Differences in soil quality indicators between organic and sustainably managed potato fields in Eastern Canada

Ecological Indicators 37 (2014) 119–130

Contents lists available at ScienceDirect

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

Differences in soil quality indicators between organic and sustainably managed potato fields in Eastern Canada Johanna E. Nesbitt a , Sina M. Adl a,b,∗ a b

Department of Biology, Dalhousie University, Halifax, NS, Canada, B3H 4J1 Department of Soil Science, University of Saskatchewan, Saskatoon, SK, Canada, S7N 5A8

a r t i c l e

i n f o

Article history: Received 2 April 2013 Received in revised form 13 August 2013 Accepted 1 October 2013 Keywords: Agro-ecosystem Bio-indicators Farming systems Organic agriculture Soil ecology Soil quality

a b s t r a c t The aim of this study was to determine if organic management of fields promoted soil quality indicators compared to sustainably managed fields following best-management practice guidelines. Using a soil quality minimum data set, conventionally and organically managed commercial potato fields in eastern Canada were compared. Microbial biomass, testate amoebae, nematodes, and microarthropods served as bioindicators, while soil pH, C:N ratio, light fraction, bulk density, and soil moisture served as the chemical and physical indicators. We also investigated whether differences in site location (different soil texture and local climate) were more or less important than field management (organic or conventional). When site location and seasonal factors were considered, the soil quality indicators were better at differentiating organic and conventional potato fields. There was no single indicator that could clearly differentiate, on its own, between the two field managements due to variability with site location or month of sampling. Microbial biomass, testate amoebae, microarthropod and soil moisture varied significantly through the growing season. The mean soil pH, C:N ratio, and moisture were significantly different between sites. However, the indicators were affected to different degrees, and differed to some extent to both “site location” and “time of sampling”. The results of this study also provide a baseline for similar soil quality evaluations in management of eastern Canada potato fields. We recommend that several indicators, including bioindicators should be used together, and that several sites should be sampled. In addition, one-time field sampling of an indicator, as it has been often practiced by growers, is likely to give false results as it does not account for variability through the growing season. © 2013 Elsevier Ltd. All rights reserved.

1. Introduction Intensification of agriculture was accompanied with increased use of synthetic fertilizers, pesticides and herbicides that has raised concerns regarding their side-effect on the environment. In reaction to this, a variety of sustainable agriculture practices have gained popularity, as well as organic agriculture as an alternative to the intensive inputs in conventional agriculture. These sustainable practices include adding organic matter to the soil, covering soil with cover crops or crop residues, reducing tillage intensity or practicing conservation tillage, using legumes within a crop rotation, implementing strip cropping, improving drainage, and avoiding compaction (Madgoff, 2001; Kennedy and Papendick, 1995). Best management practices were proposed to reduce the amount of synthetic chemicals used in conventional agriculture while maintaining acceptable levels of economic return (Hilliard and Reedyk,

∗ Corresponding author at: Department of Soil Science, University of Saskatchewan, Saskatoon, SK, Canada, S7N 5A8. Tel.: +1 306 966 6866. E-mail address: [email protected] (S.M. Adl). 1470-160X/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ecolind.2013.10.002

2000; Korol, 2004). Organic agriculture claims to be environmentally sustainable, socially just, and economically sound production practices but prohibits using most synthetic fertilizers, herbicides, and pesticides, as well as other restrictions (Lotter, 2003; El-Hage Scialabba and Hattam, 2002; Biao et al., 2003). Increasing soil biological activity in order to maintain long term soil fertility through decomposition of the organic matter are the first priorities of organic agricultural management practices (IFOAM, 2011; Fliessbach and Mader, 2000; Biao et al., 2003). In this study we compared conventional fields under best management practise to fields under organic agriculture. Soil quality is a key element in evaluating the sustainability of agriculture practices (Carter, 2002). By combining Brookes (1993) criteria, Doran and Parkin’s (1994) criteria, and Doran and Safley’s (1997) criteria, Stenberg (1999) produced a list of five essential criteria used in determining proper soil quality indicators. Because soil functions are difficult to measure, soil properties that are sensitive to change in a specific ecosystem are often used as indicators of soil quality (Stenberg, 1999; Acton and Padbury, 1993). A minimum data set is a group of soil quality indicators that are chosen based on a definition of soil quality and soil quality indicator

120

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

criteria (Larson and Pierce, 1991; Doran and Parkin, 1994; Harris et al., 1996). Modern studies have argued that physical, chemical, and biological indicators must be evaluated together in order to provide a correct assessment of soil quality (Stenberg, 1999; Wander and Bollero, 1999; Stenberg et al., 1998; Gomez et al., 2004; Schloter et al., 2003a). However, a comprehensive analysis of soil will not accurately describe soil quality unless the indicators are chosen with a specific soil function in mind, within a defined ecosystem (Janzen et al., 1992; Stenberg, 1999; Acton and Padbury, 1993). In this study, we focused on potato production within the eastern Canada Maritimes and indicators were selected accordingly. Potatoes (Solanum tuberosum L.) are the third most important food crop in the world and play an important role in feeding the world’s population (International Potato Center, 2011). Potato production in eastern Canada continues to increase and produces about 1,772,000 tonnes annually, and it is the main crop from this region (Agriculture Canada, 2011). However, potato nutrient demand on soil is high, and tuber quality requires both high organic matter and nitrogen availability. The intense use of synthetic chemicals, as fertilizer and pathogen control, in conventional potato production has also caused concern for the adjacent waterways and the surrounding environment (Patriquin et al., 1991). Sustained conventional potato monoculture as practised is leading to decreased output per hectare without substantially increasing chemical inputs, thus raising costs (Saini and Grant, 1980; Porter and McBurnie, 1996; Carter et al., 1998, 2003). To sustain soil fertility and production levels, more sustainable forms of potato production have been proposed (i.e. rotations and spring tillage) that would reduce production costs (Porter and McBurnie, 1996; Patriquin et al., 1991; Carter et al., 1998). In this study we tested the hypothesis that fields under each management practice (conventional-best management or organicmanagement) would not be different based on indicators of soil quality. The two management practices were chosen because they are claimed to be sustainable. Potato fields were chosen because of the significance of this food crop in the region and globally. Although other soil quality studies have been conducted on potato fields in Prince Edward Island (Canada) and Maine (U.S.A.) (Carter, 2002; Porter and McBurnie, 1996; Carter et al., 2003), this is the first study to include bioindicators and to compare fields across the region. It is also the first study to compare two types of sustainable practices in potato production in the region.

2. Materials and methods 2.1. Soil sampling Three conventional and three organic potato fields were chosen within a 150 km radius for a total of six field sites. All six sites were commercial farms, not experimental fields. Each organic field was located within a few kilometres of a conventional field. Fields were sampled in each of May, July, and September within one day of each other. For each field, soil samples were collected at three locations on a randomly selected diagonal transect across the field. Samples were taken in the middle of the potato hill, approximately 15–20 cm from the stem, to a depth of 10 cm. At each sampling location, separate samples were collected for nematode, testate amoebae, soil pH, C:N ratio, and soil moisture using a 2.5 cm soil corer, as described below. Similarly, separate samples for microarthropod, light fraction and bulk density were taken using a 5 cm soil corer, and microbial biomass was measured from a 1 kg composite soil sample. Soil samples were transported to the lab in a cooler where they were processed within 24 h of sampling.

