fungal death. J. Microbiol. Methods 59, 253–262. Moeskops, B., Sukristiyonubowo, Buchan, D., Sleutel, S., Herawaty, L., Husen, E., Saraswati, R., Setyorini, D., De Neve, S., 2010. Soil microbial communities and activities under intensive organic and conventional vegetable farming in West Java, Indonesia. Appl. Soil Ecol. 45, 112–120. Moore-Kucera, J., Dick, R.P., 2008. PLFA profiling of microbial community structure and seasonal shifts in soils of a Douglas-fir chronosequence. Microb. Ecol. 55, 500–511. Olsson, P.A., 1999. Signature fatty acids provide tools for determination of the distribution and interactions of mycorrhizal fungi in soil. FEMS Microbiol. Ecol. 29, 303–310. Petersen, S.O., Klug, M.J., 1994. Effects of sieving, storage, and incubation temperature on the phospholipid fatty acid profile of a soil microbial community. Appl. Environ. Microbiol. 60, 2421–2430. Phupaibul, P., Kaewsuwan, U., Chitbuntanorm, C., Chinoim, N., Matoh, T., 2002. Evaluation of environmental impact of the raised-bed-dike (Rong Chin) system along the Tha Chin River in Suphan Buri-Nakhon Pathom Provinces, Thailand. Soil Sci. Plant Nutr. 48, 641–649. Poudel, D.D., Midmore, D.J., Hargrove, W.L., 1998. An analysis of commercial vegetable farms in relation to sustainability in the uplands of Southeast Asia. Agric. Syst. 58, 107–128. Puglisi, E., Del Re, A.A.M., Rao, M.A., Gianfreda, L., 2006. Development and validation of numerical indexes integrating enzyme activities of soils. Soil Biol. Biochem. 38, 1673–1681. Puglisi, E., Nicelli, M., Capri, E., Trevisan, M., Del Re, A.A.M., 2005. A soil alteration index based on phospholipid fatty acids. Chemosphere 61, 1548–1557. Ramsey, P.W., Rillig, M.C., Feris, K.P., Holben, W.E., Gannon, J.E., 2006. Choice of methods for soil microbial community analysis: PLFA maximizes power compared to CLPP and PCR-based approaches. Pedobiologia 50, 275–280. Rerkasem, B., 2005. Transforming subsistence cropping in Asia. Plant Prod. Sci. 8, 275–287. Schroer, S., Sulistyanto, D., Ehlers, R.-U., 2005. Control of Plutella xylostella using polymer-formulated Steinernema carpocapsae and Bacillus thuringiensis in cabbage fields. J. Appl. Entomol. 129, 198–204. Shannon, C.E., Weaver, W., 1949. The Mathematical Theory of Communication. University of Illinois Press, Urbana. Sleutel, S., Vandenbruwane, J., De Schrijver, A., Wuyts, K., Moeskops, B., Verheyen, K., De Neve, S., 2009. Patterns of dissolved organic carbon and nitrogen fluxes in deciduous and coniferous forests under historic high nitrogen deposition. Biogesciences 6, 2743–2758. Turgay, O.C., Nonaka, M., 2002. Effects of land-use and management practices on soil ergosterol content in andosols. Soil Sci. Plant Nutr. 48, 693–699. Upchurch, R., Chi, C.Y., Everett, K., Dyszynski, G., Coleman, D.C., Whitman, W.B., 2008. Differences in the composition and diversity of bacterial communities from agricultural and forest soils. Soil Biol. Biochem. 40, 1294–1305. West, A.W., Grant, W.D., Sparling, G.P., 1987. Use of ergosterol, diaminopimelic acid and glucosamine contents of soils to monitor changes in microbial populations. Soil Biol. Biochem. 19, 607–612. Zelles, L., 1997. Phospholipid fatty acid profiles in selected members of soil microbial communities. Chemosphere 35, 275–294. Zelles, L., 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in the characterisation of microbial communities in soil: a review. Biol. Fertil. Soils 29, 111–129. Zhao, X.R., Lin, Q., Brookes, P.C., 2005. Does soil ergosterol concentration provide a reliable estimate of soil fungal biomass? Soil Biol. Biochem. 37, 311–317.