Fungal response from oat (Avena sativa) plants and surface residue in relation to soil aggregation and organic carbon

Abstract

The influence of soil fungi on soil organic carbon (OC) from surface residue was tested in outdoor plots in southern Ontario, Canada, 2004. Fungal hyphal length, soil aggregation, OC and light and heavy fractions of organic matter were compared with factors of plant growth (with or without oat [Avena sativa]) and surface residue (no residue, oat straw (low C:N) or corn (Zea mays) stalks (high C:N)) in a factorial arrangement. Significant increases were observed in soil OC from the oat plants, and from corn stalks compared to straw residue, in the growing season with very moist, high OC, sandy soil. In treatments with corn stalk residue, fungal hyphal length was increased with interaction from the oat plants and residue and was positively correlated with the heavy fraction organic matter along with soil OC. Fungal hyphae, plant roots and high C:N residue were all factors in soil OC increases.

Keywords:

• crop residue
• density fractions
• fungal hyphae
• macroaggregates
• mycorrhizas
• soil organic carbon

Introduction

The amount of crop residue left on the surface of the soil is considered as the most important variable in determining the potential to increase soil organic matter and thus organic carbon (OC) content (Clapp et al. 2000; Whilhelm et al. 2004). The practice of ‘no-till’ agriculture, which generates abundant surface residue, has been increasing since 1990 and is currently around 50% of cropping acreage on the Canadian Prairies (Statistics Canada 2006). There is a controversy over soil OC increases; studies in Western Canada have shown significant increases over time with no-till (Van den Bygaart et al. 2003), while studies in more humid conditions in Eastern Canada report varied results over the long term (Angers et al. 1997b).

In long-term field experiments that tested the effects of tillage, residue and fertilization on continuous corn (Zea mays) cropping, Clapp et al. (2000) found that OC increased with no-till, provided residue was returned to the surface. With various crops, residue treatments and soil types in field sites in Eastern Canada, soil OC increased in the surface 10 cm after 11 years of no-till, as compared to conventional tillage, although no significant changes occurred when OC was measured to 60 cm depth (Angers et al. 1997b). With no-till, the response of OC to carbon input is mainly in the surface 15 cm (Kern and Johnson 1993), whereas tillage can increase soil OC below 15 cm (Allmaras et al. 2004).

No-till agriculture allows greater fungal colonization, as fungi are the dominant decomposers of litter in aerobic conditions at the soil surface (Hu et al. 1995). Holland and Coleman (1987) found that a greater portion of ¹⁴C was retained in the soil when straw was surface applied rather than incorporated into the soil in a litterbag study in the field. Fungi also have slower decomposition time compared to bacteria (Parton et al. 1987) because of the 60% chitin in the cell walls (Langley and Hungate 2003) and the protein glomalin (Wright and Upadhyaya 1998). It has also been shown that slow turnover time maximizes the stability of microaggregates (<250 µm) (Tisdall and Oades 1982) and the transfer of macroaggregates (>250 µm) into stable OC (Besnard et al. 1996; Six et al. 2004).

Improved stability of OC and aggregates from increased fungal activity may explain increased conservation of OC in no-till (Six et al. 2006) as links have been established between fungal abundance and aggregation. The protein glomalin is thought to contribute to soil OC (Rillig and Allen 1999) from increased resistance of soils to decomposition (Rillig and Mummey 2006), and through increased soil aggregation (Rillig et al. 2002). In tall-grass prairie, Jastrow et al. (1998) found that mycorrhizal root colonization had the largest single effect on soil aggregation, followed by root length and microbial biomass. Reports on the interaction of mycorrhiza and decomposer fungi in field conditions and the influence of surface crop residue on growth of mycorrhizal mycelia are needed to elucidate the role of fungi in soil dynamics that can influence OC retention.

The light fraction (LF) of organic matter is recently arrived, partly decomposed plant material, and its OC content (OCLF) has been reported as representative of the amount and effect of active decomposition of residue (Janzen et al. 1992). The remaining heavy fraction (HF) of organic matter is composed of organic matter adsorbed to mineral surfaces within aggregates (Christensen 1992). The C:N ratio is thought to control the speed of decomposition where the amount of nitrogen available limits microbial growth. Conti et al. (1997) reported an increased decomposition rate with maize stubble residue added to soil with low C:N and OC levels, although no growing crop was included in the analysis. While speed of decomposition may be increased with LF, the retention of soil OC is variable. Studies where the decomposition rate varies inversely to soil OC content suggest recalcitrant residue associated with the heavy fraction organic matter, may result in greater increases in OC (Conti et al. 1997; Thomsen et al. 2008).

Oat (Avena sativa) was sown in outdoor plots with a factorial arrangement of plant (presence/absence) and surface residue (none/oat straw/corn stalks) treatments to test the relative effects of oat plants to residue, and comparative residue C:N of residue on fungal hyphal length, soil aggregation, soil OC and density fractions of organic matter. The hypotheses were: (1) plants and residue would both increase fungal hyphal length and their effects would be additive, (2) increased hyphal length would correspond to increases in soil aggregation, moisture and OC, and (3) with a lower C:N ratio, straw would increase hyphal length, OCLF and soil OC compared to corn residue.

