Soil preparation, microflora, fauna and nitrogen liberation

Summary

Principal aims:
To promote organic farming through a project which integrates knowledge about soil treat-ment, soil physics, soil microbiology, soil zoology, nitrogen dynamics and crop growth by:

1. utilising improved knowledge about the interactions between soil tillage and soil fauna to improve the advisory system with regard to soil tillage specially suited for organic farming
   
   
2. developing soil tillage as a tool to time the release of nutrients, especially nitrogen, to take place in the most intensive crop growth period
   
   
3. acquiring more knowledge about the quantity and quality of the life which occur on and within the soil as a function of soil tillage in organic farming
   
   
4. developing a mathematical simulation model, that can simulate the effect of different types of soil tillage on the micro flora, fauna and nitrogen release at varying temperature and humidity. The simulation model is intended to be a tool (among other tools) to achieve the points 1 - 3.
   
   

Project structure.
The project consisted of larger field experiments, minor experiments in the field, laboratory investigations and mathematical modelling. Three large field experiments were carried out:

1. the Kemink Excact soil tillage system.
   
   
2. a non inversion tillage carried out by aid of subsoiler tines combined with a rotovator and a drill.
   
   
3. a secondary superficial soil loosening between the rows during the growth season.
   
   

1. Kemink.
A larger system field experiment and two smaller detailed field experiments was conducted to compare the Kemink Exact Soil tillage system and its components with a traditional tillage system based on mouldboard ploughing.

The system experiment showed unambiguously that the Kemink system has a positive effect on soil struc-ture. Measurements of soil penetration resistance after the first and third year confirmed this. While soil loos-ening was successful it quickly became clear that the Kemink implements has obvious limitations on loamy soil. Despite the looser soil in the Kemink system, the total yield of three years was negatively affected by the Kemink system. There was however a positive development over time so that the yield in the Kemink system changed from being significantly lower in 1998 and 1999 to being at the same level as the yield in the traditional tillage system in 2000. For sugar beets the Kemink system gave a significantly higher yield in 2000 and this could be explained by an exceedingly strong correlation between penetration resistance and yield.

The detailed field experiment with autumn ridges as an alternative to ploughing showed that the use of ridges can reduce the nitrogen leaching potential. After setting up ridges in September 1999 the ploughed plots had an inorganic nitrogen content in 30-60 and 60-90 cm soil depth in November that was 39% and 51% higher than in plots with ridges.

This leaching reduction resulted in increased growth of the subsequent barley crop (10,0% higher dry matter yield at harvest 1999 and 27,6% higher dry matter yield at harvest 2000). The detailed field experiment with deep soil loosening in sugar beet showed that while soil loosening prior to sowing gave significant yield in-creases of 7.0% and 12,4% in 1999 and 2000 there was a negative effect of further soil loosening in the growing season.

2. Non-inversion tillage
These experiments were carried out at the organic research farm Rugballegaard near Horsens, Jutland.

The experiments were designed to test for tillage induced differences between the non-inversion tillage and common mouldboard ploughing. The experimental design was a randomised block design with four repli-cates. The samples of soil for the analysis of all the different soil physical and biological parameters were taken right next to each other in order to be representative for the same spot. In 1998/1999 samples were taken in three horizons, 0-4 cm, 8 – 12 cm and 28 –32 cm. In 2000 another horizon, 16 – 20 cm was in-cluded. In 1998/99 the experiments were carried out in winter wheat and the tillage were done in autumn. In 2000 the tillage was carried out in spring and the crop was spring barley.

For all depths examined in 1998/99, a significantly higher porosity was recorded for the soil treated with the non-inversion tillage as compared to the ploughed soil. Three weeks later this difference had disappeared for the top layer, and was reduced for the other layers. And a further five months later (in the spring,- 23 Mar), only a non-significant trend of higher porosity in the non-inversion treatment could be measured The measurements prior to the tillage treatments in the in year 2000 indicated a very loose topsoil. In both years the results di not indicated the existence of a distinct plough pan. For the layer below ploughing depth, a rather small increase in porosity was measured for the non-inversion. The trend appeared at both sampling dates following the tillage treatments but was not statistically significant.

The two tillage treatments (mouldboard ploughing or non-inverting tillage) were found to have minimal effect on net N mineralisation, C decomposition or the microbial biomass in the soil. The tillage induced changes in nitro-gen mineralisation, microbial biomass-C and N, and soil respiration was largest in the 28 – 32 cm. layer.

