Abiotic conditions and plant cover differentially affect microbial biomass and community composition on dune gradients

Soil Biology & Biochemistry 41 (2009) 102–109

Contents lists available at ScienceDirect

Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio

Abiotic conditions and plant cover differentially affect microbial biomass and community composition on dune gradients T.K. Rajaniemi a, *, V.J. Allison b, c a

University of Massachusetts Dartmouth, Biology Department, 285 Old Westport Road, North Dartmouth, MA 02747, USA Landcare Research, Private Bag 92170, Auckland 1142, New Zealand c Ministry of Agriculture and Forestry, PO Box 106 231, Auckland 1143, New Zealand b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 31 January 2008 Received in revised form 29 September 2008 Accepted 4 October 2008 Available online 31 October 2008

Dune systems are characterized by strong gradients of physical stress, with blowing sand and salt spray decreasing with distance from the ocean, and soil nutrients increasing. In this study we ask how soil microbial community composition and biomass change along transects away from the ocean, and whether these changes are regulated by abiotic stress or by resource availability. We collected bulk soil from under three plant species representative of the dune front, back, and flat: Ammophila breviligulata, Rosa rugosa, and Myrica pensylvanica. The biomass and composition of microbial communities were examined using phospholipid fatty acid (PLFA) analysis under patches of dominant vegetation, and in paired bare plots. We found that microbial biomass was strongly correlated with soil C, and thus the presence of vegetation. Community composition, on the other hand, varied with abiotic stresses, especially soil salinity. These variables in turn depended on distance from the shore, and were ameliorated in some cases by vegetation. These findings demonstrate that biomass and community composition are influenced by different environmental variables. Thus, relationships between biomass and composition are unlikely to be readily predicted on the basis of a single resource. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Coastal dunes Soil microbial community Microbial biomass Community composition Stress gradients Facilitation

1. Introduction Dunes represent an environment with strong gradients of physical stress: at the extreme of these gradients, resource limitation or physical conditions severely limit biomass accumulation. These strong abiotic gradients have been demonstrated to influence plant community composition. For example, strong winds and salt spray from the ocean reduce germination and survival of dune forbs (van der Valk, 1974). Species differ in salt spray tolerance, resulting in different species occurring on different parts of the dune (Wilson and Sykes, 1999). Low nutrient content of the substrate results in nutrient limitation on plant growth, particularly by nitrogen (Kachi and Hirose, 1983), and abundance of plant species may be associated with soil nutrient content (Houle, 1997). Further, tolerance to burial by blowing sand may also affect which species are found in different areas of the dune (Wilson and Sykes, 1999). Although not well studied, these same abiotic gradients have the potential to influence microbial community composition and biomass. In general, total microbial biomass appears to be most

* Corresponding author. Tel.: þ1 508 999 8223; fax: þ1 508 999 8196. E-mail address: [email protected] (T.K. Rajaniemi). 0038-0717/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.soilbio.2008.10.001

strongly controlled by soil carbon availability (Wardle, 1992; Yao et al., 2000; Allison et al., 2007). In dune systems, this in turn appears to be regulated by vegetation cover, with vegetation islands promoting the accumulation of soil carbon and hence microbial biomass (Sarig et al., 1999; Su et al., 2004). While carbon appears to be an effective surrogate measure of energy and thus the total biomass a system can support, composition is likely to be dependent on a multitude of environmental factors that different organisms differ in their ability to tolerate or take advantage of. In general, lower fertility sites have higher abundances of fungi relative to bacteria than do high fertility sites, possibly due to the greater ability of fungi to degrade recalcitrant materials such as lignin (Beare, 1997; Zeller et al., 2001). In addition, disturbance reduces fungal biomass, presumably by disrupting hyphal networks (Hedlund, 2002; Allison et al., 2005). Further, Allison et al. (2007) found that microbial community composition in a prairie system was dependent on phosphorus, calcium and water, and only indirectly on soil carbon. These results suggest that while the presence of plants is likely to have strong effects on microbial biomass, microbial community composition may be more strongly regulated by abiotic gradients such as disturbance and nutrient availability. These abiotic factors may, in turn, be influenced by the establishment of dune plants. For example, soil temperatures (Martinez, 2003), wind speed (Cowles,

