Soil Stresses, Quality and Care Proceedings from NJF seminar 310 Ås, April 10-12 2000
Susanne Elmholt, Bo Stenberg, Arne Grønlund & Visa Nuutinen Department of Crop Physiology and Soil Science P. O. Box 50 DK-8830 Tjele
DIAS report Plant Production no. 38 • December 2000
Publisher: Danish Institute of Agricultural Sciences Tel. +45 89 99 19 00 Research Centre Foulum Fax +45 89 99 19 19 P.O. Box 50 DK-8830 Tjele
Sale by copies: up to 50 pages 50,- DKK (incl. VAT) up to 100 pages 75,- DKK more than100 pages 100,- DKK
Subscription: Depending on the number of reports sent but equivalent to 75% of the price of sale by copies.
Photos: Reidun Aspmo & Per Schjønning
NJF-seminar Soil Stresses, Quality and Care
Different people understand differently the word soil. The soil can be looked upon as a body of nature and a part of the landscape. To many people soil is the substrate for plant growth and a prerequisite for food production and wealth. A third and perhaps growingly important view is soil as the physical ground for housings, roads and industry.
Soil is important to the whole society, not only for the agricultural sector. Soil has important ecological functions as it interacts with the dead bedrock, the atmosphere, the water, and the living organisms. Important functions are:
Biomass producer and transformer Geomembrane, filter and buffer Habitat for living organisms Raw material and building ground Cultural heritage
Even if these important roles of soil are probably widely known, the world's soil resources are degraded at an alarming rate. This is well documented by for instance the World Resources Institute (http://www.nhq.nrcs.usda.gov/WSR/).
We need to care for our prime agricultural soils, soils which provide the green in urban areas, as well as soil resources in natural landscapes.
Soil and agricultural scientists have detailed knowledge of specific functions and processes in soil. Seminar topics in NJF’s Section 1 (Soils and fertilisers) in the recent years show that soil biology and ecology have been added to the previous production-related research such as tillage and fertilisation. We are now prepared to study the whole soil and we have to show that our knowledge concerns the society and not only the farmers and gardeners. We believe that the concept of soil quality will help us to improve our communication about soils and soil use. Politicians, authorities, and farmers need a soil quality classification that makes the choices between different soil uses and managements based on soil science. Such classifications have been used in agriculture for many years. However, it is necessary to develop them and relate the classification systems and parameters used to present-day problems and technology. It is also important to use in communication modern language, as less and less people have hands-on experience with soils in an agricultural context.
This seminar intended to elucidate and discuss the present knowledge and research that can help us assess the quality of soils and to identify key properties which may be used as indicators of soil health. The seminar showed that we possess the knowledge, and that we have a good base for the further work with the soil quality concept and its practical use. On behalf of the Nordic Association of Agricultural Scientists I will thank the organising committee for a well planned and conducted seminar.
Svein Skøien, Chairman, NJF Section Soil and Fertilisers Contents
The land quality concept as a means to improve communications about soils...... 1 J. Bouma
The soil quality concept: A tool for evaluating sustainability...... 15 Douglas L. Karlen and Susan S. Andrews
Factors influencing resilience and resistance in Norwegian silt loam soils ...... 27 Tore E. Sveistrup and Trond K. Haraldsen
Soil quality with chromatography ...... 35 Solveig Buvarp Nyborg
Multi-level assessment of soil quality – linking redutionist and holistic methodologies...... 43 Per Schjønning, Lars J. Munkholm, Kasia Debosz and Susanne Elmholt
Biotic and abiotic binding and bonding mechanisms in soils with long-term differences in management...... 53 Susanne Elmholt, Kasia Debosz, Lars J. Munkholm and Per Schjønning
Multivariate techniques for presentation, interpretation and evaluation of soil quality data ...63 Mats Johansson and Bo Stenberg
Denitrification, a soil quality indicator...... 73 Mikael Pell, Kalle Svensson and Ewa Bringmark
Changes in redox potential and Fe mobilization due to waterlogging in cultivated and non- cultivated soils at Alta, Northern Norway...... 81 Christian Uhlig, Gunter Wriedt, Thomas Baumgartl and Rainer Horn
Effects of crop rotation with perennial crops on macroporosity of a clay soil...... 89 Laura Alakukku
Indication of soil degradation in strawberry fields: disappearance of earthworms...... 99 Sanna Kukkonen and Susanna Vesalo
Soil biological, chemical and physical properties in fields under different management systems...... 109 Ansa Palojärvi, Laura Alakukku, Esko Martikainen, Marina Niemi, Pekka Vanhala, Kirsten Jörgensen and Martti Esala
Effects of management practice on soil organic matter content...... 115 Tor-Gunnar Vågen
Organic wastes and soil quality...... 123 Søren O. Petersen, Kasia Debosz and Frank Laturnus Levels of structural and functional complexity in soil OM turnover as revealed by physical fractionations ...... 133 Bent T. Christensen
Management of biodiversity in arable soil by field inoculation - an example using deep burrowing earthworms...... 151 Visa Nuutinen and Jyrki Pitkänen
Land use changes and degradation of forest and soil in watersheds of Nepal – A review.....159 B.K. Sitaula, K.D. Awasthi, N.R. Chapagain, G. S. Paudel, R.P. Neupane, P.L. Sankhayan, B.R.Singh and O. Hofstad
Soil Stresses, Quality and Care: Concluding remarks from discussions in working groups and plenary sessions of NJF-Seminar no. 310 ...... 171 Susanne Elmholt, Bo Stenberg, Arne Grønlund and Visa Nuutinen The land quality concept as a means to improve communications about soils
J. Bouma Member Scientific Council for Government Policy, the Hague and Professor of Soils, Lab. Soil Science and Geology, Wageningen University, the NETHERLANDS E-mail: [email protected] Summary
Soil expertise is not communicated effectively enough to the public at large, nor to planners and politicians. Use of the land quality (LQ) concept and emphasis on soil behavior as a function of management are expected to be helpful in improving communications. Existing definitions of “soil quality” and “sustainable land management” are analyzed to derive a procedure for defining LQ indicators of sustainable land management. Land- rather than soil qualities are considered to reflect the impact of the climate and the landscape on soil behavior. Land quality is different for different types of land use and attention is arbitrarily confined here to agriculture. Simulation modeling of crop growth and solute fluxes is used to define land quality (LQ) as the ratio between yield and potential (or water-limited) yield (x 100), which defines a “yield gap LQ”. For soils with nutrient mining, a nutrient-depletion LQ is defined. The actual agro-ecological condition and its potential, both expressed by LQ ‘s for a given piece of land, is considered here as independent input into broader land-use discussions which tend to be dominated by socio-economic and political considerations. Agro-ecological considerations should not be held hostage to actual socio-economic and political considerations, which may change in the near future while LQ’s have a much more permanent character. The proposed yield-gap LQ reflects yields and risks of production as simulations are made for many years, and soil and water quality associated with the production process is taken into account. Yields and pollution risks are expressed for Dutch conditions in terms of the probability that groundwater is polluted with nitrates. The proposed procedure requires the selection of acceptable production and pollution risks, before a LQ value can be obtained. Existing definitions implicitly emphasize the field and farm level. However, LQ is also important at the regional and higher level, which, so far, has received little attention. Then, again, an agro-ecological approach is suggested when defining LQ’s as input into the planning process, emphasizing not only an independent assessment of the potential for agricultural production, but of nature conservation as well.
Keywords: agro-ecosystems, nitrate pollution, risk assessment, potential crop yields
Introduction
The condition of soils is generally not a prime concern when environmental problems are discussed in society. Global climate change due to greenhouse gasses, the “ozon hole”, solid and liquid waste disposal and, increasingly, water quality and worldwide water shortages have been more successful in catching the imagination of the public at large. These, of course, are
1 worthy objectives of concern and study, but serious concerns about the environmental degradation of soils resulting from intensive forms of land use exceeding the ecological carrying capacity have widely been reported but don’t appear to lead to alarm or action (e. g. Oldeman, 1994). Approximately 15% of the total land surface of the world is degraded as a result of adverse human action. Worldwide, about 38% of the agricultural land is affected by significant human-induced soil degradation, of which 56% is caused by water erosion, 28% by wind erosion and 7% by nutrient decline. Substantial areas of prime agricultural land are, without much opposition, permanently taken out of production when used for city expansion and construction of shopping malls or roads. How can this relative indifference be explained? Without making attempts to speculate about the psychology of public awareness, we may state that the soil science profession has been less than successful in communicating its expertise effectively to the public, to regulators and to politicians. In this paper I will attempt to briefly analyze this problem. Of course, the challenge is to improve our efforts to show that soils are important. One way to do so is to develop indicators that can quickly express the value or quality of soils. Such indicators are crucial in the modern world where attention spans are very short and where attention is increasingly attuned to attractive “ soundbites”. Soil or land quality could be an ideal indicator as will be further explored in this paper. Our colleagues in economics and sociology have used attractive indicators for years (the Gross National Product may serve as an arbitrary example here). Such indicators have so far not been defined for soils (Pieri et al., 1995).
When discussing soil quality indicators in the context of achieving an effective external communication, we must link soils and their properties with functioning and use. This not only relates to agriculture, but also to nature areas and recreational facilities. Use patterns and management practices are only acceptable when they are sustainable in the long run. We will therefore discuss the soil quality aspect in the context of sustainable land use and management.
Sustainable land management has been discussed widely within soil science and agronomy and guidelines have been proposed by FAO (1993). These guidelines list four criteria for sustainable land management relating to agriculture: (1) production should be maintained; (2) risks should not increase; (3) quality of soil and water should be maintained, and (4) systems should be economically feasible and socially acceptable. Soil quality is part of this definition and this term has, in turn, been defined in another context as: “the fitness of a specific kind of soil to function within its capacity and within natural and managed ecosystem boundaries, to sustain plant and animal productivity, maintain air and water quality and support human health and habitation” (Karlen et al, 1997). Even though it is tempting to discuss these definitions as such, we would rather focus on a joint analysis of both definitions, leading to descriptions of soil quality indicators that have relevance for sustainable land management. Overall, the objective is to develop expressions for soil quality that may serve to improve communication of soil expertise to users and stakeholders. Attention in this paper will therefore be paid to discussions on: (1) creating awareness about soils; (2) sustainable land
2 management; (3) soil quality and soil quality indicators, and (4) use of the quality concept to improve communication. The focus will be on the field level, the usual area of activity of farmers. However, the quality concept is also important at other spatial scales, which have to be considered when dealing with policy issues in regions or countries. In this broader context, soil quality has not yet been analyzed and this issue will therefore also be explored in this paper.
Creating awareness about soils
Even though many people have a strong affinity with “ the land” they live on, it is a rather abstract affinity. Soils occur in darkness below the surface of the earth and, in contrast to weather and water, are not directly visible and cannot be experienced by the senses unless exposed in a hole or in an excavation. The vertical succession of layers, which is characteristic for soils, can be shown in monoliths and can be displayed on walls of offices, schools and other buildings. In the USA, every State of the Union has a “ State Soil”, just like a “State Bird”, “State Animal” etc. The vertical succession of layers in any soil is often quite beautiful, certainly when a wide array of colors is exposed as in podzols and gley soils. Successful exhibitions of soil monoliths have been held in art galleries! But soils represent more than just pretty pictures! To really understand the significance of soils, a functional approach is needed which demonstrates functioning within a landscape context as governed by interacting physical, chemical and biological processes. More specifically, functioning relates to supplying water and nutrients to agricultural crops and various types of natural vegetation, purification of percolating wastewater and carrying loads. All these functions taken together determine the value of soils for society.
How can the functioning of soils be characterized? Yields of crops are often known, as are types of natural vegetation that occur on different soil types. Chemical, physical and hydrological properties are often known to a certain extent but often in a rather qualitative, descriptive and static manner. How can we catch the effects of interacting physical, chemical and biological processes, which determine the dynamic character of soils and their functioning in ecosystems?
Many monitoring techniques are available now to measure varying soil properties over time. Use of transducers, recorders and proximal and remote sensing techniques allows us to “ take the pulse of mother earth” (e. g. Bouma, 1999). Such techniques are, however, costly and widespread application is therefore unlikely. The advance of computer simulation of soil processes has, however, proved to be quite helpful in documenting effects of soil processes and the impact of man. (e. g. Alcamo, 1999). Some examples will be cited later in this paper. Use of computer simulation techniques to demonstrate the dynamic behavior of soils as a function of management by man is much more attractive for the modern soil user than classic soil interpretations, as presented in soil surveys, which list relative suitabilities of soils for a series of land uses. The sophisticated modern user wants a range of options to choose from
3 when coping with land-use problems. He is used to making his own choices and does not like to be presented with single “ best” solutions to problems, developed by others without his active participation. He wants researchers to define a “ window of opportunity” for any given soil, while he (or she) makes the appropriate choices.
We expect that general soil awareness can increase when soil qualities are defined in the context of characteristic “ windows of opportunity” for any given soil, and when these soil qualities are communicated effectively to various users of soil information.
Sustainable land management
In their definition of sustainable land management, FAO (1993) clearly focused on agricultural production. This can be a choice, but we should realize that land management has broader implications than agricultural production alone. This is expressed in the definition of soil quality where natural ecosystems are mentioned as well as human health and habitation. Be that as it may, it should be recognized that FAO (1993) defines sustainable land management and not sustainability as such. Emphasis is therefore on specific action by man and not on vague conceptual definitions, which is attractive. The elements of the definition are logical but they should, for better understanding, be grouped into two categories. The first three define plausible agro-ecological aspects: productivity, risks during production and quality of soil and water. The fourth category is different: economic feasibility (which also strongly effects social acceptability) is largely beyond the control of the land manager. Even though a management scheme may be sustainable from an agro-ecological point of view, it can be economically unsustainable because of poor prices for agricultural produce. This should certainly be considered but it is unwise to not explore various agro-ecological management options, even when at this point in time these options are not economically or socially feasible. Times may change: doubling of the world population, exhaustion of fossil fuels requiring growth of energy crops and needs for raw plant material to be grown in future by farmers as source materials for industry, are likely to change what may at this point in time be a negative economic outlook for agricultural production. Agro-ecological research, focused on developing future sustainable land management practices, should not be held ransom to current economic conditions.
How, then, to characterize agricultural production, production risks and the quality of soil and water? Monitoring of agricultural production systems can provide relevant information but procedures are tedious and costly and monitoring over periods of many years is needed to cover unusual weather conditions. The latter are needed to adequately express risks of production, which are associated with adverse conditions that only occur occasionally. Monitoring over such extended periods of time is simply not feasible. Fortunately, in recent years, systems analysis of agricultural production systems, using computer simulation of crop growth, has developed to the extent that methods are operational and very useful for the
4 analysis of sustainable management systems. Examples will be provided in a later section of this paper.
Soil quality
Many studies have been made about soil quality (e. g. Doran and Jones, 1996) but there is as yet not a well defined, universal methodology to characterize soil quality and to define a set of clear indicators. Doran and Jones (1996) present four physical, four chemical and three biological indicators which, according to the authors, together represent a minimal data set to characterize soil quality. But no examples are provided. Gomes et al. (1996) define six indicators and threshold values for measuring sustainability of agricultural production systems at the farm level. Implicitly, higher degrees of sustainability correspond with higher soil qualities. Other examples of soil quality studies as reported by Doran and Jones (1996) list series of soil characteristics as indicators of soil quality but none really address the broad spirit or scope of the Karlen et al. (1997) definition.
A number of general considerations can be made when defining soil quality: Any “fitness of a specific kind of soil to function” depends strongly on climatic conditions, which vary among climatic zones while the weather also varies at any given location during the year. Aside from this, soils are parts of landscapes and their functioning is strongly affected by their position in these landscapes. To consider “soil” without the climatic and landscape context, when defining quality, is not realistic. We should therefore speak about “ land quality” rather than “ soil quality” (Pieri et al, 1995), as the term “ land” expresses the broader context in which the soil functions.
“Fitness to function” will always be considered in practice in relation to other soils. Production levels in years with favorable weather may not be too different among different soils but high quality soils tend to produce well under adverse conditions when other soils don’t deliver. Of course, production figures alone are not enough: cost and quality of produce must be considered as well. When similar yields are reached at lower costs or with a higher quality, the quality of the soil may be considered to be higher as well. But here the management factor is introduced.
A potentially high-quality soil can still have low yields, resulting from poor management, while low-quality soils may have high yields due to excellent management. When defining land quality for a given type of soil, attention should be focussed on effects of a wide range of management types as applied for a given growing season. Studying a single growing season on a given farm will not provide useful information to derive representative land quality indicators. In fact, a high land quality implies that high yields may be obtained even under adverse conditions and under mediocre management. High quality land has a high resilience, which is best described as the ability to bounce back after the effects of poor management or poor weather conditions have been suffered. Low quality land does not have such resilience.
5 The above point described short-term effects of management during a growing season. For instance, compaction of structure when driving over wet land which qualifies as poor management. When the farmer would have waited a few days, compaction might not have occurred. But there is also a long-term effect of management. When poor management leads to, for instance, erosion or strong subsoil compaction, changes in the soil are permanent. Lost soil will not return and subsoil compaction is difficult to remove. Long-term management may also be favorable, for instance when increasing the organic matter content of the soil by organic manuring. Droogers and Bouma (1997) used the term genoform to describe the genetic soil type and phenoform to describe long-term effects of management in the same soil type.
“Fitness to function” is closely tied to land utilization type. The function is quite different in natural ecosystems or in agricultural production systems. When functioning is directed towards:” human health and habitation” one could also think about housing developments and recreation facilities. Here, attention will, arbitrarily, be confined to agricultural production systems, which is in line with the earlier discussion on sustainable management systems. The definition of soil quality mentions:” a specific kind of soil”, which is not further explained. We believe that it would be wise to use soil surveys and soil taxonomy systems here to define specific soil types (genoforms, as mentioned above) that are well defined, also in terms of their positions in landscapes as shown on soil maps. In the USA, the soil series would be a proper “carrier” of information. In the context of the land quality concept, however, emphasis would be on land behavior in terms of crop production, its risks and environmental side effects. The implicit hypothesis would be that each soil series, occurring in a given agro-ecological zone with a characteristic climate, has a characteristic range of production rates, risks and side effects. We may call this a characteristic “ window of opportunity”. Well expressed phenoforms of a given genoform may have significantly different windows! Also, different genoforms have not always different ranges of properties: two soils may be genetically different but may function identically. In all cases such ranges should define land quality. Again, high quality land performs well even under adverse conditions in terms of weather and management. Low quality land performs poorly even under good conditions of weather and management. The big challenge now is to quantify such broad descriptions.
The land quality concept should also indicate whether or not quality can be improved by management, either short- or long term. Is there an absolute theoretical upper level of production? If so, how can it be reached? Can it be reached without adversily affecting environmental quality? How does that level compare with the level of other soils in other areas? Such questions are likely to be asked by farmers and other land users and they need to be addressed.
