Biodiversity management of organic orchard enhances both ecological and economic profitability

Study site

The experiment was conducted in Hongyi Organic Farm, Pingyi County, Shandong Province, China (35°26′21″N, 117°50′11″E). The climate of the study area is characterized as typical temperate and monsoonal. The mean annual precipitation is 770.2 mm, with the rainfall concentrated in July (259 mm) and August (192 mm), and least precipitation in January (11 mm). The average annual temperature is 13.2 °C, with the maximum being in July (31 °C), and the minimum in January (–5 °C). The type of the soil is brown earth. The chemical and physical properties of 0–20 cm soil of OM and CM as well as cattle manure applied for the experiment are listed in Table 1.

Experimental design

Two adjoining apple orchards that located next to each other (separated by a 5 m production road) in the same field with different managements, organic management (OM) and conventional management (CM), were applied for the experiment. Each management had 3,000 m² and was divided into three plots with an area of 1,000 m². Soil type and climate were the same in OM and CM, with soil properties being quite similar. The cultivar of apples was Fuji and originated from Japan. All the trees were planted in 1998, at 3 m × 3 m distances. Soil and fruit sampling as well as earthworm, weed and pest investigations were done in the three plots as three replicates.

Details of managements for each year are shown in Table 2. The CM was treated with synthetic fertilizers, which including urea (N = 46%) and compound fertilizer of potassium sulfate (N = 15%, P₂O₅ = 15%, K₂O = 15%). The OM was fertilized with cattle manure (nutrient content is shown in Table 1). The total amount of fertilizer input in the two systems was set according to the average application rate of fertilizers in the main apple production area of Shandong Peninsula; the organic manure is slow-release fertilizer and for the purpose of compensating the nutrient loss caused by the growing grass were also taken into account. The total amount of nitrogen inputs in the two systems were controlled equivalent at the dose of 1,494 kg ha⁻¹. Non-selective pesticides, herbicides and fungicides were used in the CM just as most local orchards did and these agrochemicals contained different toxic agents such as beta-cypermethrin, parathion (Nie et al., 2005). The dose rate of the pesticides and herbicides was in accord with the instructions of the chemical products. Instead, in OM, no synthetic chemicals were applied and the EPM method was adopted to control pest; the bordeaux mixture was applied to prevent disease (Vega, Escobar & Velázquez-Martí, 2013). One native plant species (Duchesnea indica) was propagated in the spring of 2012 for suppressing harmful weeds growth, through planting branches and mowing other weeds simultaneously. Domesticated bees (Osmia excavata) were kept for pollination in OM. This experiment continued for three years (2012–2014).

Soil sampling and analysis

Five subsamples were randomly taken in each plot at 0–20 cm layer using a soil auger of 5 cm in diameter and mixed together into one soil sample. Hence, each management had three replications. The soil samples were divided into two parts. One part was air-dried and passed through a 100 mesh sieve for analysis of soil physicochemical properties, and the other part was stored at –20 °C for microbiological analysis.

Soil total nitrogen was determined by the Kjeldahl method (Bremner & Mulvaney, 1982). Soil organic carbon was measured by potassium dichromate oxidation-ferrous sulphate titrimetry method (Nelson & Sommers, 1996). Soil microbial biomass C and microbial biomass N were analyzed following the procedure of chloroform fumigation extraction (Brookes et al., 1985; Wu et al., 1990).

Amplification of 16s rDNA and sequencing analysis of bacterial communities

Soil samples collected on Jun 21st, Dec 22nd, 2013 and Jun 10th, Dec 8th, 2014 were applied for molecular analysis. The sample codes with “S” stood for summer and “W” for winter. DNA was extracted from 1 g soil samples using CTAB (Hexadecyltrimethy Ammonium Bromide) method (Konstantinos et al., 2011). Bacterial 16S rDNA at V4 hypervariable region was amplified using specific primer set 515F (5′-GTGCCAGCMGCCGCGGTAA-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) (Bates et al., 2010). Each primer contained 8–13 bp paired-end error-correcting barcodes (Table S1) (Fadrosh et al., 2014). All PCR reactions were carried out with Phusion^® High-Fidelity PCR Master Mix (New England Biolabs) and PCR products were purified with Qiagen Gel Extraction Kit (Qiagen, Hilden, Germany).