Methods

Study plots

This study is the 2nd season of a three-year study using outdoor plots in the southern Ontario municipality of Clarington (44°03' N, 78°32' W) (Manns et al. 2007). A wooden frame was divided into 24 sections, each 60 cm square, and was filled with a 12 cm depth of soil by volume. The bottom of the plots had a perforated plastic lining to prevent upward movement of water and weed growth. The soil originated from the surface (0–10 cm) of a nearby garden where compost had been surface applied over 10 years without tilling on a base classified as ‘Brighton Sand’ (Manns et al. 2007). The particle size distribution of the soil was 89% sand (fine and very fine), 11% silt,<1% clay, with pH of 7.5 (electrode measurement in 2:1 soil:deionized water) and bulk density 1.15 g cm⁻³. Each year, the total volume of soil needed to fill the frame was mixed first to reduce variance in soil conditions and provide comparable conditions for each study. Initial OC averaged 69.0±6.0 g kg⁻¹ (mean±SEM) as detailed for OC sampling (see Soil testing).

Treatments

The six treatments were presence (OT) or absence (NP) of oat plants in factorial with no residue (NR), straw residue (ST) or corn stalk residue (CS), each replicated four times and randomly assigned to the 24 sections of the frame. Oat straw from a neighboring farm and corn stalks that had stood the winter in a nearby field were mulched coarsely in a chipper-shredder resulting in segments with random length up to 10 cm and diameter of 0.3–0.6 cm for straw and 0.3–1.5 cm for corn stalks. Mulch was placed on the soil surface of designated sections, lined with a plastic netting of 2 cm mesh that allowed the mulch to be easily lifted up and weighed at harvest. The initial weight of straw and corn mulch was 533–555 and 733–750 g m⁻², respectively, representing a 3–4 cm depth. This rate of addition represents 6.0 t ha⁻¹ for straw and 8.2 t ha⁻¹ for corn residue and could accumulate on the surface after two seasons of harvest with NT in southern Ontario from observation in local fields. The mulch rate assumes a decomposition rate of 50% per year that is slightly less than the rate of ¹³C incorporation found in soil aggregates in controlled incubation experiments following wheat straw decomposition (Angers et al. 1997a).

Several days following preparation of the plots, oats (A. sativa cv. Ida) were seeded (6 May 2004) by broadcasting seed over the mulch, or lightly raking the seeds into bare soil. Plots were kept weed-free by hand weeding. Oats were thinned to a normal field density of 333 plants m⁻² (Welch 1995). Although dry weather delayed germination, supplemental watering was not needed, as higher than normal amounts of rain fell during the rest of the season.

Sampling

Soil samples were collected during creation of the plots in May before the mulch and seeds were applied. A composite sample of approximately 250 g was formed from four independent samples through a 10 cm depth for each of the 24 plots. Sampling was repeated monthly until harvest (9 August 2004). In plots with oat plants, the soil sample was taken close to the oat roots without disturbing the plants. Half of each sample was promptly stored at 4°C and the other half frozen at −20°C. At harvest the plants were cut 2 cm above the ground, dried 24 h at 80°C and weighed. The lifted mulch was washed to remove adhering soil and dried for 24 h at 80°C prior to weighing. Plant roots were loosened gently and extracted from the soil and the mass recorded following washing and air-drying. A small sample of fine roots was removed for mycorrhizal analysis.

Soil testing

Soil moisture was determined by loss of mass after oven drying at 105°C and expressed as a percentage of dry weight (Page 1982). The oven dried soil samples were passed through a 0.5 mm sieve and soil OC was determined from duplicate samples by loss on ignition at 375°C for 16 h in a muffle oven (Ball 1964). The percentage OC was calculated from the proportion of the mass loss after ignition over the dry mass prior to ignition divided by 1.8, the proportion of organic matter to OC (Page 1982) and subsequently expressed in g kg⁻¹.

The aggregate size distribution was measured by shaking 5 g of air-dried soil through a nest of sieves, sizes 4, 2, 1, and 0.25 mm, for 15 min as described by Klute (1982). Each size sieve mass was recorded and the percentage macroaggregates calculated as follows: [mass of aggregates > 0.25 mm/total dry wt of sample (5 g)]×100

The mean weight diameter index (MWD) was calculated from the aggregate size distribution as the sum of the mean diameter of each weight class×the percentage mass of each size class of air-dried aggregates for all size classes (Klute 1982). The 1–2 mm size aggregates were further analyzed for wet aggregate stability. Each sample absorbed water slowly over 10 min and was immersed in water on a 1 mm sieve with a 5 s cycle time in 20°C water for 5 min (Klute 1982). The water stability of aggregates (WSA) was expressed as a percentage of the mass after drying at 105°C for 24 h over the initial dry mass. A soil sample was tested for an effect of sand particles. Since the sand particles were less than 0.5 mm in size and not retained on the 1 mm sieve, no correction was made for primary mineral particles (Black 1965).