Considering the above, we decided to use simulation modeling of crop growth, as a function of water and nutrient regimes, as a means to characterize agricultural production of a given
6 type of soil, located within a given climatic zone. This may be either a genoform or well defined phenoforms of a given genoform. Of course, such simulations should be validated by field measurements. Models effectively integrate soil and weather as they calculate daily production values. Indeed, data are thus generated to characterize “ land” qualities. By calculating potential productions, an absolute upper yield level is obtained for any given site (at least for a given plant variety). Potential productions are based on climatic data only as well as plant parameters specific for a given species. Water and nutrient supply are supposed to be optimal while pests and diseases do not occur. Water-limited yields can also be calculated, taking into account the water that can be supplied under natural conditions to the crop at the given site by the soil, again assuming that pests and diseases do not occur. Such yields are lower because water supply is not always optimal. Of course, pests and diseases may occur but they occur independently of the site conditions that are used to calculate water- limited yields. One major advantage of simulation models is that calculations can be made for many years, providing expressions for the effect of the varying weather conditions that could never be obtained by monitoring. This, however, is only true when models have been independently validated using field measurements! Some examples will now be presented to illustrate the proposed procedure.
Land quality indicators
Exploratory studies for seven major tropical geno- and phenoforms Bouma et al. (1998) studied seven major tropical soils to illustrate use of the land quality concept for exploratory purposes (Table 1). Potential and water-limited productions were calculated in terms of “ grain equivalents”, and soil data needed for the model included estimated infiltration rates, depth of rooting and available water. The reader is referred to the source publication for details (Bouma et al., 1998; Penning de Vries et al., 1995a, 1995b). Land quality (LQ) was defined as:
LQ = (yield / potential yield) x 100.
Yield was expressed here as a calculated water-limited yield, but real yields could be used as well. Three hypothetical phenoforms were defined in this exploratory study for each of the seven genoforms in Table 1. Effects of erosion were expressed by removing the upper 40cm to 50 cm of soil, depending on soil properties. Compaction was expressed by restricting rooting to 40 cm, a depth at which a plowpan may occur. Liming expresses a potentially favorable effect of management. Acid subsoils that restrict rooting can be opened up by deep liming. Potential productions, as presented in Table 1, vary between 8 and 23 tons/ha, illustrating climatic effects in terms of radiation and temperature. LQ values were relatively high, indicating relatively high rainfall rates leading to production values that are relatively close to potential values. The Zambia soil, however, occurring in a dryer climate had a LQ of only 50. Real yields can also be used in the equation, rather than water-limited yields. Then, an impression is obtained of the yield gap for the given site and for the actual land quality of
7 the management system being used. Erosion clearly leads to reductions of LQ values, particularly in the Orthic Acrisol in China where erosion results in the occurrence of an acid subsoil near the soil surface which restricts rooting. The effects of compaction are stronger than those of erosion under the assumed conditions here, while liming has a strong effect due to deeper rooting. Values reported are relative LQ values in relation to the value for the site based on water-limited yields. An absolute LQ value can also be defined when calculated yields are compared with the highest potential production that is possible in the world (estimated to be 41.1 tons/ha). Then, LQ values are relatively low in the range of 20-39. The exploratory analysis presented here allows a rough estimate of the relative effects of some management measures, using a simple simulation model and estimates of changes in soil parameters that are associated with certain effects of management. Also, average climatic data were used here. Running the model for a number of years with real-time weather data would have given a better impression about yield stability and risks. Still, the exploratory LQ analysis presented here can, in our opinion, be useful when discussing possible effects of different types of management and in comparing different soils.
Table 1. Potential yields, in terms of grain-equivalents, for seven major soil types of the tropics and derived relative and absolute land quality (LQ) values, based on water-limited and potential yields. Absolute LQ values are derived from the maximum potential yield in the world of 41. 1 ton/ha/yr. Values are derived for the seven genoforms and for three hypothetical phenoforms, expressing the effects of erosion, compaction and liming. (after Bouma et al., 1998).
Soil Name Prot. Prod. Rel. Absolute
Type Tons Land Land
Dry Quality Quality
Matter/ Water Erosion Comp Liming Water
Ha*Yr Limited action Limited 1 Ferric Acrisol China 13 96 85 75 100 32
2 Orthic Ferralsol Indonesia 18 90 75 70 100 39
3Cambic Colombia 12 96 80 72 100 31
Arenosol
4 Ferric Luvisol Nigeria 14 90 75 55 90 32
5 Ferralic Nigeria 14 85 70 50 85 30
Arenosol
6 Orthic Acrisol China 8 90 50 85 100 20
7 Orthic Ferralsol Zambia 23 50 40 30 50 27
8 Yield gap and soil nutrient balances in Africa In a World-Bank funded study, Bindraban et al. (2000) studied land quality indicators for African soils also using potential, water-limited, nutrient-limited and actual yields. In contrast to the exploratory study of Bouma et al. (1998), they used specific data for field sites. The difference between actual and potential yield expresses the yield gap, which is considered to represent the LQ indicator. This is not unlike the approach followed by Bouma et al. (1998) except that the latter authors expressed LQ as a number between zero and one hundred and that they also defined an absolute LQ. The difference between potential and water-limited yield provides an indication what might be achieved by improved water management while the difference between water-limited and nutrient-limited yield indicates the potential impact of improved fertilization. Bindraban et al. (2000) also added a second LQ, which defines the soil nutrient balance. This is highly relevant for Africa where soil mining is rampant. The soil nutrient balance is the net difference between gross inputs and outputs of nutrients to the system and is expressed in relation to the soil nutrient stock. Of course, effects of depletion are much worse in soils with a low stock of nutrients as compared with soils having a high stock. Combining five classes for nutrient stock with five classes of N-P depletion results in six classes for the nutrient depletion LQ.
The attractive aspect of this study is its emphasis on the integrated characterization of the entire land system using quantitative and specifically defined techniques, such as simulation modeling and nutrient budgeting.
Balancing production and environmental requirements in a prime agricultural soil in the Netherlands Application of simulation techniques for a series of growing seasons, expressing climate variability, was realized in a more detailed Dutch case study in a prime agricultural soil (Droogers and Bouma, 1997; Bouma and Droogers, 1998). Three phenoforms were identified in the field and were studied. Here, only conventional arable land is shown (CONV) and land affected by biodynamic farming for a period of 60 years (BIO). The latter soil had a significantly higher organic matter content (appr. 4% versus 2%) as a result of biological management. A more detailed simulation model was used in this study, allowing calculations of crop growth and associated water and nitrogen regimes. Calculations were made for a 15- year period using real-time weather data and a large set of nitrogen fertilization rates. Results could therefore be expressed as probability curves (Figure 1) listing the probability that a certain yield would be exceeded on the vertical scale and the yield itself on the horizontal scale. Three yield curves are shown in the figure and they express the probability (never, 3% and 10% of the years) that nitrate leaching will exceed the environmental threshold value for nitrate content of groundwater. Thus, risks are expressed in quantitative terms by combining the uncertainty of yields due to variable weather in different years with the associated leaching of nitrates. One more curve is shown: the dark solid, and almost vertical, line represents potential yield.
9 1.0 1.0 Bio Conv 0.8 0.8
0.6 0.6 0.4 0.4 P (X 0.2 0.2 0.0 0 2500 5000 7500 10000 0.00 2500 5000 7500 10000 Yield (kg ha-1) Yield (kg ha-1) Exceeding N-leaching Potential never 3% 10% production Figure 1. Probabilities that yields of wheat are exceeded as a function of three probabilities that the threshold value for nitrate leaching to the groundwater is exceeded as well. Data are based on simulations for a 30-year period for a prime agricultural soil in the Netherlands (after Droogers and Bouma, 1997). The graphs in Figure 1 force the user to make choices about risks to take when balancing yield versus nitrate leaching. This is a relevant key problem in Dutch agriculture where production interests have to be balanced against environmental restraints. Again, land quality can be expressed by: (actual / potential yield) x 100. Assuming, arbitrarily, a yield level that has a probability of 20% of being exceeded and a 10% probability that nitrate leaching will exceed the threshold, a LQ value of 89 is obtained for BIO and 84 for CONV. If however, the leaching probability is reduced to 3%, LQ values become 73 and 60. If leaching is never allowed, LQ values become 61 and 33, demonstrating the higher “quality” of the BIO soil. Different LQ values are obtained when different exceedance values for yield probability are used. They can all be derived from Figure 1 as needed. Figures such as Figure 1 are suitable to allow the user to make choices. To the dismay of some, they do not provide clear-cut answers and judgements. Science provides the tools to users to allow them to exercise their responsibility. Science does not take away their responsibility but allows users, be it farmers, planners or politicians, to make more rational choices. Going back to the definition of soil quality, the procedure illustrated here to estimate LQ values, covers yields in different years, risks involved and environmental side effects in terms 10 of nitrate pollution of groundwater, which is the most important problem in Dutch soils. Elsewhere other problems may figure. This comes close to what Karlen et al. (1997) must have had in mind. LQ’s are also indicators for sustainable land management, as defined above. They define production levels over the years and risks involved as well as the quality of water. Management practices, defined in this context, relate only to nitrogen fertilization and management includes, of course, much more than that. Any LQ value indicates a yield gap, being the difference between observed yields and the potential one. A logical question is how to bridge the yield gap. To answer that question all factors, which affect real yields, have to be identified. Land quality at larger scales Land qualities, discussed so far, were derived for individual pieces of land corresponding with a certain soil series and are implicitly focused on the farmer or local land user. Attention was confined to agricultural production and, in the Dutch case study, on important aspects of environmental quality. The definition of soil quality (Karlen et al., 1997) is focused on a “specific kind of soil” and has, therefore, a built-in spatial scale dimension. But the LQ concept not only relates to plots or fields where single kinds of soil occur, but also to larger areas such as communities, regions, countries and even larger entities where many different soils occur. In that context, LQ’s are also important but questions are now asked by politicians, planners and real estate brokers, rather than farmers. How should the quality of land in a large area be judged? A soil map can be used to distinguish all soil series or soil associations that occur in the area and individual land qualities of the different soil series can be considered for different land use categories. An average land quality, weighted by relative areas occupied by the different soil series, could then theoretically be calculated. Functioning of land within natural or managed ecosystem boundaries, sustaining plant and soil productivity, maintenance of soil and water quality and support of human health and habitation” are many but not necessarily all elements of importance when dealing with land in larger areas. Infrastructure is important, but, particularly, socio-economic conditions that determine the population pressure on the land. As with the discussion on fields and farms, we advocate an approach in which the quality of land in a region is first judged in terms of its own inherent properties. Next, these properties will be only one (and as it turns out in real life, a rather minor) factor in determining the most desirable land use in a region which is highly influenced by socio-economic and political considerations and which will always be the result of tradeoffs between many conflicting land use options. The relevant question here is, then, how the quality of the land can play a role when weighing such alternative options. Clearly, restricting the attention to agriculture in this context as the only land-use type is unsatisfactory. Aside from agricultural issues, regional land-use plans are likely to deal with establishing nature areas, transport corridors and locations for housing and industrial development. Because of the relatively high capital expenditure associated with building 11 activities, LQ’s for land in a given area when considering such activities is bound to be relatively unimportant. Draining of potential building sites or adding thick layers of sand to increase the support capacity of land that has a low natural carrying capacity, is financially no problem. For nature areas and, less so, for agriculture, which are much less capital intensive, the picture is different. Local conditions of the land are very important in determining water, nutrient and temperature regimes that govern occurrence of natural vegetations but that also increasingly have an impact on types of agriculture that are ecologically balanced. The approach to take, therefore, would be to define LQ’s for agriculture and nature for land occurring in the area to be considered and to introduce this in time into the broader land-use planning process. LQ’s for agriculture have been discussed above. LQ’s for nature require a separate discussion, which is beyond the scope of this text. Modern land-use planning increasingly uses simulation modeling in the context of systems analysis to derive optimal land-use patterns in a region. This approach was recently demonstrated in Costa Rica (Bouman et al., 1999). Thus, interests of agriculture and nature would, as is often the case, not be a rest-factor left behind after land-use decisions have already been taken based on demands from housing, industry and infrastructure. Rather, these interests of agriculture and nature would be submitted at an early time allowing a significant effect on the decision making process by pointing out where prime land is located for agriculture and nature. Use of LQ indicators can be quite helpful here! Of course, the political process may still ignore this, but nobody would be able to claim afterwards that they did not know. Using the quality concept to improve communications about soils Expression of the productive capacity of land in relation to environmental requirements in terms of a single indicator, which is related to potential production, can be helpful in our opinion to communicate more effectively with stakeholders about use of land in future as compared with current conditions where emphasis is more on the properties of land rather than on its behavior. The indicator, as proposed, provides a quality measure for a given soil type, but by recognizing occurrence of genoforms and phenoforms, it acknowledges important effects of management. Thus, characteristic “ windows of opportunity” are obtained for any soil type occurring in a given agro-ecological zone. Use of actual yields provides an impression of yield gaps, while the absolute quality measure ranks the soil on a global scale. We certainly cannot preserve all land for agricultural production. However, the quality measure discussed here may help to more effectively show which land is particularly valuable and deserves more attention. Studies discussed above have an exploratory character. Before possible implementation of the scheme on a wider scale, attention should be paid to unified procedures for determining potential yields and use of simulation models, including a critical analysis of data demands. 12 Conclusions The problem of worldwide land degradation and indiscriminate use of prime agricultural land for development is insufficiently recognized by society at large. One reason is the rather ineffective manner in which soil scientists present their expertise. We suggest that creative use of the land-quality concept, leading to specific indicators, may raise public awareness of the importance of the land. Existing definitions of soil quality and sustainable land management have several elements in common. An approach is proposed here to define a land quality indicator for sustainable land management focused on agricultural land use which integrates elements of yield, risk and environmental quality using quantitative, reproducible techniques such as simulation modeling or nutrient budgeting. Well intentioned normative and descriptive approaches will not be enough. Socio- economic and political conditions are very important when defining land quality and sustainable land management. We advocate, however, a separate assessment of the agro- ecological potential of the land which should, in an early phase of discussions with all stakeholders, be introduced as independent input into broader land use discussions. The future of the land cannot be held hostage to economic conditions of the day. Land qualities, discussed in literature, have so far implicitly been focused on field and farm level and on agricultural land use, the latter increasingly within an ecologically sound context. Land qualities are also important at larger scales, such as the regional and national level and this requires additional research because a single focus on agriculture is not realistic in this case as many other forms of land use present their demands as well. References Alcamo, J. 1999. The science in diplomacy and the diplomacy in science. Environment Science and Policy 2, 363-368. Bindraban, P. S., Verhagen, A., Uithol, P.W.J. & Henstra, P., 2000. A land quality indicator for sustainable land management: the yield gap. The case of sub-saharan Africa. World Bank Institute report 106. AB-DLO Wageningen, The Netherlands. Bouma, J., 1999. New tools and approaches for land evaluation. The Land 3. 1, 3-10. Bouma, J. & Droogers, P., 1998. A procedure to derive land quality indicators for sustainable agricultural production. Geoderma 85, 103-110. Bouma, J., Batjes, N.H. & Groot, J.J.R., 1998. Exploring land quality effects on world food supply. Geoderma 86, 43-59. 13 Bouman, B.A.M., Jansen, H.G.P., Schipper, R.A., Nieuwenhuyse, A., Hengstdijk, H. & Bouma, J., 1999. A framework for integrated biophysical and economic land use analysis at different scales. Agric. Ecosystems and Environment 75, 55-73. Droogers, P. & Bouma, J., 1997. Soil survey input in exploratory modeling of sustainable land management practices. Soil Sci. Soc. Amer. J. 61, 1704-1710. Doran, J. W. & Jones, A.J. (eds), 1996. Methods for assessing soil quality. SSSA Special Publication 49. Soil Sci. Soc. America, Madison, USA. 410 pp. Food and Agricultural Organization (FAO), 1993. FESLM: An international framework for evaluating sustainable land management. World Resources Report 73. FAO, Rome, Italy. 