T.K. Rajaniemi, V.J. Allison / Soil Biology & Biochemistry 41 (2009) 102–109

1899), and sand accretion (Martinez, 2003) are lower under shrub canopies. Soil nitrogen, while highly variable, is higher beneath nitrogen-fixing shrubs (Alpert and Mooney, 1996; Shumway, 2000), while vegetation cover promotes accumulation of carbon irrespective of plant species (Lichter, 1998; Su et al., 2004). In this paper, we examine shifts in microbial community composition and biomass along a foredune gradient. Our study area consisted of a barrier dune system with a single, low dune crest. Soil was collected from three zones: the dune front (nearest the ocean); the dune back (just behind the crest); and the flat area behind the dune itself. The biomass and composition of microbial communities were examined both in patches of dominant vegetation and in paired bare plots. These paired comparisons enable us to tease apart the effects of plant islands versus abiotic stressors on microbial community composition and biomass. We hypothesized that microbial biomass would be dependent on plant biomass rather than position on the dune, as soil C is an excellent predictor of microbial biomass. In contrast, we hypothesized that microbial community composition would depend on the abiotic environment, with any plant effects due to plant-mediated amelioration of abiotic stress. 2. Materials and methods 2.1. Study area This research was conducted at Waquoit Bay National Estuarine Research Reserve in Falmouth, Massachusetts (N4133 500 , W70 300 2200 ). The mean air temperature and precipitation are 9.8  C and 1039.5 mm, respectively (averages from 1949 to 1970 and 1976 to 1982, respectively;WorldClimate.com). The study site is a barrier dune system, separating the Atlantic Ocean from Waquoit Bay. The dune system consists of a single, low dune crest, 2 m or less in height and 5–20 m from the high tide line. A largely flat area behind the crest extends 100–200 m to wetlands or the bay. In some areas, there is a second dune crest on the bay-side shore, but these were not included in sampling. The soil throughout the dune system is a coarse sand with low organic matter content (mean total carbon 0.14%, maximum observed 0.93%). All sampling sites were located on Hooksan series soil, a very deep, excessively drained soil typical of vegetated sand dunes on the coast. The area shows a typical loose zonation of plant species. Ammophila breviligulata Fern., American beachgrass, is a dunestabilizing grass dominating the dune front. On the dune back, beachgrass is interspersed with patchy shrubs, primarily the rose Rosa rugosa Thunb. In the flat area behind the dune, grass abundance declines, and the dominant shrub switches to the nitrogenfixing bayberry (Myrica pensylvanica Mirbel) (species referred to hereafter by genus names). Forbs, including beach pea (Lathyrus japonicus Willd.) and seaside goldenrod (Solidago sempervirens L.), occur at low density throughout the dune system. Wormwood (Artemisia caudata Michx., a forb) and beach plum (Prunus maritima Marsh., a shrub) are also common in the dune flat. 2.2. Sampling and laboratory analyses Samples were collected in a series of paired plots in July and August 2006. On ten transects, each 120 m apart, densely vegetated plots (1 m2) were paired with sparsely vegetated or bare plots, for each of the three zones. Along each transect, a dune front plot with dense Ammophila was paired with a plot with sparse Ammophila (entirely bare areas were rare in this zone); a dune back plot beneath a large Rosa shrub was paired with a nearby plot empty or nearly empty of vegetation, and a dune flat plot beneath a large Myrica shrub was paired with a nearby bare area. Bare areas were located at least 1 m from the dense grass or the shrubs,

103

and the plot pairs were located at the same distance from the dune crest whenever possible. Distance from the dune crest was recorded for each plot (as a negative value for plots in front of the dune crest). In each plot, percent cover of each plant species was determined visually, soil cores were collected, and a series of environmental variables were measured. The temperature was recorded on a sunny day within 2 h of solar noon at the soil surface and at a depth of 30 cm. Soil moisture was measured by time domain reflectometry with a MiniTrase system (SoilMoisture Equipment Corp., Santa Barbara, CA). Soil moisture was measured two days after a rainfall, when moisture in these typically dry, sandy soils was expected to vary the most. Finally, salt spray was measured with salt spray traps (following Cartica and Quinn, 1980). A piece of cheesecloth was suspended in a wooden frame 20 cm above the soil surface for 24 h, and a 10 cm  10 cm section from the center of the cheesecloth was removed for analysis. The cheesecloths from each salt spray trap were immersed and swirled in 100 ml of deionized water. Salinity of the water was then measured with a conductivity meter (SevenEasy, Mettler Toledo, Columbus, OH). Three soil cores (1.7 cm in diameter and 30 cm deep) were collected from random positions within each plot and composited. Soil samples were kept on ice for several hours until returned to the lab, then refrigerated (4  C). The soil samples were analyzed for available N, total C and N, and salinity. Plant-available nitrogen  (NHþ 4 and NO3 ) was extracted from a 5-g subsample of field-moist soil in 50 ml of 2 M KCl; extracts were analyzed for ammonium and nitrate at the Soil and Plant Nutrient Laboratory at Michigan State University. Nitrate in extracts was analysed on a Lachat Flowinjection-analyser (FIA; Lachat Instruments, Loveland, CO) using the cadmium reduction method, while ammonium was analysed on an FIA using the salicylate method. A portion of each soil sample was ground in a Spex 8000M mixer mill (SPEX CertiPrep, Metuchen, NJ) and analyzed for total C and total N by dry combustion by the Stable Isotope/Soil Biology Laboratory of the University of Georgia Institute of Ecology. Finally, soil salinity was determined by combining 20 g of field-moist soil with 100 ml of distilled water, shaking for 30 min, and measuring conductivity of the solution after settling (Rowell, 1994). Microbial biomass and community composition in soil samples were assessed using phospholipid fatty acids (PLFAs). Phospholipids are integral components of cell membranes, and are metabolized rapidly upon cell death: as a result, PLFAs reflect viable biomass (Frostegård and Bååth, 1996). Further, specific signature PLFAs are associated with subsets of the microbial community, and thus reveal shifts in microbial community composition (Vestal and White, 1989; Tunlid and White, 1992). Lipids were extracted from refrigerated soil, immediately upon receipt at the PLFA laboratory. Lipids were extracted from 10 g of field-moist soil, in a singlephase mixture of chloroform, methanol, and phosphate buffer (0.05 mol L1, pH 7.4) in a ratio of 1:2:0.8, by an adaptation of the method described by Bligh and Dyer (1959). After 2 h, water and chloroform were added to separate the mixture into polar and nonpolar fractions, and total lipids were extracted from the nonpolar chloroform phase. The PLFAs were separated from other lipid classes by using silicic acid column chromatography (Vestal and White, 1989; Zak et al., 1996). FAME 19:0 was added as an internal standard at a concentration of 70 ng/ml, and the PLFAs subsequently methylated by using a mild-alkaline methanol– toluene solution. We assume that any loss of the internal standard during the methylation step occurs at the same rate as loss of extracted PLFAs. Methylated PLFAs were dissolved in ethyl acetate, then separated and identified on an Agilent 6890 GC (Agilent Technologies, Santa Clara, CA), equipped with a 25-m BP-5 column. Helium was