2.2. Field sites The first organic management field (O1) is an easily drained fine sandy loam Charlottetown series soil, but has a small percentage of easily drained fine sandy loam Alberry soil as well. The farm has been under cultivation since the mid 1900s, but was converted to organic agriculture over a seven year period (1993–2000) and was certified in 2000. Parasol 50% (copper hydroxide, fungicide) was applied twice, Bluestone (copper sulphate, fungicide) was applied twice, and Entrust (spinosad, insecticide) was applied once during the 2004 growing season. Potato tops were physically cut off at the end of the season. The first conventional management field (C1) is an easily drained fine sandy loam of the Charlottetown series. The land has been under potato cultivation since 1942. The potato rows were seeded alongside a band fertilizer of NPK-Mg. Lorox (linuron, herbicide) was applied once, Manzate (mancozeb, fungicide) was applied six times, Ridol-Bravo was applied once, Bravo (chlorothalonil, fungicide) was applied three times, and mineral oil was applied five times throughout the 2004 growing season. Reglone (diquat dibromide (37.3%), desiccant/herbicide) was applied twice at the end of the season as a top kill. The second organic management field (O2) is a poorly draining clay loam Washburn series soil. The farm has been in cultivation since 1980 and was certified organic in 1987. The rotation used in this field is manure, clover, potatoes, and then mixed vegetables. Parasol was applied three times and Entrust was applied twice during the 2004 growing season. A propane flamer was used at the end of the season to clean up late blight, as a final Colorado Potato Beetle control and as a top-kill. The second conventional management field (C2) is on a poorly drained silty loam of the Interval series. The land has been under cultivation since the early 1800s. The rotation includes corn, brassica, and potatoes. Chemical applications to the field did occur but were not recorded. The third organic management field (O3) is found on light brown sandy loam of the Torbrook series with good to excessive drainage. The farm has been under organic cultivation since 1988. The rotation consists of potatoes, two years of mixed vegetables, followed by a green manure of oats and peas which are left in through the winter and harrowed under in the spring. Floating row covers are used to speed the early stages of growth, and to avoid pests and disease. The third conventional management field (C3) is found on sandy loam Truro series soil with good to fair drainage. Admire (imidacloprid, insecticide) and an N–P–K fertilizer were banded in furrow when the crop was seeded. An N fertilizer was also broadcast on the crop mid-growing season. Sencor (herbicide) was applied once, Bravo was applied three times, and Tatoo C (fungicide) and Cymbush (cypermethrin, insecticide/miticide) were applied once throughout the growing season. Reglone was applied once as a top kill. 2.3. Soil quality indicator measurements Microbial biomass C was measured using chloroform fumigation-extraction according to standard procedures (Paul et al., 1999). On each sampling occasion, a composite sample of approximately 1 kg of soil was taken from each field. The soil was sieved using a 2.83 cm diameter sieve and organic particles larger than 3 mm were removed by hand. The fumigated and unfumigated extracted filtrates were stored in 50 mL falcon tubes at −20 ◦ C until the chloroform labile C analysis was analysed with a LECO CNS auto-analyser. Testate amoebae were stained and enumerated using standard procedures (Adl et al., 2006a) from three 2.5 cm × 10 cm deep soil

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

cores taken at each sampling site of every field (3 samples per field). One soil composite was obtained from each sampling site and three subsamples of 1 g placed into each of three test tubes. This resulted in three 1 g subsamples per soil sample for a total of nine test tubes per field. The slides were dried on a slide dryer and examined under one drop of glycerine with a cover slide. Three 4 cm long transects were observed at 400× on a Zeiss compound microscope. The number of stained testate amoebae were enumerated along each transect to obtain abundances from line transects (Krebs, 1998). Nematodes were extracted and enumerated according to standard procedures (Coleman et al., 1999) from three 2.5 cm × 10 cm soil cores per sampling location (3 sampling sites per field) and combined into one composite soil core for each sampling site. Three subsamples of known weight (4–8 g) were extracted per sampling location and stored in 5% formalin. Nematodes were enumerated with a Nikon inverted microscope at 100× and functional groups were identified at 400×. Microarthropods were extracted and enumerated according to standard procedure (Coleman et al., 1999). Two 5 cm × 10 cm deep soil cores were taken per field sampling site (3 samples per field), weighed and placed into an extraction cup. The microarthropods were collected in 95% ethanol and enumerated in a Petri dish with a Nikon dissecting microscope. The samples were enumerated at 30× and identified at 80× to sub-order (mites) or to family (collembola). Soil pH, light fraction (LF), bulk density (BD) and gravimetric water content were obtained according to Robertson et al. (1999). The soil C:N ratio was measured according to Elliott et al. (1999) from six composite field samples that were air dried, ground to <0.5 mm, and weighed to the nearest 10−4 g. Three replicates (approximately 0.3 g) of each sample was placed in a LECO CNS Analyser and the output was in % C g−1 dry soil and % N g−1 dry soil. 2.4. Statistical analysis Each soil quality indicator was initially assessed individually and descriptive statistics were performed. The soil quality indicators were divided by date sampled, separated by field site and means were plotted in SigmaPlot v. 8.0. Because nematode, microarthropod and Collembola were further divided by functional group, sub-order and family, a randomized block design was used to test the effects of management practice, month sampled, their interaction, and location, on (1) nematode functional groups, and (2)

121

mite sub-orders plus Collembola families (GLM Multivariate, SPSS v. 13.0, type III sum of squares). All organic and all conventional field soil indicator means were averaged for each date and were plotted in SigmaPlot v. 8.0. Data were further analysed using a General Linear Model (GLM) randomized block design (GLM Multivariate, SPSS v. 13.0). The between-subject effect of location (block), management practice, month of sampling, and the interaction between management practice and month were tested (type III sum of squares). Estimated marginal means were determined using Fisher’s least significant difference (LSD), calculated at the p = 0.05 level. Post hoc multiple comparison tests were carried out for month and location only, using Tukey’s honestly significant difference (HSD) test. Univariate test results were expressed for significant soil quality indicators if the GLM suggested that the between-subject effects differed. Soil quality indicators did not meet the univariate assumption of normality, and could not be transformed. Pillai’s trace is robust to violations of GLM assumptions, especially violation of similar variance–covariance matrices, and was therefore used to establish significance in this study (Johnson and Field, 1993). Light fraction was not included in the GLM because it was only measured in May and September. The relationship between the biological and environmental soil quality indicators was further explored using canonical correspondence analysis (CCA) (MVSP v. 3.1). Microbial biomass, testate amoebae abundance, nematode abundance, and microarthropod abundance were compared to the environmental indicators (soil pH, C:N ratio, bulk density and soil moisture). Also, the correlation between soil quality indicators, management practice and month was analysed. Several other statistical approaches were attempted, but could not be used because the data violated assumptions of the tests. These included discriminant analysis, logistic regression, multivariate analysis of variance, and several ordination techniques other than canonical correspondence analysis.