125 pp. Gomes, A.A., Swete kelly, D.E., Syers, J.K. & Coughlan, K.J., 1996. In: Methods for Assessing Soil Quality. Doran, J.W. & Jones, A.J. (Eds). SSSA Special Publication no. 49, 401-409. Karlen, D.L., Mausbach, M.J., Doran, J.W., Cline, R.G., Harris, R.F. & Schuman, G.E., 1997. Soil Quality: A Concept, Definition and Framework for Evaluation. Soil Sci. Soc. Amer. J. 61, 4-10. Oldeman, L. R. 1994 The global extent of soil degradation. In: Soil resilience and sustainable land use. Greenland, D.J. & Szabolcs, I. (Eds). CAB International. Wallingford, 99-118. Penning de Vries, F. W. T., van Keulen, H. & Rabbinge, R., 1995a. Natural resources and the limits of food production. In: Bouma, J. et. al. (Eds). Ecoregional approaches for sustainable land use and food production. Kluwer Acedemic Press. Dordrecht, The Netherlands. pp 65-89. Penning de Vries, F. W. T., van Keulen, H. & Luyten, J.C., 1995b. The role of soil science in estimating global food security in 2040. In: Wagenet, R.J., Bouma, J. & Hutson, J.L. (Eds). The role of soil science in interdisciplinary research. Soil Sci. Soc. Amer. Spec. Publ. 45: 17-37, Madison, USA. Pieri, C., Dumanski, J., Hamblin, A. & Young, A., 1995. Land Quality Indicators. World Bank Discussion Papers 315. World Bank, Washington, D. C. USA. 51 pp. 14 The soil quality concept: A tool for evaluating sustainability Douglas L. Karlen and Susan S. Andrews USDA-Agricultural Research Service, National Soil Tilth Laboratory, 2150 Pammel Drive, Ames, Iowa 50011-4420, US E-mail: [email protected] Summary Evolution of the soil quality concept in North America and its adoption around the world are briefly reviewed. Simply defined as “how the soil is functioning” within a field, across farms, or within entire watersheds, the soil quality concept is discussed in relation to physical, chemical, and biological indicators that provide the actual measures needed to examine soil management effects such as the stresses that cultivation imposes. Various methods being used to monitor and assess soil quality, including user-friendly scorecards and development of indices, are discussed. Steps associated with the development of soil quality indices are outlined. They include (1) identification of appropriate indicators for various soil functions and/or land uses, (2) selection of an appropriate minimum data set, and (3) development of scoring functions that can be used to facilitate integration of the soil physical, chemical, and biological measurements in an efficient and meaningful manner. We emphasize that soil quality is not an end in itself, but rather that it should be used as a concept for evaluating the combined physical, chemical, and biological effects of various soil management practices. The development and use of soil quality indices as tools for assessing the sustainability of all land management decisions is strongly recommended. Keywords: soil health, conservation tillage, sustainable agriculture, soil management, land use planning, decision aids Evolution of Soil Quality and Soil Health Warkentin and Fletcher (1977) were among the first to suggest developing the concept of soil quality, specifically addressing its relationship to intensive agriculture. They stressed that it was not possible to obtain a quality parameter by defining pure soil as can be done for pure water and that there was no concept relating amounts of soil components to quality. With regard to land capability classes, they argued that those designations provided a quality concept based upon limitations rather than a concept based upon positive potential, and that the latter was needed for intensive agriculture. Warkentin and Fletcher (1977) suggested four criteria as the basis for a future concept of soil quality. They stressed recognizing the (i) increased range of uses for soil resources, (ii) various concerned public groups, (iii) changing priorities and demands of society, and (iv) human or institutional context. Although written for a conference nearly 25 years ago, they pointed out that soil resources were being called upon for (a) recycling and waste assimilation, (b) food and fiber production, and (c) aesthetics and leisure use. These multiple demands brought with them increased public awareness, changing priorities, and different cultural and institutional values with regard to soil 15 resources. The theme for this conference, “Soil Stresses, Quality, and Care” suggests we are continuing to struggle with these same issues. This broader concept of soil quality was not introduced in the North American literature until the mid-1980s. The primary emphasis before that time, with regard to soil resource management, was simply the effects of erosion on productivity (Pierce et al., 1984). However, in the mid- to late 1980s, several reports and books brought attention to the degradation of agricultural soils and its implications for sustainable agriculture and environmental health. In Canada, a report by the Senate Standing Committee on Agricultures launched the subject of soil degradation into the sphere of political interest (Gregorich, 1996). Although this and other similar reports succeeded in sounding the alarm, they were typically short on scientific evidence to support their sometimes dramatic claims. However, there was a strong incentive to focus federal soil science research on soil quality, and in 1990 the Soil Quality Evaluation Program (SQEP) was established in Canada under the broader National Soil Conservation Program. During the same period, Larson and Pierce (1991) functionally defined soil quality, and suggested ways to evaluate it and how it changes due to soil management practices. They defined soil quality as the capacity to function within the ecosystem boundaries and to interact positively with the environment external to that ecosystem. They were also among the first to propose a quantitative formula for assessing soil quality. Very quickly, soil quality was interpreted as a more sensitive and dynamic way to document a soil’s condition, response to management changes, and resilience to stresses imposed by natural forces or human uses. This new paradigm provided the impetus for the Rodale Institute Research Center to sponsor a workshop on “Assessment and Monitoring of Soil Quality” in Emmaus, PA (Haberern, 1992). The consensus among Workshop participants was that soil quality should not be limited to soil productivity, but should encompass environmental quality, human and animal health, and food safety and quality. Interest among policymakers, natural resource conservationists, scientists, and farmers increased rapidly after the U.S. National Academy of Sciences published the book entitled Soil and Water Quality: An Agenda for Agriculture (National Research Council, 1993) and stated that there was a definite need for more holistic soil quality research. This interest resulted in several symposia (Doran et al., 1994; Doran and Jones, 1996) producing several definitions, functions, and uses for which soil quality should be assessed (Doran and Parkin, 1994). In response, Dr. L.P. Wilding, 1994 president of the Soil Science Society of America (SSSA), appointed a 14-person committee (S-581) with representatives from all divisions. Their charge was to define the concept of soil quality, examine its rationale and justification, and identify the soil and plant attributes that would be useful for describing and evaluating soil quality. In the June 1995 issue of Agronomy News, the committee reported that the simplest definition for soil quality is “the capacity (of soil) to function”. An expanded version presents soil quality as “the capacity of a specific kind of soil to function, within natural or 16 managed ecosystem boundaries, to sustain plant and animal productivity, maintain or enhance water and air quality, and support human health and habitation” (Karlen, et al., 1997). The committee reported they had struggled with several different words such as replacing “capacity” with “fitness”. However, because of the interdisciplinary nature of the concept, choice of words became much more difficult than anyone imagined. A similar reaction occurred when the committee suggested using soil quality and soil health interchangeably as suggested in the Canadian report entitled “The Health of Our Soils” (Acton and Gregorich, 1995). Throughout the remainder of the 1990s, soil quality research and technology transfer activities throughout the U.S. moved rapidly in several different directions. Soil quality test kits (Liebig et al., 1996; Sarrantonio et al., 1996), practical, farmer-based scorecards (Romig et al., 1996) and soil resource management programs (Walter et al., 1997) were developed. Soil quality indicator evaluations (Karlen et al., 1999a; Liebig and Doran, 1999) and spatial extrapolation techniques (Smith et al., 1993) were studied. Doran et al. (1996) examined the broader linkages between soil quality (or soil health) and sustainability. Finally, various soil quality indexing approaches (Andrews, 1998; Andrews et al., 1999; Hussain et al., 1999; Jaenicke and Lengnick, 1999; Karlen et al., 1998; Wander and Bollero, 1999) were pursued. The indexing projects were carried out at several different scales and therefore with various degrees of accuracy (Fig. 1). Related Indexing Projects SQ Index Farming Systems Natural Resource SQ test kit Inventory SQ scorecard Predicted Accuracy Predicted Spatial Scale Figure 1. Scale and accuracy tradeoffs associated with soil quality indexing projects. Another milestone with regard to the recent evolution of the soil quality concept was the 1994 reorganization of the USDA-Soil Conservation Service, renamed the Natural Resources Conservation Service (NRCS). This resulted in the creation of the Soil Quality Institute (SQI) by that Agency. Their mission was to “cooperate with partners in the development, 17 acquisition and dissemination of soil quality information and technology to help people conserve and sustain our natural resources and the environment”. By emphasizing outreach and communication, the SQI has been successful in achieving their goal (http://www.statlab.iastate.edu/survey/SQI/sqihome.shtml). Evolution of the soil quality concept in the U.S. has not been without controversy (Sojka and Upchurch, 1999). A legitimate concern is that to date, assessments have generally focused on crop production and ecological functions despite the intention to address multiple soil functions. We suggest that this occurred primarily because the technical disciplines of the people who were among the first to begin examining the concept were primarily soil biology, soil fertility and plant nutrition, and ecology. However, the need to develop a consensus on the proper means to assess soil quality from an environmental perspective was clearly identified by Sims et al. (1997). They stressed the need for soil scientists to take a proactive role in framing, from all perspectives, the debate on soil quality and environmental issues. This includes developing new approaches for quantifying environmental risks posed by soils in agricultural and nonagricultural settings. From a global perspective, contaminant levels and their effects have been more central to the soil quality debate in Canada and Europe (Singer and Ewing, 2000). However, a recent review suggests that many German-language publications are continuing to struggle with how to differentiate the soil quality concept from the numerous definitions and attributes associated with soil fertility phenomena (Patzel et al., 2000). Singer and Ewing (2000) stated that increasingly, contemporary discussion of soil quality includes the environmental cost of production and the potential for reclamation of degraded soils. They continue, stating that “reasons for assessing soil quality in an agricultural or managed system may be somewhat different than reasons for assessing soil quality in a natural ecosystem. In an agricultural context, soil quality may be managed to maximize production without adverse environmental effect, while in a natural ecosystem, soil quality may be observed as a baseline value or set of values against which future changes in the system may be compared.” The major challenge, however, is that determining soil quality requires one or more value judgments and since we still have a lot to learn about soil resources, these issues can not be easily addressed. Sojka and Upchurch (1999) articulated the impact of those value judgments extremely well when they concluded that “our children and grandchildren of 2030 will not care whether we crafted our definitions or diagnostics well. They will care if they are well fed, whether there are still woods to walk in and streams to splash in - in short, whether or not we helped solve their problems, especially given a 30-yr warning.” Whether or not early adopters of the soil quality concept agree or disagree with Sojka and Upchurch’s objections to premature “institutionalizing soil quality” and other factors outlined in their review, we contend that the same long-term goal has been the foundation sustaining all of the efforts to develop and promote the concept. 18 Soil Quality Indices as Tools for Assessing Sustainability A fundamental principle associated with all of the soil quality index projects that we have been associated with is that any assessment process must recognize both inherent and dynamic components. The inherent characteristics are those determined by the basic soil forming factors: parent material, climate, time, topography, and vegetation (Jenny, 1941). They determine why two soils (A and B) will always be different (Fig. 2). Soil A Soil Quality Soil B Time Aggrading Sustaining Soil Quality Degrading To baseline Time Figure 2. Inherent soil quality for soil A and B, and trends identified by indexing. 19 The dynamic fluctuation associated with soil quality assessment results from a combination of current or past land use and anthropogenic management decisions. A second caveat is that soil quality indexing should only be used as a tool to identify positive, negative, or neutral trends associated with specific sets of management practices imposed on a specific soil resource. There are no “magic” or perfect scores or ratings. The sole purpose for developing a soil quality index is to help visualize the integrated effects that land use decisions are having on the physical, chemical, and biological soil properties or processes. AGRICULTURAL SUSTAINABILITY INDEX Environmental Economic Social Quality Sustainability Viability Index Index Index ENVIRONMENTAL QUALITY INDEX Soil Air Water Quality Quality Quality Index Index Index SOIL QUALITY INDEX Physical Chemical Biological Factors Factors Factors Figure 3. Hierarchy of agricultural indices showing soil quality as one of the critical foundations for sustainable land management. Figure 3 illustrates how we view soil quality indices as just one component of a more complete environmental quality index. It also shows that the environmental quality index is one part of an agricultural sustainability index to help land managers make decisions that are the best when examined for their economic, environmental, and social impacts. 20 Current Projects Focused on Measuring Soil Quality Understanding conceptual linkages between soil quality, environmental quality and agricultural sustainability is a first step toward recognizing that soil is a finite, dynamic and living resource that acts as a fragile interface between agriculture and the environment (Doran et al., 1996). Undoubtedly, this understanding was among the factors leading to this conference on “Soil Stresses, Quality, and Care”. However, as our discussions will certainly document, soil quality cannot be measured directly. It must be inferred by measuring changes in various soil attributes or attributes of the ecosystem (Seybold et al., 1997). These "indicators" of soil quality should (i) encompass ecosystem processes and relate to process oriented modeling, (ii) integrate soil physical, chemical, and biological properties and processes, (iii) be accessible to many users and applicable to field conditions, (iv) be sensitive to variations in management and climate, and where possible, (v) be components of existing soil data bases (Doran and Parkin, 1994). They should also be easily measured and reproducible (Gregorich et al., 1994) and sensitive enough to detect changes in the soil resource as a result of human use or degradation (Arshad and Coen, 1992). It would be impossible to use all ecosystem or soil attributes as indicators of soil quality. Thus, a minimum data set (MDS) consisting of selected chemical, physical, and biological soil properties has been suggested for assessment (Larson and Pierce, 1994). One indicator included in almost every published MDS is some measure of soil organic matter (SOM) (e.g., Doran and Parkin, 1994; Gregorich et al., 1994; Larson and Pierce, 1994). Several different SOM fractions, including microbial biomass, water-soluble organic matter, particulate organic matter, and humus or stabilized organic matter, have been included in the various MDS. SOM is one of the more useful indicators of soil quality, because it interacts with numerous soil components. It greatly affects soil physical properties like infiltration, aeration, water retention, aggregate formation, bulk density, and soil temperature. SOM has also been shown to influence biological and chemical properties such as soil pH, buffer capacity, cation exchange, nutrient mineralization, agricultural chemical sorption, the spectral environment for plant seedlings, and the diversity and activity of soil organisms (Doran and Jones, 1996) Descriptive or qualitative indicators have also been used to assess soil quality. Among these are soil crusting or surface sealing, rills, gullies, ripple marks, sand dunes, salt crusts, and standing or ponded water (Arshad and Coen, 1992). Others found that farmers often describe their soil using terms such as loose, soft, crumbly, loamy, earthy smelling, darkly colored, massive, lumpy, or dense (Romig et al., 1995). They also tend to rely on what their senses tell them about soil quality and how it affects tillage and yield more than any specific plant response or soil measurement. In our soil quality studies (Fig. 1), the primary effort has focused on measuring numerous soil physical, chemical and biological properties and using various approaches to select the most appropriate ones for the assessment being made. For Major Land Resource Area (MLRA) 21 evaluations, factor analysis was used to construct “soil quality factors” consisting of several different indicators (Brejda et al., 2000) using data from the U.S. Natural Resource Inventory. Discriminant analysis was then used to identify the factors and indicators most sensitive to land use. The NRCS-SQI Farming Systems Project is developing a spreadsheet tool that predicts soil quality changes based on management practices at the field to farm scale. For our plot- and field-scale studies (Andrews et al., 1999), soil quality indexing has generally involved three primary steps: (1) choosing indicators for a minimum data set (MDS) either by “expert opinion” or using principal component analysis (PCA); (2) transforming indicator scores; and (3) combining the indicator scores into the index (Fig. 4). We have selected this approach so that when management goals focus on sustainability rather than just crop yields, the soil quality index (SQI) can be used as one component nested within a hierarchy of agricultural indices (Fig. 3). MINIMUM DATA SET (MDS) PHYSICAL CHEMICAL BIOLOGICAL PCA-chosen or Expert Opinion-selected variables 1 1 1 SCORING FUNCTIONS 0 0 0 SQI = ƒ (transformed MDS) Figure 4. Protocol for transformation of indicator values into a soil quality index. 22 We are currently using this soil quality indexing protocol to compare organic and conven- tional management practices for vegetable and crop production in the Central Valley of California, to compare long-term ridge-tillage and conventional tillage practices for corn production on deep loess soils, and to quantify transitional effects of converting the deep loess conventional tillage watersheds to either no-tillage or contour strip cropping systems. Data are also being collected to quantify soil quality impacts of grazing cattle in a collaborative study between U.S. and New Zealand scientists. Preliminary results from the California study indicate that (i) an “expert opinion” choice of indicators can be skewed by disciplinary biases and using statistical methods may eliminate that source of error; (ii) use of PCA to triage large data sets can effectively represent variation in the total data set and it appears to select a MDS as well or better than the expert opinion; and (iii) organic and low input plots consistently received higher SQI scores compared with the conventional treatments (Karlen et al., 1999b). Similar evaluations are planned for the other projects. Finally, a soil quality indexing protocol (Andrews, 1999) is being developed to help interpret data collected using the Soil Quality Test Kit (Liebig et al., 1996; Sarrantonio et al. 1996). The primary challenge associated with this multi-agency effort is to create a soil quality scoresheet that is transferable across soil and climatic regions as well as among management practices and priorities. Major tasks for this work-in-progress include: (1) delineate inherent soil quality regions across the U.S., (2) identify indicator ranges for each region, soil, or crop, (3) assign weighting factors to account for management priorities, and (4) calibrate and validate the scoresheet by testing in each region. Data is continuing to be compiled, but thus far, the scoresheet methodology appears to be adequately representing overall soil quality in an easily interpretable format for the tested regions. Summary and Conclusions Factors associated with the evolution of the soil quality concept were examined and use of the concept as a tool for assessing sustainability was reviewed. The fact that soil quality cannot be measured directly has been addressed by identifying and selecting appropriate indicators and sets of indicators for use in developing soil quality indices. Initial results suggest the process can be successful at several different scales and with appropriate accuracy for each of those scales. Additional efforts to develop soil quality indices, especially for non-crop production functions are desperately needed. Hopefully, with the increased environmental quality emphasis of many European studies, an identifiable outcome of this conference on “Soil Stresses, Quality, and Care” will be indexing protocols that truly facilitate solving the soil resource problems that our children and grandchildren will face if we fail to take appropriate action as we begin the 21st Century. 23 References Acton, D. F. and L.J. Gregorich. 1995. The Health of Our Soils – Toward Sustainable Agriculture in Canada. Centre for Land and Biological resources Research, Research Branch, Agriculture and Agri-Food Canada, Ottawa, ON. Andrews, S.S. 1998. Sustainable agriculture alternatives: ecological and managerial implications of poultry litter management alternatives applied to agronomic soils. PhD. Dissertation. University of Georgia, Athens, GA., USA. Andrews, S.S. 1999. Regional scoresheets for interpreting the soil quality test kit. p. 219. Annual Meeting Abstracts. ASA-CSSA-SSSA, Inc., Madison, WI. Andrews, S.S., L. Lohr, and M.L. Cabrera. 1999. A bioeconomic decision model comparing composted and fresh litter for winter squash. Agricultural Systems. 61(3):165-178. Arshad, M.A., and G.M. Coen. 1992. Characterization of soil quality: Physical and chemical criteria. Am. J. Altern. Agric. 