104

T.K. Rajaniemi, V.J. Allison / Soil Biology & Biochemistry 41 (2009) 102–109

used as the carrier gas, at a constant rate of 2.7 ml min1. Sample injection was at a split ratio of 15:1, with an inlet temperature of 250  C and a flame ionization detector (FID) temperature of 280  C. The oven temperature was initially 150  C, ramped at 4  C min1 until reaching 250  C. PLFAs were identified by retention time in comparison to known standards and quantified by comparison to the internal standard. Fatty acid nomenclature is in the form of A:BuC, where ‘A’ is the number of C atoms in the chain, ‘B’ is the number of double bonds, and ‘C’ is the position of the double bond from the methyl end of the molecule; cis geometry is indicated by the suffix ‘c’. The prefixes ‘i’, ‘a’, and ‘me’ refer to iso, anteiso, and midchain methyl branching, respectively, with ‘cy’ indicating a cyclopropyl ring structure. 2.3. Statistical analyses Before analysis, data for wind speed, soil moisture, soil C, and total soil N were natural-log transformed to improve normality and homogeneity of variance. For correlations and regressions (see below), values for salt spray were adjusted because salt spray was measured on three different days (due to a limited number of traps available), and days varied in windiness. The mean and standard deviation for salt spray on each measurement day were calculated, and each salt spray value was converted to a z-score by subtracting the day’s mean and dividing by standard deviation. All figures show back-transformed values. Microbial biomass was calculated as the sum of the individual PLFAs (nmol g1 soil). The composition of the soil microbial community was summarized using a principle components analysis (PCA) on the relative mole abundances of PLFAs in each sample. PCA was conducted using CANOCO v 4.5 (ter Braak and Smilauer, 2002) and ordination diagrams were drawn with CanoDraw 4 (ter Braak and Smilauer, 2002). Soil microbial community composition was also summarized as the ratio of fungi:bacteria, with the PLFAs 18:2u6,9c and 18:1u9c attributed to fungi, and PLFAs 14:0, i15:0, a15:0, i16:0, 16:1u7c, 10Me16:0, i17:0, a17:0, cy17:0, 18:1u7c, 10Me18:0 and cy19:0 attributed to bacteria. Shifts in bacterial groups were summarized by the Gram þve:Gram –ve ratio, with the PLFAs i15:0, and i16:0 attributed to Gram positive, and cy17:0, cy19:0, 16:1u7c and 18:1u7c attributed to Gram negative bacteria (designations based on data from Frostegård et al., 1993; Zak et al., 1996; Zelles, 1999). We used two-way ANOVA to test for effects of zone (dune front, back, or flat), and presence/absence of vegetation (vegetated or bare plots) on environmental variables and soil microbial community variables. When zone effects were significant, multiple comparisons were made with the Tukey test. For salt spray, unadjusted values were used in ANOVA, but date of collection was included as a random factor in the ANOVA model. We tested for correlations of each environmental variable with each microbial community variable. Because many environmental variables were correlated with each other, we also used stepwise regression to identify the environmental variables that best explained microbial community composition. Soil total N and C were highly correlated (R ¼ 0.945, P < 0.001), so total N was not included as an independent variable in stepwise regression. 3. Results As planned, bare plots had much lower vegetation cover than vegetated plots (P < 0.001; Fig. 1A). Shrub-dominated plots in the dune back and flat had greater cover than grass-covered dune front plots. The presence of vegetation also affected each of the environmental variables measured, with shrubs having a greater effect than grass on most variables. Soil temperature at the surface and at 30 cm depth was lower in vegetated plots than bare plots