3. Results 3.1. Bio-indicators Results of the bioindicator measurements and test of significant difference by GLM are summarized in Table 1. Microbial

Table 1 Soil quality indicator means (with standard error) in organic and conventional fields by month, and results of the general linear model. Microbial biomass (mg C g−1 )

Testate amoebae (g−1 )

Nematodes (g−1 )

Microarthropods (g−1 )

pH

C:N

Bulk density (g cm−3 )

Soil moisture (% g−1 )

Month May July September

154.93 (20.9) 266.73 (20.9) 306.26 (26.4)

65,571 (8233) 46,011 (8232) 38,787 (10,413)

19 (5) 15 (5) 10 (7)

194 (396) 343 (396) 3815 (501)

6.05 (0.2) 6.06 (0.2) 6.07 (0.2)

8.38 (0.7) 10.54 (0.7) 11.07 (0.9)

0.470 (0.03) 0.453 (0.03) 0.437 (0.04)

22.95 (1.3) 19.53 (1.3) 24.67 (1.6)

May July September

116.56 (20.9) 138.53 (20.9) 153.27 (20.9)

66,487 (8232) 43,927 (8232) 31,502 (8232)

6 (5) 11 (5) 8 (5)

410 (396) 365 (396) 1205 (396)

5.80 (0.2) 5.41 (0.2) 5.37 (0.2)

11.96 (0.7) 10.97 (0.7) 11.45 (0.7)

0.550 (0.03) 0.533 (0.03) 0.583 (0.03)

16.18 (1.3) 14.01 (1.3) 17.47 (1.3)

***

NS

NS

*

*

*

**

***

Month NS

**

**

NS

**

NS

*

NS

NS NS

NS NS

*

NS

NS NS

Location

NS

NS

NS

NS

**

**

NS

**

Indicators

Organic

Conventional

General linear model Management practice Management practice × month

Shown are means (standard error) of the interactions between management practice and month. Degrees of freedom (d.f.): management practice d.f. = 1, month d.f. = 2, management practice × month d.f. = 2, location d.f. = 2, Error d.f. = 9, Total d.f. = 17. Type III sum of squares: NS (not significant). * p ≤ 0.05. ** p ≤ 0.01. *** p ≤ 0.001.

122

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

500

350

B

A

300

Chloroform Labile C ug g-1 dry soil

400

250 300 200 200 150 100

100

0

50

C

D

Testate Amoebae g-1 dry soil

100000

70000

80000

60000

60000

50000

40000

40000

20000

30000

0

20000

F

E

Nematode Abundance g-1 dry soil

25 30 20 20

15

10 10 5

Microarthropod Abundance g-1 dry soil

0 6000

0

G

May July September

H

Organic Conventional

4000 3000

4000 2000 1000

2000

0 0

-1000

O1

C1 O2

C2

O3

C3

Field and Month

May

July

September

Month

Fig. 1. Bioindicators measured in organically managed fields (O1, O2, O3) and in conventional fields (C1, C2, C3). Pooled data of all organic (filled circle) or conventional (clear circle) fields for each month.

biomass, as implied from chloroform labile carbon, increased over the growing season in every field (Fig. 1a). Fields under organic management had the highest chloroform labile C means in the soil, throughout the growing season, compared to conventional

fields (Fig. 1b). Mean testate amoebae and nematode abundance was more variable between field sites through the growing season (Fig. 1c and d). When combining all organic and all conventional field sites, mean abundance for both decreased over the growing

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

123

Table 2 Nematode functional group abundance means (with standard error) with respect to the interaction effects of the randomized block design general linear model. Nematodes (g−1 )

Bacterivores

Fungivores

Root feeders

Omnivores

Predators

Organic Conventional

12 (3) 5 (2)

0 (0) 0 (0)

2 (1) 3 (1)

0 (0) 0 (0)

0 (0) 0 (0)

May July September

10 (3) 9 (3) 6 (3)

0(0) 0 (0) 2 (1)

1 (1) 4 (1) 2 (1)

0 (0) 0 (0) 0 (0)

0 (0) 0 (0) 0 (0)

×May ×July ×September

17 (4) 12 (4) 8 (5)

0 (0) 1 (0.2) 0 (0)

1 (2) 2 (2) 2 (2)

0 (0) 0 (0) 0 (0)

0 (0) 0 (0) 0 (0)

×May ×July ×September

4 (4) 6 (4) 4 (4)

0 (0) 0 (0) 1 (0.3)

2 (2) 5 (2) 3 (2)

0 (0) 0 (0) 0 (0)

0 (0) 0 (0) 0 (0)

7 (3) 8 (3) 11 (3)

0 (0) 0 (0) 1 (0.2)

4 (1) 1 (1) 2 (1)

0 (0) 0 (0) 0 (0)

0 (0) 0 (0) 0 (0)

Management practice

Month

Management practice × month Organic

Conventional

Location A B C

Means were estimated with Fisher’s “least significant difference” with P = 0.05, type III sum of squares.

season (Fig. 1e and f). There was no significant difference in testate amoebae mean abundances between organic and conventional fields (Fig. 1d). Mean nematode abundance in the organic fields was higher than in conventional fields in every month sampled, but this was not statistically significant. Among the nematode functional groups, bacterivore, fungivore, root feeder, predator, and omnivore. Means and standard errors were analysed by randomized block design (Table 2). Nematode functional groups did not significantly differ between organic and conventional fields, sampling date, the interaction between management and date sampled, or between locations (Tables 2 and 3). Mean microarthropod abundance was higher in September in most field sites sampled except in C3 (Fig. 1g), and this was significantly higher for organic fields (Fig. 1h, Tables 1 and 4). Mean abundance of mite suborders and Collembola families did not significantly differ between

organic and conventional fields, months sampled, interaction between management practice and month, or between locations (Table 5). 3.2. Physical and chemical indicators When organic fields were combined, soil pH was fairly stable over the growing season, but in the conventional fields the mean soil pH decreased slightly through the growing season (Fig. 2a and b, Table 1). The mean C:N ratio was consistently higher in the conventional fields than in the organic fields, but increased over the growing season in the organic fields (Fig. 2c and d). Light fraction was measured at the beginning and at the end of the growing season, in May and September. Mean light fraction weight remained significantly higher in the organic fields compared to conventional

Table 3 Effect of management practice, month sampled, their interaction, and location on nematode functional groups. Effect

Statistic

Value

F

Error d.f.

p-Value

Significance

Management practice Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

0.754 0.25 3.05 3.05

3.05 3.05 3.05 3.05

5 5 5 5

0.123 0.123 0.123 0.123

NS NS NS NS

Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

1.201 0.06 10.69 10.30

1.78 2.97 4.28 12.36

12 10 8 6

0.170 0.050 0.025 0.004

NS

Management practice × month Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

1.192 0.09 7.53 7.08

1.75 2.42 3.01 8.49

12 10 8 6

0.1777 0.089 0.066 0.011

NS NS NS

0.903 0.26 2.26 1.93

0.99 0.97 0.90 2.32

12 10 8 6

0.498 0.518 0.569 0.168

NS NS NS NS

Month *

NS **

*

Location Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

Abbreviations and symbols: d.f. (degrees of freedom), type III sum of squares were used, NS (not significant), *** p ≤ 0.001, 1, 2, 3, 4 order of the effect’s contribution to the overall model. * p ≤ 0.05. ** p ≤ 0.01.