7:25-31. Brejda, J.J., T.B. Moorman, D.L. Karlen, and T.H. Dao. 2000. Identification of regional soil quality factors and indicators: I. Central and Southern High Plains. Soil Sci. Soc. Am. J. 64:2115-2124. Doran, J.W., D.C. Coleman, D.F. Bezdicek, and B.A. Stewart. 1994. Defining Soil Quality for a Sustainable Environment. SSSA Spec. Publ. No. 35, Soil Sci. Soc. Am., Inc. and Am. Soc. Agron., Inc., Madison, WI. Doran, J.W., and T.B. Parkin. 1994. Defining and assessing soil quality. p. 3-21. In: J.W. Doran et al., (ed.) Defining Soil Quality for a Sustainable Environment. SSSA Spec. Publ. No. 35, Soil Sci. Soc. Am., Inc. and Am. Soc. Agron., Inc., Madison, WI. Doran, J.W., and A.J. Jones. 1996. Methods for Assessing Soil Quality. SSSA Spec. Publ. No. 49, Soil Sci. Soc. Am., Inc., Madison, WI. Doran, J.W., M. Sarrantonio, and M.A. Liebig. 1996. Soil health and sustainability. p. 1-54. In: D.L. Sparks (ed.) Adv. Agron., Vol. 56., Academic Press Inc., San Diego, CA. Gregorich, E.G., M.R. Carter, D.A. Angers, C.M. Monreal, and B.H. Ellert. 1994. Towards a minimum data set to assess soil organic matter quality in agricultural soils. Can. J. Soil Sci. 74:367-385. Gregorich, E.G. 1996. Soil Quality: A Canadian perspective. Proc. Soil Qual. Indic. Worksh. Feb 8-9, 1996. Ministry of Agric. and Fisheries, and Lincoln soil Quality Res. Cntr. Lincoln Univ., Christchurch, NZ. Haberern, J. 1992. Coming full circle – The new emphasis on soil quality. Am. J. Altern. Agric. 7:3-4. Hussain, I., K.R. Olson, M.M. Wander, and D.L. Karlen. 1999. Adaptation of soil quality indices and application to three tillage systems in southern Illinois. Soil Till. Res. 50:237- 249. Jaenicke, E.C. and L.L. Lengnick. 1999. A soil-quality index and its relationship to efficiency and productivity growth measures two decompositions. Am. J. Agric. Econ. 81:881-893. 24 Jenny, H. 1941. Factors of soil formation, a system of quantitative pedology. McGraw Hill, N.Y. Karlen, D.L., M.J. Mausbach, J.W. Doran, R.G. Cline, R.F. Harris, and G.E. Schuman. 1997. Soil quality: A concept, definition, and framework for evaluation. Soil Sci. Soc. Am. J. 61:4-10. Karlen, D.L., J.C. Gardner, and M.J. Rosek. 1998. A soil quality framework for evaluating the impact of CRP. J. Prod. Agric. 11:56-60. Karlen, D.L., M.J. Rosek, J.C. Gardner, D.L. Allan, M.J. Alms, D.F. Bezdicek, M. Flock, D.R. Huggins, B.S. Miller, and M.L. Staben. 1999a. Conservation reserve program effects on soil quality indicators. J. Soil Water Conserv. 54:439-444. Karlen, D.L., S.S. Andrews, and J.P. Mitchell. 1999b. A soil quality index for vegetable production. p. 219. Annual Meeting Abstracts. ASA-CSSA-SSSA, Inc. Madison, WI. Larson, W.E., and F.J. Pierce. 1991. Conservation and enhancement of soil quality. p. 175- 203. In: J. Dumanski et al. (ed.) Evaluation for sustainable land management in the developing world. Vol. 2: Technical papers. Proc. Int. Worksh., Chiang Rai, Thailand. 15-21 Sept. 1991. Int. Board for Soil Res. and Management, Bangkok, Thailand. Larson, W. E. and F. J. Pierce. 1994. The dynamics of soil quality as a measure of sustainable management. p. 37-51. In: J.W. Doran et al., (ed.) Defining Soil Quality for a Sustainable Environment. SSSA Spec. Publ. No. 35, Soil Sci. Soc. Am., Inc. and Am. Soc. Agron., Inc., Madison, WI. Liebig, M.A., J.W. Doran, and J.C. Gardner. 1996. Evaluation of a field test kit for measuring selected soil quality indicators. Agron. J. 88:683-686. Liebig, M.A., and J.W. Doran. 1999. Impact of organic production practices on soil quality indicators. J. Environ. Qual. 28:1601-1609. National Research Council. 1993. Soil and water quality: An agenda for agriculture. Natl. Acad. Press, Washington, DC. Patzel, N., H. Sticher, and D. Karlen. 2000. Soil fertility – Phenomenon and Concept. J. Plant Nutr. Soil Sci. 163:000-000 (in press). Pierce, F. J., W. E. Larson, and R. H. Dowdy. 1984. Soil loss tolerance: Maintenance of long-term soil productivity. J. Soil and Water Conserv. 39:136-138. Romig, D.E., M.J. Garlynd, R.F. Harris, and K. McSweeney. 1995. How farmers assess soil health and quality. J. Soil Water Conserv. 50:229-236. Romig, D.E., M.J. Garlynd, and R.F. Harris. 1996. Farmer-based assessment of soil quality: A soil health scorecard. p. 39-60. In: J.W. Doran and A.J. Jones (eds.) Methods for Assessing Soil Quality. SSSA Spec. Publ. No. 49, Soil Sci. Soc. Am., Inc., Madison, WI. Sarrantonio, M., J.W. Doran, M.A. Liebig, and J.J. Halvorson. 1996. On-farm assessment of soil quality and health. p. 83-105. In: J.W. Doran and A.J. Jones (eds.) Methods for Assessing Soil Quality. SSSA Spec. Publ. No. 49, Soil Sci. Soc. Am., Inc., Madison, WI. Seybold, C.A., M.J. Mausbach, D.L. Karlen, and H.H. Rogers. 1997. Quantification of soil quality. p. 387-404. In: R. Lal et al. (ed.) Soil Processes and the Carbon Cycle. CRC Press, Inc., Boca Raton, FL. 25 Sims, J.T., S.D. Cunningham, and M.E. Sumner. 1997. Assessing soil quality for environmental purposes: Roles and challenges for soil scientists. J. Environ. Qual. 26:20- 25. Singer, M.J., and S. Ewing. 2000. Soil Quality. p. G-271-G-298. In: M. E. Sumner (Ed.-in- Chief) Handbook of Soil Science. CRC Press, Boca Raton, FL. Smith, J.L., J.J. Halvorson, and R.I. Papendick. 1993. Using multiple-variable indicator kriging for evaluating soil quality. Soil Sci. Soc. Am. J. 57:743-749. Sojka, R.E. and D.R. Upchurch. 1999. Reservations regarding the soil quality concept. Soil Sci. Soc. Am. J. 63:1039-1054. Walter, G., M.M. Wander, and G.A. Bollero. 1997. A farmer-centered approach to developing information for soil resource management: the Illinois soil quality initiative. Am. J. Altern. Agric. 12:64-72. Wander, M.M. and G.A. Bollero. 1999. Soil quality assessment of tillage impacts in Illinois. Soil Sci. Soc. Am. J. 63:961-971. Warkentin, B.P., and H.F. Fletcher. 1977. Soil quality for intensive agriculture. p. 594-598. Proc. Int. Sem. on Soil Environ. and Fert. Manage. in Intensive Agric. Soc. Sci. Soil and Manure, Natl. Inst. of Agric. Sci., Tokyo. 26 Factors influencing resilience and resistance in Norwegian silt loam soils Tore E. Sveistrup1 and Trond K. Haraldsen2 1Planteforsk, The Norwegian Crop Research Institute, Holt Research Centre, N-9292 Tromsø, NORWAY E-mail: [email protected] 2Jordforsk, Centre for Soil and Environmental Research, N-1432 Ås, NORWAY Summary Norwegian loamy soils show wide variability in soil resilience and resistance to stress induced by management in the crop production. Soils from southern and northern Norway have been studied. Amongst the soil parameters investigated the development of macroporosity and roots seemed to be the most useful tools as indicators for the soil resilience and resistance, and thus also as indicators for the soil quality in crop production. In the field the soil structure and morphology was studied with special emphasis on macroporosity and root development, and soil samples were collected for physical, chemical and morphological laboratory studies. Cultivation of a silt loam soil low in clay content from the north lead to collapse of the soil structure, which was present in the non-cultivated soil of the same origin. This strongly reduced the root penetration and water infiltration. Comparison of structure stability in paired non-cultivated and cultivated soils of same origin from the south, showed increased structural stability in the cultivated soil. The main reason of this difference seemed to be the presence and continuous creation of tubular macropores in the soil from the south, and a complete lack of such pores in the soil from the north. Cultivation of silt loam soil more rich in clay did not show the same difference between soils from south and north. The main difference seemed to be a drying of the soil from the north, which kept up the resistance of the soil to the forces induced by agricultural practice. Keywords: macro porosity, silt loam, soil resilience 27 Introduction Norwegian loamy soils show wide variability in soil resilience and resistance to stress induced by present day management in the crop production. Soil climate, biological activity and cultivation history are main factors leading to the present day status of the soils. Standard soil physical investigations do give to poor or often nearly no information about continuity of the macropores and thus also the conditions for the transport processes taking place in the soil and for the root development, which are crucial for the soil resilience and the resistance to the cropping. Studies of the soil morphology have therefore been conducted together with standard soil physical investigations at different locations in Norway to get improved information about the conditions for and about the transport processes and the root development and thus also of the soil resilience and resistance. Materials and Methods Materials A set of soils of comparable origin and parent material but from different climatic regions of Norway has been studied. The studies comprise of soil profile investigations on vertical and on horizontal sections at different soil depth, which was followed up by detailed laboratory studies of selected soil samples. Special emphasis was given to studies and description of the soil morphology, tubular macropores and the root development. Silt loam soils, low in clay on an alluvial floodplain, from Tana in northern Norway, 70° N, (Sveistrup 1992; Sveistrup et al. 1995) and from a one event lacustrine sediment at Vansemb in southern Norway, 60°N, have been compared. At both places forest soils (never cultivated) and cultivated soils are studied. At Vansemb the soils of an afforested pasture are studied too. The cultivation has taken place ≈ 60-70 years at both places. At the time of investigation a seven-years old ley was present at Tana and newly sown spring cereals at Vansemb. Marine silt loam soils, higher in clay content than the former soils, also from northern and southern Norway, have been studied in the same way. The soils were located at Alta, 70° N, and at Landvik, 58°N. The soils at both locations are under cultivation. The Alta soil has been cultivated for ≈ 60 years with hey and grass for silage production and for cattle grazing. The Landvik soil has for several centuries been cultivated in rotation with annual crops and perennial grasses. At Alta there was a four-year-old ley at the time of investigation and at Landvik ploughed land after vegetables the former year. Methods At each site the soil investigations and sampling took place on vertical and horizontal soil surfaces at different soil depth. Special emphasis was put to recording of structure and soil morphology, macroporosity and root development. Horizontal soil surfaces of size 50 cm x 50 cm were prepared and cleaned carefully by scraping, brushing and blowing away loose soil 28 material to visualise the features to be recorded as well as possible. The structural system was recorded in detail. Macropores with diameter > 1 mm were classified into diameter classes and counted on the same horizontal sections. In general, special emphasis was put to the related distribution of soil features such as: location of roots according structural development, locations with signs of water flow etc. During the field investigations (the macro scale) soil samples were collected for laboratory studies. The laboratory studies were divided into chemical, physical (methods see referred literature), and morphological investigations. The morphological studies were divided into meso scale studies where undisturbed samples were studied under a binocular stereo microscope keeping the three dimensional view intact, and micro scale studies where selected soil samples were impregnated and thin sections of ±30 µm were studied under a polarization microscope (Bullock et al. 1985, Murphy 1986). Results and Discussion The silt loam soils low in clay, Vansemb and Tana (table 1) The climate at Tana is colder and the summer shorter than at Vansemb. The summer period (June-August) is about 4 ° C colder in Tana. Somewhat more precipitation is falling at Vansemb than at Tana. Except for the plough layer of the Vansemb soil, which in former time was added clayey material to improve soil fertility, the soils are low in clay content, 7% or less. The bulk density was highest in the cultivated soil at both sites. The afforested Vansemb soil showed lowest bulk density of all the compared soils for both horizons. Cultivation has also decreased the airfilled porosity; at Vansemb nearly to the half, at Tana to about 1/5 of non-cultivated soil. In the forest soil no tubular macropores created by earthworms (diameter > 2 mm) were found on the horizontal soil sections at any of the sites. In the afforested Vansemb soil and in the cultivated soil at the same place, more than 100 earthworm channels per m2 were found. In the cultivated Tana soil no traces of activity of burrowing animals were detected. At Vansemb several tubular macropores with diameter 1-2 mm were found in all the soil profiles, including the forest soil which never had been cultivated. This is expressed as area of tubular macropores in table 1. No such macropores were found in the Tana forest soil. The structural development, both macro and micro, showed a very distinct platyness in the top soil at Vansemb and in the Tana forest soil. In the cultivated Tana soil there was also a platy macro structure, but no such micro structural development was found, just a massive soil mass. The rooting depth was deepest in the Vansemb soil, down to 50-60 cm in the cultivated and more than 1 m at the other sites. Below ploughing depth and below 10-15 cm in non- cultivated soil, almost all roots were located to tubular macropores or between structural aggregates. In the non-cultivated Tana soil roots were found down to about 40 cm. The roots deeper than 5-10 cm’s depth were mainly found between the platy structural aggregates. 29 In the cultivated Tana soil the roots did not reach deeper than 10 cm and mostly stopped in the upper few cm of the soil. The content of organic carbon was low when reaching a depth of 10 cm for the non-cultivated soils at both sites. The pH of the forest soils was comparable for the two sites, while the base saturation in the forest soil was highest in the Tana soil. Soils of this origin and texture have by nature a weak skeleton. In their natural state a platy macro and micro soil structure is created by winter frost and the desiccation of the soil created by the frost. This was clearly seen in the forest soils at both sites. By cultivation and traffic load this structure was destroyed, and in the Tana soil replaced by massive coarse platy structure. This soil has got a strongly increased and high bulk density, decreased porosity, low water infiltration capacity (Haraldsen & Sveistrup 1994) and as a result a shallow and poor root development. The Tana silt loam soil has a low resilience and a low resistance to agricultural use. In the Vansemb soil there has also been an increase in bulk density due to cultivation though less than in the Tana soil, and also a decrease in porosity. However, these negative results due to cultivation have been compensated by introduction of deep burrowing earthworms. In the channels created by the earthworms traces of water transport were recorded. Roots were following these macropores and not only interaggregate space as in the cultivated Tana soil. Therefore the rooting depth was much deeper in the Vansemb soil. The physical and chemical state of the non-cultivated soils is from nature more favourable in the Tana soil than in the Vansemb soil. The main reason leading to the different direction of development in soil resilience and resistance as a consequence of cultivation, seems to be the introduction of earthworms at Vansemb following the improvements in soil conditions due to supply of nutrients and lime and incorporating organic matter into the soil. Since this has been done at both sites, the climatic conditions and especially higher temperatures seem to be the reason of earthworm establishment in the Vansemb cultivated soil and not in the Tana cultivated soil. The silt loam soils medium in clay, Landvik and Alta (table 2) The summer temperature, from June to August, is about 4° C colder at Alta than at Landvik. The precipitation is much higher at Landvik, also during the summer months. The clay content of the soils is 13-15 % except for the deeper Alta soil. The bulk density of the plough layers in the cultivated soils is quite similar for the two sites, while it is much higher in the forest soil from Alta. Below the plough layer the bulk density of the two soils from Alta is comparable, and much higher than for the soil at the same depth from Landvik (Sveistrup et al., 1997). The porosity and the airfilled porosity give the same picture as for the bulk density. The number of tubular macropores is much higher in the Landvik soil. Here was also found much present day earthworm activity. At Alta no activity of present burrowing animals was found. 30 Table 1. Soil and climatic parameters in different depth in soils under different management at Vansemb and Tana. Vansemb Tana Forest Afforested Cultivated Forest Cultivated pasture Texture, sand-silt-clay, % 10-14 cm 22-77-1 8-86-6 8-77-14 26-67-7 48-46-6 25-34 cm 26-73-1 21-81-3 17-80-3 74-23-3 65-31-3 Bulk density, Mg m-3 10-14 cm 1.12 1.01 1.33 1.14 1.44 25-34 cm 1.28 1.22 1.39 1.34 1.49 Porosity, m3m-3 10-14 cm 0.53 0.55 0.46 0.58 0.46 25-34 cm 0.51 0.50 0.50 0.52 0.47 Airfilled porosity, m3m-3 10-14 cm 0.12 0.14 0.07 0.34 0.06 25-34 cm 0.13 0.09 0.08 0.26 0.07 Tubular macropores, (2-4 mm) number m-2 10-14 cm 0 29 nd1) 00 25-34 cm 0 119 156 0 0 Tubular macropores, (1-4 mm) area cm2 m-2 10-14 cm 2 7 nd 0 0 25-34 cm 2 12 18 0 0 pH, H2O 10-14 cm 4.8 4.6 6.1 4.8 5.5 25-34 cm 4.8 4.9 5.9 5.1 5.5 Organic C, g 100 g-1 10-14 cm 0.3 0.5 1.7 0.8 2.0 25-34 cm 0.2 0.4 0.3 0.4 0.4 CEC sum cations, cmol kg-1 10-14 cm 4.1 7.7 7.0 2.6 4.0 25-34 cm 2.8 6.0 1.6 1.7 2.2 BS sum cations, % 10-14 cm 2 4 64 22 92 25-34 cm 3 5 32 45 87 Macro structure 10-14 cm platy/blady platy/blady clods platy platy/massive 25-34 cm massive massive/platy platy platy massive Micro structure 10-14 cm lent.banded lent.banded nd lent.platy massive 25-34 cm single grains banded single grains nd nd Rooting depth, cm >100 >100 50-60 35-40 <10 Temp. year, °C, normal 4.2 -0.7 Temp. June-August, °C, normal 14.6 10.5 Precip. year, mm, normal 670 455 Precip. June-August, mm, normal 210 137 1) nd: not detected 31 Table 2. Soil and climatic parameters in different depth in soils under different management at Landvik and Alta. Landvik Alta Cultivated Forest Cultivated Texture, sand-silt-clay, % 10-25 cm 26-60-14 4-81-15 3-83-14 25-45 cm 14-73-14 1-75-24 4-83-13 Bulk density, Mg m-3 10-25 cm 1.02 1.57 0.94 25-34 cm 1.23 1.69 1.60 Porosity, m3m-3 10-14 cm 0.57 0.39 0.62 25-34 cm 0.51 0.35 0.39 Airfilled porosity, m3m-3 10-14 cm 0.20 0.05 0.19 25-34 cm 0.12 0.04 0.05 Tubular macropores, number m-2 25-34 cm 685 ≈ 350 ≈ 150 Tubular macropores, area cm2 m-2 25-34 cm 80 <1 <1 pH, H2O 10-14 cm 6.0 6.2 6.0 25-34 cm 6.2 7.4 6.1 Organic C, g 100 g-1 10-14 cm 3.4 3.0 4.9 25-34 cm 0.3 0.2 0.4 CEC sum cations, cmol kg-1 10-14 cm 19 22 21 25-34 cm 11 16 7 BS sum cations, % 10-14 cm 39 67 66 25-34 cm 23 87 50 Macro structure 10-14 cm clods platy/blady clods 25-34 cm platy platy lent. platy Micro structure 10-14 cm nd1) lent. platy blocky 25-34 cm nd lent. platy lent. platy Rooting depth, cm >80 >100 >100 Temp. year, normal 6.9 0.3 Temp. June-August, normal 15.4 11.4 Precip. year, normal 1230 435 Precip. June-August, normal 276 160 1) nd: not detected 32 The area of the tubular macropores on horizontal soil sections was much higher at Landvik since the diameter of the pores here also was bigger. The structural development was of the same type, platyness at both sites. At Alta the platy structural aggregates were more complete, more lenticularly shaped and also finer in size. This development of a platy structure seems partly to be guided by a sedimentary layering, which is transformed into a platy soil structure. The sedimentary layering was most distinct in the Alta soil. The rooting depth was deep in all soils. Also here the roots below ploughing depth or below ≈ 10 cm in the forest soil were located to tubular macropores or between structural aggregates. The results of the soil chemical analysis show comparable results for all the soils except for base saturation, which was higher in the Alta soil than in the Landvik soil. A resistant soil structure is by nature build up in the Alta soil. The structural aggregates resist for a great part the cultivation and the forces induced by traffic, or they get restored by natural, structural building processes. The climate at Alta allows the soil to dry up quite well during summer and together with the winter freezing of the soil, a lenticular macro- and microstructure is developed. Micromorphological investigations showed how water has been passing in the aggregate interspace and created zones depleted for iron and clay along the surface of the aggregates. The bulk density of the plough layer at Alta has decreased compared to the forest soil and is nearly the same as in the plough layer of the Landvik soil. Also the porosity in the plough layer has changed in a positive direction in the point of view of plant growth. The tubular macropores of the Alta soil are probably relict since no indication of new formation of such pores was observed, but they are still acting as passages for water flow and root penetration. In the Landvik soil a high amount of earthworm channels is created and new are constantly being formed. These two soils are keeping up a high soil resilience and resistance to agricultural practice. In the Landvik soil where the precipitation is quite high, the high amount of tubular macropores drain the soil and give passages to the roots. Here the frost action plays a minor role. In the Alta soil with the relative low precipitation, and no earthworm activity, the drying processes during summer and during the winter frost in the soil seem to be the main factors stabilizing the soil structure. Conclusions Investigations of soils from northern and southern Norway have shown that the resilience and resistance of silt loam soils low in clay content were strongly influenced by the formation of tubular macropores. Where such pores have been created, cultivation of the soil may increase the resilience and resistance of the soil to external forces. Where such pores have not been created, due to climatic or other factors, the soil may become less resistant to stress. For silt loam soils more rich in clay, the resistance against external forces and restoration capacity after disturbance might be quite high. This is especially so if the soil has the possibility to dry 33 up during summer and during winter frost and thus create passages for water and air and roots, even when tubular macropores are not created. Macropore development and/or rooting depth seem to be useful tools as indicators for the resilience and resistance of soil to external forces. References Bullock, P., Fedoroff, N., Jongerius, A., Stoops, G. & Tursina, T., 1985. Handbook for Soil Thin Section Description. Wolverhampton, Waine Research Publ. 152 pp. Haraldsen, T.K. & Sveistrup, T.E., 1994. Effects of cattle slurry and cultivation on infiltration in sandy and silty soils from northern Norway. Soil & Tillage Research 29, 307-321. Murphy, C.P. 1986. Thin section preparation of soils and sediments. AB Academic publishers. Berkhamsted, Hertz, Great Britain. 