Fungal hyphal length and mycorrhizal colonization

A soil suspension was made by combining 1 g of thawed soil with 100 ml autoclaved water at high speed in a blender for 15 s. The soil solution with suspended organic matter was decanted off from sand grains that remained in the bottom of the beaker 10 s after stirring and stored at 4°C. Measurements were adjusted for the moisture percentage in the soil where fresh/frozen soil was used for testing.

Hyphal length is a measure of the density of fungal growth in the soil (Kabir et al. 1997) and was estimated from hyphae stained with lactophenol aniline blue in a soil suspension. A constant volume of soil mycelia is considered dead (West 1988), inactive or simply not growing (Walker 1975). Phenolic aniline blue effectively colored viable hyphae, but not ‘ghost’ hyphae, and estimated hyphae length of in vitro fungal cultures (West 1988). The distinctly blue stain adheres to β-glucan linkages in fungal cell walls of living hyphae (Paul and Clark 1982). The length of each piece of blue hyphae was estimated in 10 ml of suspension with three drops of stain in the glass Petri dish with 1.2 cm gridlines under 20× magnification with a stereoscopic microscope. For each treatment, the hyphal lengths were summed for triplicate samples and averaged to give the total length in cm g⁻¹ soil dry wt. The length estimation of individual hyphae was confirmed by microscope photos.

Mycorrhizal colonization on oat roots was enumerated with root staining by ink/vinegar stain following an adaptation of the method of Vierheilig et al. (1998) as detailed in Manns et al. (2007). Evidence of hyphae, vesicles or arbuscules that occurred on each 1 cm piece on a single view of the microscope at 100× magnification was scored as a percentage occurrence of the total number of inspected pieces (approximately 40 per treatment) (Gryndler et al. 2002).

Density fractionation of organic matter

Light fraction organic matter is described as plant matter in the early stages of decomposition that is high in carbon, lipids and lignin (Lavelle and Spain 2001) and was found to be correlated to the microbial respiration rate in response to management inputs (Janzen et al. 1992). The heavy fraction (HF) is humic and mineral material or complexes that have a much higher density and thus can be separated from LF by specific gravity (Lavelle and Spain 2001).

A 5 g sample of field moist soil from each plot was wet-sieved to separate organic matter from sand. The organic matter was further separated into LF and HF with LUDOX^TM HS-30 colloidal silica (Sigma Aldrich, Oakville, Ont., Canada), 1.2 g cm⁻³ specific gravity (van den Pol-van Dasselaar and Oenema 1999), following the method of Meijboom et al. (1995). The LF and HF were calculated as the % dried mass of the fraction divided by the initial 5 g of whole soil adjusted for moisture content to dry weight. The LF and HF sediment were further tested for % OC content by weight as discussed previously for soil testing. The % OC in each fraction was expressed as OCLF and OCHF in g kg⁻¹.

Residue organic carbon and nitrogen

The straw and corn residue were assessed for OC, and organic nitrogen at planting and harvest sampling. The harvest residue was placed in 0.5 mm netting to contain all material for washing, dried at 80°C for 24 h and ground in a Wiley mill prior to analysis. Residue OC and organic nitrogen were determined with a LECO FP428 autoanalyzer using the Dumas method for nitrogen (Carter 1993).

Soil OC mass balance

To account for the changes in soil OC, a mass balance was computed to compare the amount of OC potentially available from plant photosynthesis based on published carbon fixation rates of Paul and Kucey (1981), to the actual changes in OC in the residues and soil in each treatment. To calculate the gross carbon from photosynthesis, the carbon fixation rate of 7.6 mg OC per g of shoot per hour (Paul and Kucey 1981) was applied to the fresh plant weight at harvest over an average time of 30 days to take into account reduced biomass during the early growing period and reduced photosynthesis during plant senescence.

The amount of carbon that could potentially have been accumulated in plant root growth, respiration and exudates was calculated from the allocation percentages of 19%, 28% and 0.5%, respectively (Paul and Kucey 1981). A shoot:root ratio of 2.5 was used to estimate root biomass, as this was the approximate ratio of the dried plant: root mass, and also the ratio observed for oats in field studies (Bolinder et al. 2007). The actual changes in OC from residue that could have contributed to soil OC were also calculated from the measured biomass and organic carbon and nitrogen content at harvest and planting (g dry wt). Fungal biomass was calculated from measured hyphal length over the season using 0.33 g cm⁻³ as a conversion factor that assumes 45% OC composition of fungi (Berg et al. 1998) and an average 14-day turnover time in the soil.

Statistics

Results of the 2×3 factorial treatments were examined with two-way ANOVA model (Statistica, Statsoft) to detect significant differences due to main effects of the plant (factor 1) – presence/absence and surface residue (factor 2) – no residue/oat straw/corn stalks. Significant differences among the three residue treatments were further tested with Fisher's Least Significant Difference (LSD) post-hoc test. Straw and corn residue results were combined for analysis as ‘residue’ where there were no significant differences among the individual treatments. Normality was ascertained with the Ryan-Joiner test. The Pearson product-moment method was used to assess correlation among the soil variables. Linear regression further analyzed the amount of variance explained by the dependence of one variable on another where biological associations were implied.

