Given the attention drawn since several decades by arbuscular mycorrhizal fungi (AMF) as potential biological alternatives to chemicals in a low-input agriculture, much effort has been spent in the investigation of mechanisms influencing the dynamics inside AMF communities. In the present study we evaluated the influence of different crop rotations on the AMF soil community, after a 50 y long-term field experiment established at Martonvásár, Hungary. Four types of crop rotation were chosen for sampling: corn monocropping, corn-alfalfa, corn-wheat, and corn-spring barley-peas-wheat. Community composition of AMF in soil was analyzed with a molecular approach amplifying a portion of 28S rDNA. The crop rotation practice didn’t show an influence on identity of the species composing AMF assemblages, but on the other hand seemed to affect positively the true diversity, defined as number of MOTUs present in the communities.
Arbuscular mycorrhizal fungi (AMF) are obligate root symbionts which establish associations with the majority of land plant species [
For a profitable capitalization of these symbiota a better understanding of which mechanisms influence the dynamics inside an AMF community is therefore required. Particular attention has been given to the factors provoking reduction of AMF diversity with consequent limitation of complementary benefits provided by different AMF species or genotypes. Chemical fertilization and, above all, high phosphorus levels do not favor a wide AMF community because these conditions can suppress root colonization and thus growth of AMF [
Crop rotation is an agronomic practice that exhibits general beneficial aspects associated with maintenance or improvement of soil fertility, reduction in erosion potential and in the build-up of pests and decreased reliance on agricultural chemicals. The effects of crop rotation on the AMF communities are still partially investigated and most of the available data concern morphological aspects as rate of roots colonization and spore abundance [
In the present work we propose to compare, by means of molecular tools, the effect of different crop rotation practices on the biodiversity and structure of the resident soil AMF communities.
The sampling site was located at Martonvásár (47˚21'N, 18˚49'E), Hungary, where an experimental field was established in 1958 by the Agricultural Research Institute of the Hungarian Research Academy of Sciences. The area was marked by continental climate with an annual rainfall average up to 539 mm and an annual temperature average of 10.6˚C (data between 1958 and 2008). The soil was classified as humus loam of the chernozem type with forest residues (21% clay, 48% silt and 18% sand content), slightly acidic in the ploughed layer (pH 6 - 7), with poor supplies of available phosphorus and good supplies of potassium. The experimental area was split into small parcels (14 × 7 m) for the different agricultural land use systems in relation to the crop rotation type. Each parcel was characterized by a conventional annual tillage of 20-cm depth and by receiving no fertilization year after year. Weeds and insects were controlled by pesticide treatments as Force (tephlutrine 1.2 lha−1), Acenit A 880 EC (acetochlore 2.0 lha−1), and a combination of Motivell (nicosulfuron) + Cambio (dicamba + bentason) + Dach HC surfactant (0.8 + 2.0 + 0.6 ha−1) used since 1995 as pre-sowing, pre-emergence, and post-emergence treatments, respectively.
Four crop rotation systems were chosen for sampling: maize monoculture (C1), maize and legume (C3: 3 y alfalfa and 5 y maize), maize and wheat (C5: 2 y maize and 2 y wheat), Norfolk type rotation (C7: 1 y maize, 1 y spring barley, 1 y pea, 1 y wheat). Two parcels for each rotation system, at least 40 m distant, were used for sampling.
Three soil cores (~5 × 5 × 30 cm) were randomly collected from each parcel on June 2012. In total 24 soil samples were collected (4 rotation systems × 2 parcels × 3 soil cores).
DNA extraction from 0.5 g of soil samples was performed using a FastDNA Spin for Soil Kit (Q-BIOgene, Heidelberg, Germany) according to the manufacturer’s protocol.
