Soil fertilization leads to a decline in between-samples variability of microbial community δ13C profiles in a grassland fertilization experiment

Abstract

Gas chromatography combustion isotope ratio mass spectrometry (GC-C-IRMS) was used to measure the (13)C/(12)C ratios of PLFAs at natural abundance levels from a temperate grassland nitrogen (N) and phosphorus (P) factorial fertilization experiment in northern Greece. In each plot two rhizosphere samples were derived centred around individual Agrostis capillaris and Prunella vulgaris plants. It was hypothesized that the isotopic signal of microbes that preferentially feed on recalcitrant litter such as fungi would be modified by fertilization more strongly than that of opportunistic microbes using labile C. Microbial community δ(13)C was affected by both P and N fertilization regime and plant species identity. However, we have been unable to detect significant nutrient effects on individual groups of microbes when analyzed separately in contrast to our original hypothesis. Intra-treatment variability, as evaluated from Hartley's F(max) tests in the five first PCA components axes as well as the size of the convex hulls in PCA scoreplots and Mahalanobis distances, was considerably higher in the non-fertilized controls. Moreover, a significant relationship was established between the change in PLFA abundances and their respective changes in δ(13)C for the aggregate of samples and those simultaneously fertilized with N and P. We conclude that use of compound specific isotope analysis in the absence of labelling represents a valuable and overlooked tool in obtaining an insight of microbial community functioning.

Figure 1. Isotopic signal of key microbial groups.

δ¹³C value (±S.E.) of main microbial groups in the fertilization experiment as affected by N and P fertilization practices and plant species identity. The only significant effect was that of plant species on fungi. The two dotted lines have been drawn for reference purposes for δ¹³C signals equal to −30 and −35, respectively. Abbreviations stand for: C-control +N-N fertilized samples +P-P fertilized samples +NP = NP fertilized samples, Ac-Agrostis capillaris, Pv-Prunella vulgaris.

Figure 2. Ordination diagrams of the isotopic signal of PLFA signatures.

Principal component analysis distance scoreplots and respective convex hulls of PCA scores grouped according to N and P fertilization regime. In all plots the x axis illustrates PCA scores of the first component whereas the y axis illustrates PCA scores of the (a) second; (b) third; (c) fourth and (d) fifth component, respectively. The five axes explained a cumulative 72.36% of variance. Abbreviations stand for: C-control +N-N fertilized samples +P-P fertilized samples +NP = NP fertilized samples.

Figure 3. Modification of PLFA abundance vs isotopic signal diagram.

Scatter plot of the relationship between percent change of individual PLFA concentrations and Δ¹³C in the fertilized treatments compared to the non-fertilized controls. Open circles: N fertilized samples, triangles: P fertilized and squares: NP fertilized samples. The thin continuous line is the best fit least-squares line for the NP fertilized samples. The thick continuous line is the least-squares trendline for all three fertilization treatments when compared to the control. Note, non-significant trendlines for N and P fertilized samples were not drawn.