| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Crop and Soil Sci. and Dep. of Microbiology, Oregon State Univ., Corvallis, OR 97331-3804 USA
bottomlp{at}ucs.orst.edu
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Abbreviations: F, fumigated soil • FDA, fluorescein diacetate • SMBC, soil microbial biomass carbon • SOM, soil organic matter • TKN, total Kjeldahl N • TOC, total organic carbon • UF, unfumigated soil
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
A better understanding is needed of how winter cover crop residues may contribute to accretion of SOM and development of soil structure, while, at the same time, how they can be mineralized in amounts sufficient to reduce the inorganic fertilizer N requirements of the subsequent summer crop. As a prelude to this study, we hypothesized that changes in soil structural properties brought about by the use of cover crops might change the interactions between microorganisms and their substrates in the soil fabric. We hypothesized that cover crop-induced soil structural changes might be reflected in the size distribution of soil aggregates, and that heterogeneous distribution of microbial biomass and associated activities among the size-classes of aggregates might allow us to identify sites where activities and biomass levels correlate directly, as well as sites where biomass is high but activities are low. With this information in hand, we would be in a position to begin dissecting, at the microsite level, the soil structure–nutrient availability–predation complex that controls growth and turnover of soil microbial communities.
Several studies have shown that soil microorganisms and their activities are heterogeneously distributed across aggregate-size classes (Gupta and Germida, 1988; Seech and Beauchamp, 1988; Beauchamp and Seech, 1990; Vargas and Hattori, 1990; Miller and Dick, 1995a, 1995b; Franzluebbers and Arshad, 1997). While some studies have shown greater microbial biomass and higher activities in macroaggregates compared with microaggregates (Gupta and Germida, 1988; Miller and Dick, 1995a; Franzluebbers and Arshad, 1997), others have reported similar and even higher levels of activities in microaggregates compared with macroaggregates (Seech and Beauchamp, 1988; Miller and Dick, 1995b; Jastrow et al., 1996).
The objectives of the present study were to examine the influence of a legume [red clover (Trifolium pratense L.)] and a nonlegume [triticale (x Triticosecale Wittmack)] cover crop on soil aggregation and on the spatial distribution of microbial biomass and associated activities among the different aggregate-size classes at seedbed preparation and at the end of the growing season.
|
Materials and methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Field treatments included three winter cover crop treatments in a vegetable crop rotation that alternates two summer crops, sweet corn (Zea mays L. cv. Jubilee) and broccoli (Brassica oleracea L. Botrytis group cv. Gem). The three winter cover crop treatments were: (i) winter fallow (no cover crop), (ii) red clover (cv. Kenland), and (iii) triticale (x triticosecale Wittmack cv. Celia). The cover crops were established by the relay method, in which they were seeded (85 kg ha-1 for triticale and 25 kg ha-1 for red clover) under the summer crop in late July to take advantage of irrigation. The strategy behind the relay procedure was to have the cover crops established prior to the onset of fall rains. Hereafter, these treatments will be referred to as fallow, legume, and cereal, respectively. Each year, cover crops were incorporated into the soil by rototilling in mid April, followed by seedbed preparation in early mid May, and the summer crop of sweet corn or broccoli was planted mid to late May. Weed control involved preplant applications of 2.24 kg ha-1 EPTC (S-ethyldipropylthiocarbamate) for sweet corn; no herbicide was used for broccoli.
Experimental Design
The experimental design was a randomized complete block split-plot with four replications. The plots were 18 m x 9 m with winter cover crop treatments as the main plots and N rate as the subplots. The main plots were divided into three equal subplot areas of 54 m2. Nitrogen fertilizer (urea) rates were zero, medium (56 and 140 kg of N ha-1 for sweet corn and broccoli, respectively), and recommended (224 and 280 kg of N ha-1 for sweet corn and broccoli, respectively). Because of manpower and budgetary constraints, soil samples were taken only from the intermediate N rate subplots.
Soil Sampling
In early June 1995 and 1996 (at seedbed preparation), and in September 1995 (after broccoli harvest) and September 1996 (after sweet corn harvest), soil samples were collected from each replicate plot. Approximately 10 samples of soil were taken with a shovel to a depth of 20 cm at 1-m intervals along two transects within each replicate. A composite sample of the soil from each replicate was prepared, and the samples were transported to the laboratory in large plastic storage bags. Soil was sampled to 20-cm depth because seedbeds were prepared by tillage to that depth. June samples were taken to evaluate soil properties at the time of seedbed preparation, prior to growing the summer crop. The September soil samples were taken immediately prior to the onset of fall rains to evaluate soil properties after the summer growing season.