Sequencing libraries were generated using TruSeq^® DNA PCR-Free Sample Preparation Kit (Illumina, San Diego, CA, USA) following manufacturer’s recommendations and index codes were added. The library was sequenced on an Illumina MiSeq platform and 300 bp paired-end reads were generated and then merged using FLASH (Magoc & Salzberg, 2011). Quality filtering on the raw tags were performed to obtain the high-quality clean tags according to the QIIME (Caporaso et al., 2010) and UCHIME algorithm (Edgar et al., 2011) was used to remove the chimera sequences in order to obtain the effective tags. Sequences with ≥97% similarity were assigned to the same OTUs using UPARSE software (Edgar, 2013). Representative sequence for each OTU was screened for further taxonomic annotation with the GreenGene Database (DeSantis et al., 2006) based on RDP classifier (Wang et al., 2007). The estimated species richness was demonstrated with rarefaction curves, Chao 1, Shannon’s index and relative abundance.

Earthworm abundance evaluation

Earthworms in the soil were qualified through hand-sorting method trimonthly in each experimental year. Three soil blocks were investigated in each plot at one sampling time. The size of the soil blocks was 30 cm (length) × 30 cm (width) × 20 cm (depth). The earthworms collected were brought to the laboratory to be identified and counted.

Weed biodiversity assessment

Weed communities were assessed in mid-July of each year. Three quadrats of 1 m × 1 m size within inter-row area were randomly identified in each plot of OM and CM, assessed the number and coverage of each species. Biodiversity indices including Simpson index (D), Shannon–Wiener diversity index (H′), and Pielou index (Hill, 1973; Pielou, 1969) were calculated by the following Eqs. (1)–(3): (1)${\displaystyle D=1-\sum _{i=1}^s{\left(N_i∕N\right)}^2}$ (2)${\displaystyle H^{\prime }=-\sum _{i=1}^{s}PilnPi}$ (3)${\displaystyle E=H^{\prime }∕lnS.}$

Monitoring of phototactic pest dynamics

Three frequency trembler lamps with short wave (330 nm) were installed equidistantly in the three plots of OM to capture the phototactic pests. The light-traps had photosensitive automatic switch which turned the light on at night and off in the daytime. We collected the captured pests every morning from May to October in 2012 to 2014. The captured pests were categorized as scarab beetles, moths, and other small insects and then weighed separately.

Apple yields assessment

Measurements of yields were conducted at harvest time in each year. Ten trees were randomly selected in each plot of OM and CM, weighed all of the fruits and we calculated the apple yield per plant. The yields of both OM and CM were obtained by average yield per plant multiplying planting density.

Pesticide residue determination of fruits

Although none chemicals were used in the OM, organic apples were also tested by the same multi-residue analysis as CM. A total of 191 kinds of pesticides and herbicides categories were assayed by gas chromatography or liquid chromatography according to different chemical components (Table S2). The method and examination standard for testing pesticide residue were based on China national food safety standard-Maximum residue limits for pesticides in food (GB2763-2014).

Economic benefits assessing

The total inputs and outputs of two apple managements were recorded in details, with the output–input ratio being calculated individually. The apple price was fixed according to the different quality grades. The CM apples were all sold in the local market, while part of the OM ones with good appearance were sold through online shop and bulk orders at high price. Meanwhile, a part of OM apples were sold at low price due to the small fruit weight. The packing charges as well as the labor cost of packing and selling were deducted.

Statistical analysis

The annual mean values used in the figures were calculated from the quarterly mean values. All data were analyzed using one-way analysis of variance (ANOVA), and the differences between CM and OM were tested by two-sample Student’s t-test and the differences between experimental years were tested by LSD (least significant difference) test with SPSS 16.0. Differences between managements were identified as significant if P < 0.05 and highly significant if P < 0.01. Figures were performed using Sigma Plot 10.0 (Aspire Software Intl. Ashburn, VA, USA). All the data of the results were means ± standard error.