Results

Main effects

Surface residue had a significant effect on the % macroaggregates and soil OC at harvest sampling, while the oat plant effects on both variables were dependent on the plant growth stage (Two-way ANOVA, Table 1). Average rainfall over the growing season (350 mm) resulted in abundant growth of the plants and fungal hyphae. Soil moisture is not detailed further in this account.

Table 1. Two-way model ANOVA (oats×residue) p values of main effects of the oat plant and residue on soil variables along with interaction term. Post-hoc comparison among residue treatment means by LSD test; No residue (NR), oat straw (ST) with low C:N, corn stalks (CS) with high C:N.

	Two-way ANOVA			LSD comparison of means
	Plant	Residue	Interaction	ST-NR	CS-NR	ST-CS
July
Macroaggregates (%)	0.022*	0.173	0.365	0.069	0.518	0.218
Organic carbon (g kg^−1)	0.974	0.062	0.853	0.040*	0.040*	1.000
Fungal hyphal length (cm g^−1)	0.248	0.139	0.035*	0.250	0.051	0.376
August
Macroaggregates (%)	0.187	0.042*	0.714	0.031*	0.025*	0.920
Organic carbon (g kg^−1)	0.014*	0.045*	0.858	0.110	0.330	0.015*
Fungal hyphal length (cm g^−1)	0.007**	0.454	0.040*	0.371	0.232	0.753
May to August
Change in OC (g kg^−1)	0.343	0.020*	0.450	0.013*	0.950	0.015*
*and **denote ANOVA p values are marked significant at 0.05 and 0.01, respectively.

Fungal hyphal length

Fungal hyphal length averaged 27.6±2.7 cm g⁻¹ soil in the soil control throughout the experimental period. In all other plots, hyphal length increased to a range of 40–60 cm g⁻¹ soil through June and July. In August hyphal length then decreased to levels observed for bare soil in May, in all plots except with the combination of oat plants and residue where hyphal length was maintained or increased (Figure 1). There was a significant effect of the oat plant (p=0.007) to maintain hyphal length only when accompanied with interaction (p=0.04) of the plant with both straw and corn stalk residue (Table 1).

Figure 1.  Fungal hyphal length (cm g⁻¹ soil) measured over time in plots with treatments of no plant (NP) or oat plants (OT) and with no residue (NR), straw residue (ST) of low C:N, or corn residue (CS) of high C:N from initial planting (May) to harvest (August) 2004. May figures were averaged across all treatments. Error bars represent SEM (n=4).

Soil aggregates

The oat plants significantly decreased aggregate size (p=0.02) in July (Figure 2). The % macroaggregates was subsequently increased by both straw (p=0.03) and corn stalk (p=0.025) residue in August over the no residue control in plots with and without oat plants. There was a significant correlation of % macroaggregates to hyphal length only in treatments with both corn stalk residue and oat plants (r ²=0.99, p=0.006, n=4).

Figure 2.  Percentage of soil macroaggregates measured over time in plot treatments with no plants (NP) or oats (OT) and with no residue (NR), straw residue (ST) with low C:N, or corn residue (CS) with high C:N from initial planting (May) to harvest (August) 2004. May figures were averaged across all treatments. Plots marked A are significantly greater (p=0.05) than other treatments at August sampling, and B are significantly less (p=0.05) than other treatments at July sampling (n=4) by two-way ANOVA.

The water stability of the aggregates (WSA) declined from May to June (t-test, p<0.001) from an average of 82.4±3.5% to 55.3±1.4%, which was attributed to the mixing of the soils in the creation of the plots. From June to August WSA increased to an average of 76.9±1.3% in all plots which approached the May value, but treatment affects were not significant (data not shown). The linear regression of %WSA = 0.14×hyphal length + 71.27 (r ²=0.24, p=0.015, n=24) represented the strongest relationship between two variables observed over all plots. The weighted distribution of the aggregate size classes (MWD), calculated from the August distribution of soil aggregates, was correlated to soil OC (r ²=0.41, p=0.023) and hyphal length (r ²=0.38, p=0.03) over all treatments with oat plants (n =12).

Organic carbon (OC)

Soil OC increased in all plots over the three-month growing season, with the greatest increase in the last four weeks (Figure 3). Soil OC in plots with residue (with and without oat plants) was increased significantly with straw (p=0.04) and corn stalk (p=0.04) residues compared to no residue by LSD comparison of means at July sampling (Table 1). There was a significant positive effect of the oat plants (p=0.01) on soil OC at August sampling. In addition, plots with corn stalk residue had significantly greater soil OC than with straw residue (p=0.045) over all treatments, while the soil plots were intermediate in soil OC between straw and corn stalk residue. When the changes in soil OC from May to August with residue and oat plants were adjusted for the changes in the no residue plots, there was a significant reduction in soil OC with straw residue compared to no residue (p=0.01) and to corn stalk residue (p=0.02) with and without oat plants (Table 2).