Soil Sieving and Aggregate Distribution
Before sieving, large soil clods in field-moist soil samples were gently crushed by hand. To reduce the impact of air drying on microbiological properties of the soil, it was laid out on brown paper in a cold room at 4°C for 5 to 7 d. After this period of drying, the soil had reached a gravimetric water content that ranged between 100 and 140 g kg-1. The field moisture content of the samples collected in 1995 and 1996 ranged between 120 and 205 g kg-1.
Aggregates were prepared by placing portions (200 g) of the cold air-dried soil on nested sieves (20-cm diam.) mounted on a Tyler Ro-Tap sieve shaker (Combustion Engineering Inc., Mentor, OH). The shaker operates at one speed, producing approximately 200 to 250 oscillations min-1. Soil was sieved for 3 min into the following aggregate-size classes: <0.25, 0.25 to 0.5, 0.5 to 1.0, 1.0 to 2.0, and 2.0 to 5.0 mm. All samples of aggregate-size classes were stored at 4°C. Results of preliminary experiments indicated that sieving for 3 min was sufficient to separate the different aggregate-size classes (data not shown). The aggregates retained in each sieve were weighed. The percentage of soil in each aggregate-size class was calculated after excluding the material >5.0 mm. Soil collected in June 1995 was used only for determination of aggregate size-class distribution. In September 1995, the microbial biomass and activities measurements were conducted on composited aggregate samples that were prepared by mixing equal amounts of aggregates from the same aggregate size-class prepared from each replicate plot of a given field treatment. The aggregate preparations produced from soil recovered in June and September 1996 were treated separately as field replications. Aggregates and whole soil samples were stored in polyethylene bags at 4°C for 
7 d prior to determinations of biomass and readily mineralizable C and N. Enzyme assays were conducted on aggregate samples within 2 to 3 wk of preparation.
Soil Chemical Analyses
Whole soil and aggregate samples were ground with a mortar and pestle to pass a 0.25-mm sieve. Total organic C was determined by direct combustion to CO2 in a DC-80 Dohrmann carbon analyzer (Dohrmann Inc., Santa Clara, CA) equipped with an infrared detector. Total Kjeldahl N (organic N and NH+4) was determined as described by Bremner and Mulvaney (1982). Total organic C was determined for soil samples collected in September 1995 and June and September 1996. Total Kjeldahl N was determined only for soil samples collected in June and September 1996.
Soil Physical Analyses
Particle-size analyses were conducted on five aggregate-size classes and whole soil samples recovered from the fallow and legume treatments (June and September 1996) using the pipette method (Gee and Bauder, 1986). All samples were analyzed without any pretreatment for removal of organic matter.
Microbiological Analyses
With the exception of September 1995, when readily mineralizable N was not determined, the following properties were measured for whole soil and soil aggregate-size classes.
Soil Microbial Biomass Carbon
Soil microbial biomass C was determined by the chloroform–fumigation–incubation method as described by Horwath and Paul (1994). Plant debris and roots were meticulously removed from the aggregate samples before determination of SMBC. The water content of dried soil samples (20 g) was raised with distilled water to 280 g kg-1 soil, and samples were incubated in the dark for 4 d at 25°C. One half of the samples were then fumigated (F) under vacuum at room temperature for 48 h in desiccators containing 20-mL volumes of ethanol-free chloroform (HPLC grade, Mallinckrodt, St. Louis), while the unfumigated controls (UF) were kept at 25°C. After fumigation, chloroform was removed by repeated evacuations and flushings of the desiccators, F and UF soil samples were transferred to canning jars (500-mL volume) containing a scintillation vial with 10 mL of 0.3 M KOH. The jars were incubated in the dark for 10 d, and the amount of CO2–C evolved was determined by titration. Soil microbial biomass C was determined from the difference between CO2–C evolved by the F and UF controls multiplied by a correction factor of 2.44. Fumigated soil (pH 6.1–6.3) was not reinoculated with nonsterile soil because preliminary studies indicated that reinoculation was not necessary.