Improvement of soil characteristics

As is shown in Table 3, from 2012 to 2014, soil organic carbon content (SOC) of 0–20 cm layer in OM remarkably increased, from 15.6 g kg⁻¹ to 22.7 g kg⁻¹. SOC of CM increased in the first year of survey but decreased from 2013 to 2014, which might be caused by longtime application of synthetic fertilizers as well as herbicides, resulting in soil fertility degradation and instability. The differences of soil features between the two different systems were striking in the first two years of the experiment, and highly significant in the third year (P < 0.01). For instance, SOC of OM was 2.1 times higher than that of CM in 2014 (Table 3). Total nitrogen (TN) content of OM was higher than that of CM and the difference was obvious (P < 0.05) in 2014 by a factor of 1.97. However, it was not statistically significant in 2012 and 2013 (P > 0.05). The same as SOC, TN of CM also elevated at the beginning then decreased later on, which might be caused by nitrogen loss of ammonia volatilization and nitrate nitrogen leaching under low C/N ratio condition.

As is shown in Table 3, microbial biomass carbon (MBC) of OM increased substantially from 220 mg kg⁻¹ to 376 mg kg⁻¹, which was significantly higher than that of CM in 2014 (P < 0.05). MBC of CM was 134 mg kg⁻¹ and 164 mg kg⁻¹, respectively in 2013 and 2014. Microbial biomass nitrogen (MBN) also displayed an obvious growth in OM, increased by 102% from 2013 to 2014. MBN from CM kept at a markedly lower level, being 30% of that of OM in 2014 (P < 0.05).

Soil bacterial community

A total of 1,137,310 PE reads were obtained from the 24 soil samples derived from CM and OM in both summer and winter time of 2013 and 2014. At the 3% phylogenetic distance level, most rarefaction curves reached saturation, indicating that the majority of the OTUs had been covered (Fig. 1). As is shown in Fig. 2, the dominant bacterial phyla of the total soil samples were Proteobacteria (57%) and Actinobacteria (10%), with others including Acidobacteria (6%), Gemmatimonadetes (6%), Bacteroidetes (5%), Planctomycetes (4%), Firmicutes (3%), Verrucomicrobia (2%), Chloroflexi (2%), TM6(2%) and so on. Soil bacterial abundance of CM was significantly richer than that of OM in Actinobacteria(except summer 2013), Gemmatimonadetes (except winter 2013) and Chloroflexi (winter 2013) at phyla level (P < 0.05). By contrast, OM samples were evidently more abundant in Bacteroidetes (winter 2013, summer 2014), Gemmatimonadetes (winter 2014), TM6 (winter of both year). However, none significant differences (P > 0.05) were found at other sampling times (Fig. 2).

Despite richer relative abundance for several phyla, CM soils were found to be poorer in bacteria community as a whole. Except at the winter of 2013, Shannon’s diversity index of CM was lower than that of OM at the other three sampling times, notably in winter 2014 with a gap of 7% (Fig. 3). The detected rhizobia of the soil samples included Devosia, Burkholderia, Rhizobium, Cupriavidus, Mesorhizobium and Ochrobactrum (Fig. S1). The sum of these six genera’s relative abundances was a little bit higher in OM than that in CM at the four sampling periods separately, however the differences were not prominent (P > 0.05). Besides, the relative abundance of some other functional bacteria genus such as Pseudomonas, Flavobacterium, Gemmata, are higher in OM than those in the CM.

Earthworm abundance

The density of earthworms in OM was consecutively increased in the first two years, from 8 m⁻² to 365 m⁻² and peaked at the level of 369 m⁻² at the beginning of the third experimental year. However, it declined to 273 m⁻² later (Fig. 4), which may be owing to the density feedback under limited carrying capacity. By contract, the CM held fewer earthworms around 6∾33 m⁻², which remained at an exceptionally lower level (P < 0.01). In 2014, OM had an average of 20 times more earthworms than CM, and the largest gap of density was more than 50 times appeared in the spring of 2014 (Fig. 4).

