|
|
||||||||
a UMR Environnement et Grandes Cultures I NA P-G/INRA, 78850 Thiverval-Grignon, France
b Unité de Science du Sol INRA, 45160 Ardon, France
* Corresponding author (Louis-Marie.Bresson{at}grignon.inra.fr)
| ABSTRACT |
|---|
|
|
|---|
Abbreviations: 2D, two dimensions MSW, municipal solid waste
| INTRODUCTION |
|---|
|
|
|---|
Northern Paris basin is especially affected by these problems. Most of the landscape is covered by a loess deposit where soils are prone to stucture degradation because of their low clay and organic matter content (<15 and <1.5%, respectively) (Eimberck, 1990). Development, over the last decades, of farming systems where spring crops cover a significant part of the land, lead to increasing structure degradation. As a result, flooding and pollution events began to be clearly identified by farmers and by communities, as a main environmental issue (Papy and Boiffin, 1988; Martin et al., 1997).
Concomitantly, more emphasis has been placed on the recycling of MSWs through their agricultural use as fertilizers or organic amendments. Since organic amendments improve aggregate stability (e.g., Tisdall and Oades, 1982), MSW might be a good substitute for traditional farmyard manure which is no longer available in intensive cropping areas.
Beneficial effects of sewage sludge application have been widely acknowledged, even though Metzger and Yaron (1987) pointed out that controversial results could be found in the literature. Literature is scarce on the effects of compost on soil structure. In their review on the changes in soil physical properties because of organic waste applications, Khaleel et al. (1981) presented 17 papers, only two dealing with MSW composts (Mays et al., 1973; Epstein et al., 1976). They did not find a significant difference between the various types of wastes. Regardless of waste type, both long-term and short-term studies indicated a significant linear relationship between reduction in bulk density and increase in soil organic C. The decrease in bulk density has been ascribed to the dilution effect resulting from the mixing of the soil with less dense organic material added (Khaleel et al., 1981; Tester, 1990). Structural changes resulting from interactions between added organic matter and soil material were seldom suggested, but a close examination of the data published in the literature supports this hypothesis. Changes in soil structure (Guisquiani et al., 1995), as well as, changes in macro and mesoporosity (Pagliai et al., 1981; Guisquiani et al., 1995) have been observed in thin sections and cannot be explained by a mixing effect. This was corroborated by the increase in aggregate stability because of organic waste application (Pagliai et al., 1981; Gerzabeck et al., 1995) or application of humic acid extracts (Canarutto et al., 1996) even though such a stabilizing effect might not be observed in rather stable soils (Guidi et al., 1988).
Based on the studies reported above, we hypothesized that MSW composts may have beneficial effects on soil structure. Also, there is little information available which deals with effects on hydraulic conductivity (Felton, 1995), crust formation, runoff, or erosion, even though the main assessments reported above (i.e., bulk density decrease and aggregate stability increase) lead to the conclusion that compost application might enhance water infiltration, slow down crust formation, delay runoff, and reduce erosion. This might not apply to the highly unstable soils of the European loess belt where seedbeds often slump because of aggregate coalescence (Bresson and Boiffin, 1990; Kwaad and Mücher, 1994; Bresson and Moran, 1995), which is an important process of structure degradation together with aggregate slaking or microcracking (Le Bissonnais et al., 1989).
The aim of the study was to determine if MSW compost application decreases surface structure degradation, i.e., crusting and slumping, on a highly unstable silt loam soil, and to test its possible consequences on runoff and erosion. The experiment involved repacked seedbeds and simulated rainfall. The focus was on (i) the description of structural changes both at the soil surface and within the seedbed using macroscopic and microscopic observation methods (Bresson and Boiffin, 1990), (ii) the impact of applied compost on processes of structure degradation, and (iii) the related changes in runoff and sediment production at local scale. The initial stages of degradation were emphasized, because the purpose of compost application is to slow down soil surface degradation until seedlings emerge and crop cover establishes, to prevent further degradation, to delay runoff, and to reduce erosion.
| MATERIALS AND METHODS |
|---|
|
|
|---|
|
Compost was carefully hand mixed with the soil material at a rate of 15 g kg-1 (dry matter), which increased the organic C content by 3.6 g kg-1. This rate is similar to a 50 Mg ha-1 application rate mixed within the 0- to 25-cm ploughed layer. The amended soil and a control soil (no MSW added) were left for 1 wk in a cool chamber at 4°C.
Surface Morphological Evaluation and Runoff Monitoring
Runoff trays 50 by 50 by 15 cm with buffer zone 15-cm wide (Bradford et al., 1987; Le Bissonnais and Singer, 1992; Fox and Le Bissonnais, 1998) were filled with control soil and amended soil. The repacked seedbeds were slightly compacted up to a bulk density in the 1.10 to 1.20 Mg m-3 range to get a structure similar to the structure of seedbeds observed in the field. This was achieved by lifting the side of the tray by about 5 cm up and letting it to fall back, and repeating this operation twice for each side. The seedbed was carefully raked to a depth of 3 cm to homogenize the distribution of the largest clods on the surface. Three replicates were made for the control soil and for the amended soil.