Readily Mineralizable Carbon
Readily mineralizable C was calculated as the amount of CO2–C evolved by the UF controls of the SMBC determinations.
Readily Mineralizable Nitrogen
Readily mineralizable N was calculated from the difference between the amounts of NO3–N found in the UF controls of the SMBC determination after incubation and the initial amounts found in those samples before the addition of water (time zero). The amounts of NH4–N found in the samples before and after incubation were always below the limit of accurate detection (<2.0 mg NH4–N kg-1 soil), and they were not considered in the calculations of the total amount of N mineralized.
Enzyme Activities
ß-Glucosidase activity and FDA hydrolysis were determined for the soil samples collected in September 1995 and in June and September 1996. ß-Glucosidase activity was determined by the method of Tabatabai (1994). FDA hydrolysis was determined by a modification of the procedure described by Zelles et al. (1991). One gram of soil and 20 mL of a 60 mM sodium phosphate buffer (pH 7.6) were placed in a 125-mL Erlenmeyer flask and shaken at 100 rpm on a rotary shaker at 25°C. After 15 min, 100 mL of a 4.8-mM FDA solution were added, and the suspension was shaken for an additional 1.75 h. The reaction was stopped by addition of 20 mL acetone. The suspension was centrifuged for 5 min at 4300 x g, filtered, and filtrate absorbance was measured spectrophotometrically at 499 nm.
Results of all microbiological, chemical, and physical analyses are reported on an oven-dry-weight basis, determined by drying the soils for 24 h at 105°C.
Statistical Analysis
Data were analyzed using the SAS statistical package (SAS Inst., Cary, NC). Within each season the data were analyzed using a repeated measures analysis of variance, aggregate size being the repeated term (SAS Inst., 1989). The appropriate covariance structure was determined using a sphericity test. This test determines whether it is acceptable to treat the data as a univariate analysis, and as a result, whether the experiment can be modeled as a split-plot. For those parameters that did not satisfy the sphericity-test, significance levels were determined using the Greenhouse and Geiser adjustment (Freund et al., 1986). Because the soil samples collected in September 1995 were composited, the structure of the repeated measures ANOVA was different (the block effect was not considered in the analysis). In each sampling time, whole soil samples were analyzed as a sixth level in the repeated term.
For the purpose of comparing sampling time, only data collected in June and September 1996 were considered. Data collected in September 1995 were inappropriate because replicates had been composited. A repeated measures factors analysis of variance was used, the two repeated terms being sampling time (with two levels, June and September) and aggregate size (with six levels, five corresponding to each aggregate-size class plus one whole soil sample). Sources of variation included winter cover crop treatment, aggregate size, sampling time, and their interactions. Main effects means were separated using a
.
At each sampling time, covariance analyses were used to model the relationships among the variables evaluated. Individual analyses using the SAS mixed-models procedure (SAS Inst., 1995) were conducted using one of the variables as the covariate and the other as the response, after accounting for the design structure (treatments, blocks, aggregate size).
|
Results |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
|
|
|
|
|
Readily Mineralizable Carbon
Not only did whole-soil readily mineralizable C levels change significantly (P < 0.05) as a function of cover crop treatment in June and September 1996 (Table 1), the levels of readily mineralizable C also differed significantly (P < 0.001) as a function of aggregate size at both sampling times. In September 1995, the lowest level of readily mineralizable C was found in the 2.0- to 5.0-mm aggregate-size class, regardless of treatment (Fig. 2)
. Furthermore, significant differences were observed among <2.0-mm size classes of the fallow and cereal treatments.
|
0.01). In June 1996, all aggregate-size classes >0.5 mm recovered from the fallow treatment showed significantly greater levels of readily mineralizable C than aggregates <0.5 mm, whereas in September 1996, no significant differences were found in the C levels across aggregate-size classes recovered from the fallow treatment (Fig. 2). In the case of the legume cover crop treatment, the 0.5- to 1.0-mm size class showed significantly more readily mineralizable C than other size classes in June (Fig. 2), whereas in September, significantly greater levels of readily mineralizable C were found in the size class 0.5 to 1.0 mm of the cereal and in both the 0.5- to 1.0- and 1.0- to 2.0-mm size classes of the legume treatment.