149 pp. Sveistrup, T.E. 1992. Morphological and physical properties of virgin and cultivated silty and sandy soils from Finnmark, Northern Norway. Norwegian Journal of Agricultural Sciences 6, 45-58. Sveistrup, T.E, Marcelino, V. & Stoops, G., 1995. Effects of slurry application on the microstructure of surface layers of soils from northern Norway. Norwegian Journal of Agricultural Sciences 9, 1-13. Sveistrup, T.E., Haraldsen, T.K. & Engelstad, F., 1997. Earthworm channels in cultivated clayey and loamy Norwegian soils. Soil & Tillage Research 43, 251-262. Acknowledgements The results presented originate mainly from the project “Processes in macropore systems in Norwegian silt loam soils” financed by The Norwegian Research Council, Norwegian Centre for Soil and Environmental Research and The Norwegian Crop Research Institute, Holt Research Centre. Results from other projects are also included. 34 Soil quality with chromatography Solveig B. Nyborg Hardanger-Midthordland Forsøksring, N-5610 Øystese, NORWAY E-mail: [email protected] Summary The American scientist Dr Ehrenfried Pfeiffer (1899 -1961) developed an application of chromatography in regard to testing the biological quality of soil, compost and plant products. The chromatogram reveals differences in soils which may show almost identical values on a traditional chemical soil analysis, but that may differ greatly in biological and microbiological characteristics. The chroma-test rests upon the properties of specially manufactured filter paper discs, in which individual fractions of different substances are separated by capillary action. The resulting picture allows interpretation according to distinct differences in colours and patterns, related to differences in molecular weight, size and condition of the substances. For interpretation purposes the chromatogram is compared with a standard series. Since 1995 I have studied and practised chroma-testing in Austria and Norway, and have produced several hundred chromas of different Norwegian soils. The soils tested have a wide range of quality, from soils which are very poor in microbial life and ability to build up humus, to those which are very rich in microbial life and humus. To me it appears that soil quality depends more on farming practices than on given soil types and textures. Soil samples coming from organic farms often show better chromas than conventionally cultivated fields. My conclusion is that soil quality testing by means of chromatography is a useful and important supplement to traditional chemical soil analysis. Even though chromatography is a qualitative test, it provides fast in depth information about diversity of microlife as well as the total microbial activity. A conventional microbial count is not only expensive, but also takes from days to weeks, depending on the objectives. Whereas the chroma methodology is simple, and by far not as time consuming. The chroma test provides an overview of the soils biological condition and one can derive conclusions from it on how to improve the soil’s fertility. In field experiments the chroma-test shows the impact of different treatments on soil quality in a direct and detailed manner. Keywords: chromatography, soil quality testing, microbial life, humus 35 Introduction Chromatography has been used in urine analysis since 1944, but in the 1950’s Dr. Ehrenfried Pfeiffer, an American microbiologist, dipl. agronomist and biochemist, started to apply this method to test soil and compost quality (Pfeiffer, 1960).The chromatogram reveals differences in biological and microbiological characteristics of soils, where traditional chemical soil analysis may show almost identical chemical properties. These differences - in turn - are experienced by farmers as differences in plant health, yields and product quality. Later during his research Dr. Pfeiffer also used the chroma method for differentiating the quality of food or vitamins, which can be almost identical by chemical standards but differ greatly in biological and qualitative properties. Qualitative in this context describes the total content, as well as the properties of ‘bioactive’ substances. Today chromatography is a well- known test method in many different contexts. In 1995, I was introduced to soil quality testing by means of filter-paper-chromatography (according to Dr. Ehrenfried Pfeiffer) by Mrs. Uta Lübke from Austria. Siegfried and Uta Lübke have used chromatography in their research for nearly 30 years. Their work has been focusing on field studies in combination with scientific research. Uta and Siegfried Lübke have been aiming for holistic insights into soil functions and life cycles in nature, in all their work, and have therefore always resorted to methods of analysis that will require as little destruction of a natural condition as necessary. This is one reason they found chromatography being such a useful method to supplementing their other testing methods. The goal of their work has always been the development of an organic cultivating system, which builds up and maintains a microbially rich, well-balanced and fertile soil. The findings of their determined research is the rather well known “CMC-method” (Controlled Microbial Composting) and a cultivation system called “Humusmanagement”. The CMC-method is acknowledged by European governments, which has led to mandatory guidelines for composting facilities as well as research projects in cooperation with universities (Witzenhausen), municipalities (e.g. Zürich and Hamburg) and federal states (e.g. Tirol and canton Zug). Currently there is a 10 years research project being conducted in cooperation with an agricultural school to show that Humusmanagement has a long term effect. In this project Chromatography has already proven to be a valuable tool for monitoring soil improvements. The microbiological test conducted in this project has only given an indication for an improvement of the test fields, whereas chromatography has already revealed distinct differences (personal message from Angelika Lübke, 2000). What I learned in Austria, opened my eyes for the importance of soil microbial life in regard to nutrition and the health of plants, and - in a wider aspect - the health of animals and humans. I wanted to learn more about the subject, especially about the methods used to test 36 soils, compost and plant quality. Since 1995 I have studied and practised chroma-testing in Austria and Norway, and have done several hundred chromas of different Norwegian soils. Materials and Methods The Chromatogram is a qualitative test. It rests upon the properties of specially manufactured filter paper, by which individual fractions of a soil, compost or plant extract are separated through capillary action. The filter paper is prepared with silver nitrate, which makes the different fractions visible. The resulting picture allows interpretation through distinct differences in colours and patterns, related to differences of the diluted substances in molecular weight, size and condition. The interpretation can be done by anyone who studies the pictures carefully, comparing the actual chromatogram with a standard-series. One of them has been published and interpreted in Voitl & Guggenberger (1986). A short summary of the method A hole is punched through the middle of the filter paper disc. A wick is placed in the center hole so that it touches the bottom of a little petri dish (diam. 3,5 cm). Pencil marks are marking the 4 cm and 6 cm distances from the center hole, to indicate how far the solutions should spread, when absorbing from the petri-dish. The small petri-dish is placed in the center of a larger petri-dish with a diameter of 9 cm, to keep the filter paper level. For the first preparation silver nitrate is poured into the small petri dish and the solution is allowed to spread into the disc until it reaches the 4 cm mark. Then the filter paper is removed for drying, before a new wick is inserted into the center hole. The soil extract (or compost or plant extract) has been prepared well in advance of paper activation. About 5 grams of soil is placed in a 100 ml Erlenmeyer flask and 50 ml 1% NaOH solution is added. The flask is twirled for complete mixture and has to extract 6 hours before decanting. The extract is then poured into the small petri-dish. In a next step the petri-dishes with the filter paper on top, are placed in a chroma-box with a glass lid. A temperature of 30°C and 80 % relative humidity are necessary to allow for the soil solution to spread, until it reaches the 6 cm mark. A reaction between the soil solution and the silver nitrate takes place and, after drying, colours and patterns will develop within two or three days, when the filter paper is exposed to daylight. Every sample is prepared twice. Once on filter paper quality no.1 (Whatman Filter paper 1 Qualitative Circles 150mm Ø, Cat No 1001 150), and once on filter paper quality no.4 (Whatman 4 Qualitative Circles 150mm Ø, Cat No 1004 150), which has a larger poresize than no.1. The picture on filter paper no.4 (right hand side on the pictures below) can be considered a ‘magnification’ of picture of no.1, where some of the details are “blown up”. 37 Filter paper no. 4 absorbs more of the filtrate than paper no. 1, and therefore a higher density of substances will create a visible difference between the two paper qualities. The chromatogram is interpreted by looking at the shape, patterns and colours of the different zones (the center, the middle and the outer zone) which are then compared to standards. In principle the chroma will show more complex patterns and colours the more ‘complex’ the extract of soil, compost or plants. The ‘make-up’ of a soil extract indicates the variety of biological activity of the sample, because every species produces different organic-biological compounds. The soil type and mineral texture has relatively little direct influence on the appearance of the chroma. For a soil you can “read” various important information out of the chroma: In general, how fertile and vital the soil is. In special (very roughly described): - An overview of the bacterial and fungal activity (patterns, colours and shapes of points in the middle zone of the chroma). - The soil’s content of crude organic material (shape of outer zone). - The soil’s capacity to decompose organic material, and to build up colloidal stable humus structures. - The capacity to withhold excess nutrients during the winter (colour and shape of outer zone or absence of outer zone). - How well the clay minerals and the organic substances in soil are linked together /aggregation (colour and size of center zone). - Soil structure (shape of spikes from the center to the edge). The relationships between chroma features and soil properties were first demonstrated and described by Pfeiffer (1960). More about the technique and the interpretation of chromas can be found in Voitl & Guggenberger (1986) and Nyborg (1999). Results and Discussion During the last 5 years I have, out of my own interest, collected soil samples from different districts in Norway. Even though I have not had the opportunity to do any systematic research or trials by now, some of the results of this testing should be of interest in the context of this seminar. The soils tested have a wide range of quality, from nearly ”dead” soils, which are very poor in microbal life and ability to build up humus, to those which are very rich in microbial life and humus. It is my impression that the soil quality depends more on farming practices than on soil types and textures. Often – but not always – the soil samples show a better quality on organically than on conventionally cultivated fields. 38 When a chromatogram indicates a healthy and microbially active soil, I often call the farmer who has sent the soil sample. And regardless of soil type and the results of chemical soil analysis I always get the answer that the soil in question is an especially fertile and high yielding one. Some examples of chroma-tests on Norwegian soils: Figure 1. This is the chroma of an organic-biologically farmed field on sandy soil in Jæren. It shows a clear white, center zone, which indicates good integration between organic and mineral material in the soil. The structures/patterns in the inner and middle zones appear nearly three- dimensional and show a wide variety of bacterial and fungal activity. In the outer zone one can see diffuse spots of a light-brown colour (clouds), which means the soil contains “friable-humus”, where the nutrients are tied into ‘short chemical chains’ and are easily accessible for the plants. In other words, this chroma picture shows an active and fertile soil, which is rich in humus. Figure 2. This is a chroma of an extensively farmed pasture on coarse sand with a naturally high pH (>8) in Vestvågøy, Lofoten. The meadow is fertilized with small amounts of sheep manure and the yields are low. The grayish colours indicate a low content of organic material. The picture lacks the white, center zone, which means there is little aggregation between organic and mineral matter in the soil. The inner and middle zone are poor in patterns, have a diffuse grayish-brown colour, and points of a similar shape and size, which represent a microflora with only little diversity. Lack of clouds in the outer zone (also called protein zone) show that there is no recent turnover of organic matter. The chroma shows evidence for a very unfertile soil, poor in microbial life, organic matter and humus. Figure 3. This is a chroma of the same soil type as fig.2 and from the same area. In this field crop rotation has been practiced for years, and because it is located near a barn, it has received much more manure. The farmer is very content with the production and the product quality. One can notice a white center, fine structures and clouding in the outer zone. There is much nitrogen available. This is a chroma of a microbially active soil, where the microlife has built up permanent humus aggregates, which contribute to its high fertility. Figure 4. This is a chroma-test from an intensively and conventionally managed vegetable field on clay soil in Lier. Cauliflower has been grown for 40 years, at two harvests a year. This picture shows no whitecenter, in fact the center is dark, which indicates an uncontrolled mineralization process. The colours are grayish because of a low content of organic matter. The structures and patterns show a poor microbial life and the lack of clouding in the outer zone (protein zone) indicate lack of nutrient turnover. The straight radial lines are a sign for limited fungal activity and mineral matter, which has not been integrated into the soil whole. The soil is quite ‘dead’ regarding the microbial life, and the aggregation is poor. 39 Figure 5. This soil sample is taken in a cauliflower field nearby the one shown in fig. 4. This field has also been farmed intensively in the past, but when the sample was taken, it had already been in organic production for two years. The chroma of this soil is considerably better than the one in fig. 4. It shows finer patterns, indicating a richer microbial life (especially the fungi). A white center is starting to form and light clouds in the protein zone show a nutrient turnover. Figure 6. This is a chroma from an organically farmed pasture on a poorly transformed peat soil in Aust-Agder. This picture is typical for a poorly drained high organic soil. The dark brownish-black protein zone shows crude organic material, probably in an anaerob condition. In the inner and middle zone there are no signs of microbal activity, and there is no aggregation. The diffuse brownish colour of the chroma indicates that there is a turnover of the organic matter to a certain degree, but the excess nutrients are not transferred into stable soil aggregates. Figure 7. Here is a chroma-test for a pasture on peat soil, which is far more transformed and of higher biological quality. The outer ring shows a high content of crude organic material. In this soil there is fungal and bacterial activity, but the picture lacks the white center zone, which would appear if the soil had a good, stable aggregation. The peat soils I tested have hardly ever shown a white center. The sharp contrast between the inner and the middle zone is typical for pastures fertilized with undigested (raw / uncomposted) manure. Figure 8. This is a conventionally treated pasture in Midthordland, lying on sandy loam with a high content of organic matter. The left side of the figure shows a nearly identical pattern as the peat soil on fig.7, but the right side shows some more activity both among the fungi and the bacteria. Humus aggregation is still poor and the left side shows the application of raw manure. 40 Figure 9. This sample is taken 20 meters away from the sample in fig.8. It is exactly the same soil type, but here the farmer has been producing herbs organically for the last 2-3 years. The manure is composted together with herb residues and grass cuttings. A crop rotation has been practiced together with the use of green manure. There is much more activity in the inner and middle zones, there is a beginning white center, which indicates good aggregation, and the outer zone is “breaking up”, but still shows a high content of crude organic material. If the farmer has recently applied manure or mulched green manure, this dark outer-zone does not signify a problem. If not, it is a sign that organic turnover and humus building is not optimal. As an adviser in the Norwegian Association of Agricultural Advisers, with organic farming as chief occupation, I have found the chroma-test technique very helpful in practical agricultural advising. The test results, like the examples above, help to explain many of the practical problems in the fields the samples are taken from, - such as loss of nutrients, soil erosion, drought problems, weed, insect or disease problems or problems with product quality. This test technique also gives a quick response to altered cultivation practices, like the samples from the cauliflower fields above (fig. 4 and 5). The chroma-test has actually shown differences in soil and product quality on demonstration fields only a few weeks after a treatment. In a potato field, where different amounts of compost were applied in June ‘98, and chromatests taken August ‘98 the results already showed distinct differences in quality (Nyborg, 1998). There are still problems in the application of the chroma-test. First, the test is simple but a bit time-consuming. Normally the test will not be suitable to be carried out by the adviser himself, but the adviser should be able to interpret the chroma to make full use of it. Second, the test is to a certain degree subjective and the interpretation can differ somewhat from person to person depending on experience. It is important to calibrate ones interpretation with others who practice chroma-testing regularly. Today we have the data-equipment to do so in an easy way, scanning the chroma-pictures and sending them by e-mail. Third, there is variation in the chroma picture during the year, depending on variation in soil temperature, soil moisture and cultivation practices as well as the condition of the plants cultivated. For the interpretation of the chromas, the soil conditions at the time of sampling must be considered. Samples that are to be compared have to be taken under the same conditions. Conclusions Soil quality testing by means of chromatography has proven to be a useful and important supplement to traditional chemical soil analysis. The method is simple but time-consuming. The test provides a quick overview biological condition of the soil and may suggest how to improve the fertility of the soils. In soil research the chroma-test will show the impact of different treatments on soil quality in a very direct and detailed manner. 41 References Lübke, A., 2000. Personal messages. Nyborg, S.B., 1998. Effekt av kompost på jord og produktkvalitet i tidligpotet. Rapport (unpublished), Hardanger-Midthordland Forsøksring, Øystese, Norway. Nyborg, S.B., 1999. Rundbildekromatografi- som redskap for kvalitetstesting av jord og kompost. Hardanger-Midthordland Forsøksring. Øystese. 