Rainfall simulation with deionized water was performed using a 361 needles, 1 by 1 m simulator, located 6 m high. A 5-mm mesh grid was placed 1.5 m below the simulator to gain a better homogeneity of the rain. Rainfall characteristics included a mean drop diameter of 2.5 mm and 80% of terminal velocity (Le Bissonnais et al., 1995). Rainfall intensity was measured before and after every run, and its variation was <5% so that no statistical design was used. Runoff trays were set at a 5% slope and subjected to a 19 mm h-1 simulated rainfall for 60 min. In the Paris basin, such an event has a 3 yr return period and events of 20 mm h-1 during 15 min occur every year.
The morphological evolution of the soil surface was monitored using photographs and descriptions according to the method suggested by Boiffin (1986) and Bresson and Boiffin (1990).
Runoff was collected continuously during the rainfall simulation and aliquots were sampled under continuous agitation for sediment concentration. For statistical analysis, variances of both treatments were compared using a F-test, then the probability of significance was computed using a two-tailed unpaired Student's t-test.
Soil Structure Monitoring
Microstructure boxes, 10 by 30 by10 cm, with free drainage and runoff were used for microstructure studies. These boxes can be directly impregnated after the experiment, avoiding disturbance resulting from handling during sampling. Boxes were filled using the same procedure as for runoff trays, and two replicates were used. Microstructure boxes were then set with the same slope angle and subjected to the same simulated rainfall for 60 min. The 0- to 75-mm upslope part of the boxes was covered as to protect it from rainfall and maintain the soil surface at the initial state. The following 75 to 150 mm and 150 to 225 mm were covered after 4 min and 15 min of rainfall simulation, respectively. The remaining 225 to 300 mm were left exposed up to the end of the simulation, i.e., 60 min. This timing resulted from the previous monitoring of the soil surface macroscopic morphology described above. This procedure aimed to obtain the initial state, two development stages of the structural crust and the first stage of the depositional crust, all within the same box. The procedure did not account for lateral translocation of detached particles because of the small size of the box used. It is relevant here, however, because the first stages of surface degradation mainly depend on aggregate breakdown and slumping processes, rather than on long-range translocation. Microstructure boxes were air dried and impregnated with a polyester resin. Four thin sections, 60 by 130 mm, were prepared for each replicate and observed using a polarizing microscope with 8 to 300 times magnification. Crusts were characterized using the diagnostic features suggested by Valentin and Bresson (1992)(1998).
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
|
Soil Microstructure
The initial structure of the control microstructure boxes consisted of loosely packed aggregates, which resulted in a high interaggregate porosity (Fig. 2a)
. Intraaggregate porosity was rather high and consisted of 30-µm wide polyconcave voids. At the primary particle level, the microstructure was quite unusual (Fig. 3a)
. The fine particles (<20 µm, i.e., fine silt and clay fraction) were not sufficient to form a continuous phase where coarse grains (mostly ranging between 50 and 150 µm) could be embedded. The fine particles were distributed either as coatings around the grains or as 50- to 100-µm microaggregates. The solid phase was not continuous in two dimensions (2D), and the three-dimensional continuity might not involve fine particles at all, but coarse grains instead. The consequences of such a microstructure on aggregate stability are expected to be: (i) a high continuity within pore space, which should reduce air entrapment and therefore aggregate slaking (Le Bissonnais et al., 1989; Bresson and Valentin, 1994; Le Bissonnais, 1996), and (ii) a low cohesion when wet, which should enhance aggregate coalescence (Bresson and Boiffin, 1990; Kwaad and Mücher, 1994; Bresson and Moran, 1995). After 1.3 mm of cumulative rainfall, meniscus-like bridges formed between aggregates a few millimeters below the surface (Fig. 2b). These features indicate that at this incipient stage, the structural crust formed because of an illuvial process which is common in soils wet before rainfall (Boiffin and Bresson, 1987; Le Bissonnais et al., 1989; Bresson and Cadot, 1992). After 4.8 mm of cumulative rainfall, the stuctural crust was well developed (Fig. 2c). Aggregate coalescence was involved rather than illuviation, as evidenced by the typical convexconcave to vesicular shape of the voids (Fig. 3b) (Bresson and Boiffin, 1990; Bresson and Valentin, 1994). This means that, in such an unstable soil, rapid structural collapse prevented further illuviation (Bresson and Cadot, 1992). The coalescence process affected the whole seedbed (Fig. 2c) and the aggregate framework was 2D continuous. Packing voids were polyconcave at the bottom and convexities developed towards the transition with the structural crust. A thin depositional crust overlaid the structural crust in microdepressions between surface aggregates. The deposited material was poorly sorted and microbedded, which is typical of the first stage of depositional crusts (Bresson and Valentin, 1994). After 19 mm of cumulative rainfall, overall macroporosity was very low and consisted of convexconcave to vesicular voids, which shows that coalescence developed throughout the seedbed (Fig. 2d). The depositional crust had spread over 30% of the surface. The deposited material was clearly microbedded, which could be related to variations in the hydrologic behavior of the soil surface (Mücher and De Ploey, 1977; Mücher et al., 1981; Boiffin and Bresson, 1987; Bresson and Boiffin, 1990). Since characteristics of the simulated rainfall remained constant, these variations might be ascribed to variations of surface morphology such as connection between puddles or silting up of puddles.