Most of the microfungi ( e.g. filamentous fungi) had highest spore density in the upper soil layers. This vertical distribution remained after both tillage treatments. However, there was a decrease in the density in the upper layer after ploughing. One of these fungi, Cladosporium spp, showed a dramatic decrease in the 0-4 cm layer after ploughing, most likely caused by mixing with other soil layers. This tendency was less pronounced after the non-inversion tillage. These observed effects remained during the sampling periods, i.e from October to March in 1998/99 and from March to mid May in 2000.

Generally, the two types of tillage did not vary in immediate effect on the nematodes in the field nor was any differences observed a month later. This is in accordance with the results of the laboratory study on nema-todes.

The population density before tillage was exceptionally high, especially in September 1998, as compared to typical population sizes previously recorded in both natural and agricultural soils (estimate including interpolated values for non-sampled strata: 175.000. m2 ). Both tillage methods resulted in a reduction of the collembolan density immediately after tillage to about a third in 1998 and about half of the initial density in the spring 2000. The total density through the entire profile remained low through the autumn and spring 1998-99 and throughout the spring 2000. Significant effects of tillage treatment on the total collembolan population were not found at any time during the two experimental periods. However, significant differences related to tillage type were found be-tween the vertical distribution of the Collembola population in both periods. Conventional ploughing resulted in an inversion of the original, vertical distribution pattern. In comparison, the non-inverting tillage type did not cause distinct changes in the vertical distribution pattern.

The number of mites as composite group showed different reactions in response to tillage in the two different experimental fields. In 1998-99 a negative effect on the mite fauna was not apparent at the sampling immedi-ately after tillage but was realised later and recorded at the following sampling date. However, in the spring 2000 an immediate reduction of the initial density was observed already at the sampling carried out few days after tillage. The different groups of mites did not react in the same way to tillage. Thus, although the number of most mite species were reduced in response to tillage, the number of acaridid mites increased strongly immediately after tillage in 1998. This population peak was followed by a steep relapse in density. Generally, both conven-tional ploughing and non-inverting tillage seem to be harmful to the mite fauna. However, a higher number of mites was found in the uppermost layers of that part of the field which was treated with non-inverting tillage com-pared to the part treated with conventional ploughing.

The density of earthworm was very high both years with 350 – 550 individuals m-2. This corresponds to up to approximately 900 individuals m-2 for the entire soil profile. Both tillage treatments caused a marked de-crease in the earthworm density, and there was not found any significant differences between the two meth-ods. The non-inversion method appeared especially detrimental to the earthworm fauna, which is most likely due to the rotovator. In 1998/99 the worms in the upper 0-4 cm were especially vulnerable to this treatment and were reduced by 70%. Also in 1998/99 ploughing caused a significant increase (>90%) of the earthworm density in the 8 – 12 cm layer. The treatments were less detrimental to the earthworms in 2000 and the de-crease in density was not significant. No differences between the two treatments were found this year.

In 1998/99 the grain yield was lowest in the non-inversion tillage plots, and the difference were right around the significance level. If a very weeded block was exclude from the statistical analysis as an outlayer, the difference was significant, but it was not significant if al blocks were included in the analysis.

An analysis for correlation between the measured parameters was carried out by aid of multivariate statistics. Both years before the treatments a strong correlation between several of the measured microbiological pa-rameters and both the mite and collembola communities were found. Similarly, before tillage treatments, a strong correlation between the soil physical parameters and both the micro fungi flora, the Collembola com-munity and the mite community was found. This correlation was weaker in autumn 1998. Several of the measured parameters had a co-variation with depth, e.g. the pore size distribution. All correlations in the system were broken up completely by tillage, and were not re-established during the sampling periods.

3. The secondary superficial soil loosening between the rows
The experiments were both years carried out in winter wheat at the organic research farm Rugballegaard. In 1999, the experiment was placed in the ploughed plot of the non-inversion tillage experiments. The loosen-ing was carried out with a Kress hoe with loosening tines mounted. The soil was loosened down to approxi-mately 8 cm. In 1999, the loosening was carried out on the 18 May, and in 2000 three experiments was made with different timing.

The loosened plot produced a significantly lower yield in 1999. The plant samples during the growth season showed no significant differences in the dry matter production, but there was a clear tendency towards a higher uptake of nitrogen of about 4 kg N ha-1. This is in line with the soil samples, which showed clear indi-cations of higher mineral nitrogen content in the 0-4 cm and 8 – 12 cm layers about 14 days after soil loosen-ing. This difference in soil mineral nitrogen had disappeared by the following soil samplings. In 2000, no sig-nificant differences due to soil loosening could be demonstrated on the grain yield. In stead, there was a significant increase in grain nitrogen content after the first soil loosening. This gave an increase of 5 kg N ha-1 in the grains.