29 pp. Pfeiffer, E., 1960. Chromatography Applied to Quality Testing. Bio-Dyn. Farm & Gard. Assn., Stroudsburg, Pa. 44 pp. Voitl, H. & Guggenberger, E., 1986. Der Chroma-Boden-Test. Orac,Wien. 181 pp. 42 Multi-level assessment of soil quality – linking redutionist and holistic methodologies Per Schjønning, Lars J. Munkholm, Kasia Debosz and Susanne Elmholt Danish Institute of Agricultural Sciences, Department of Crop Physiology and Soil Science, Research Centre Foulum, P.O. Box 50, DK-8830 Tjele, DENMARK E-mail: [email protected] Summary Soil quality is often used as a qualitative, general term. However, quantification is an important feature of the scientific approach to nature. On the other hand, addressing specific soil parameters as indicators of soil quality includes a reduction of the whole soil system. Therefore, results obtained by specialized methodologies ought to be evaluated by methods integrating the soil characteristics in situ. In this presentation, results are given from an investigation of the tilth of two differently managed loamy soils. One of the soils had been managed for decades with a forage crop system (labeled FCS), which included fertilization with farmyard manure, while the other had been grown with a continuous cereal system (labeled CCS), receiving no input of organic matter. In the field, the structure of the top 30- cm soil layer was described visually (spade analysis) and by studying the fragmentation behavior (soil drop test). Further, the field measurements included determination of soil strength by a torsional shear box method. In the laboratory, shear strength was determined on bulk soil sampled in metal cylinders, and tensile strength was estimated from crushing tests of individual, differently sized aggregates. The FCS soil appeared porous, with crumbs as structural units, while the CCS soil was compact with blocks as structural units. The soil drop test yielded the highest degree of fragmentation for the FCS soil. The torsional shear box method showed the CCS soil to have the highest bulk soil strength. This was confirmed by the laboratory shear annulus method. Finally, the tensile strength measurements revealed a much higher strength of 8-16, 4-8 and 2-4 mm dry aggregates from the CCS soil as compared to the FCS soil, while 1-2 mm aggregates were strongest in the FCS soil. This indicates a higher friability for the FCS soil, which is in accordance with the soil behavior in the field tests. In conclusion, the quality of the FCS soil – as evaluated by its mechanical behavior – was found to be higher than that of the CCS soil. An important result is the good correlation between the integrating field methods and the differentiating laboratory methods. This means that the quantifying, redutionist scientific approach is not conflicting with the ‘holistic’ descriptions in the field. Keywords: farming systems, soil structure, soil strength, soil fragmentation, ease of tillage, field methods, laboratory methods, methodologies 43 Introduction In a review of the development of research in organic agriculture, Niggli & Lockeretz (1996) mention the contrasting opinions concerning the most relevant scientific approach when addressing alternative farming systems. They highlight the need for ’short-term, user- oriented, highly applied research’ as well as ’the long, slow search for a better understanding of the fundamental natural processes on which any agricultural system rests’. In this context, it is of great importance to secure a good communication between scientists and the non- scientific approach of the NGO’s. The scientists normally use methodologies which demand a ‘reduction’ of the soil system before the specific analysis can take place (reductionism) while the NGO’s involved in the promotion of the alternative systems are focussing on evaluation of the whole system (holism). Research in organic farming systems should therefore preferably include the linking of quantitative, scientific measures of system characteristics and the farmers qualitative impression and judgement of the same characteristics (Harris & Bezdicek, 1994; Romig et al., 1995). This paper reports some of the results obtained in a case study investigation of soil strength and fragmentation characteristics at two long-term differently managed soils in Denmark. An organically managed field was referenced by a conventionally managed counterpart of similar pedological origin. The two soils are identical to those labeled as Soil Pair II by Elmholt et al. (2000) [this issue]. Quantitative scientific laboratory methods were supplemented with descriptive methods in the field in order to evaluate the conclusions drawn from the classical methods. A full description of the investigation is given by Schjønning et al. (submitted) and Munkholm & Schjønning (submitted). Materials and Methods Soils studied The full investigation (Schjønning et al., submitted; Munkholm & Schjønning, submitted) included three groups of soils, each group with two or three differently managed fields. In this paper, only one of these groups will be considered (labeled group III in Schjønning et al. (submitted), group 2 in Munkholm & Schjønning (submitted) and Soil pair II in Elmholt et al. (2000) [this issue], respectively). Both soils within this group were sandy loams developed on till plains from the Weichselian glacial stage and may be classified as Oxyaquic Agriudolls / Glossic Phaeozems according to the USDA / FAO classification systems, respectively (Krogh & Greve, 1999). Consult Table 1 in Elmholt et al. (2000) [this issue] for information on basic soil characteristics. The management characteristics of the soils in investigation are described in detail by Schjønning et al. (submitted). One of the soils belonged to a dairy farm with a five-year crop rotation including a two-year grass ley (labeled FCS [forage crop system]) and received farmyard manure. It was compared to a neighboring soil managed for at least twenty years 44 with no animal manure application and grown continuously with small grain cereals or rape (Brassica napus L.) (labeled CCS [continuous cereal system]). Both soils had been managed according to conventional tillage practices including mouldboard ploughing. Hierarchical strategy for analyses As mentioned by Elmholt et al. (2000) [this issue], the full investigation was a multidiscip- linary approach, including physical as well as microbiological characterization of the soils. This may be designated as horizontal interdisciplinarity, Figure 1. At the same time, especially the soil physical characteristics were assessed by several methodologies, ranging from visual evaluations in the field to measurements on single, dry aggregates in the laboratory. We suggest to label this approach vertical interdisciplinarity, Figure 1. DISCIPLINES (horizontal interdisciplinarity) Physics Microbiology and Chemistry Microbial Structural Mechanical Pore activity and bonding and properties characteristics biomass binding agents Evaluations, field Measurements, field Undisturbed soil, lab. Remoulded soil, lab. Aggregates, lab. (vertical interdisciplinarity) LEVEL OF INVESTIGATION Figure 1. The analytical strategy included combinations of research disciplines (horisontal interdisciplinarity) as well as analyses of soil characteristics at different levels of ‘reduction’ of the research object (vertical interdisciplinarity). Measurements Sampling and field measurements took place in the spring in an autumn-sown winter wheat or spelt (Triticum aestivum L.) as detailed by Elmholt et al. (2000) [this issue]. Only short descriptions of specific methodologies will be given in this paper and these will be found within the ‘Results and Discussion’ section. Consult Schjønning et al. (submitted) and Munkholm & Schjønning (submitted) for a detailed description of methodology. 45 Results and Discussion Visual soil evaluation The visual evaluation of soil in the field was performed as described by Munkholm (2000) (Figure 2, left). It revealed, that the soil in the CCS field was compact with a blocky structure and of a firm consistency even when moist (Figure 2, right). The FCS soil was on the other hand porous with a crumbly structure, when both wet and dry. The ease/difficulty of digging and sampling in the two differently managed fields further confirmed these observations. Figure 2. The visual soil evaluation involves studies of the top 30 cm soil (left). The CCS soil appeared dense, with a blocky soil structure (right). Notice the compact plough pan below 20 cm depth. The soil drop test Science is all about the quantification of the observations, as opposed to general empiricism. The question is therefore whether the visual soil evaluation could be quantified in some way by scientific tests. The soil drop test in a reproducible way can quantify how the soil fractionates at a certain energy input. The test is described by Schjønning et al. (submitted) and consists of letting an undisturbed soil sample (a cube measuring c. 7 x 8 x 11 cm) drop to the ground from a height of exactly 75 cm (Figure 3, left) thereby in principle simulating a soil tillage process. The soil fractions are then transferred to a nest of sieves from which the aggregate size distribution can be determined (Figure 3, right). The figure shows that the CCS soil fragmented only poorly (many large, intact aggregates) while the FCS soil was more friable which indicates that this soil would be a more easily tilled soil, for example for a seedbed. 46 Figure 3. Performing the soil drop test (right) provides information of the soil fragmentation behaviour in the field (right). Shear strength – field measurements The aggregate size distribution is a relative number. In order to understand and describe a system, a quantification of scientifically well-defined parameters in absolute numbers is required. We therefore decided to try and measure the forces that determine soil fractionation. To do so we used the torsional shear box method in the field (Payne & Fountaine, 1952). This enables the determination of the forces, per unit area, that bind soil particles together. A cylindrical box is forced into the ground, a specific load is applied (there are measurements at several different loads) and the box is turned around. The peak strength when soil fails is measured (Figure 4, left). Figure 4, right, reveals that the highest shear strength was measured in the CCS soil which is in accordance with the results from the previously mentioned level of analysis (the soil drop test). 47 Figure 4. The torsional shear box method (left) gives information about soil strength in absolute terms (right). Shear strength – laboratory measurements Laboratory measurements provide control of the test conditions better than in the field. Undisturbed soil core samples of 100 cm3 were therefore taken in the field and drained to – 300 hPa matric potential. Using methodology in principle equivalent to that of the field tests, the forces counteracting shearing were measured at loads ranging to above 100 kPa (Schjønning, 1986) (Figure 5, left). Figure 5, right, shows that it was possible to reproduce the soil properties measured in the field. The laboratory tests also showed the CCS soil to have the highest shear strength. There is, however, a difference in the level of shear strength between the two methods. This may be due to the fact that the laboratory measurements were applied to samples taken after harvest; despite the fact that the soil cores had approximately the same water content as in the spring, they were sampled 4-5 months after the field measurements. Another important potential cause is the different character of the soil failure generated by the two methods. With the field method the failure is along natural planes of least resistance in the soil whereas with the laboratory method the soil is forced to fragment along predefined soil horizons. This means that forces between as well as within aggregates will contribute to shear strength. 48 Figure 5. The determination of bulk soil strength in the laboratory (left) allows a good control of measuring conditions and provides information (right) analogous to that obtained by the torsioinal shear box method. Tensile strength and soil friability The shear strength measurements indicated that the bulk soil strength was in agreement with the fractionation pattern obtained with the drop test, i.e. the FCS soil fragmented into smaller aggregates, corresponding to a lower cohesive force in the soil. However, the drop test also showed that only the large clods fragmented. In this well-structured FCS soil, the smallest aggregates had such a large cohesive strength that the result of the test was a broad distribution of aggregate sizes and not only a collection of smaller particles. In order to understand the results of the drop test (and similar results from the visual evaluation concerning the friable consistency of crumbs) further differentiation (further reduction of the research objective) is required. In the project this was achieved by measuring tensile strength on individual air-dried aggregates. The hypothesis is that the measurement of the strength of several sizes of aggregates enables a quantification of the friability perceived with the visual evaluation and the drop test in the field. A friable soil is defined as a soil where large aggregates have a low tensile strength and small aggregates a relatively large strength (Utomo & Dexter, 1981). The analysis consists of measuring the tensile strength of air-dried individual aggregates in a compression test, using a mechanical press in the laboratory (Dexter & Kroesbergen, 1985; Figure 6, left). Figure 6, right, shows that the large aggregates (right-hand side of the x-axis) in the CCS soil had a larger tensile strength than those of the FCS soil. This is in line with the indications from the drop test (see Figure 3, right). Figure 6 also shows that the opposite is true for small aggregates (1-2 mm); small aggregates in the 49 FCS soil had a larger strength than in the CCS soil. The friability of the soil is subsequently quantified as the coefficient of the slope in the double logarithmic depiction of aggregate size and tensile strength (Figure 6, right). A numerically large slope (high strength of small and low strength of large aggregates) is therefore an expression of a highly friable soil, which was the case for the FCS soil. Figure 6. Determination of tensile strength of individual air-dried aggregates (right) means a rather extreme reduction of the soil system. However, a high friability (a steep slope) was found for the FCS soil, which also appeared friable in the field (right). Conclusions and perspectives In combination, the results presented here revealed a high ease of tillage in the FCS soil; when mechanically disturbed, the bulk soil fragmented into smaller sizes ideal for a good seedbed. In contrast, the CCS soil required a larger energy input to fractionate and there was a tendency for the soil to be pulverized into inconveniently small particles. There is not necessarily a conflict between a holistic and a redutionist methodology, and our results showed a good correlation between the integrating field methods and the differentiating laboratory methods. The results further indicated that sophisticated analytical methodologies (redutionist research) are essential for quantification and understanding of soil behavior. The redutionist methods, however, should always be used and interpreted in the larger (holistic) context. 50 More research is needed in the further uncovering of the mechanisms responsible for a ‘good’ soil structure considering the crucial effects for optimal soil functioning. We need a better understanding of how the soil ecosystem in a diversified farm management system develops to give the most optimal conditions for the soil processes of importance for sustainable farming. Such work may well follow the above design and include the application of integrating, holistic methods as well as differentiating, redutionist investigations of causal relations in the ecosystem. References Dexter, A.R. & Kroesbergen, B., 1985. Methodology for determination of tensile strength of soil aggregates. J. Agric. Eng. Res. 31, 139-147. Elmholt, S., Debosz, K., Munkholm, L.J. & Schjønning, P., 2000. Biotic and abiotic binding and bonding mechanisms in soils with long-term differences in management. Proceedings from a workshop in Aas, Norway, 10-12 April 2000. DIAS Report 38, 2000. The Danish Institute of Agricultural Sciences. Tjele, Denmark, 53-61. Harris, R.F. & Bezdicek, D.F., 1994. Descriptive aspects of soil quality/health. In: Doran, J.W., Coleman, D.C., Bezdicek, D.F. & Stewart, B.A. (eds.), Defining Soil Quality for a Sustainable Environment. Soil Sci. Soc. Amer. Special Publication 35, ASA, Madison, WI, pp. 23-35. Krogh, L. & Greve, M.H., 1999. Evaluation of World Reference Base for Soil Resources and FAO Soil Map of the World using nationwide grid soil data from Denmark. Soil Use Manage. 15, 157-166. Munkholm, L.J., 2000. The spade analysis - a modification of the qualitative spade diagnosis for scientific use. DIAS report no. 28, Plant Production. Danish Institute of Agricultural Sciences, Foulum, 40 pp. Munkholm, L.J. & Schjønning, P., (submitted). Long-term effects of fertilization and crop rotation on a humid sandy loam. II. Soil mechanical characteristics. European J. Soil Sci. Niggli, U. & Lockeretz, W., 1996. Development of research in organic agriculture. In: Østergaard, T. (ed.), Fundamentals of Organic Agriculture. Proceedings, Vol. 1, 11th IFOAM International Scientific Conference August 11-15, 1996, Copenhagen. Payne, P.C.J. & Fountaine, E.R., 1952. A field method of measuring the shear strength of soils. J. Soil Sci. 3, 136-144. Romig, D.E., Garlynd, M.J., Harris, R.F. & McSweeney, K., 1995. How farmers assess soil health and quality. J. Soil Water Conserv. 50, 229-236. Schjønning, P., 1986. Shear strength determination in undisturbed soil at controlled water potential. Soil Till. Res. 8, 171-179. Schjønning, P., Elmholt, S., Munkholm, L.J. & Debosz, K., (submitted). Soil quality aspects of humid sandy loams as influenced by different long-term management. Agric. Ecosyst. Environ. Utomo, W.H. & Dexter, A.R., 1981. Soil friability. J. Soil Sci. 32, 203-213. 51 Acknowledgements The technical assistance of Palle Jørgensen, Michael Koppelgaard and Jørgen M. Nielsen is gratefully acknowledged. We also thank the farmers for providing their fields for our studies. This work was financed by the Danish Environmental Research Programme and was performed in the context of the Danish Research Centre for Organic Farming. 52 Biotic and abiotic binding and bonding mechanisms in soils with long-term differences in management Susanne Elmholt , Kasia Debosz, Lars J. Munkholm and Per Schjønning Danish Institute of Agricultural Sciences, Department of Crop Physiology and Soil Science, Research Centre Foulum, P.O. Box 50, DK-8830 Tjele, DENMARK E-mail: [email protected] Summary During the last decades Denmark has experienced a growing interest in low-input farming systems like organic farming. These systems rely on a high soil fertility to maintain nutrient availability and plant health. Soil aggregation contributes to this fertility, because it is crucial to soil porosity, aeration and infiltration of water. This paper reports a study of two pairs of differently managed, neighboring fields. The aim was to elucidate long-term effects of the different farming systems on physical and biological variables with influence on bonding and binding mechanisms of soil aggregation. Each pair consists of an organically grown dairy farm soil, based on a forage crop rotation system, including grass (Org-FCS(G)) and a conventionally managed soil. One of the conventional farms has a forage crop rotation with annual cash crops and no grass (Conv-FCS(NG)) and one has been grown continuously with small grain cereals and rape (Conv-CCS). Our results indicate that the Org-FCS(G) soils stimulate biotic soil aggregating agents as measured by extracellular polysaccharides (EPS) and hyphal length measurements, respectively. Generally, the Conv-CCS soil, which relies exclusively on synthetic fertilisers and cereal production, offered poor conditions for the biotic binding and bonding agents. Nevertheless this soil contained a large amount of stable macro- aggregates. This is explained by the physical results, which indicated that the strong macro- aggregation was due to clay dispersion and cementation processes rather than to biotic processes. Keywords: soil structure, farming systems, stability, hyphal lengths, polysaccharides, microbial biomass, dispersible clay Introduction Denmark experiences a growing interest in low-input management systems like organic farming. These systems cannot be manipulated by agrochemicals. Nutrient availability and plant health rely on a high fertility of the soil and a proper physical environment for plants and microorganisms. The Danish Research Center for Organic Farming (DARCOF) was established in 1996 with the aim to co-ordinate Danish research and development for organic farming. A number of research projects were initiated with the intention to facilitate a conversion from conventional to organic farming, while encouraging a sustainable development of the economic, ecological and social aspects of agriculture. Among these projects is a study on 'Soil fertility and soil tilth as influenced by organic farming practice and 53 soil tillage'. One purpose of this project was to elucidate long-term effects of distinctly different farming systems on the interactions between physical and biological properties, which play a role in soil aggregation. Soil aggregation is crucial for ensuring a desirable soil structure for plant growth (i.e. good aeration, infiltration of water, soil root contact and a low resistance against root penetration). The hierarchical model of soil aggregation, presented by Hadas (1987) and Dexter (1988), assumes that a range of different mechanisms will combine primary particles (clay, silt, sand) and organic matter into floccules, micro-aggregates (63-250µm) and gradually larger macro- aggregates (>250µm) (Figure 1). There are typically 103 ± 1 particles of a given hierachical order in a single particle of the next higher order (Dexter, 1988). Figure 1. Conceptual model of soil aggregation, illustrating micro-aggregates , small macro-aggregates and large macro-aggegates (based on Dexter, 1988). According to theories on soil aggregating mechanisms (Tisdall & Oades, 1982; Degens, 1997), these can be divided into three groups: 1) The 'persistent bonding agents', a range of humic compounds, which are associated with metal ions. This group is not addressed in the present paper as it is considered less susceptible to soil management than the other groups. 