Figure 3.  Soil OC (g kg⁻¹dry soil) comparison over time in plots with no plants (NP) or oats (OT) and with treatments of no residue (NR), straw residue (ST) with low C:N or corn residue (CS) with high C:N from initial planting (May) to harvest (August) 2004. May figures were averaged across all treatments. Error bars represent SEM (n=4).

Table 2. Means and SEM in g m⁻² of plant and root biomass, and estimated contributions to changes in soil OC from May to August 2004 in each treatment with oat plants with residue treatments; no residue (NR), straw (ST) of low C:N or corn stalk (CS) with high C:N.

		Oat plant plots
	NR	ST	CS
Soil OC mass balance	g m^−2	g m^−2	g m^−2
Mean harvest plant biomass (fresh wt)	1959	1519	1707
SEM	(206)	(47)	(317)
C fixation per hr (7.6 mg g^−1 shoot h^−1)*	15	12	13
C fixation 12 h day^−1	179	139	156
C fixation 30 days	5360	4156	4669
Mean harvest root biomass (S:R 2.5)	784	608	683
Root biomass C (19% of fixed C)	1018	790	887
Amount decomposed (50%)	509	395	444
Root exudates (0.5% of fixed C)	27	21	23
Mycorrhiza (1% of fixed C)	54	42	47
Total OC from plant to soil	590	457	514
OC loss from residue		122	141
OC assimilated from residue (50%)		61	70
Total OC to soil	590	518	584
Total change in soil OC (May to August)	1625	611	944
SEM	(184)	(305)	(205)
Changes in soil OC in addition to changes in soil alone
Oat plants (OT)	467	−414	293
No oat plants (NP)	0	−322	145
*Carbon fixation and allocation rates by Paul and Kucey (1981); Fixation rate: 7.6 mg of OC per gram shoot per hour; Root allocation: 19% biomass, 28% respiration, 0.5% exudates.

Residue decomposition and carbon mass balance

The initial C:N ratios for straw and corn were 31 and 74, respectively. There was no difference in the total residue mass loss (g m⁻²) or residue OC loss (g m⁻²) between straw and corn stalk residue (Table 3). However, by percentage mass, corn stalks had less residue loss (p<0.001%), OC loss (p=0.001) and % nitrogen loss (p<0.001) than straw residue.

Table 3. Comparison of means±SEM of residue decomposition by loss of mass, OC and Nitrogen in % and mass (g m⁻²), with no plants (NP) or oat plants (OT) between straw (ST) with low C:N and corn (CS) with high C:N. Two-way model ANOVA p values give a comparison of variables between straw and corn residue (n=16).

		NP-ST		NP-CS		OT-ST		OT-CS		ANOVA
Plant/residue		Means	SEM	Means	SEM	Means	SEM	Means	SEM	p
Mass residue loss	g m^−2	213.50	10.22	214.08	19.28	236.47	5.78	224.61	23.42	0.73
OC loss	g m^−2	115.22	8.50	116.25	6.31	122.00	5.86	140.61	18.83	0.40
Nitrogen loss	g m^−2	4.00	0.19	0.78	0.08	3.92	0.28	0.92	0.14	<0.001***
Percentage
Residue loss	%	49.47	2.12	33.98	3.09	53.92	0.95	35.41	3.59	<0.001***
OC loss	%	58.40	4.03	38.23	2.07	60.89	2.85	45.89	5.84	<0.001***
Nitrogen loss	%	64.63	2.90	19.56	2.41	62.10	3.83	22.08	3.82	<0.001***
Residue CN		37.39	2.08	56.95	2.67	33.24	2.31	51.95	6.37	<0.001***
***p significant at 0.001.

Surface residue decomposition, plant roots, root exudates and fungal mycelia all contribute to increases in soil OC. The OC loss from residue averaged 123.5±9.9 g m⁻² (Table 3), which could represent a gain of 60 g m⁻² OC in the soil (< 10% of soil OC increase) after allowing for 50% loss of OC in microbial respiration (Table 2). The fresh root biomass of the oat plants at harvest was estimated to be from 600–800 g m⁻². Root biomass OC could range up to 1000 g C m⁻² based on the OC fixation input from harvested shoot weights in this study (Table 2). Calculated plant root biomass could account for 76% and 86% of the potential additions to soil OC for treatments, with and without residue, respectively. Root exudates based on the oat plant biomass would be less than 30 g m⁻² (< 3%). Fungal biomass calculated from hyphal length in our study could not have contributed more than 7 g m⁻² (< 1%) OC over the season.

Organic matter density fractions

There were no significant differences between treatments in %LF and %HF or the OC in each fraction. The OCLF and OCHF averaged 13.4 and 45.0 g kg⁻¹, respectively. There was little change in the average %HF between sampling times (31.1±4.0% and 29.6±3.1% in July and August, respectively). However, average %LF decreased (p<0.001) from 14.5±2.2% in July to 7.2±1.2% in August. In plots with oats in corn stalk residue (n=4), there was negative correlation between LF and HF (r ²=0.96, p=0.015). The %LF was negatively correlated to hyphal length (r ²=0.94, p=0.03) and % macroaggregates (r ²=0.89, p=0.05), while %HF was positively correlated to both variables. There were significant linear regressions of %HF on hyphal length (Hyphal length = 0.0648×%HF + 5.9518, r ²=0.78, p=0.004) and OC (OC = 0.0648×%HF + 5.9518, r ²=0.63, p=0.02) in all treatments with corn stalk residue (n=8) (Figure 4).