Readily Mineralizable Nitrogen
Statistically significant differences were observed in the amounts of readily mineralizable N in whole soil samples from the three treatments, and the amounts measured at the two sampling times changed as a function of treatment (Table 1). Although no significant differences could be detected in the amounts of N mineralized in the fallow treatment at the two sampling times, the mineralizable N determined in the cereal and legume treatments in June were greater than those measured in September. Distinct patterns of readily mineralizable N were found across aggregate-size classes at each sampling time (Fig. 3)
. In June, mineralizable N varied significantly in response to cover crop treatments (P < 0.01), aggregate size (P < 0.0001), and their interaction (P < 0.01). With the exception of the 0.25- to 0.5-mm size class, significantly smaller amounts of mineralizable N were found in all aggregate sizes of the fallow treatment than in the corresponding aggregates from both cover crop treatments (Fig. 3). For all treatments, the largest amount of mineralizable N was found in the 2.0- to 5-mm size class.
|
Soil Enzyme Activities
In 1996, a significant cover crop effect was observed on the rates of ß-glucosidase activity measured in June (P < 0.05) and September (P < 0.01) for whole soil samples (Table 1). The rates determined in aggregates and whole soil from the legume and cereal treatments averaged 37 and 30% higher than the fallow treatment (Table 1 and Fig. 4)
. With the exception of aggregates from the fallow treatment in September 1996, the rates of ß-glucosidase activity differed significantly (P < 0.01) among aggregate-size classes. The interaction between treatment and aggregate size was not significant in any of the three periods evaluated. The distribution patterns of activities across aggregate-size classes also changed between June and September 1996 (the interaction between sampling time and aggregate-size class was statistically significant, [P < 0.001]). In June, aggregates recovered from cover crop treatments showed the lowest rates of activity in the <0.25 and 2.0- to 5.0-mm size classes, whereas in September the lowest activities were found in the 2.0- to 5.0-mm size class.
|
|
|
Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Heterogeneous distributions of microbial biomass and associated activities among aggregate-size classes have been observed previously (Gupta and Germida, 1988; Seech and Beauchamp, 1988; Beauchamp and Seech, 1990; Miller and Dick, 1995a and 1995b; Franzluebbers and Arshad, 1997). Conceptually, there are at least two ways to think about this phenomenon. First, variation in microbial biomass among aggregate-size classes might be a reflection of differences in their structural properties that selectively or generally restrict the transfer of nutrients. Alternately, variations in microbial biomass might imply that the structural variation among the aggregates restricts predatorial grazing to certain aggregate-size classes but not to others. By sampling soil on different occasions during the same cropping year, we gained some insights into the dynamic nature of the interactions between microorganisms and the different size-classes of soil aggregates. For example, recent studies indicate that microorganisms inside of microaggregates (<0.25 mm) are biologically active and significantly involved in processing soil C (Jastrow et al., 1996). In this context, we observed that the levels of both SMBC and mineralizable N in September 1996 had declined to 
70 and 50% of their June values, respectively, in both the <0.25 and 0.5- to 1.0-mm size classes (Fig. 1b and 1c). During the same time interval, SMBC levels increased by 30% in the 2.0- to 5.0-mm size classes of both the cereal and legume treatments, while the level of mineralizable N declined by 50 to 60%. Further studies are needed to determine whether microbial biomass growth and turnover rates differ in these different aggregate-size classes, and whether changes in these rates occur between the June and September sampling times. Nonetheless, it is no trivial experimental challenge to obtain proof that microstructural features are sufficiently different among aggregate-size classes to differentially influence microbe x substrate x predator interactions.
Another particularly intriguing phenomenon was the observation that the profiles of readily mineralizable C and N among aggregate-size classes did not coincide at either the June or September 1996 sampling times, and that a significant negative relationship existed between readily mineralizable N and ß-glucosidase activity. These results might be explained if cover crop residue-derived substrates of different quality were to exist in the different aggregate sizes and influence the properties of the mineralization processes. For example, in September 1996 the 2- to 5-mm size class of the legume treatment had the highest level of readily mineralizable N and one of the lowest levels of readily mineralizable C. In contrast, on the same date, the 0.5- to 1.0-mm size class of the cereal treatment had the highest level of readily mineralizable C and one of the lowest levels of readily mineralizable N. As a consequence, net rates of N mineralization may have been influenced by the differences in the gross rates of either N immobilization or mineralization. Further studies are required, perhaps using isotope dilution procedures (Hart et al., 1994) to determine if this idea has any merit.