in the dune back and flat, but vegetation had no effect on temperature in the dune front (zone by vegetation interaction, P < 0.05; Fig. 1B and C). Soil moisture was highest in the dune flat, farthest from shore (P < 0.001); the presence or absence of vegetation had no consistent effect on soil moisture (Fig. 1D). Salt spray aboveground and soil salinity declined with increasing distance from shore, from the dune front to the dune flat. Both shrubs reduced aboveground salt spray relative to bare plots; shrubs in the dune back reduced soil salinity as well (zone by vegetation interaction, P < 0.05 for salt spray and P < 0.001 for soil salinity; Fig. 1E and F). Available N tended to decline from the dune front to the dune flat in bare plots, but tended to increase along the same gradient in vegetated plots (zone by vegetation interaction, P < 0.05; Fig. 1G). Total N and total C were greater in the dune back and flat than in the dune front. Vegetation in the dune back and especially the dune flat increased total N and C, while dune front vegetation had no effect (zone by vegetation interactions, P < 0.001; Fig. 1H and I). The soil microbial community also responded to the different zones, and to the presence and absence of vegetation. Total microbial abundance was similar in bare plots in all three zones, but increased under shrubs in the dune back and flat (zone by vegetation interaction, P < 0.001; Fig. 2A). Of the measured environmental variables, total PLFA was most highly correlated with total soil C (R ¼ 0.921, P < 0.001; Table 2). Although soil salinity, surface temperature, and plant cover were all significant, when added to the stepwise regression model these variables only explained an additional 6% of the variation in total PLFA (Table 3). Microbial community composition also varied significantly. The first three PCA axes explained 44.8%, 12.8%, and 11.4% of the variation in the soil microbial community. The microbial communities collected from the dune front fall to the left of PCA axis 1, while samples collected from the dune flat fall to the right of PCA axis 1 (Fig. 3). The community in the dune front is characterized by higher relative abundances of the Gram negative bacterial PLFAs 16:1u7c and 18:1u7c (Table 1). Microbial communities from the dune front also fall lower on PCA axis 2 (Fig. 3), and are associated with somewhat higher relative abundances of the saprophytic fungal PLFA 18:2u6,9c (Table 1). Consistent with the patterns seen in the PCA plot, zone and presence of vegetation had interactive effects on the first PCA axis describing the community (P < 0.05), with values increasing under vegetation in the dune back but decreasing under vegetation in the dune front and flat (Fig. 2B). The second PCA axis also responded to vegetation zone (P < 0.01; Fig. 2C). Values on the third PCA axis did not respond to either vegetation zone or presence of vegetation (not shown). Neither factor affected the fungi:bacteria ratio (Fig. 2D), but the ratio of Gram positive to Gram negative bacteria was greater in the dune back and flat than in the dune front (P < 0.001; Fig. 2E): there were higher relative abundances of Gram negative bacteria closer to the water. Values on the first PCA axis were most strongly correlated with soil salinity (Table 2). This variable explained 36.3% of variation in axis 1 scores; a model that also includes distance from the dune crest and soil moisture (variables only weakly correlated with one another) explained 63.1% of variation (Table 3). Very little of the variation in axis 2 scores could be explained by the measured environmental variables. A model including temperature at 30 cm depth and soil moisture, the only variables correlated with axis 2 score, explained only 15.2% of variation (Table 2 and 3). None of the environmental variables were correlated with the fungi:bacteria ratio (Table 2). The ratio of Gram positive to Gram negative bacteria was correlated with distance from the dune crest, soil moisture, soil surface temperature, soil salinity, and soil C and N (Table 2). Together, the first three of these variables explained 39.3% of the variation in Gram ratio (Table 3).

zone *** veg*** zone x veg*** c

b 80

B

a

60 40 20

D

zone *** veg*** zone x veg** a

b

b

20 15 10 5

salt spray (mS m-1) available N (mg/kg)

G

1200

zone ** veg*** zone x veg* a

F

ab

900

b

600 300 0 1.2

zone ns veg ns zone x veg*

H

0.9 0.6 0.3 0

soil moisture (g g-1 soil)

25

zone *** veg*** zone x veg* a b b

30 20 10

0.04

zone *** veg ns zone x veg ns b

0.03

ab a

0.02 0.01 0

soil salinity (mS m-1)

30

40

105

0

0

0

E

50

suface temp (°C)

100

1600

soil total N (g kg-1)

C

temp at 30 cm (°C)

A

vegetation cover (%)

T.K. Rajaniemi, V.J. Allison / Soil Biology & Biochemistry 41 (2009) 102–109

0.003

zone *** veg** zone x veg*** a b

c

1200 800 400 0 zone *** veg*** zone x veg*** b 0.002

b

a 0.001

0

I

soil total C (g kg-1)

dune front dune back dune flat 0.05

zone *** veg*** zone x veg*** b

0.04 0.03 0.02

b

vegetated bare

a

0.01 0

dune front dune back dune flat Fig. 1. Effects of vegetation zone and presence/absence of vegetation on environmental variables. Significance of factors in two-way ANOVA: *P < 0.05, **P  0.01, ***P  0.001. When vegetation zone effects are significant, zones with significantly different values (Tukey test) are labeled with different letters. Error bars are one standard error of the mean; n ¼ 10 for each treatment.