Figure 4.  Linear regression of heavy fraction organic matter, expressed as a % of total dry soil mass on hyphal length (g m⁻²) (?) y = 6.0583x - 128.42, r ²=0.7851, p=0.89 and soil organic carbon (g kg⁻¹) (?) y = 0.0648x + 5.9518, r ²=0.6301, p=0.79 in all plots with corn stalk residue (with and without oat plants) at harvest (n=8).

Discussion

Significant effects

The study highlighted the different effects of surface residue and plant roots on major variables (hyphal length, WSA, aggregation and soil OC), and with plant growth stages. Surface residue significantly increased soil OC in July, while at harvest the dried oat plants had a greater positive effect on soil OC. There was a significant interaction term between residue and plant on fungal hyphal length at both sampling times suggesting that an interaction of mycorrhizal and saprophytic fungi may have contributed to the increases in soil OC. Variable effects of saprohytic fungi and arbuscular mycorrhiza interactions have been observed in previous studies (Fracchia et al. 1998), although changes were only related to germination, hyphal growth and community species composition of fungi. Interaction (Milleret et al. 2009) and additive (Andrade et al. 1998) effects of mycorrhiza and plant roots have been connected to the stability of aggregates but no published literature was found on links of fungal interaction to soil OC.

Hyphal length

The separate effects of saprophytic fungal hyphae and mycorrhizal hyphae were not varied directly in this study, but were separated by treatments, by their growth pattern over the season and by diameter observed in the previous study. The types of fungi were separated within plot treatments with no plant plus residue (saprophytic) and with oat plants plus no residue (mycorrhizal and saprophytic). Plots with oat plants plus straw and corn residue showed a sustained increase in fungal hyphal length from May to August, 2004, whereas plots with only residue or plants, experienced high fungal hyphal length in July that decreased to that of the base level of bare soil at harvest sampling.

The changes in hyphal length could also indicate a change in the type of fungi, with fast growing saprophytic fungi dominant in June and July, and slow growing mycorrhizal fungi dominant at August sampling. Saprophytic fungal growth is at a maximum with warm temperature, which corresponds to the highest fungal hyphal length recorded in July in this study. A high population of very small diameter hyphae (< 5 µm) was also found in July in our previous study for which oats were grown with and without straw surface residue in the same soil, resulting in ratios of hyphal fragments by diameter (< 5 µm:>5 µm) of 80:20 in July and 45:55 in August (Manns et al. 2007). Assuming large hyphae were mycorrhizal (Ayres and Boddy 1986; Schreiner and Bethlenfalvay 2003) an increase in mycorrhizal hyphae, which are more resistant to decomposition and which grow at a slow and steady rate over the season, would correspond to a peak of mycorrhizal growth in September previously observed in studies (Biederbeck and Campbell 1971; Kabir et al. 1997).

Our observed interaction of oat plants and residue on fungal hyphal length suggests there is greater total hyphal length when mycorrhizal and saprophytic fungi co-exist in a complex environment. Assuming that 50% of hyphae were from mycorrhizas in August, as in the previous season (Manns et al. 2007), the significantly higher hyphal length, which only occurred with interaction of residue and oat plant, may represent a synergy of mycorrhizal and saprophytic fungi or beneficial effects of surface residue on mycorrhizal growth. Current literature using isolated microorganisms in controlled studies indicates mycorrhizal and decomposer fungal interactions can be either positive or negative, depending on the species (Dighton et al. 1987).

Klironomos and Kendrick (1995) found increased mycorrhizal fungal hyphae in the presence of saprophytic decomposers of surface litter in simulated forest microcosms resulting from reduced predation by soil microarthropods on mycorrhizal hyphae. Milleret et al. (2009) found that earthworms reduced fungal biomass and noted that grazing on fungal hyphae can either stimulate growth or reduce fungal biomass. A recent greenhouse study by Ortiz-Ceballos et al. (2007) reported that mulch decreased arbuscular mycorrhizal fungi, unless worms were also present. Mycorrhizal fungal hyphae (Boddy et al. 1989) and plant roots (Dommergues and Krupa 1978) act as sinks for ammonium that is released from saprophytic fungi after decomposition of the residue. Improved nutrient cycling in the soil food web from surface residue could also explain the increased growth of mycorrhizal and saprophytic fungi in concert.

Mycorrhizal colonization

Mycorrhizal fungal colonization on the oat roots averaged 88%, but no significant treatment effects were evident. The colonization of mycorrhizal fungi on oat plant roots may not correspond directly to hyphal length (Schenck 1982) since mycelial development varies with species and soil conditions (Bethlenfalvay and Linderman 1992). The calculated OC that could be retained in the observed length of fungal hyphae was only 1% of the increase in soil OC, so mycorrhizal hyphae could only be an indirect factor in increasing soil OC.