124

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

8

6.4

A

B 6.2

Soil pH

6

6.0 5.8

4 5.6 5.4

2

5.2 0 16

5.0

D

C

12

C:N Ratio

14 11

12 10

10 8 6

9

Organic Conventional

4

8 2 7

0

E 0.004

Light Fraction g-1 dry soil

F

May July September

0.0025

0.0020

0.003 0.0015 0.002 0.0010 0.001

0.0005

0.0000

0.000

Bulk Density g cm-3

1.4

G

H 1.1

1.2 1.0

1.0

0.8 0.6

0.9

0.4 0.2

0.8

0.0

O1

C1 O2

C2

O3

C3

Field and Month

May

July

September

Month

Fig. 2. Physical and chemical indicators in organically managed fields (O1, O2, O3) and in conventional fields (C1, C2, C3). Pooled data of all organic (filled circle) or conventional (clear circle) fields for each month.

fields, with a slight but not-significant difference, over the growing season (Fig. 2e and f). Over the growing season, mean bulk density decreased from May to July to September in all fields except C2 (Fig. 2g). Conventional sites had a consistently higher mean bulk

density than the organic fields (Fig. 2h). Soil moisture at each sampling time showed the organic fields retained significantly higher mean soil moisture content than the conventional fields at the time of sampling (Table 1).

0 (0) 10 (6) 0 (0) 184 (81) 94 (81) 84 (93) 12 (20) 10 (20) 29 (23) 39 (157) 306 (157) 194 (179) A B C Location

Means were estimated with Fisher’s “least significant difference” with p = 0.05, type III sum of squares.

0 (0) 32 (21) 20 (24) 319 (107) 315 (107) 276 (121) 0 (0) 23 (9) 20 (10)

47 (12) 0 (0) 0 (0) 0 (0) 0 (0) 45 (222) ×May ×July ×September Conventional

3.4. Management effect on soil quality indicators

145 (112) 269 (112) 105 (128)

94 (158) 103 (158) 292 (158) 0 (0) 40 (29) 49 (29) 269 (151) 60 (151) 456 (151)

0 (0) 24 (29) 0 (0) 170 (151) 260 (151) 606 (191) 0 (0) 0 (0) 39 (16) 0 (0) 59 (222) 973 (281) Management practice × month ×May Organic ×July ×September

Soil quality indicators for May, July, and September were significantly different among each other for several indicators. Univariate tests revealed that microbial biomass (p = 0.005), testate amoebae abundance (p = 0.017), microarthropod abundance (p = 0.001) and soil moisture (p = 0.028) significantly differed between the sampling months (Table 1). Microbial biomass carbon differed significantly between the beginning of the growing season and the mid and end of the season; testate amoebae abundance was significantly different between May and September, while neither of these months were significantly different than July; and microarthropod abundance differed significantly between the beginning and mid growing season, and the end of the season (Table 6). When the indicators were considered as a single value for soil quality, they were found to be significantly different (F = 20.64, p < 0.001, Table 7) for each of May, July, and September. The effect of month of sampling is therefore significant when considering the results. Soil quality was also significantly different among the three paired sites (F = 8.93, p < 0.006, Table 7), indicating a site effect. Therefore, management effects on soil quality indicators must be greater than both location and month effects to be statistically useful.

44 (51) 134 (51) 86 (58)

0 (0) 79 (72) 168 (72) 0 (0) 41 (115) 142 (115) 0 (0) 21 (28) 52 (28)

0 (0) 21 (9) 0 (0)

0 (0) 0 (0) 279 (91) 0 (0) 0 (0) 539 (145) 0 (0) 0 (0) 549 (200)

24 (28) 0 (0) 6 (35)

0 (0) 10 (6) 0 (0) 0 (0) 21 (81) 341 (93) 12 (20) 10 (20) 29 (23) Month

May July September

344 (140) 15 (128)

0 (0) 30 (157) 509 (180)

13 (8) 15 (7)

23 (9) 0 (0) 20 (10)

345 (95) 262 (87)

220 (107) 160 (107) 531 (121)

0 (0) 31 (21) 25 (24)

8 (18) 29 (17)

183 (100) 163 (91)

47 (112) 52 (112) 420 (128

180 (73) 61 (66) 10 (18) 24 (16)

0 (0) 0 (0) 0 (0)

6 (51) 40 (51) 114 (58)

93 (45) 83 (41) 0 (0) 7 (5)

Entomobrydae Hypogasturidae Isotomidae Prosigmata Mesostigmata Euoribatida Oribatida

125

3.3. Season and location effects on bioindicators

Management practice Organic Conventional

Collembolan families Mite suborders

Table 4 Mite suborders and collembolan families abundance means (with standard error) with respect to the interaction effects of the randomized block design General Linear Model.

Smithuridae

Onychuridae

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

Light fraction was measured only twice, at the beginning and t the end of the growing season, it is thus treated separately from the other indicators. The results show a significant difference between the three paired locations (F = 7.83, p < 0.029, Table 8). All other comparisons were not significant. This result, and the previous results indicating a location and month effect on the indicators, suggest that both local climate and soil have a strong effect on the soil ecology and thus on the indicators. When the other soil quality indicators were considered as a single value, and the soil quality was evaluated over the growing season, organic and conventional potato fields were found to be significantly different (F = 21.01, p < 0.05, Table 7). In particular, the means of six of the eight soil quality indicators included in the GLM significantly differed between the organic and conventional fields sampled (Table 1). Microbial biomass (p < 0.000), microarthropod abundances (p = 0.045), soil pH (p = 0.016), and soil moisture (p < 0.000) were significantly higher in all organic fields (Tables 1 and 6 and Figs. 1 and 2). Soil C:N ratio (p = 0.043) and bulk density (p = 0.017) were significantly lower in the organic fields when compared to the conventional fields (Tables 1 and 6 and Fig. 2). The soil quality was also significantly different when considering both the organic and conventional fields at the time of sampling. Not only did the soil quality of the conventional fields differ from the soil quality of the organic fields over the entire season but the soil quality differed significantly between months (F = 6.44, p < 0.015, Table 7). When considering the indicators that describe soil quality, microarthropod abundance was the only variable that differed significantly among each field and each sampling month (p = 0.017, Table 1). Canonical correspondence analysis (CCA) was used to explore the relationship between soil quality indicators, management practice and months. Soil quality bioindicators were plotted against the soil quality environmental indicators, months, and organic or conventional management, using a CCA Biplot (Tables 9 and 10, Figs. 3 and 4). The biological soil quality indicators are represented by “species points”, or weighted averages, along an environmental gradient, with arrows for the environmental soil quality indicator; the longer the arrows, the more strongly correlated they are to the ordination axes and the closer to the centre of the bi-plot,

126

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

Table 5 Effect of management practice, month sampled, their interaction, and location on mite and collembolan abundances. Effect