2) The 'transient bonding agents', which 'glue' (bond) together primary soil particles into micro-aggregates. They consist primarily of extracellular polysaccharides (EPS) and are produced by bacteria, fungi and plants. 3) The 'temporary binding agents', which enmesh (bind) primary particles and micro-aggregates to larger aggregates and are assumed to play their main role in macro-aggregation. They consist of fungal hyphae and plant roots. In order to elucidate the effect of soil management on the interaction between soil physical and biological variables it is important to use a multidisciplinary approach, integrating these different disciplines of soil science. Therefore both physical and biological variables that indicate bonding and binding (groups 2 and 3, cf. above) mechanisms were assessed. In this paper we present some of our results on the physical condition of the soils (aggregate stability, soil porosity), on indicators of biotic binding (hyphal lengths) and bonding (EPS) as well as abiotic bonding mechanisms (clay dispersion/cementation). Results on microbial biomass are presented to indicate the living conditions for soil microorganisms. 54 Materials and Methods Four differently managed arable soils, Soil Pair I and Soil Pair II, were used. The soils are located on the island of Sealand and a few relevant chemical and physical characteristics of the soils are shown in Table 1. A more detailed description of physical, chemical and management characteristics can be found in Schjønning et al. (submitted). Table 1. Selected chemical and physical characteristics of the four soils. For further characteristics and details on methodology and analyses of variation, cf. Schønning et al., in press. ‘Forage Crop System’ (FCS) or ‘Continuous Cereal System’ (CCS). Crop rotation with Grass (G) or with No Grass (NG). Soil Pair I Soil Pair II FCS(G) FCS(NG) FCS(G) CCS Soil type Sandy loam Sandy loam Sandy loam Sandy loam Clay (< 2µm), % 20 21 17 19 Organic matter, % 3.9 3.5 3.5 2.4 pH (CaCl2) 6.7 7.1 6.2 6.1 Bulk density, g cm-3 1.35 1.35 1.36 1.49 Soil Pair I (SP I) In SP I, one dairy farm soil had been grown organically (Org) since 1951 with high amounts of animal manure (estimated yearly 233 kg N/ha from slurry, composted farmyard manure and grazing) and a diversified crop rotation, based on grass/clover leys and cereals (forage crop system with grass, FCS(G)). The Org-FCS(G) soil had an estimated yearly input of incorporated dry matter of 5.6 t/ha. This soil was referenced by a conventional dairy farm having a forage crop rotation with no grass (Conv-FCS(NG)). The crop rotation of the Conv- FCS(NG) soil consisted of annual cash crops only, including beets for sugar production. The dairy farm characteristics of this soil were primarily reflected in a rather high amount of animal manure application (estimated yearly 261 kg N/ha from slurry). The Conv-FCS(NG) soil had an estimated yearly input of incorporated dry matter of 4 t/ha. Soil Pair II (SP II) In SP II, the dairy farm soil had been grown organically since 1958 and was managed nearly identically to the Org-FCS(G) soil of SP I (same owner). This soil has a forage crop rotation too but lacked grazing and had only one year of grass/clover ley in the rotation. The Org- FCS(G) soil had fairly low, estimated yearly inputs of animal manure (128 kg N/ha from slurry and composted farmyard manure) and high inputs of incorporated dry matter (4.5 t/ha). It was referenced by a conventionally managed soil (Conv). This soil had not received animal manure for a minimum of twenty years and had been grown continuously with small grain cereals and rape (continuous cereal system, CCS), generally without mulching of plant residues (estimated yearly input of dry matter 0.8 t/ha). Despite this, the input of nitrogen to the Conv-CCS soil was higher than to the Org-FCS soil (estimated yearly input from synthetic fertilisers 159 kg N/ha). 55 Soil sampling and analysis The time and strategy of soil sampling aimed to avoid/minimise effects of temporal variations in the studied variables caused by short-term effects of e.g. tillage, fertilisation, crop rotation, rhizodeposition and drying/wetting of the soil. The four fields were grown to winter wheat at the time of sampling, Triticum aestivum L. ssp. spelta in the organically managed SP I-FCS and SP II-FCS fields and Triticum aestivum L. ssp. vulgare in the conventionally managed SP I-FCS(NG) and SP II-CCS fields. For each soil nine sampling points were identified as the intersection of a 10 x 10m grid, from which undisturbed soil blocks were collected at 6-13 cm depth. This depth was chosen to avoid soil, which had been directly disturbed by seedbed preparation and sowing operations. The soil blocks (~ 650 cm3) were retrieved by carefully hammering a metal shovel sideways into a soil ‘wall’ and cutting the block from the bulk soil by pressing down a metal plate. The soil blocks from each of the nine sampling points were placed in plastic boxes with dimensions corresponding exactly to nine block units and covered with a plastic lid. A slight air exchange to the surrounding atmosphere secured oxic conditions in the boxes. Undisturbed soil cores (100 cm3) were collected in metal cylinders forced into the soil by means of a hammer. After removal of the soil-filled cylinder from the bulk soil, the end surfaces were trimmed with a knife and mounted with plastic caps to protect the soil from mechanical disturbance and evaporation. Two replicate samples were collected at each of the nine sampling points in each field. The soil (both blocks and cores) was stored at 2oC until analysis for soil porosity, WAS, clay dispersibility and microbial biomass. The soil for measurements of EPS and hyphal lengths was air-dried immediately after sampling and sieved into sizes 4-8mm, 0.50-1mm and 0.063-0.25mm. All results are expressed on a dry weight basis (gravimetric determination of water content by drying the soil for 24 hours at 105oC). Soil porosity: The core samples in the metal cylinders were weighed before saturating them on sandboxes with capillary water from beneath. The cores were drained to –100hPa and oven-dried (105oC, 24 hs). The weight of each sample was recorded at each matric potential before and after drying. Results are given as the percentage of soil pores in relation to the soil volume, either as total porosity or macro-porosity (pores > 30µm). Wet aggregate stability (WAS): Core subsamples of approx. 45 g from each of the nine blocks of each soil were gently passed through an 8 mm sieve and taken to a 250 µm sieve installed in a wet-sieving apparatus (Yoder, 1936). Thirty seconds of initial capillary wetting were followed by 2 min of vertical movement of the sieve (stroke length 32mm, 38 strokes/min). Stable aggregates remaining on the sieve (macro-aggregates) were transferred quantitatively to a beaker, water was evaporated at 80oC and the soil further dried at 105oC. Finally the soil was dispersed by end over end shaking (24 hours with a 0.002 M Na4P2O7 solution) and poured through the 250µm sieve. Primary particles >250µm were weighed following drying at 105oC. Results are given as gram wet stable aggregates per gram soil. 56 Clay dispersibility: Subsamples of 1.5 g from each soil block of a given soil, drawn from the 8 mm sieved samples described in the section on WAS, were weighed into cylindrical centrifuge bottles and applied with 50 ml of 0.002 M Na4P2O7 solution. The bottles were shaken end over end (33 rpm) for 24 hours. Clay dispersed from the soil samples was determined from the turbidity of a clay-holding suspension, siphoned off the shaking bottles after a specified time period (Pojasok & Kay; 1990; Watts et al., 1996), in this case 24hs. Correction was made for primary particles >2 mm in the given soil. Results are given as dispersed clay in relation to the clay fraction of the sample. Hyphal length: Subsamples from each soil block of a given soil were taken representatively from the air-dried, fractionated samples of aggregates 4-8mm, which were sieved (<2 mm) before subsampling. Hyphal lengths were essentially determined according to procedures by West (1988): Subsamples were dispersed in 0.0033 M sodium hexametaphosphate and the suspension was sieved (38 µm) to remove clay and silt. The remaining material was blended, diluted, stained for 2 hours with calcofluor white, filtered (Nuclepore 110659, black polycarbonate, diameter 25mm, mesh width 0.8µm) and examined at a magnification of 200 x with an epifluorescent microscope, using the gridline intersect method (Olson, 1950). Results are given as meter of hyphae per gram soil. Extracellular polysaccharides (EPS): An easy extractable carbohydrate fraction was extracted from air-dry aggregates of 4-8mm according to the method described by Ball et al. (1996). The following modifications were used: The air-dried aggregates were shaken with hot water (80o C) using 1: 6 soil extractant ratio (wt/vol.) for 16 h. Before extraction, highly soluble substances and floating plant material were removed by shaking the soil with distilled cold water (20°C). The carbohydrate content of the hot water extract after centrifugation (5800 x g, 10 min) was analysed using a reaction with thymol in strongly acid solution. The results are given as mg EPS-C per kg soil. Microbial biomass: Microbial biomass C was determined in field-moist soil by the fumigation-extraction method (Vance et al., 1987). Three core samples from each of the nine soil blocks were sieved (2 mm) and mixed thoroughly. Soluble C and chloroform labile C (C solubilized by CHCl3 during an 18 h fumigation period) were extracted with 0.5 M K2SO4 in a soil:solution ratio of 1:4 (wt:vol). Organic C in all 0.5 M K2SO4 extracts was determined by an automated UV persulphate oxidation procedure using a Dorhmann DC-180 Carbon Analyzer (Wu et al., 1990). Biomass C was calculated as soluble C in fumigated minus soluble C in non-fumigated soil using 0.45 as the kc-factor (Kaiser et al., 1992). The results are given as mg biomass C per kg soil. Results and Discussion Figure 2 presents a range of physical and biological results from the four soils. Figure 2A shows the mechanical stability of the soil, WAS, expressed as the amount of wet-stable macro- 57 aggregates (g) per g soil. For Soil Pair I, the macro-aggregate stability is higher in the ORG- FCS(G) soil than in the CONV-FCS(NG) soil. A: Wet aggregate stability, (g agg. >250 µg/g soil) B: Hyphal length (m/g soil) 1.0 25 0.8 20 0.6 15 0.4 10 0.2 5 0.0 0 Pair I Pair I Pair II Pair II Pair I Pair I Pair II Pair II FCS(G) FCS(NG) FCS(G) CCS FCS(G) FCS(G) FCS(G) CCS C: Extracell. polysaccharides (mg C/kg soil) D: Microbial biomass C (mg/kg soil) 600 750 500 600 400 450 300 300 200 100 150 0 0 Pair I Pair I Pair II Pair II Pair I Pair I Pair II Pair II FCS(G) FCS(NG) FCS(G) CCS FCS(G) FCS(NG) FCS(G) CCS E: Soil porosity (m3 /100 m3 soil) F: Clay dispersibility (g disp. clay/g clay) 50 0.5 40 0.4 30 0.3 20 0.2 10 0.1 0 0.0 Pair I Pair I Pair II Pair II Pair I Pair I Pair II Pair II FCS(G) FCS(NG) FCS(G) CCS FCS(G) FCS(NG) FCS(G) CCS Figure 2. Physical and biological characterisation of the two Soil Pairs (I and II). Pair I, FCS(G): Organically managed, forage crop rotation with grass. Pair I, FCS(NG): Conventionally managed, forage crop rotation without grass. Pair II, FCS(G): Organically managed, forage crop rotation with grass. Pair II, CCS: Conventionally managed, cereal crop system. All results are shown as mean values for the samples from the nine grid points, the bars giving the standard error of the mean. In 2E the bottom stacks show the volume of pores <30 µm and the top stacks show the macro-porosity (pores > 30µm). In Soil Pair II, however, WAS is higher in the CONV-CCS soil than in the ORG-FCS(G) soil. This indicates that bonding and especially binding agents - which are supposed to play the larger role in macro-aggregation - are more abundant in the ORG-FCS(G) soil of Pair I and in the CONV-CCS soil of Pair II. Figure 2B shows that the fungal hyphae, which are supposed to be one of the major agents of macro-aggregation, are about twice as abundant in the two organically managed soils (both being FCS soils with grass in the rotation) as compared with the conventionally managed soils. The lowest values were found in the CCS soil. It should be noted, however, that Schjønning et al. (submitted) report results on the soil ergosterol content, which indicate a higher fungal biomass in the SP I-FCS(NG) soil than in the SP I-FCS(G) soil. For the SP II soils both the hyphal length 58 measurements and the soil ergosterol contents show higher results for the FCS soil than for the CCS soil. The results on hyphal measurements reported here are in accordance with Tisdall & Oades (1982). They put forward the hypothesis that an increase in the frequency of grass will increase the percentage of organic carbon in the soil and - what is interesting in this context - that this will primarily affect the temporary binding agents, i.e. the roots and hyphae. Both organically managed soil have a crop rotation with grass leys and they have higher percentages of organic matter than their conventional reference soils (Table 1). 750 600 450 300 y = 10.1*x + 331 EPS, mg C/kg R2 = 0.4500 150 *P=0.0001 0 0 5 10 15 20 25 30 Hyphal lengths, m/g FCS(G) SP I FCS(NG) SP I FCS(G) SP II CCS SP II Figure 3. Correlation between hyphal lengths and EPS, expressed as a linear regression for all samples. According to Tisdall & Oades (1982), the bonding agents will be less susceptible to changes in crop rotation and organic matter. Figure 2C shows the results for the EPS, which are considered important bonding agents. The pattern is the same as for the hyphae although the differences between the organically and the conventionally managed soils are less pronounced. At grid point level a significant correlation was found between hyphal lengths and EPS (Figure 3), which is also in accordance with the theory of Tisdall & Oades (1982). Figure 2D shows the microbial biomass C, which is regarded an important indicator of soil quality (Doran & Parkin, 1994). Like the results on biotic binding and bonding agents (Figure 2B and C), these results also indicate that the conditions for microbial life are poor in the CCS soil as compared with the other three soils. This may be related to poorer soil porosity in the CCS system, especially regarding macro-pores >30 µm (top stacks in Figure 2E). At sample level our results show a significant correlation between hyphal lengths or EPS, respectively, on the one hand and total soil porosity on the other (results not shown). All the biological results as well as the pore characteristics indicate the lowest soil quality in the CCS soil. Bonding as well as binding agents appear to have poor conditions, indicating that biotic soil aggregation is poor in this soil. Nevertheless the WAS results showed a high 59 amount of stable macro-aggregates (Figure 2A). Figure 2F shows clay dispersibility for the four soils, expressed as the amount of clay released from soil aggregates during a prolonged shaking procedure of 24 hours with Na4P2O7. The biological results combined with the fact that clay is more easily dispersed from the CCS soil than from the FCS soils indicates that macro-aggregation in this soil depends on abiotic rather than biotic processes. A large dispersibility means that the clay is loosely bound to the soil aggregates and that the soil is susceptible to dispersion (slaking) and cementation during wetting and drying processes. Re- orientation and hardening of the dispersed clay minerals may eventually lead to a dense and mechanically strong soil. In situ studies of the soil structure confirmed this and revealed that the CCS soil had a very firm blocky structure whereas the FCS(G) soil had a more porous and crumbly structure in the studied soil layer (Munkholm, 2000). In an agronomic sense this may cause problems in preparing a proper seed bed in the CCS soil. The macro aggregates, which were retrieved by wet sieving from this soil, are therefore small, dense blocks rather than porous crumbs as discussed by Schjønning et al. (submitted). This seems to be the reason why the CCS soil is a poorer habitat for soil organisms than the FCS soils. Conclusions Our results stress the need to integrate biological, physical and chemical methodologies to increase our understanding of soil aggregation mechanisms. The results on EPS and hyphal length measurements indicate that the FCS(G) soils stimulate biotic bonding and especially binding mechanisms. Generally, the CCS soil, the management of which is based exclusively on synthetic fertilisers and cereal production, offered poor conditions for the biotic binding and bonding agents. The strong macro-aggregation of this soil was ascribed instead to clay dispersion and cementation processes. References Ball, B.C., Cheshire, M.V., Robertson, E.A.G. & Hunter, E.A., 1996. Carbohydrate composition in relation to structural stability, compactibility and plasticity of two soils in a long-term experiment. Soil Till. Res. 39, 143-160. Degens, B.P., 1997. Macro-aggregation of soils by biological bonding and binding mechanisms and the factors affecting these: a review. Aust. J. Soil Res. 35, 431-459. Dexter, A.R., 1988. Advances in Characterization of Soil Structure. Soil Tillage Res. 11, 199- 238. Doran, J.W. & Parkin, T.B., 1994. Defining and assessing soil quality. In: Doran, J.W., Coleman, D.C., Bezdicek, D.F. & Stewart, B.A. (eds), Defining Soil Qyality for a Sustainable Environment. Soil Science Society of America Inc., American Society of Agronomy,Inc, Madison, Wisconsin, 3-21 Hadas, A., 1987. Long-term tillage practice effects on soil aggregation modes and strength. Soil Sci. Soc. Am. J. 51, 191-197. 60 Kaiser, E.A., Mueller, T., Joergensen, R.G., Insam, H. & Heinemeyer, O., 1992. Evaluation of methods to estimate the soil microbial biomass and the relationship with soil texture and organic matter. Soil Biol. Biochem. 24, 675-683. Munkholm, L.J., 2000. The spade analysis - a modification of the qualitative spade diagnosis for scientific use. DIAS report no. 28, Plant Production. Danish Institute of Agricultural Sciences, Foulum, 40 pp. Olson, F.C.W., 1950. Quantitative estimates of filamentous algae. Trans. Am. Mic. Soc. 69, 272-279. Pojasok, T. & Kay, B.D., 1990. Assessment of a combination of wet sieving and turbidimetry to characterize the structural stability of moist aggregates. Can J. Soil Sci. 70, 33-42. Schjønning, P., Elmholt, S., Munkholm, L.J. & Debosz, K., (submitted). Soil quality aspects of humid sandy loams as influenced by different long-term management. Agric. Ecosyst. Environ. Tisdall, J.M. & Oades, J.M., 1982. Organic matter and water-stable aggregates in soils. J. Soil Sci. 33, 141-163. Vance, E.D., Brookes, P.C. and Jenkinson, D.S., 1987. Microbial biomass measurements in forest soils: The use of chloroform fumigation-incubation method in strongly acid soils. Soil Biol. Biochem. 19, 697-702. Watts, C.W., Dexter, A.R., Dumitru, E. & Arvidsson, J., 1996. An assessment of the vulnerability of soil structure to destabilization during tillage. Part I. A laboratory test. Soil Till. Res. 37, 161-174. West, A.W., 1988. Speciment preparation, stain type, and extraction and observation procedures as factors in the estimation of soil mycelial lengths and volumes by light microscopy. Biol. Fert. Soils 7, 88-94. Wu, J., Joergensen, R.G., Pommerening, B., Chaussod, R. & Brooks, P.C., 1990. Measurement of soil microbial biomass C - an automated procedure. Soil Biol. Biochem. 22, 1167-1169. Yoder, R.E., 1936. A direct method of aggregate analysis of soils and a study of the physical nature of erosion losses. J. Amer. Soc. Agron. 28, 337-351. Acknowledgements The study was funded by The Danish Research Center for Organic Farming (DARCOF). Jørgen M. Nielsen, Anette Clausen, Palle Jørgensen and Michael Koppelgaard are thanked for skilful technical assistance. 