In 1999 the soil loosening treatment performed in winter wheat in mid May was found to stimulate net N miner-alisation in the soil. The increase of plant available N did, however, not affect N uptake or plant growth. This was probably due to damage of the soil loosening treatment on the winter wheat.

In 1999 the sampling of soil microarthropods continued in the ploughed plots in order to look for possible effects of the superficial soil loosening. The Collembola density increased from a relatively low level in March to about 55.000 m-2 in June, which corresponds to approximately 100.000 m-2 in the entire soil profile. There was not found any significant effects of soil loosening.

Similarly, there was not found significant effects of soil loosening on the earthworm density.

In stead, there was a significantly negative effect on the density of surface living spiders, which were re-duced by 30%. The first sampling demonstrated the acute mortality while the second sampling (still at a low level) demonstrated a weak re-invasion capacity.

4. Explanatory experiments in laboratory and field
A number of explanatory experiments were carried out. The aims of these experiments were both to support the interpretation of the results from the field experiments, and to provide input knowledge to the mathematical simulation model. Only some of the experiments are mentioned in this summary.

a)
In accordance with field results, the simulation of soil tillage in the laboratory experiment showed no effect on total net N mineralisation. The simulation of tillage did, however, release nitrogen from a protected (inactive) pool in the soil. Results obtained by use of different methods for disturbance of soil structure and the microbial biomass also indicated that soil tillage mainly caused a short-term release of N from the microbial biomass in the soil. Other pools of soil organic N such as partly decomposed plant material may also give a low and long-term contribution to N release caused by tillage. These mechanisms may in part explain the effects of decreased soil organic pools and thus soil fertility found of soil tillage.

b)
We have investigated the effects of soil water potential on the performance and survival of two important soil invertebrate groups, earthworms and Collembola. As a model for earthworms, we used Aporrectodea caliginosa, which is one of the most common species in agricultural soil. In laboratory experiments we have described how this species respond to decreases in soil water potential, with respect to reproduction, growth and survival. For Collembola we have investigated the effects of drought on survival, and the underlying physiological mechanisms forming the basis for drought survival in these animals.

c)
The effect of distance between resources for nematodes. Decomposition activity increased at shorter distances between contrasting plant resources within the range 6-50mm and the microbial activity was less controlled by nitrogen shortage with small distances between the resources. The organic amendment that caused a 3-4 times increase in nematode number at the resource within a few weeks did not have any effect on nematode number down to 6mm from the amendment, however. Since the two methods of tillage can not be expected to give different soil loosening at or below the 10mm scale, the model study therefore support the lack of immediate response in the field study.

d)
The food preferences of arthropod predators. Laboratory investigations with Collembolas as food for a predatory beetle, Bembidion lampros (known to participate in aphid control) were carried out. The five investigated species of Collembola were found to have different values as food for the beetle, which indicates a possible connection between the quality of the Collembola community and the fauna of predatory beetles.

5. The mathematical simulation model
A mathematical simulation model (a food web model) that can simulate the interactions in the degrader food web of Danish agricultural soil dependent on temperature, humidity, pore size distribution and tillage. This food web model has been coupled with the FASSET-model, which is being used by scientists at Research Centre Foulum. The food web model is a population dynamical model that is driven by the populations’ de-mand for prey, and is regulated by the amount of prey captured.

The food web model utilises an extended “predator – prey” concept. A predator is just an organism that feeds on something else. Similarly, a prey is something that is being eaten or taken up by a “predator”. Thus, organic matter can be regarded a prey, and bacteria or micro fungi can be regarded predators. This concept has not previously been tried out in simulations of the degradation of organic matter. The possible advantage is that the simulation model does not operate with organic matter pools, which cannot be measured in the field. The disadvantage is, presently, that the organic matter C/N-ration does not change with time

Nitrogen can be released in two ways in the model: 1) If an organism/population acquires food with a higher nitrogen content than it demands, the exceeding nitrogen will be excreted, and 2) the content of nitrogen in the food that is used for respiration is excreted. These mechanisms work for all organisms in the food web. Thus, the liberation of nitrogen is exclusively simulated by as biologically drive mechanisms, and includes the impact of both micro-flora and fauna.

The model also includes a simulation of the pore size distribution as a controlling mechanism in relation to the vertical distribution of organisms in the soil profile.