Value

Management practice Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root Month Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root Management practice × month Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root Location Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

F

Error d.f.a

p-Valueb 0.327 0.327 0.327

0.983 0.02 47.42 47.42

5.27 5.27 5.27 5.27

1 1 1 1

1.632 0.00 154.22 152.45

0.99 2.18 0.00 33.88

4 2 0 2

1.711 0.00 70.67 68.02

1.32 1.65 0.00 15.11

4 2 0 2

0.433 0.443

0.284 0.21 0.00 1.10

18.00 18.00 18.00 9.00

4 2 0 2

0.975 0.979

NS NS NS NS

0.572 0.361

NS NS NS NS

0.029

NS NS NS

0.064 NS NS NS NS

0.562

1,2,3,4

order of effect’s contribution to the overall model; type III sum of squares. NS, not-significant; a Degrees of freedom. b Significance. Table 6 Means and standard errors of the soil quality indicators with respect to management practice, month, and location, using the randomized block generalized linear model. Indicators

Microbial biomass (mg g−1 )

Management practice 242.64 (13.2) Organic 136.12 (14.7) Conventional

Testate amoebae (# g−1 )

Nematodes (# g−1 )

Micro-arthropods (# g−1 )

Soil pH

C:N ratio

Bulk density (g cm−3 )

Soil moisture (%)

6.06 (0.1) 5.53 (0.1)

9.99 (0.5) 11.46 (0.4)

0.905 (0.02) 1.107 (0.01)

22.38 (0.8) 15.89 (0.7)

50,100 (5200) 47,305 (4700)

15 (3) 9 (3)

1450 (250) 660 (230)

Month May July September

135.74 (14.7)a 202.63 (14.7)b 229.77 (16.8) b

66,000 (5800) a 45,000 (5800)ab 35,100 (6600) b

12 (4) a 13 (4) a 9 (4) a

300 (280) a 354 (280) a 2510 (320) b

5.93 (0.1) a 5.74 (0.1) a 5.72 (0.1) a

10.17 (0.5) a 10.76 (0.5) a 11.26 (0.6) a

1.096 (0.02) a 1.066 (0.01) a 1.160 (0.02) a

19.57 (0.9) a 16.77 (0.9) b 21.07 (1.0) ab

Location A B C

171.40 (14.7) a 202.00 (14.7) a 194.74 (16.8) a

48,900 (5800) a 44,200 (5800) a 53,000 (6600) a

11 (4) a 10 (4) a 14 (4) a

1020 (280) a 1200 (280) a 950 (320) a

5.16 (0.1) a 6.40 (0.1) b 5.82 (0.1) a

10.19 (0.5) a 12.83 (0.5) b 9.16 (0.6) a

0.984 (0.03) a 1.022 (0.03) a 1.042 (0.05) a

15.70 (0.9) a 21.52 (0.9) b 20.19 (1.0) ab

Means estimated using Fisher’s least significant difference with p = 0.05. Values within the same column labelled with the same letter are not significantly different from each other, according to Tukey’s honestly significant difference calculated at p = 0.05.

Fig. 3. Canonical correspondence analysis biplot of soil quality indicators, management practices and month sampled.

Fig. 4. Canonical correspondence analysis biplot of management practices and month sampled.

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

127

Table 7 Effect of management practice, month sampled, their interaction, and location on soil quality, with respect to the interaction effects of the randomized block design general linear model. Effect

Statistic

Value

F

Error d.f.

p-Value

Significance

Management practice Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

0.994 0.01 84.04 84.04

21.01 21.01 21.01 21.01

2 2 2 2

0.046 0.046 0.046 0.046

*

Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

1.961 0.00 135.66 97.49

20.64 15.28 8.48 36.56

6 4 2 3

0.01 0.009 0.110 0.007

**

Management practice × month Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

1.893 0.00 52.78 42.29

6.44 5.33 3.30 15.86

6 4 2 3

0.015 0.059 0.257 0.022

*

1.922 0.00 55.86 38.85

8.93 6.45 3.49 14.57

6 4 2 3

0.006 0.042 0.245 0.025

**

* * *

Month **

NS **

NS NS *

Location Pillai’s trace Wilk’s lambda Hotelling’s trace Roy’s largest root

Abbreviations and symbols: d.f. (degrees of freedom), type III sum of squares were used, NS (not significant), * p ≤ 0.05. ** p ≤ 0.01.

1,2,3,4

*

NS *

order of the effect’s contribution to the overall model.

Table 8 Effect of management practice, month sampled, their interaction, and location on the soil quality indicator light fraction. Light fraction (g LF g−1 dry soil)

F

d.f.

p-Value

Significance

Organic Conventional

0.00182 (0.0003) 0.00080 (0.0003)

5.77

1

0.061

NS

May September

0.00137 (0.0003) 0.00124 (0.0003)

0.09

1

0.776

NS

0.00194 (0.0004) 0.00169 (0.0005) 0.00080 (0.0004) 0.00079 (0.0004)

0.08

1

0.787

NS

0.00246 (0.0003) 0.00087 (0.0003) 0.0060 (0.004)

7.83

2

0.029

*

Effect Management practice

Month

Management practice × month Organic ×May ×September Conventional ×May ×September Location A B C

Abbreviations and symbols: d.f. (degrees of freedom), type III sum of squares were used, NS (not significant). * p ≤ 0.05.

the smaller the deviation from the grand mean of all the environmental variables (Jongman et al., 1995). When the organic and conventional fields, and May, July and September scores were separated, the organic, conventional, and July scores were shown to be multicolinear (Table 9 and Fig. 3). When organic and conventional variables were combined into a single variable called ‘management practice’, and May, July, and September were combined into a single variable called ‘month’ in CCA biplot B, all multicolinearity was

eliminated (Table 9 and Fig. 4). The eigenvalues, percentages, cumulative percentages and species–environment correlations of both biplots are shown in Table 9. Both CCA biplots illustrate the same correlations and relationships between variables. The occurrence of testate amoebae close to the centre of the plot indicates they were not sufficiently affected by management practise or by any of the other variables to be useful indicators in this study. The microarthropods indicate they are strongly responsive

Table 9 Canonical correspondence analysis of soil quality indicators, filed site management practices and month sampled. CCA Biplot, Fig. 3

Eigenvalues Percentage Cumulative percentage Indicator-environment correlations Multicollinearity detected

CCA Biplot, Fig. 4

Axis one

Axis two

Axis one

Axis two

0.044 52.715 52.715 0.732 Organic, conventional, July

0 0.398 53.113 0.505

0.035 41.833 41.833 0.654

0 0.212 42.045 0.329

128

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

Table 10 Intraset correlations between environmental variables and constrained site scores of Figs. 3 and 4. CCA Fig. 3

CCA Fig. 4

Scores

Envi. axis 1

Envi. axis 2

Scores

Organic Conventional May July September % moisture Bulk density pH C:N ratio

0.076 −0.076 −0.436 −0.226 0.792 0.229 −0.581 0.195 0.379

0.379 −0.379 −0.725 0.825 −0.057 0.193 −0.159 0.194 −0.263

Mngt pract

0.011

0.242

Month

0.887

−0.231

0.155 −0.583 −0.134 0.212

−0.569 −0.320 −0.502 −0.305

to the local climate, season and month of sampling, as well as soil pH, moisture and C:N ratio, but negatively affected by higher bulk density. The seasonal effect dominates their abundance dynamics and they are therefore not useful to distinguish between the two management practices in this study. The microbial biomass estimated from chloroform labile C also responded in the same way as the microarthropods, but to a lesser extent, and higher biomass correlated more strongly with organic management. The nematode abundance was strongly affected by management practise and higher abundance correlated more with organic management. The light fraction weight was higher in organic plots (Table 6 and Fig. 2). The environmental indicators also responded to month of sampling, showing higher moisture, pH and C:N ratio towards September, but lower bulk density in September in part caused by potato tuber mounding. Overall the organic fields correlated with higher pH, moisture, nematode abundance, micro-arthropod abundance, microbial biomass, and lower bulk density and C:N ratio. The conventional fields under best management practise correlated with higher bulk density, higher C:N ratio, lower pH, and lower abundance of bioindicators.