4. Discussion

4.1. Controls on microbial biomass

As expected, the dune system in this study provided a strong gradient of environmental stress, with greater surface temperatures, aboveground salt spray, and soil salinity, and lower plantavailable nitrogen nearer the shore. Vegetation, especially shrubs which dominate the dune back and flat, ameliorated these stresses, cooling the soil, blocking wind and salt spray, and promoting accumulation of soil nutrients and organic matter. Here, we investigated whether established vegetation affects the soil microbial community, either directly or indirectly through its effects on environmental variables.

As predicted, the presence of plants had a strong effect on microbial biomass, most probably through vegetation effects on soil carbon. This result is consistent with strong effect of soil C on microbial biomass in other systems (Wardle, 1992; Yao et al., 2000; Allison et al., 2007). Other studies in coastal dunes systems have also shown that vegetation enhances microbial biomass (Sarig et al., 1999; Su et al., 2004), and at least one has attributed this effect to higher carbon under shrubs (Su et al., 2004). Although variation in soil carbon can be partially explained by plant cover, other factors, some potentially related to plant species,

T.K. Rajaniemi, V.J. Allison / Soil Biology & Biochemistry 41 (2009) 102–109

zone *** veg*** zone x veg***

20

c vegetated

15 b

10

bare

a 5 0

C

zone *** veg ns zone x veg*

0.06

b

0.04

b

0.02 0 a

-0.02

0.02

PCA axis 2 score

B

PCA axis 1 score

A

total PLFA (ng g-1soil)

106

-0.04

ab

b

0.01 0

a

-0.01 -0.02

zone ns veg ns zone x veg ns

0.4

E Gram pos:neg ratio

D

fungi:bacteria ratio

-0.06

zone ** veg ns zone x veg ns

0.3 0.2 0.1 0

dune front dune back dune flat

0.8 0.6

zone *** veg ns zone x veg ns b b a

0.4 0.2 0

dune front dune back dune flat

Fig. 2. Effects of vegetation zone and presence/absence of vegetation on soil microbial community. Significance of factors in two-way ANOVA: *P < 0.05, **P  0.01, ***P  0.001. When vegetation zone effects are significant, zones with significantly different values (Tukey test) are labeled with different letters. Error bars are one standard error of the mean; n ¼ 10 for each treatment.

seem to contribute as well. This study cannot explicitly identify species effects, but the significant zone  vegetation interactions suggest that some plant traits may be important. For example, soil carbon was similar in bare and vegetated plots in the Ammophila zone, despite their differences in plant cover. Also, Myrica plots had much higher soil carbon than Rosa plots with only a small increase in cover (Fig. 1). Plant biomass or productivity, especially of roots, might be a better predictor of soil C (Bardgett et al., 1999). ‘‘Bare’’ plots in the dune front were usually sparsely vegetated with Ammophila, and most likely contained rhizomes and roots from

0.06 dune front, bare dune front, veg dune back, bare dune back, veg dune flat, bare dune flat, veg

0.04

PCA axis 2

0.02 0 -0.02 -0.04 -0.06 -0.08 -0.1

-0.05

0

0.05

0.1

PCA axis 1 Fig. 3. Results of principle components analysis (PCA) on PLFA data.

nearby plants. Weak relationships between plant biomass and carbon have been noted before (e.g. Allison et al., 2007), and may reflect differences in rates of decomposition of different components of the biomass (Gill et al., 1999). Species-specific differences in root exudates might also contribute to patterns in soil carbon; several studies have detected plant-species specific effects on microbial biomass that are not accounted for by plant biomass (Wardle and Nicholson, 1996; Bardgett et al., 1999). 4.2. Controls on microbial composition The composition of the soil microbial community was most strongly affected by abiotic environmental factors, particularly soil salinity. The main shift in microbial community composition along the dune system appears to be towards higher relative abundances of Gram negative bacteria closer to the water. Other aspects of the environmental gradient, such as soil moisture, also appear to affect microbes independently of plant cover in this system. Direct effects of the abiotic environment have previously been reported by Bardgett et al. (1999), with nitrogen shown to influence microbial composition independently of plant productivity. Vegetation may indirectly influence environmental variables, and thus contribute to microbial community shifts. In this study, plots vegetated with Myrica had lower levels of salinity. A previous coastal dune study also found that changes in soil salinity associated with vegetation resulted in changes in microbial composition, although in that study the soils under plants had higher salinity than in bare areas (Sarig et al., 1999). The significant effect of distance from the dune crest on PCA axis one scores and Gram ratio, even when other environmental

T.K. Rajaniemi, V.J. Allison / Soil Biology & Biochemistry 41 (2009) 102–109

107

Table 1 Effect of relative abundance of PLFAs on PCA axis positions. PLFA

Functional group

Eigenvector (PCA axis 1)

Eigenvector (PCA axis 2)

Eigenvector (PCA axis 3)

14:0 15:0 16:0 18:0 a15:0 i17:0 16:1u7t a17:0 i15:0 i16:0 16:1u7c cy17 18:1u7c cy19:0a 10Me16:0 10Me18:0 16:1u5c 18:2u6,9c 18:1u9c

Genericc Genericc Genericc Genericc Bacteriab,c Bacteriab,c Bacteriaa Bacteriab,c Gram positiveb,c Gram positiveb,c Gram negativeb,c Gram negativeb,c Gram negativeb,c Gram negativeb,c Desulfobacterd/actino Actinomycetesc AMFe Saprophytic fungib,c Saprophytic fungia