|
|
The initial packing of amended microstructure boxes was slightly looser, and aggregates were slightly better sorted. Compost fragments appeared mainly as fibrous particles and organic microaggregates interspersed between soil aggregates (Fig. 3c). Almost no change in structure was observed after 1.3 mm of cumulative rainfall (Fig. 2e). After 4.8 mm of cumulative rainfall, the structural crust was very thin, about 1 mm thick (Fig. 2f), and consisted of typical bridges composed of bare silt particles (Fig. 3d). No depositional crust could be observed. Crusting developed much slower than in the control soil, and the process was typically an illuviation process instead of a coalescence process. The underlying material did not slump. This suggests that compost application lead to a greater cohesion of the wet soil material, which in turn, maintained the aggregate framework and allowed the illuviation process to develop. Such an impact of organic matter on soil rheological properties has been already suggested (Bresson and Boiffin, 1990). After 19 mm of cumulative rainfall, the structural crust was slightly thicker (about 2.5 mm), and exhibited some coalescing features at the bottom (Fig. 2g). The underlying seedbed remained clearly aggregated, even though less loosely packed, and displayed a sharp transition with the crust. A depositional crust developed in the microdepressions of the surface which remained rather rough. In some places, depositional material was observed below the surface. It differed significantly from silt illuviation by a clearly bedded microstructure. This internal depositional crust indicates the persistency of some macroporosity at the soil surface even after runoff had begun.
The stabilizing effect observed cannot be mistaken for the increase in bulk density commonly reported in the literature and usually ascribed to a mixing effect, i.e., the dilution of the soil material by a less dense organic material (Khaleel et al., 1981). The above description of porosity deals with a low magnification, which means that compost particles and soil are not distinguished. In turn, this description cannot account for the mixing effect but for the interaction effect instead, i.e., the change in soil structure.
The slowing down of aggregate coalescence induced by compost application probably results from an increase in wet cohesion. Chenu and Guérif (1991) found that polysaccharides greatly increased tensile strength of dry clay aggregates, and this effect might also be true in wet aggregates, as shown for polyvinyl alcohol (PVA) (Williams et al., 1967). Entanglement of primary particles or microaggregates by fungal hyphaes has been shown to play a great role in sludge amended soils (Metzger et al., 1987), but microscopic evidence of such a process could not be found in the experiment reported here, possibly because of its short duration (1 wk).
Hydrological Behavior
On the control runoff trays, runoff started after 2.5 mm of cumulative rainfall (Table 2) and reached a steady state after about 13 mm of cumulative rainfall. Final runoff rate was 17.1 mm h-1, i.e., a runoff coefficient of 90.2%, which is greater than the values measured in the field at the same scale on a similar soil material (Le Bissonnais et al., 1995; Le Bissonnais et al., 1998). On the amended runoff trays, compost application changed the hydrological behavior (Table 2). Runoff started significantly later (9.2 vs. 2.5 mm of cumulative rainfall) and drainage persisted (3 mm h-1) at the end of the simulation. These results are consistent with the observed change in structure, i.e., delay of crust formation and prevention of slumping.
|
Compost application decreased soil loss from 54.6 to 18.3 g (Table 2), mainly by decreasing the flush of sediment related to the incipient runoff. As indicated by microstructure observation, particles detached from the top aggregates when rainfall had started were allowed to illuviate a few millimeters below the surface where they were protected against further entrainment by surface runoff.
| CONCLUSIONS |
|---|
|
|
|---|
This study shows that the effect of compost application on soil bulk density reported in the literature was not a mixing but an interaction effect. Interaction occurred within 1 wk after compost application but its nature and the way it prevented coalescence are still to be established.
| ACKNOWLEDGMENTS |
|---|
Received for publication December 12, 2000.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
B. Augeard, L. M. Bresson, S. Assouline, C. Kao, and M. Vauclin Dynamics of Soil Surface Bulk Density: Role of Water Table Elevation and Rainfall Duration Soil Sci. Soc. Am. J., January 25, 2008; 72(2): 412 - 423. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. M. Bresson, C. J. Moran, and S. Assouline Use of Bulk Density Profiles from X-Radiography to Examine Structural Crust Models Soil Sci. Soc. Am. J., July 1, 2004; 68(4): 1169 - 1176. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Journal of Environmental Quality |
||||