Amplification of LSU fragments (portion of 28S rDNA) was achieved by means of a nested approach with Phusion High Fidelity DNA Polymerase (Fermentas), using LR1 (5ʹ-GCATATCAATAAGCGGAGGA-3ʹ) and NDL22 (5ʹ-TGGTCCGTGTTTCAAGACG-3ʹ) as outer primers [
Sequence similarities were determined using the BLASTn sequence similarity search tool provided by GenBank. Sequences were also checked for chimeras using the “chimera.slayer” command in Mothur v.1.33.3 [
LSU sequences were aligned through the CIPRES web-portal with MAFFT on XSEDE and clustered by Mothur into Molecular Operational Taxonomic Units (MOTUs) at the conventional 97% similarity level, adopted for the definition of a microbial “species” [
MEGA 4.0 software assessing Kimura-2p model as distance method and 1000 replicates of non-parametric bootstrap-ping was used to construct a neighbor-joining (NJ) consensus tree. Rarefaction curves, non-parame- tric richness indices (ACE and Chao1) and Shannon diversity index were estimated with Mothur in order to analyze the α-diversity of AMF community in each crop rotation trial.
The obtained data about MOTU distribution were subjected to principal component analysis (PCA) and common components coefficients, eigenvalues and the proportion of the total variance expressed by each single MOTU were calculated. For the analyses the sequences were clustered also in MOTUs with 94% and 90% as cutoff of similarity to highlight possible effects at taxonomic rank above species level. The scree plot was used to select the components most relevant for the ordination analysis. Correlations between MOTUs and each principal component were calculated, and those having an absolute value >0.5 were considered relevant [
Canonical Correspondence Analysis (CCA) was run using PAST version 2.16 to relate the abundance of phylospecies to environmental variables. In the assay the pH, humus percentage, total nitrogen, calcium, phosphorus and potassium content of the soil were taken in account.
DNA was extracted successfully from all the soil samples with an average yield of 36 ng·ul−1. All soil DNA samples gave positive PCR products after the nested amplification. Overall 480 clones were sequenced and, after BLASTn analyses, 394 AMF sequences (82%) were obtained for further analyses. The primer 28G1-28G2 proved to be highly specific because only 1% of sequences was found to correspond to non-AMF species. According to the BLASTn results 100% of AMF sequences belonged to the former Glomus group A, corresponding to the family of Glomeraceae in the classification by Redecker et al. [
AMF sequences were clustered, after editing, in 26 MOTUs by Mothur, ranked according to the abundance of sequences (
Considering the relative abundance of the MOTUs and their position in the phylogenetic tree, 43% of sequences belonged to Rhizophagus-Sclerocystis clade (former Glomus Group Ab), 31% to Funneliformis-Sep- toglomus clade (former Glomus Group Aa) while the remaining 23% to Glomus species of uncertain affiliation clustering basal in the Glomeraceae. The latter are very common in environmental studies [
Other MOTUs well represented, as MOTU 04 related to Glomus cf. diaphanum, MOTUs 05 and 06 related to Glomus species of uncertain affiliation, and MOTU 07 related to Septoglomus viscosum, didn’t show a distribution among all the treatments (considering both parcels) neither were found to be characteristic of specific rotation systems.
A Canonical Correlation Analysis was performed to verify whether the differences in composition and structure of the AMF assemblages could be due to the soil properties. However, no significant correlation was found (P = 0.8). A Principal Component Analysis (PCA) was performed to investigate a correlation between crop rotation types and relative AMF communities detected. When the analysis was carried out using the MOTUs with 97% of similarity cutoff as input, no correlation was found (