0.007 0.055 0.474 0.149 0.026 0.033 0.059 0.043 0.089 0.309 0.511 0.121 0.474 0.198 0.126 0.091 0.216 0.135 0.107

0.007 0.008 0.002 0.025 0.018 0.031 0.117 0.032 0.270 0.021 0.027 0.228 0.218 0.385 0.276 0.245 0.003 0.523 0.501

0.002 0.001 0.331 0.065 0.000 0.026 0.101 0.038 0.011 0.143 0.192 0.130 0.147 0.053 0.304 0.328 0.152 0.569 0.479

a b c d e

Zak et al., 1996. Frostegård et al., 1993. Zelles, 1999. Coleman et al., 1993. Olsson, 1999.

variables are included in the model, suggests that at least one important aspect of the environmental gradient was not measured in this study. Rate of sand deposition is one possible candidate: sand deposition has been shown to decrease with increasing distance from the shore, and affects plant growth and survival (Maun and Perumal, 1999; Wilson and Sykes, 1999). There may also be species-specific effects of plants on the microbial community that were not captured by the environmental measurements. Scores on PCA axis one are increased by the presence of vegetation in Ammophila and Myrica plots, but decreased in Rosa plots, a pattern not seen for any environmental variables (Figs. 2 and 3). Both mycorrhizal fungi and rhizosphere bacteria are known to have host-specific growth responses (Bever et al., 1996; Westover et al., 1997). All three dominant dune species are mycorrhizal (T. Rajaniemi and S. Morse, unpublished data), and may support different fungal species. In addition, Myrica associates with nitrogen-fixing actinomycetes (Fimbel and Kuser, 1995), and pathogenic fungi infecting Ammophila have been identified (van der Putten et al., 1990; de Rooij-van der Goes, 1995). The ratio of fungi to bacteria was not related to any of the variables measured here. Previous studies have shown that the fungi:bacteria ratio is related to productivity (Beare, 1997; Zeller et al., 2001). Productivity is low throughout the portion of the dune gradient considered in this study, so little variation might be expected in the ratio of fungi to bacteria.

4.3. Comparing controls on biomass and composition While soil carbon was the primary determinant of microbial biomass in this study, it had considerably less influence on microbial community composition. Different controls on biomass and community composition may in part explain inconsistent relationships between community biomass and composition revealed in the ecological literature (Waide et al., 1999; Mittelbach et al., 2003). Such relationships have been most closely studied in plant communities, in which production is limited by nitrogen availability (Vitousek and Howarth, 1991). Nitrogen availability also has strong effects on plant species composition and diversity, as revealed by observational studies (Gross et al., 2000; Stevens et al., 2004) and fertilization experiments (Tilman, 1987; Gough et al., 2000). However, there is a great deal of variability in the relationship between productivity and composition (Waide et al., 1999; Mittelbach et al., 2003), suggesting that composition may be controlled by factors other than resource availability. In support of this idea, a meta-analysis shows that individual species’ responses to nitrogen are often dependent on geography, neighboring species identity, or other environmental factors (Pennings et al., 2005). In this dune system, we have demonstrated that while soil microbial biomass is most strongly correlated with the availability of soil C and N (Table 2), community composition is not. Instead,

Table 2 Correlation coefficients (R-values) for correlations between environmental variables and microbial community variables. Environmental variable

Total PLFA

PCA axis 1

PCA axis 2

fungi:bacteria

Gram pos:neg

Distance from dune crest Plant cover Soil surface temperature Soil temp. at 30 cm Wind speed Soil moisture Salt spray Soil salinity Available N Total N Total C

0.440*** 0.654*** 0.419*** 0.463 0.256* 0.019 0.530*** 0.748*** 0.361** 0.919*** 0.921***

0.605*** 0.126 0.486*** 0.408*** 0.069 0.167 0.154 0.611*** 0.022 0.418*** 0.490***

0.240 0.238 0.215 0.331** 0.071 0.220 0.197 0.297* 0.232 0.232 0.220

0.006 0.127 0.107 0.209 0.008 0.172 0.043 0.036 0.221 0.051 0.031

0.289* 0.058 0.423** 0.226 0.093 0.371** 0.090 0.411** 0.105 0.438*** 0.326*

*P < 0.05, **P  0.01, and ***P  0.001.