4. Discussion 4.1. Soil management effect on soil quality indicators Soil quality indicators were meant to be practical and fast measures to determine health or fertility levels of soils. When chosen well, indicators can be used to monitor field sites and to make management decisions. One needs to exercise caution as there are studies that exerted huge effort only to conclude the wrong bioindicator was selected (Geissen and Kampichler, 2004). The indicators selected for this study were chosen because they had been proposed as useful indicators in the literature. However, as this study shows, not all indicators are useful for the same purpose. Some indicators were better used for certain comparisons, but not others. For example, some indicators may discriminate between large differences between sites (such as a forest and an agriculture field), but not smaller differences between similar fields under agriculture. In addition, the level of taxonomic resolution used in bio-indicators may be more significant than abundance measures typically used. Furthermore, our results show that, predictably, indicators fluctuate with month of sampling through the growing season and with site location. Therefore, for soil quality indicators a one-time measurement does not suffice, reducing the practicality of such measures. Using an aggregate of indicators, with sufficient sampling through the growing season, as well as multivariate statistical analysis of the data provide better indicators of field management effects on soil. Our results show that our minimum data set of indicators, when used together, could differentiate between organic and conventional best-practice potato fields.

% moisture Bulk density pH C:N ratio

Envi. axis 1

Envi. axis 2

When used on their own, the indicators provided mixed results, some showing no significant difference while others detecting differences (Table 1): individually, six of the nine soil quality indicators included in the minimum data set differed between organic and conventional fields. Some indicators responded more to site location or month than field management. For nematode functional groups, mite suborders, or collembolan family, the abundances were not significantly different, using a variety of statistical measures (Tables 2–5). However, we note not all statistical techniques were consistent, so that comparing several is useful. Better discrimination between organic and conventional best-practice fields was obtained when the fields were averaged by management (Tables 6 and 7). The indicators detected significant overall differences between organic and best-practice management, and were less affected by month of sampling or site location. The multivariate CCA results provided a better visualization of correspondences among the indicators (Table 10 and Figs. 3 and 4). 4.2. Interactions among indicators and with management Overall, the general linear model and the canonical correspondence analysis showed fields under organic management had higher pH, soil moisture, litter light fraction, and lower C:N ratio and bulk density than the conventional fields. Given the emphasis on the role of the soil food web in decomposition and mineralization in organic agriculture, it was expected that organic management would promote soil conditions that were favourable for biological activity (Neher, 1999a,b; Fliessbach and Mader, 2000). However, conventional fields under best management practice were not significantly different from the organic fields for testate amoebae and nematode abundances. The organic fields supported higher microarthropod abundance but the significant microarthropod effect is due to the elevated abundance in the organic fields in September. Only microbial biomass was significantly higher in the organic fields, and this was probably affected by the lower bulk density and lower C:N ratio in these fields. The organic field mean C:N ratio ( = 9.99) and the conventional field mean C:N ratio ( = 11.46) were below 20:1 (Table 6), which may indicate the predominance of the bacterial energy channel (Ferris and Matute, 2003). Dendooven et al. (2000) also reported significant difference between their low conventional and organic soil C:N values (CON = 8 and ORG = 5). Their study suggests that the low C:N ratio could be attributed to differences in C availability, to differences in microbial biomass C:N ratios, to differences in N dynamics, or to differences that cannot be reflected in the measurements of organic and conventional practices (Dendooven et al., 2000). Lower bulk density suggests that the soil in the organic fields was favourable to biological activity (Harris et al., 1996). When used as a physical soil quality indicator in comparisons of organic and conventional soil bulk density, results in other studies have differed. In a field experiment by Bulluck et al. (2002) organic and synthetic fertility

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

amendments were compared. The organic fertilizer, comprised of cotton-gin trash, composted yard waste and cattle manure showed a decrease in bulk density, while the synthetic fertilizer did not (Bulluck et al., 2002). In a study of three groups of organic and conventional fields, organic field bulk density was shown to be significantly higher in the first group, not significantly different in the second group, and significantly lower in the third group compared to the conventional counterpart (Schjønning et al., 2002). Schjønning et al. (2002) suggest that their results indicate that soil compaction due to extensive traffic may reach the same levels in organic fields as is currently experienced in conventional fields. Higher microbial biomass and microarthropod abundance in organic fields may indicate a difference in the quality (C:N ratio) or availability of plant litter and compost amendments, since both are strongly related to resource quality and availability (Wardle and Lavelle, 1997; Wardle et al., 1999). The lower soil C:N ratio (higher quality) in organic fields would support an increased microbial biomass, however microarthropods generally thrive on lower quality organic matter (higher C:N ratio) (Wardle and Lavelle, 1997; Georgieva et al., 2005). Microarthropod abundance is known to be heavily influenced by resource quality and availability, but has also varied with management practice (Behan-Pelletier, 1999, 2003; Mebes and Filser, 1998; Wardle et al., 1999). In other comparisons of organic and conventional fields, as in this study, both testate amoebae and nematode abundances have been shown not to differ significantly (Foissner, 1997; Neher, 1999b). Testate amoebae abundance has been cited as a poor indicator of agriculture management practices differences, while nematode abundances were expected to differentiate between differing management practices, months and provinces (Georgieva et al., 2005). Our study agrees with the literature, that the majority of the nematodes found in organic soils are bacteriovores, and the remaining population is made up of root feeders (Neher, 1999b). Bacteriovores (both nematodes and protozoa) have been suggested as better indicators of bacterial activity, substrate quality, and nutrient release in the soil than direct measurement of bacterial populations (Georgieva et al., 2005). If decomposition and nutrient mineralization had been more efficient in the organic systems, indicated by higher microbial biomass and microarthropod abundance, the abundance of bacterivorous nematodes and testate amoebae should have been significantly higher as well. Neher (1999b) suggests that the nematode communities in organic and conventional fields are too similar, supporting the view that even with the lack of inorganic inputs, the organic soil food web composition is not as different from a conventional soil food web as was once thought. Although testate amoebae, nematode, and microarthropod abundance are standard bioindicators that have traditionally been enumerated (Behan-Pelletier, 1999; Foissner, 1999; Neher, 2001), species richness and diversity may be more useful indicators of soil quality and soil health (Naeem, 2002; Adl et al., 2006b). However, it is most likely that the fields under best-management practice are not all that different biologically from fields near-by under organic management. After discussions with the growers, it became evident that the organic fields were managed with more frequent tillage than the fields under best-management practice. We therefore sought a notillage potato field under organic management in the area. We compared our data to the same quality indicators in the no-till organic potato plot for the month of September. The results suggest that without tillage, the plots under organic management have both more functional diversity and higher abundance of indicator organisms, and are significantly different from the best-management practice fields (Nesbitt, unpublished). This is consistent with previous observations on the effect of tillage on soil diversity (Adl et al., 2006b).