Through identification of samples, we found that earthworms in the two managements mainly included Amynthas heterochaetus and Drawida japonica. Earthworms were classified into detritivores and geophages (Lee, 1985) based on whether feeding on rich humus or not, reflecting characteristics of composition under the two managements. It was obvious that due to the rich organic matter input, the density of detritivores in OM was extremely higher than that of CM (P < 0.01) (Fig. S2A), while the difference of geophages’ density was quite smaller between the two managements (P < 0.01) (Fig. S2B).

Weed biodiversity

In CM, weeds were extirpated by chemical herbicides. In OM, however, due to the strong clonal propagation capacity, D. indica grew faster than harmful weeds in early spring, so it became obviously dominate. The relative cover of D. indica increased substantially by 3.3 folds from 2012 to 2014 and reached 72% (Fig. 5). Simpson index (D), Shannon–Wiener index (H′), and Pielou index (E) of weed communities showed the same decreasing trend, reducing by 38%, 54% and 17% respectively from 2012 to 2014 (Fig. 5). Nevertheless, declines of Simpson index and Pielou index were not remarkable between the first two years but noteworthy between the last two years. Reduction of Shannon–Wiener index was significant among the three years (Fig. 5).

Dynamics of nocturnal phototactic pests

In CM, pests were controlled by chemical broad spectrum pesticides. In OM, continual trapping and monitoring of nocturnal phototactic pests was carried out for the whole growth seasons of 2012–2014. The captured phototactic pests mainly included Lepidoptera, Coleopteran, Orthoptera, and Hemiptera. The amount of total daily captured pests displayed a generally declining tendency, so did the scarab beetles and moths (Fig. 6). According to our monitoring results, the average weight of total daily captured pests per light dropped from 21 g day⁻¹ to 13.7 g day⁻¹ by 35% reduction from 2012 to 2014 (Figs. 6A–6C). The average number of daily captured scarab beetles had greatly shrunk from 14 day⁻¹ to 2 day⁻¹, or by 86% decrease (Figs. 6D–6F). The peak number of scarab beetles population at outbreak has decreased sharply from 83 day⁻¹ to 23 day⁻¹, by 72% decrease, against the time node of zero-value being shifted forward from August to July (Figs. 6D–6F). The trapped moths involved many species, which mainly had two or three activity peaks in summer and the beginning of autumn. The average weight of daily captured moths declined from 11.9 g day⁻¹ to 9.2 g day⁻¹, with a reduction of 23% (Figs. 6G–6I).

Yields and economic benefits

The apple yield of OM was lower than that of CM in each of the three years of the experiment, by 16.7%, 20.2%, and 12.2% respectively (Fig. 7). The difference of yield between the managements was significant in 2012 (P < 0.05) and not notable in the last two years (P > 0.05). The yield of the OM had a slow but steady rising tendency, while the yield of CM showed a downward trend after the rise in the second year.

As is shown in Table 4 (see Table S3 for calculating details of input and output), CM invested mainly on synthetic fertilizers and pesticides, and cost more in terms of labors due to the high pesticides spray frequency and artificial pollination. Although OM was somewhat labor-intensive, especially in cattle manure fertilization and weed control, it saved some labor because of physical pest controlling and natural honeybee (O. excavata) pollination. In addition, OM cost more in purchasing equipment and setting paper bags. Therefore, the total inputs of two systems were approximately the same.

As for food security, a detection report for 191 items of chemical pesticides and herbicides residue (Table S2) indicated that organic apples can meet EU’s organic food standards with none pesticides residue found. Residues of pesticides including chlorbenzuron, chlorpyrifos and fungicide namely tebuconazole were found in conventional apples. Accordingly, as shown in Table S3, the organic apples possessed higher marketable price with an average price of $2.03 kg⁻¹ and were mainly sold to high-consuming customers in large cities such as Beijing, Shanghai and Guangzhou, resulting in relatively higher economic outputs. In contrast, the conventional apples were sold in the local market at a lower price with an average of $0.56 kg⁻¹. In consequence, OM displayed much higher economic benefits, with the output–input ratio being 3.67 or 103% higher than that of CM.