108

T.K. Rajaniemi, V.J. Allison / Soil Biology & Biochemistry 41 (2009) 102–109

Table 3 Results of stepwise regression on environmental variables on microbial community variables. Independent variable

P

Adjusted R2

ln soil C Soil salinity Surface temp Plant cover

<0.001 <0.001 0.001 0.026

0.845 0.883 0.897 0.904

PCA axis 1

Soil salinity Dist from crest ln soil moisture

0.002 <0.001 <0.001

0.363 0.451 0.631

PCA axis 2

Temp 30 cm ln soil moisture

0.004 0.030

0.095 0.152

Gram pos:neg ratio

Surface temp ln soil moisture Distance from crest

0.041 <0.001 0.001

0.165 0.258 0.393

Dependent variable Total PLFA (nmol g

1

)

R2 value after each variable is added to the regression model (e.g. for total PLFA, a model with only ln soil C has R2 ¼ 0.845; a model with all four independent variables has R2 ¼ 0.904).

composition is more strongly influenced by soil salinity than by soil C and N. These results are consistent with previous field and greenhouse experiments showing that microbial biomass responds strongly to plant biomass or carbon, while composition responds to other environmental factors (Bardgett et al., 1999; Allison et al., 2007). In soil systems, soil carbon is often highly correlated with other variables, and thus initial analyses may erroneously suggest community composition shifts are driven by soil carbon. When examining the effect of perturbations on soil communities, it is important to tease apart the multitude of environmental variables which may potentially affect community composition. This study and others (Wardle and Nicholson, 1996; Bardgett et al., 1999) also suggest an important effect of plant species identity. While the current study cannot completely separate species effects from environmental effects, it does demonstrate that the controls on composition cannot be assumed to be the same as controls on biomass. Acknowledgements We thank Chuck Talley for assistance with field work and Gaye Rattray for help with PLFA extractions. The research was supported by an NZFRST post-doctoral fellowship to VJA. References Allison, V.J., Miller, R.M., Jastrow, J.D., Matamala, R., Zak, D.R., 2005. Changes in soil microbial community structure in a tallgrass prairie chronosequence. Soil Science Society of America Journal 69, 1412–1421. Allison, V.J., Yermakov, Z., Miller, R.M., Jastrow, J.D., Matamala, R., 2007. Using landscape and depth gradients to decouple the impact of correlated environmental variables on soil microbial community composition. Soil Biology & Biochemistry 39, 505–516. Alpert, P., Mooney, H.A., 1996. Resource heterogeneity generated by shrubs and topography on coastal sand dunes. Vegetatio 122, 83–93. Bardgett, R.D., Mawdsley, J.L., Edwards, S., Hobbs, P.J., Rodwell, J.S., Davies, W.J., 1999. Plant species and nitrogen effects on soil biological properties of temperate upland grasslands. Functional Ecology 13, 650–660. Beare, M.H., 1997. Fungal and bacterial pathways of organic matter decomposition and nitrogen mineralization in arable soils. In: Brussard, L., Ferrera-Cerrato, R. (Eds.), Soil Ecology in Sustainable Agricultural Systems. CRC Press, Boca Raton, FL, pp. 37–70. Bever, J.D., Morton, J.B., Antonovics, J., Schultz, P.A., 1996. Host-dependent sporulation and species diversity of arbuscular mycorrhizal fungi in a mown grassland. Journal of Ecology 84, 71–82. Bligh, E.G., Dyer, W.L., 1959. A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology 37, 911–917. Cartica, R.J., Quinn, J.A., 1980. Responses of populations of Solidago sempervirens to salt spray across a barrier beach. American Journal of Botany 67, 1236–1242. Coleman, M.L., Hedrick, D.B., Loveley, D.R., White, D.C., Pye, K., 1993. Reduction of Fe(III) in sediments by sulphate-reducing bacteria. Nature 361, 436–438.