Mycorrhizal hyphae have been found to contain a protein glomalin which is water-insoluble and heat stable (Rillig et al. 2002) and contains a high percentage OC (Rillig and Allen 1999). Rillig and Mummey (2006) maintained mycorrhizas influence root and shoot growth directly and soil aggregates indirectly from altered soil water and rhizodeposition. Rillig et al. (2002) found an indirect effect of hyphae on glomalin and root length that explained significant amounts of variation in WSA, while the direct effect of hyphae on WSA was minimal.

Glomalin can be linked to soil OC indirectly from its association with increased aggregation (Rillig et al. 2002) and through slow decomposition of fungal hyphae (Rillig et al. 2003). A large greenhouse study comparing five mycorrhizal species and nine plant species with 10 replications found significant interaction of plant and fungal species on increasing 1–2 mm diameter water stable aggregates and a positive correlation of hyphal length with %WSA (Piotrowski et al. 2004). The plant roots had high density, although glomalin-related soil protein was not measurable because of high background levels in the soil. The interaction effects of mycorrhiza with plant roots may arise from effects on root physiology such as reduced root weight, specific root length and fineness (Rillig et al. 2002) or modification of root exudates (Fracchia et al. 1998).

Soil aggregation

Fungal hyphal length significantly correlated to the % macroaggregates only in the treatment of oat plants with corn stalk residue with limited sample size (n=4). However, the MWD expressed a positive correlation of aggregate size to hyphal length and OC over all plant treatments that is in agreement with previous reports that hyphal length increased OC in association with increased aggregate size (Hu et al. 1995; Beare et al. 1997; Angers et al. 1997b). Fungi have been associated with increased soil aggregation in grasslands (Tisdall and Oades 1979) and with winter rye (Lolium perenne) planted into sorghum residue in NT (Beare et al. 1997) for which there is a high abundance of mycorrhizal fungi.

Oats have a lower amount of mycorrhizal association compared to grasses, winter rye or corn, which may have contributed to the lack of correlation in the other treatments between hyphal length and% soil macroaggregates in this study compared to reported literature. Bossuyt et al. (2001) found that saprophytic fungal biomass from wheat straw decomposition was essential to development of macroaggregates in 14 days in a microcosm study. The authors also suggested that only a threshold amount of fungal hyphae was necessary for aggregate formation and this may relate to the absence of a correlation in our high organic matter soil.

The significant correlation of fungal hyphal length to WSA over all plots indicates that fungal hyphae were a significant factor in the resistance of aggregates to decomposition at harvest sampling. This supports the work of current authors on a strong relationship between mycorrhizal hyphal length and WSA. Van der Heijden et al. (2006) tested varied species of mycorrhizal inoculant and seedlings with sterile high quality soil in microcosms and found mycorrhizal colonization increased soil aggregation and %WSA. Andrade et al. (1998) ascertained a close relationship of fungal hyphal length to %WSA in controlled microcosm experiments growing Sorghum bicolor for 10 weeks with 43 µm mesh dividing hyphae from roots of S. bicolor. The relationship to WSA was highest with roots plus mycorrhiza, followed by roots alone, mycorrhiza alone, and the soil control in that order (Andrade et al. 1998). Piotrowski et al. (2004) also found a positive correlation of %WSA to hyphal length and root biomass researching the difference in five arbuscular mycorrhizal species and nine plant species in a greenhouse study over 11 months. In a microcosm experiment, Milleret et al. (2009) found no effect on water stability of aggregates individually from individual effects of mycorrhiza Glomus intraradices and roots of Allium porrum, but WSA increased significantly when there was significant interaction of mycorrhiza and roots.

The decrease in the percentage of macroaggregates in this study from the oat plants in July, followed by an increase in aggregate size from residue in August, is consistent with the effect of high exudation rates early in the growing season, which diminish as the plant dries (Whilhelm et al. 2004). The action of an abundance of roots limits soil aggregation as the disturbance and bacterial action stimulate decomposition (Jastrow et al. 1998). At plant senescence, root exudation stops and is replaced by saprophytic fungal decomposers of root material (Wamberg et al. 2003) that would contribute to the increases in hyphal length and OC from the oat plants.

Soil OC

The increase in soil OC from the oat plants may be related to high root mass stimulated by the high rainfall. Plants roots have been shown to increase OC in controlled studies. In a pot study under growth chamber conditions, Gale and Cambardella (2000) found that a greater amount of ¹⁴C was retained in the soil from oat roots (42%) compared to oat leaves left to decompose on the surface (16%). Those authors suggested that a portion of root OC is rapidly released into the soil after plant senescence, followed by increases in stabilized, protected, slow decomposition forms (Gale and Cambardella 2000). The authors observed rapid initial decomposition with 80% of coarse roots decomposed within 90 days. Increases in SOC across a chronosequence of prairie restoration had the closest association to very fine roots (Jastrow et al. 1998), possibly from mycorrhizal associations. Therefore, it would be possible for a large volume of very fine roots from the oat plant growth to turnover during the 90-day growing season and be measurable in the soil OC by harvest. This could explain the plant effect on soil OC in our very moist conditions, where soil OC increased primarily between July and August sampling.