Although soil enzyme activities have been used for many years as indicators of soil biological activity, it has always been unclear to what extent they reflect the current physiological status of the microbial population, as opposed to the history of past microbial colonization. In this study, the greatest rates of enzyme activities tended to be found in the smallest aggregate-size classes (<0.25 and 0.25–0.50 mm) of the September samples, which also possessed the lowest levels of SMBC and mineralizable N. One possible interpretation of this observation is that the SMBC decline in these aggregate-size classes between the June and September samplings is brought about by the action of lytic enzymes such as proteases, lipases, and phosphatases produced under nutrient-limited stress in the microaggregate environment. Further work is needed that will address the question of the accessibility of the microaggregate-containing microbial population to nutrients present in the macroaggregate soil structure, as well as research into the extent of the role that cryptic growth plays in the existence of the microaggregate community.
Some interesting observations were made in this study regarding the relative benefits of using nonlegume and legume cover crops in this cropping system for sequestering C and N in soil. There were trends for both TOC and TKN levels to be consistently lower in soil removed from the cereal treatment, which indicates that the use of triticale as a cover crop may promote mineralization of soil organic matter. Secondly, while TKN levels were consistently greater in all aggregate-size classes of the legume treatment than in either fallow or cereal treatments (Table 3), readily mineralizable N increased significantly with increasing aggregate size in the legume treatment (Fig. 3). Further studies are needed to determine whether the N sequestered in the smaller aggregate-size classes of the legume treatment is more recalcitrant than in the larger ones, or whether different sizes of aggregates support different gross rates of mineralization and immobilization of N.
On a final note, we wish to emphasize that establishing and managing cover crops in the Pacific Northwest is not a trivial matter. Fall weather conditions can be extremely variable, and establishment and final yield of the cover crop can be erratic (Burkett et al. 1997). For example, in 1995 and 1996 the above-ground dry matter yields of the cover crops averaged 1640 and 593 kg ha-1 for the cereal, and 1376 and 897 kg ha-1 for the legume treatment, respectively. Furthermore, below ground biomass production was not measured despite the likelihood of it influencing the properties we measured. In this particular cropping system we cannot overlook the fact that the alternating summer crops of broccoli and sweet corn are fertilized differently and produce different amounts and quality of crop residues. Despite our interest in identifying useful soil quality properties for this cropping system, it is going to be difficult to distinguish between temporal fluctuations of soil properties from changes that truly reflect treatment effects and that may be of a more permanent nature.SAS Institute 1989; SAS Institute 1995
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Received for publication July 6, 1998.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. J. Franzluebbers and J. A. Stuedemann Early Response of Soil Organic Fractions to Tillage and Integrated Crop-Livestock Production Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 613 - 625. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. B. Villamil, G. A. Bollero, R. G. Darmody, F. W. Simmons, and D. G. Bullock No-Till Corn/Soybean Systems Including Winter Cover Crops: Effects on Soil Properties Soil Sci. Soc. Am. J., September 20, 2006; 70(6): 1936 - 1944. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. C. Kaspar, T. B. Parkin, D. B. Jaynes, C. A. Cambardella, D. W. Meek, and Y. S. Jung Examining Changes in Soil Organic Carbon with Oat and Rye Cover Crops Using Terrain Covariates Soil Sci. Soc. Am. J., May 23, 2006; 70(4): 1168 - 1177. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. M. Sainju, T. H. Terrill, S. Gelaye, and B. P. Singh Soil Aggregation and Carbon and Nitrogen Pools under Rhizoma Peanut and Perennial Weeds Soil Sci. Soc. Am. J., January 1, 2003; 67(1): 146 - 155. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Schutter and R. P. Dick Microbial Community Profiles and Activities among Aggregates of Winter Fallow and Cover-Cropped Soil Soil Sci. Soc. Am. J., January 1, 2002; 66(1): 142 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. N. Duddleston, P. J. Bottomley, A. Porter, and D. J. Arp Effects of Soil and Water Content on Methyl Bromide Oxidation by the Ammonia-Oxidizing Bacterium Nitrosomonas europaea Appl. Envir. Microbiol., June 1, 2000; 66(6): 2636 - 2640. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