The analysis was repeated using MOTUs with 94% and 90% of similarity cutoff to seek a possible effect of crop rotation system on the AMF communities at a taxonomic rank above species level. In both cases the PCA and the hierarchical clustering did not highlighted a correlation between treatments and the related AMF assemblages (data not shown). The lack of correlation could be related to the findings of several studies where, beside niche-based mechanisms [
The rarefaction curves (
M1 | M2 | M3 | M4 | M5 | M6 | M7 | M8 | M9 | M10 | M11 | M12 | M13 | M14 | M15 | M16 | M17 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
C1-1 | 10.4 | 14.6 | 2.1 | 0 | 0 | 0 | 33.3 | 0 | 0 | 16.7 | 4.2 | 4.2 | 0 | 2.1 | 0 | 0 | 0 |
C1-2 | 5.7 | 11.3 | 3.8 | 15.1 | 0 | 47.2 | 0 | 5.7 | 0 | 0 | 9.4 | 0 | 0 | 0 | 0 | 1.9 | 0 |
C3-1 | 0 | 14 | 4 | 4 | 18 | 0 | 8 | 10 | 0 | 4 | 0 | 18 | 14 | 0 | 0 | 0 | 0 |
C3-2 | 2.3 | 11.6 | 0 | 44.2 | 27.9 | 0 | 0 | 2.3 | 0 | 0 | 0 | 0 | 0 | 0 | 4.7 | 0 | 2.3 |
C5-1 | 35.2 | 3.7 | 5.6 | 0 | 20.4 | 0 | 1.9 | 5.6 | 0 | 0 | 3.7 | 0 | 0 | 3.7 | 1.9 | 3.7 | 0 |
C5-2 | 7.5 | 30 | 2.5 | 0 | 0 | 0 | 2.5 | 22.5 | 0 | 0 | 5 | 0 | 0 | 15 | 10 | 2.5 | 0 |
C7-1 | 0 | 2.1 | 68.1 | 0 | 2.1 | 0 | 2.1 | 0 | 0 | 12.8 | 0 | 0 | 4.3 | 0 | 0 | 2.1 | 4.3 |
C7-2 | 37.3 | 13.6 | 3.4 | 10.2 | 0 | 0 | 1.7 | 0 | 28.8 | 0 | 0 | 0 | 0 | 0 | 0 | 1.7 | 1.7 |
parcel of maize-legume rotation (C3-1) all the MOTUs predicted by the non-parametric richness estimators ACE and Chao1 had been collected by the sampling effort. On the contrary in the Norfolk type rotation (C7) less than the 60% of MOTUs predicted resulted represented by the sequences analyzed.
The true diversity, measured as effective number of taxa [
noculture (C1), with 8.2 - 10.7 MOTUs predicted, to Norfolk type rotation (C7), with 15 - 15.5 MOTUs predicted, highlighting an effect of the number of crops in succession on the AMF communities. Higo et al. [
One explanation could be given considering a certain host plant preference by AMF [
Crop rotation | Observed MOTUs | Estimated MOTUs | O/Ea | Diversity index | ||||||
---|---|---|---|---|---|---|---|---|---|---|
ACE | CHAO | Average ACE-CHAO | H' | |||||||
C1-1 | 10 | 11.18 | 10.25 | 10.71 | 0.93 | 2.19 | ||||
C1-2 | 8 | 8.39 | 8.00 | 8.20 | 0.98 | 1.89 | ||||
C3-1 | 10 | 10.00 | 10.00 | 10.00 | 1.00 | 2.32 | ||||
C3-2 | 8 | 11.82 | 9.00 | 10.41 | 0.77 | 1.80 | ||||
C5-1 | 12 | 12.91 | 12.20 | 12.55 | 0.96 | 2.30 | ||||
C5-2 | 10 | 14.82 | 13.00 | 13.91 | 0.72 | 2.19 | ||||
C7-1 | 9 | 17.75 | 12.33 | 15.04 | 0.60 | 1.58 | ||||
C7-2 | 9 | 19.00 | 12.00 | 15.50 | 0.58 | 1.85 |
aO/E = No. MOTUs observed divided by No. MOTUs estimated (average value between the two non-parametric richness estimators ACE and Chao1).
field used for sampling. Such agrotechnical regime could have facilitated the durability of AMF inocula in the soil, even in absence of the “optimal” host for some of them. From this point of view the Norfolk type rotation could have been favored in the accumulation of more diversity both for the higher number of crops in succession and for the shorter time between a crop and the following.
In conclusion, the present study represents the most extended molecular analysis on the effect of different crop rotations on the AMF assemblages in the soil. According to the molecular evidence, crop rotation practice seems to affect the AMF community. The differences in species composition and structure between AMF communities, dominated largely by widespread generalists such as Rhizophagus irregularis and Funneliformis mosseae, appeared due to stochastic reasons. Nevertheless a clear trend regarding the “true diversity”, decreasing from Norfolk type to monoculture AMF community, was detected, suggesting the number of crops in rotation among the factor drivers of AMF community shaping.
This study was supported by grants from the National Research Council (OTKA K101878) and from KTIA (AIK_12-1-2012-0012).