129

5. Conclusion Using a combination of indicators that included physical, chemical and biological parameters provided a better evaluation of the field sites, than if a single or fewer indicators are selected. Indicators varied through the growing season so that multiple sampling times are necessary to infer any conclusions on field management. Bioindicators under best-management-practice were not that different from the organic management fields. We suggest the increased tillage frequency in the organic fields is responsible for preventing a recovery of the biodiversity and organism abundance. The differences in physical parameters between both management practices are most likely due to the increased organic matter in the fields under organic practice. This increased abundance of some bioindicators but not to the point of being consistently significant. Acknowledgements This research was supported by a NSERC grant to S.M.A. We thank the Dalhousie University Faculty of Graduate Studies, the School for Resource and Environmental Studies, and the Soil Ecology Society for their financial support and travel awards to J.E.N. We thank the potato growers that co-operated with this study, provided access to fields and information about field management: Raymond Loo, Lori Robinson, Karen Davidge, Gordon Harvey, Norbert Kungl, Kris Pruski. References Acton, D.F., Padbury, G.A., 1993. A conceptual framework for soil quality assessment and monitoring. In: Acton, D.F. (Ed.), A Program to Assess and Monitor Soil Quality in Canada: Soil Quality Evaluation Program Summary (Interim). Centre for Land and Biological Resources Research Contribution No. 93-49. Research Branch, Agriculture Canada, Ottawa, ON, pp. 2–7. Adl, M.S., Acosta-Mercado, D., Lynn, D.H., 2006a. Protozoa. In: Carter, M.R., Gregorich, E.G. (Eds.), Soil Sampling and Methods of Analysis. , 2nd ed. Canadian Society of Soil Science, Boca Raton, pp. 455–470. Adl, M.S., Coleman, D.C., Read, F., 2006b. Slow recovery of soil biodiversity in sandy loam soils of Georgia after 25 years of no-tillage management. Agriculture Ecosystem and Environment 114, 323–334. Agriculture Canada, 2011: website http://www.agr.gc.ca/eng/industry-marketsand-trade/statistics-and-market-information/by-product-sector/horticulture/ horticulture-canadian-industry/sector-reports/statistical-overview-ofcanadian-horticulture-2009-2010-1-of-8/statistical-overview-of-canadianhorticulture-2009-2010-4-of-8/?id=1319484750457 (accessed December 2011). Behan-Pelletier, V.M., 1999. Oribatid mite biodiversity in agroecosystems: role as bioindicators. Agriculture, Ecosystems and Environment 74, 411–423. Behan-Pelletier, V.M., 2003. Acari and collembola biodiversity in Canadian agricultural soils. Canadian Journal of Soil Science 83, 279–288. Biao, X., Xiaorong, W., Zhuhong, D., Yaping, Y., 2003. Critical impact assessment of organic agriculture. Journal of Agriculture and Environmental Ethics 16, 297–311. Brookes, P.C., 1993. The potential of microbiological properties as indicators in soil pollution monitoring. In: Schulin, R., Desaules, A., Webster, R., von Steiger, B. (Eds.), Soil Monitoring: Early Detection and Surveying of Soil Contamination and Degradation. Birkhauser Basel, Basel, CH, pp. 229–254. Bulluck, L.R., Brosius, M., Evanylo, G.K., Ristaino, J.B., 2002. Organic and synthetic fertility amendments influence soil microbial, physical and chemical properties on organic and conventional farms. Applied Soil Ecology 19, 147–160. Carter, M.R., 2002. Soil quality for sustainable land management: organic matter and aggregation interactions that maintain soil functions. Agronomy Journal 94, 38–47. Carter, M.R., Kunelius, H.T., Sanderson, J.B., Kimpinski, J., Platt, H.W., Bolinder, M.A., 2003. Productivity parameters and soil health dynamics under long-term 2-year potato rotations in Atlantic Canada. Soil and Tillage Research 72, 153–168. Carter, M.R., Sanderson, J.B., MacLeod, J.A., 1998. Influence of time of tillage on soil physical attributes in potato rotations in Prince Edward Island. Soil and Tillage Research 49, 127–137. Coleman, D.C., Blair, J.M., Elliott, T.E., Wall, D., 1999. Soil invertebrates. In: Robertson, G.P., Coleman, D.C., Bledsoe, C.S., Sollins, P. (Eds.), Standard Soil Methods for Long-Term Ecological Research. Oxford University Press, Oxford, UK, pp. 349–377. Dendooven, L., Murphy, E., Powlson, D.S., 2000. Failure to simulate C and N mineralization in soil using biomass C-to-N ratios as measured by the fumigation extraction method? Soil Biology and Biochemistry 32, 659–668.