Cowles, H.C., 1899. The ecological relations of the vegetation on the sand dunes of Lake Michigan. Botanical Gazette 27, 95–117. 167–202, 281–308, 361–91. de Rooij-van der Goes, P.C.E.M., 1995. The role of plant-parasitic nematodes and soil-borne fungi in the decline of Ammophila arenaria (L.) Link. New Phytologist 129, 661–669. Fimbel, R.A., Kuser, J.E., 1995. Competitive and mutualistic interactions between pitch pine, bayberry, and their symbionts. Soil Science 160, 69–76. Frostegård, A., Bååth, E., Tunlid, A., 1993. Shifts in the structure of soil microbial communities in limed forests as revealed by phospholipid fatty acid analysis. Soil Biology & Biochemistry 25, 723–730. Frostegård, A., Bååth, E., 1996. The use of phospholipid fatty acid analysis to estimate bacterial and fungal biomass in soil. Biology & Fertility of Soils 22, 59–65. Gill, R., Burke, I.C., Milchunas, D.G., Lauenroth, W.K., 1999. Relationship between root biomass and soil organic matter pools in the shortgrass steppe of Eastern Colorado. Ecosystems 2, 226–236. Gough, L., Osenberg, C.W., Gross, K.L., Collins, S.L., 2000. Fertilization effects on species density and primary productivity in herbaceous plant communities. Oikos 89, 428–439. Gross, K.L., Willig, M.R., Gough, L., Inouye, R., Cox, S.B., 2000. Patterns of species density and productivity at different spatial scales in herbaceous plant communities. Oikos 89, 417–427. Hedlund, K., 2002. Soil microbial community structure in relation to vegetation management on former agricultural land. Soil Biology & Biochemistry 34, 1299– 1307. Houle, G., 1997. No evidence for interspecific interactions between plants in the first stage of succession on coastal dunes in subarctic Quebec, Canada. Canadian Journal of Botany 75, 902–915. Kachi, N., Hirose, T., 1983. Limiting nutrients for plant growth in coastal sand dune soils. Journal of Ecology 71, 937–944. Lichter, J., 1998. Primary succession and forest development on coastal Lake Michigan sand dunes. Ecological Monographs 68, 487–510. Martinez, M.L., 2003. Facilitation of seedling establishment by an endemic shrub in tropical coastal sand dunes. Plant Ecology 168, 333–345. Maun, M.A., Perumal, J., 1999. Zonation of vegetation on lacustrine coastal dunes: effects of burial by sand. Ecology Letters 2, 14–18. Mittelbach, G.G., Scheiner, S.M., Steiner, C.F., 2003. What is the observed relationship between species richness and productivity? Reply. Ecology 84, 3390– 3395. Olsson, P.A., 1999. Signature fatty acids provide tools for determination of the distribution and interactions of mycorrhizal fungi in soil. FEMS Microbiology Ecology 29, 303–310. Pennings, S.C., Clark, C.M., Cleland, E.E., Collins, S.L., Gough, L., Gross, K.L., et al., 2005. Do individual plant species show predictable responses to nitrogen addition across multiple experiments? Oikos 110, 547–555. Rowell, D.L., 1994. Soil science: methods and applications. Wiley, NY. Sarig, S., Fliessbach, A., Steinberger, Y., 1999. Soil microbial biomass under the canopy of coastal sand dune shrubs. Arid Soil Research and Rehabilitation 13, 75–80. Shumway, S.W., 2000. Facilitative effects of a sand dune shrub on species growing beneath the shrub canopy. Oecologia 124, 138–148. Stevens, C.J., Dise, N.B., Mountford, J.O., Gowing, D.J., 2004. Impact of nitrogen deposition on the species richness of grasslands. Science 303, 1876–1879. Su, Y., Zhao, H., Li, Y., Cui, J., 2004. Carbon mineralization potential in soils of different habitats in the semiarid Horqin sandy land: a laboratory experiment. Arid Land Research and Management 18, 39–50. Tilman, D., 1987. Secondary succession and the pattern of plant dominance along experimental nitrogen gradients. Ecological Monographs 57, 189–214. ter Braak, C.J.F., Smilauer, P., 2002. CANOCO software for canonical community ordination (version 4.5). Microcomputer Software. Ithaca, NY, 500 pp. Tunlid, A., White, D.C., 1992. Biochemical analysis of biomass, community structure, nutritional status, and metabolic activity of microbial communities in soil. Soil Biochemistry 7, 229–262. van der Putten, W.H., Maas, P.W.T., Van Gulik, W.J.M., Brinkman, H., 1990. Characterization of soil organisms involved in the degeneration of Ammophila arenaria. Soil Biology & Biochemistry 22, 845–852. van der Valk, A.G., 1974. Environmental factors controlling the distribution of forbs on coastal foredunes in Cape Hatteras National Seashore. Canadian Journal of Botany 52, 1057–1073. Vestal, J.R., White, D.C., 1989. Lipid analysis in microbial ecology: quantitative approaches to the study of microbial communities. Bioscience 39, 535–541. Vitousek, P.M., Howarth, R.W., 1991. Nitrogen limitation on land and in the sea: how can it occur? Biogeochemistry 13, 87–115. Waide, R.B., Willig, M.R., Steiner, C.F., Mittelbach, G., Gough, L., Dodson, S.I., et al., 1999. The relationship between productivity and species richness. Annual Review of Ecology and Systematics 30, 257–300. Wardle, D.A., 1992. A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soil. Biological Reviews of the Cambridge Philosophical Society 67, 321–358. Wardle, D.A., Nicholson, K.S., 1996. Synergistic effects of grassland plant species on soil microbial biomass and activity: implications for ecosystem-level effects of enriched plant diversity. Functional Ecology 10, 410–416. Westover, K.M., Kennedy, A.C., Kelley, S.E., 1997. Patterns of rhizosphere microbial community structure associated with co-occurring plant species. Journal of Ecology 85, 863–873. Wilson, J.B., Sykes, M.T., 1999. Is zonation on coastal sand dunes determined primarily by sand burial or by salt spray? A test in New Zealand dunes. Ecology Letters 2, 233–236.

T.K. Rajaniemi, V.J. Allison / Soil Biology & Biochemistry 41 (2009) 102–109 Yao, H., He, Z., Wilson, M.J., Campbell, C.D., 2000. Microbial biomass and community structure in a sequence of soils with increasing fertility and changing land use. Microbial Ecology 40, 223–237. Zak, D.R., Ringelberg, D.B., Pregitzer, K.S., Randlett, D.L., White, D.C., Curtis, P.S., 1996. Soil microbial communities beneath Populus grandidentata grown under elevated atmospheric CO2. Ecological Applications 6, 257–262.

109

Zeller, V., Bardgett, R.D., Tappeiner, U., 2001. Site and management effects on soil microbial properties of subalpine meadows: a study of land abandonment along a northsouth gradient in the European Alps. Soil Biology & Biochemistry 33, 639–649. Zelles, L., 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in the characterisation of microbial communities in soil: a review. Biology & Fertility of Soils 29, 111–129.