From our OC mass balance calculations, plant root biomass accounted for the dominant portion of OC additions, with a small difference accounted for by residue (< 10%), root exudates (< 3%) and fungal hyphae (< 1%). Thus, plant roots and residue could explain the OC increases in oat plant plots after adjustment for changes in the soil controls. However, the increases in soil OC without oat plants of up to 1000 g m⁻² must represent another factor that is not accounted for in our study. It is possible that soil algae were capable of fixing up to 127.4 g m⁻² OC in conditions with high levels of phosphorous, organic matter (Lavelle and Spain 2001) and soil moisture that occurred in the four weeks preceding harvest.

Residue C:N

Plots with corn stalk residue had higher soil OC compared to plots with straw residue despite a higher C:N than straw residue and no difference in the amount of OC (by mass) released from both residues to the soil. Corn stalk residue also has a higher lignin content than straw that could be stabilized directly into slowly decomposing organic matter pools (Carter and Stewart 1996) represented by HF in this study. Our results are in contrast to Conti et al. (1997) who reported that lower C:N in corn stubble from fertilization resulted in an increased decomposition rate, microbial activity and OCLF, without change in humic acids or total C and N. Their study concluded that higher OCLF also mineralized more rapidly, and did not contribute to the synthesis of stable forms such as humic acids, that may involve lignin or phenolic polymers (Conti et al. 1997). In our study, the positive correlation of HF to hyphal length and OC in the case of corn residue was not found with straw residue suggesting the recalcitrance of the residue to decomposition was more important than the amount of residue recently decomposed as LF in maintaining aggregation and OC in the sandy soil and wet conditions.

Density fractions

Heavy fraction represents slow decomposition into more stable forms rather than the LF commonly associated with plant root exudates or residue decomposition that are more labile. The contrasting trend observed between %LF (negative) and %HF (positive) correlations to hyphal length and macroaggregates in plots with oat plants plus corn stalk residue, indicates that the transfer of fungal hyphae, corn residue and/or oat roots into macroaggregates and soil OC was more closely associated with HF than LF. This agrees with Grandy and Robertson (2007) who found heavy fraction (>1.6 g cm⁻³) had the highest OC accumulation in the largest aggregate size class when examining soil from the top 0–5 cm in varied ecosystems. Increases in soil OC in agricultural systems were highest with perennial alfalfa, followed by no-till (55% and 43%) over 12 years, with average OC accumulation in the average range of 30 g C m⁻² g⁻¹ for the mid-western United States, and HF accounting for 82% of increased OC with no-till (Grandy and Robertson 2007). In contrast, other reports found increases in %LF from mulching (Hassink 1995; Bending et al. 2000), and correlation to high organic matter (Six et al. 2004). Our study may vary from published data due to the density separation at 1.2 g cm⁻³.

The use of high organic matter sandy soil, conducive to decomposer and mycorrhizal fungal growth, and high rainfall to drive root growth and decomposition provided ultimate conditions, along with low variance, to obtain measurable increases in soil OC. The monthly changes verify that the soil OC slowly increased to a peak at the end of the oat season, where root decomposition was at a maximum. The increases in soil OC support the previous season of this plot study, where only with surface applied residue, soil OC increased in one season (Manns et al. 2007).

In a recent review of literature, Six et al. (2006) maintained that conservation of OC in no-till was due to fungal mediated aggregate stability with increased resistance to decomposition. This theory is supported by our study results with corn residue, and we suggest that the interaction of plant roots and mycorrhiza was enhanced by surface residue resulting in increased stability of aggregates and soil OC. Understanding the role of fungi in soil OC and determining a sustainable amount of surface residue for no-till agriculture are important factors for addressing climate change.

Conclusions

In contrast to field scale studies that show soil OC only increases in the long-term with surface residue, this study found significantly greater soil OC with corn stalk residue compared to straw residue surface applied over the growing season of the oat crop. The correlation of %HF with hyphal length and soil OC with corn residue suggested slower decomposition of mycorrhizal hyphae, corn residue and oat roots were factors in the increased soil OC. Soil OC increased with corn stalk residue compared to straw residue, although corn stalks have a higher C:N than straw, and the amount of OC decomposed or measured in the LF of soil was similar in both treatments. Soil aggregation was increased from both straw and corn stalk residue after seed set of the oat plants but the oat plants were more important than residue for the increase in soil OC at plant senescence. The increased fungal hyphal length with interaction of the oat plants and surface residue could be explained from prolonged growth of saprophytic and mycorrhizal hyphae from the plant and residue environment together. Our results indicate plant roots combined with high C:N residue on the soil surface results in the greatest support for soil OC in the effort to reduce atmospheric CO₂ through agricultural management.

Acknowledgements

The support of the Symons Trust Fund for Canadian Studies at Trent University is gratefully acknowledged for making this study possible. The assistance of Professor Bev Kay, Land Resource Science, University of Guelph, on the study design and analysis and manuscript editing was very much appreciated.