130

J.E. Nesbitt, S.M. Adl / Ecological Indicators 37 (2014) 119–130

Doran, J.W., Parkin, T.B., 1994. Defining and assessing soil quality. In: Doran, J.W., Coleman, D.C., Bezdiceck, D.F., Stewart, B.A. (Eds.), Defining Soil Quality for a Sustainable Environment, No. 35. SSSA Special Publication, Madison, WI, pp. 3–21. Doran, J.W., Safley, M., 1997. Defining and assessing soil health and sustainable productivity. In: Pankhurst, C.E., Doube, B.M., Gupta, V.V.S.R. (Eds.), Biological Indicators of Soil Health. CAB International, New York, NY, pp. 1–28. El-Hage Scialabba, N., Hattam, C. (Eds.), 2002. Organic agriculture, environment and food security. Environment and Natural Resources Service Sustainable Development Department, No. 4. Food and Agriculture Organization of the United Nations, Rome. Elliott, E.T., Heil, J.W., Kelly, E.F., Monger, H.C., 1999. Soil structural and other physical properties. In: Robertson, G.P., Coleman, D.C., Bledsoe, C.S., Sollins, P. (Eds.), Standard Soil Methods for Long-Term Ecological Research. Oxford University Press, Inc., Oxford, UK, p. 92. Ferris, H., Matute, M.M., 2003. Structural and functional succession of the nematode fauna of a soil food web. Applied Soil Ecology 23, 93–110. Fliessbach, A., Mader, P., 2000. Microbial biomass and size-density fractions differ between soils of organic and conventional agricultural systems. Soil Biology and Biochemistry 32, 757–768. Foissner, W., 1997. Protozoa as bioindicators in agroecosystems, with emphasis on farming practices, biocides, and biodiversity. Agriculture, Ecosystems and Environment 62, 93–103. Foissner, W., 1999. Soil protozoa as bioindicators: pros and cons, methods, diversity, representative examples. Agriculture, Ecosystems and Environment 74, 95–112. Geissen, V., Kampichler, C., 2004. Limits to the bioindication of Collembola in environmental impact analysis: a case study of forest soil-liming and fertilization. Biology and Fertility of Soils 39, 383–390. Georgieva, S., Christensen, S., Stevnbak, K., 2005. Nematode succession and microfauna–microorganism interactions during root residue decomposition. Soil Biology and Biochemistry 37, 1763–1774. Gomez, E., Garland, J., Conti, M., 2004. Reproducibility in the response of soil bacterial community-level physiology profiles from a land use intensification gradient. Applied Soil Ecology 26, 21–30. Harris, R.F., Karlen, D.L., Mulla, D.J., 1996. A conceptual framework for assessment and management of soil quality and health. In: Doran, J.W., Jones, A.J. (Eds.), Methods for Assessing Soil Quality, No. 49. SSSA Special Publication, Madison, WI, pp. 61–82. Hilliard, C., Reedyk, S., 2000. Agricultural Best Management Practices. Water Quality Matters Publication. Prairie Farm Rehabilitation Administration, Agriculture and Agri-Food Canada, Available from: http://www.agr.gc.ca/ pfra/water/facts/agribtme.pdf (accessed 16.03.06). IFOAM, 2011. International Federation of Organic Agriculture Movements. The IFOAM Norms for Organic Production Processing, Available from: http://www.ifoam.org/about ifoam/standards/norms.html (accessed 24.12.11). International Potato Center, 2011. About Potato. Centro Internacional de la Papa (CIP), Lima, Peru, Available from: http://www.cipotato.org/potato/potato.htm (accessed 23.06.11). Janzen, H.H., Larney, F.J., Olson, B.M., 1992. Soil quality factors of problem soils in Alberta. In: Proceedings of the Alberta soil Science Society Workshop, Lethbridge, AB. Johnson, C.R., Field, C.A., 1993. Using fixed-effects model multivariate analysis of variance in marine biology and ecology. Oceanography and Marine Biology Annual Review 31, 177–221, Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for Biologists. Cambridge University Press, Inc., Cambridge, UK. Jongman, R.H.G., Ter Braak, C.J.F., Van Tongeren, O.F.R., 1995. Data Analysis in Community and Landscape Ecology. Cambridge University Press, Inc, Cambridge, UK. Kennedy, A.C., Papendick, R.I., 1995. Microbial characteristics of soil quality. Journal of Soil and Water Conservation 50, 243.

Korol, M., 2004. Fertilizer and pesticide management in Canada, vol. 1, no. 3. Farm Environmental Management in Canada: Agriculture and Agri-Food Canada. Krebs, C.J., 1998. Ecological Methodology, 2nd ed. Addison Wesley Longman publications, Menlo Park, CA, pp. 620. Larson, W.E., Pierce, F.J., 1991. Conservation and enhancement of soil quality. In: Elliott, C.R. (Ed.), Evaluation for Sustainable Land Management in the Developing World, vol. 12(2). IB-SPRAM, Bangkok, pp. 175–203. Lotter, D.W., 2003. Organic agriculture. Journal of Sustainable Agriculture 21, 59–128. Madgoff, F., 2001. Concepts, components, and strategies of soil health in agroecosystems. Journal of Nematology 33, 169–179. Mebes, K.-H., Filser, J., 1998. Does the species composition of Collembola affect nitrogen turnover? Applied Soil Ecology 9, 241–247. Naeem, S., 2002. Ecosystem consequences of biodiversity loss: the evolution of a paradigm. Ecology 83, 1537–1552. Neher, D.A., 1999a. Soil community composition and ecosystem processes. Agroforestry Systems 45, 159–185. Neher, D.A., 1999b. Nematode communities in organically and conventionally managed agricultural soils. Journal of Nematology 31, 142–154. Neher, D.A., 2001. Role of nematodes in soil health and their use as indicators. Journal of Nematology 33, 161–168. Patriquin, D., Edwards, L., Gimby, M., Jobin, P., Samson, R., 1991. Seven case studies in sustainable agriculture. Final Report to the Science Council of Canada. Paul, E.A., Harris, D., Klug, M.J., Ruess, R.W., 1999. The determination of microbial biomass. In: Robertson, G.P., Coleman, D.C., Bledsoe, C.S., Sollins, P. (Eds.), Standard Soil Methods for Long-Term Ecological Research. Oxford University Press, Inc., Oxford, UK, pp. 291–319. Porter, G.A., McBurnie, J.C., 1996. Crop and soil research, Chapter II. In: Alford, A.R., Drummond, F.A., Gallandt, E.R., Groden, E., Lambert, D.A., Liebman, M., Marra, M.C., McBurnie, J.C., Porter, G.A., Salas, B. (Eds.), The Ecology, Economics, and Management of Potato Cropping Systems: A Report of the First Four Years of the Maine Potato Ecosystem Project, Bulletin 843. University of Maine: Maine Agricultural and Forest Experimental Station, pp. 8–62. Robertson, G.P., Sollins, P., Ellis, B.G., Lajtha, K., 1999. Exchangeable ions, pH, and cation exchange capacity. In: Robertson, G.P., Coleman, D.C., Bledsoe, C.S., Sollins, P. (Eds.), Standard Soil Methods for Long-Term Ecological Research. Oxford University Press Inc., Oxford, UK, pp. 106–114. Saini, G.R., Grant, W.J., 1980. Long-term effects of intensive cultivation on soil quality in the potato-growing areas of New Brunswick (Canada) and Maine (U.S.A.). Canadian Journal of Soil Science 60, 421–428. Schjønning, P., Emholt, S., Munkholm, L.J., Debosz, K., 2002. Soil quality aspects of humid sandy loams as influenced by organic and conventional long-term management. Agriculture, Ecosystems and Environment 88, 195–214. Schloter, M., Dilly, O., Munch, J.C., 2003a. Indicators for evaluating soil quality. Agriculture, Ecosystems and Environment 98, 255–262. Stenberg, B., 1999. Monitoring soil quality of arable land: microbiological indicators. ACTA Agriculturae Scandinavica, Section B, Soil and Plant Science 49, 1–24. Stenberg, B., Pell, M., Torstensson, L., 1998. Integrated evaluation of variation in biological, chemical and physical soil properties. Ambio 27, 9–15. Wander, M.M., Bollero, G.A., 1999. Soil quality assessment of tillage impacts in Illinois. Soil Science Society of America Journal 63, 961–971. Wardle, D.A., Lavelle, P., 1997. Linkages between soil biota plant litter quality and decomposition. In: Cadisch, G., Giller, K.E. (Eds.), Driven by Nature: Plant Litter Quality and Decomposition. CAB International, New York, NY, pp. 107–124. Wardle, D.A., Nicholson, K.S., Bonner, K.I., Yeates, G.W., 1999. Effects of agricultural intensification on soil-associated arthropod population dynamics, community structure, diversity and temporal variability over a seven-year period. Soil Biology and Biochemistry 31, 1691–1706.