61 Multivariate techniques for presentation, interpretation and evaluation of soil quality data Mats Johansson and Bo Stenberg Department of Microbiology, Swedish University of Agricultural Sciences, Box 7025, SE-750 07 Uppsala, SWEDEN E-mail: [email protected] Summary The complex nature of soil quality allows it to be assessed only if the physical, chemical and biological components are evaluated simultaneously. Univariate correlations and correlation matrices are normally used to study the relation between variables or indicators. However, these correlations will not reveal the structure of the variation of all soil quality indicators needed to assess soil quality. Thus, we need an integrated approach to evaluate a set of relevant soil-quality indicators. The commonly used index approaches have drawbacks, as an index is not directly related to any specific function or indicator, which causes problems when interpreting the reasons for a high or low index. Multivariate analyses have the capability of giving information regarding the relation between indicators and can therefore be an alternative to index approaches as it also reveals the functional structure of the soil. In this paper, our experiences using multivariate techniques for the presentation, interpretation and evaluation of soil quality data will be presented. They include i) the integrated evaluation of the relationship between soil quality indicators and their influence on the formation of soil quality groups, ii) the evaluation of the capacity of soil quality indicators to separate soils with different N and sewage sludge amendments, and iii) the relationship between soil quality indicators and productivity potentials. From these experiments it is concluded that principal component analysis and partial least-square regression can give interpretable quality groupings and separations between soils. Discriminant function analysis is a powerful technique to identify important variables for group distinctions. It is shown that microbiological indicators have a straightforward relevance, especially when soil functions of concern are related to cycling of nutrients or degradation of chemicals. Together with their integrative response and sensitivity to changes in the soil environment this strongly suggests that microbiological indicators should be included in a set of variables to assess soil quality. Keywords: discriminant function analysis, partial least square regression, principal component analysis, productivity, sewage sludge, soil microbiology, soil quality index 63 Introduction Soil quality assessment must be based on relevant soil-quality indicators. No single indicator can be found that will describe and quantify all aspects of soil quality. This is obvious, as soil quality is governed by the complex ecological interaction of the soil physical, chemical and biological components. The key is to find those attributes which are important for soil quality and which respond rapidly to changes. However, the complex nature of soil quality allows it to be assessed only if the physical, chemical and biological components are taken into account and evaluated simultaneously. In order to overcome the problem of complex interactions in the soil ecosystem, we need an integrated approach to evaluate a set of variables. The index approach is a commonly suggested strategy to evaluate a set of quality indicators. Doran & Parkin (1994) proposed such an index as a function of soil-quality elements. The elements encompassed productivity, environmental qualities, and food and animal health. Each element was a mathematical function of the physical, chemical and biological status of the soil. The status was related to relevant quality indicators through mathematical modelling. This kind of index is not coupled specifically to an indicator, or even element, which causes problems when interpreting the reasons for a high or low index. With the aim to simplify complex soil functions and making them easier to interpret, multivariate tools could be valuable. These tools have the capability of revealing the information regarding the relation between indicators. This is the functional structure approach. Multivariate analysis is a collective term used for a number of statistical techniques aiming at the description of complex environments, all having the capability of considering a number of more or less related variables simultaneously. A commonly used technique is the principal component analysis (PCA). It is designed to reduce a large number of variables (indicators) into a small number of indices called the principal components, which are linear combinations of the original values. These principal components represent most of the variation in all of the variables used. The principal components are calculated using least square fit. Use of two or three principal components enables a graphic visualisation to which we may project the different objects and create a two or three-dimensional score plot. By integrating a number of quality indicators in a PCA the distribution of the indicators in loading plots will represent the functional structures of a system. As the co-ordinates of a specific soil in the corresponding score plot relate that soil to the indicators, its function compared to other soils is revealed. Thus, different quality elements can be evaluated by defining the ideal co-ordinates for a quality element. This technique facilitates the use of one model for more than one element. As discussed above, PCA is a technique suitable to visualise the data as a first step in the evaluation process. However, there may be other, more specific, questions that need to be 64 addressed. For instance, we might want to know what set of indicators is the most suitable for the discrimination between a desired and a non-desired soil quality. Discriminant function analysis (DFA) is, unlike PCA, focused on the ability of different variables to discriminate between two or more naturally occurring (or a priori) defined groups. A large number of variables may initially be analysed and those able to distinguish between treatments could be selected in future analyses. The principal components could also be used to calibrate empirical models for the prediction of dependent variables. Prediction models of for example C- and N-yields can be performed with the combined use of PCA and multiple linear regression, a technique called principal component regression (PCR). It is also possible to use partial least squares (PLS) regression. As with PCR the principal components are used in the PLS analysis, but the extraction of each component is rotated to maximise the predictive ability of the dependent variables. In this way, fewer components may be needed for the same predictions and the models are thus more stable. The methods facilitate graphical visualisation of soil functions as well as the quality of a specific soil by plotting score or loading vectors against each other. On these graphical plots, ideal or favourable score areas could be determined (Elliot, 1997). This is equivalent to the index approach, where a threshold or an ideal index could be determined for different soil functions. The difference is that the loading plot corresponding to the score plot reveals an oversight of the influence of the indicators on the outcome. The aim of this paper is to describe our experiences using multivariate techniques to present, interpret and evaluate soil quality data. These include i) the integrated evaluation of the relationship between soil quality indicators and their influence on the formation of soil quality groups, ii) the evaluation of the capacity of soil quality indicators to separate soils with different N and sewage sludge amendments, and iii) the relation between soil quality indicators and productivity potentials. Materials and Methods Soil quality indicators and the formation of soil quality groups For this study two sets of data have been used in PCA to study the relation between soil variables and the formation of soil quality groups. The first represented natural variation with 26 mineral soil samples from all over Sweden selected to have as different soil properties as possible (Stenberg et al., 1998b). The variation and structure in mainly the chemical and microbiological component, together with soil texture, were studied. The second set was from a long term soil structure field experiment on a weakly structured soil at Sundby (59°42’N, 16°40’E), Västerås, in central Sweden, sampled 1995 after two 4- years crop rotations (Stenberg et al., 2000). Treatments with and without 6.5 Mg CaO ha-1 lime for structural improvement at initialisation of the experiment, and stubble cultivated or mouldboard ploughed to 12 and 25 cm, respectively, were tested on primarily microbiological 65 and soil structure indicators, as well as on basic chemical properties. The microbiological indicators were those described below. In the data set representing natural variation were also included organic carbon (org-C, %), total nitrogen (tot-N, %), pH and ammonium lactate (AL) soluble P, K, Ca and Mg (mg 100g-1). In the soil structure experiment soluble cat-ions were not analysed. However, on dry aggregates tensile strength (Y, N mm-2) was analysed, and on wet aggregates dispersed clay after shaking (T, mg g-1). From Y and T an aggregate stability index (AGS) was calculated. In addition, dry bulk density (BD, g cm-3), water content (WC, g 100g-1), total porosity (TPV, %), and equivalent pore diameter classes (<1, 1-5, 5-30, 30-100, 100-600, >600µm) were measured on undisturbed soil cores. In the field, soil penetration 2 resistance (Pen, kPa) and soil air permeability with unbroken (Kaid, µm ) and broken (Kaiu, µm2) pores were assessed. Separation of different N and sewage sludge amendments Two Swedish long-term field experiments with sewage sludge amendments in combination with different nitrogen regimes were used in this study (Johansson et al., 1999). The field experiments were situated close to Lund (Igelösa) and Malmö (Petersborg), in the southern part of Sweden (56°N, 13°E). The addition of sludge was equivalent to 0 (A), 1 (B) and 3 (C) ton ha-1 year-1 (dry matter). The additions were performed every 4th year (1981, 1985, 1989 and 1993), resulting in additions of 0, 4 and 12 Mg ha-1, respectively. The amendments were applied after harvest, but prior to autumn ploughing. Treatments with and without sludge were complemented with mineral fertiliser. The additions of nitrogen were 0, half and full rate according to crop requirements. At half and full nitrogen rates, the recommended rates of phosphorous and potassium were also applied. Soil quality indicators and productivity potentials In this study the soil data and annual mean temperature (Temp), precipitation (P) and potential evapotranspiration (E) from the 26 samples - representing natural variation as outlined above - were coupled to a production index (Stenberg, 1998). The production index was assessed from a pot experiment growing Italian Ryegrass in a climate chamber. Two harvests were taken; one after 34 days (H1) and one after another 40 days (H2). The roots were sampled after the second harvest. The samples where analysed for total C and N. Simple correlations were compared with PCR and PLS analyses. Microbiological analyses Basal respiration (B-resp) and substrate-induced respiration (SIR) were determined through the evolution of CO2 with a respirometer using the method and equipment described by Nordgren (1988). The SIR was measured after a 10-day incubation period. At the time of SIR, each soil sample received a mixture consisting of glucose, (NH4)2SO4 and talcum evenly mixed into the soil. Nitrogen mineralisation capacity (N-min) was analysed by an anaerobic incubation method (Stenberg et al., 1998a). Ammonium in the supernatant of centrifuged samples was analysed photometrically. Potential ammonium oxidation rate (PAO) was assayed as accumulated nitrite using a short incubation, chlorate inhibition technique 66 - (Torstensson et al., 1992). The rate was calculated by linear regression of accumulated NO2 . The potential denitrification activity (PDA) was assayed according to the short incubation, C2H2 inhibition method described by Pell et al. (1996). The denitrification rates were calculated by non-linear regression of accumulated N2O over time using a product formula for growth-associated product formation (Stenström et al., 1991). The method also gained the specific growth rate of denitrifiers (µPDA). Phosphatase activity was analysed by the p- nitrophenyl phosphate method described by Sjökvist (1993). Both acidic (Aci-P) and alkaline (Alk-P) phosphatase activities were determined, at pH 6.5 and 9 respectively. Statistical analysis Multivariate statistics were performed using the Unscrambler (CAMO A/S, Trondheim, Norge) and Statistica (Statsoft, Inc) software packages. All variables were centred and scaled to equal variance by dividing them by their own standard deviation. Standard and stepwise discriminant function analyses (DFA), based on Mahalanobis distance, were performed in order to determine which variables discriminated between the a priori defined groups of sludge and nitrogen treatments. Results and Discussion Soil quality indicators and the formation of soil quality groups In the natural variation samples, no distinct groups were formed. However, there was a clear separation between aggregated and single grained soils (Fig. 1). Of course clay content had a large influence on this separation, but also PDA, SIR, Alk-P, PAO and soluble cat ions (Fig.1B). This indicates the importance of soil texture and structure for the microbiological status of a soil. Known differences in crop sequence and fertilizer strategies were not reflected by the PCA’s. Soils with ley and organic manures in their rotations where scattered all over the score plots (Fig. 1A). On the other hand the most intensively cropped soils were gathered close to each other (circled in Fig. 1A). Relatively high pH-values, low organic matter content and intermediate clay-content characterise these soils. Apparently the organic matter in these soils was more susceptible to degradation as microbial activities expressed per amount organic-C , were highest there. This was probably caused by the intense cropping in combination with a nutrient status, pH, and structure and climate regimes (all but two soils are located in the southernmost part of Sweden) which are all favourable for microbial activity. 67 4 0.6 (7) 3 cSilt (16) (23) (12) 0.4 2 (5) (9) (6) (21) 1 (15) sg (22) 0.2 Temp (4) pH B-res (10) (2) Alk-P C/N 0 (1) (3) SIR qCO2 (18) PAO (20) (17) (26) 0.0 P/E -1 (11) PDA (8) Ca-AL -2 (13) (14) -0.2 K-AL a (24) -3 Principal Component 2 (27%) 2 Component Principal Clay Mg-AL Corg N (25) 2 (27%) Component Princopal -0.4 tot -4 (19) -5 -0.6 -6 -4 -2 0 2 4 6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 0.2 0.3 0.4 0.5 Principal Component 1 (45%) Principal Component 1 (45%) Figure 1. The corresponding score (A) and loading (B) plots of the first two principal components explaining 72% of the total variation in the 26 soils representing natural variation. Microbiological variables are expressed per amount organic C, others per amount dry matter. Soils in grey include ley in the crop rotation; sg = single grain; a = aggregated. Most intensively cultivated soils are circled. From Stenberg et al. (1998b). In the soil structure experiment at Sundby, four principal components separated well between both limed and unlimed treatments, cultivation depth 12 and 24 cm, and sample depth (0-12 and 12-24 cm) (Fig. 2). In the first component the stubble cultivated upper layer was well separated from the lower layers and somewhat from the mouldboard ploughed upper layer (Fig. 2A). This separation was possibly derived from the concentration of residues to the upper layer of the topsoil caused by shallower tillage. Higher values in variables related to the organic matter (including heterotrophic microbial activities), aggregate stability and pores <1 and >30 µm, which were found there. Interestingly, the corresponding adverse difference between the lower layers could not be detected, indicating a total increase in organic matter and microbial activity down to 24 cm. However, penetration resistance and water content separated the lower layers in the two tillage treatments in component 3 and 4 (component 4 not shown). In the second component pH, PAO, µPDA and PDA separated between limed and unlimed samples of all depth (Fig. 2A and B). Thus after 8 years, lime did not have any measurable effects on the soil structure. However, yields were slightly higher in the stubble- cultivated treatment, especially in combination with lime. These effects were pronounced in 1994 when soil surface crusts were formed soon after drilling followed by a long dry period. These circumstances inhibited seed germination in all treatments, but the problem was reduced in those which were stubble-cultivated and limed. 68 6 0,6 B µPDA A pH 0,4 PAO 4 qCO2 N-min 0,2 2 <1 B-res WC N-tot 0,0 BD C-org AGS>30 Pen 0 SIR -0,2 Kaid Kaiu Component 2 (22%) 2 Component -2 (22%) 2 Component -0,4 PDA -4 -0,6 -6-4-202468 -0,4 -0,2 0,0 0,2 0,4 Component 1 (40%) Component 1 (40%) 3 0,6 C D Pen 2 0,4 1 AGS>30 0,2 Kaiu µPDA 0 pH qCO2Kaid 0,0 BD PAO -1 C-orgB-res -0,2 N-minN-totSIR -2 PDA Component 3 (11%) Component Component 3 (11%) <1 -3 -0,4 WC -4 -0,6 -6-4-202468 -0,4 -0,2 0,0 0,2 0,4 Component 1 (40%) Component 1 (40%) Figure 2. The corresponding score (A and C) and loading (B and D) plots of the first three components explaining 73% of the total variation in the Sundby soil structure experiment. All variables are expressed per amount dry matter. Open symbols = with lime; Filled symbols = no lime; G= stubble cultivated, sampling depth 0-12 cm; I= stubble cultivated, sampling depth 12-24 cm; ▲= mouldboard ploughed, sampling depth 0-12 cm; ▼= mouldboard ploughed, sampling depth 12-24 cm. From Stenberg et al. (2000). Separation of different N and sewage sludge amendments In most cases it was possible to divide the different sludge and nitrogen treatments into groups. The first principal component from the PCA analysis generally described organic carbon and total nitrogen, which means that these parameters describe the major variation (30-38%) in the data set (not shown). The second principal component mainly accounted for pH effects and described 16-30% of the total variation (not shown). DFA of the experimental sites after 16 years of sludge amendment (Fig. 3) revealed that Aci-P was the strongest contributor to the ability to divide the different sludge application rates into groups. Other important variables were Alk-P, PAO and Tot-N for Igelösa, and pH, PAO and B-resp for Petersborg. 69 At Igelösa all three sludge treatments were significantly different (p<0.05), but at Petersborg it was not possible to distinguish between (A) and (B). It was not possible to distinguish between the different nitrogen treatments at Igelösa. At Petersborg, however, all three treatments were different (not shown). The most important variables were N-min, PAO, Aci-P and Tot-N. 3 Petersborg 3 Igelösa B-resp 2 2 1 Aci-P 1 Aci-P Org-CTot-N SIRTot-N PDA 0 Org-C 0 N-min Root 2 PAO PDA pH Alk-PpH -1 -1 B-resp N-min Alk-P SIR -2 -2 PAO -3 -3 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 Root 1 Root 1 Figure 3. Biplot of canonical centroid scores and loadings from discriminant function analysis of the two experimental sites after 16 years of sewage sludge amendment. Treatments were 0 (G), 1 (I) and -1 -1 3 (▲) ton sludge ha year . Bars indicate SD. From Johansson et al. (1999). The differences between the two soils were reflected in the way Petersborg reacted to lack of nitrogen fertilisation. The more active Igelösa soil could more or less compensate for this lack of nitrogen, but this was not the case for Petersborg. This was further strengthened by the incapability of DFA to distinguish between any of the different nitrogen treatments at Igelösa. However, PCA revealed similarities between the two experimental sites. Although the absolute activities in the two soils were different they responded similarly to sludge amendments, as revealed by their scores (not shown), and the microbiological parameters were distributed in the same pattern, as revealed by their loadings (not shown). The DFA was a powerful technique enabling identification of the different sludge treatments and the establishment of the most influential parameters for this differentiation. Soil quality indicators and productivity potentials In this study the variation in soil variables measured on the 26 natural variation samples were coupled to C- and N-uptake in roots and in H1 and H2. To put an emphasise on the microbiological indicators potential as qualitative indicators of the availability of the organic matter they were expressed on an organic C content instead of a dry matter basis. For all soil variables and combinations of soil variables, the best correlations were found to H2 and in all cases better to the N-uptake than to the C-uptake. The best simple correlation was between total N-uptake in H2 and total N-content in soil (r2=0.80). The independent variables selected for the best PCR and PLS models were in all cases org-C, tot-N, C/N, coarse silt, PDA, N-n, 70 SIR and P/E (humidity). PCR and PLS modelled N-uptake in H2 slightly better than tot-N (r2 = 0.97 and 0.95 respectively). PCR used three components and PLS only two. This made PLS the easier model to interpret. The first component in the PLS-model could be assigned as a quantitative component as it mainly explained org-C, tot-N coarse silt and P/E. The second component could be assigned as a qualitative component as it mainly explained the microbiological indicators and C/N (Fig. 4). According to this, soils with normal microbial activity were grouped along a regression line. Soils with high, and for fertility more favourable, microbial activity, were grouped above the line (circle a), and vice versa (circle b). As visualised by Fig. 4, the relation between samples and indicators can immediately be interpreted, as scores and loadings are plotted in the same graph in a biplot. 0.6 Figure 4. Bi-plot of the best PLS model showing the SIR N-minPDA 5 relationship between variables (loadings) and 26 soils 0.4 7 representing natural variation (scores) and the 178 a 16 18 22 R1C 0.2 correlation to harvest (H and R) variables. The first 15 14 H1CH2C 10 6 R2NH2N two components explain 74% of all H and R 3 1 H1N 0.0 23 variables on the average. Microbiological variables 24 9 12 P/E Ntot cSilt 19 13 4 2 are expressed per amount organic C, others per -0.2 21 Corg amount dry matter. For circles a and b refer to Loadings (Comp. 2) 26 -0.4 20 discussion. The line is the regression line of 25C/N b 11 uncircled soils. From Stenberg (1998). -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 Conclusions Loadings (Comp. 1) The use of different multivariate techniques has proven to be a fruitful approach capable of answering several different questions regarding soil quality issues. Interpretable quality groupings and separations between soils are revealed by PCA and PLS. While PCA is unbiased, PLS has the feasibility to exclusively relate groupings towards a chosen soil quality element. DFA is powerful when it comes to the identification of important variables for group distinctions. Microbial indicators are not commonly used in standardised soil test packages. However, their integrative response to the soil environment and their sensitivity to changes strongly suggest that they should be included in a set of variables to assess soil quality. The microbiological indicators have a straightforward relevance, especially when soil functions of concern are related to cycling of nutrients or degradation of pollutants. 71 References Doran, J. W. & Parkin, T. B. 1994. Defining and assessing soil quality. In: Doran, J. W., Coleman, D. C., Bezdicek, D. F. & Stewart, B. A. (eds.) Defining soil quality for a sustainable environment. 35, American Society of Agronomy Special Publication, Madison, pp. 3-21. Elliot, E. T. 1997. Rationale for developing bioindicators of soil health. In: Pankhurst, C. E., Doube, B. M. & Gupta, V. V. S. R. (eds.) Biological Indicators of Soil Health. CAB International, Wallingford, UK, pp. 49-78. Johansson, M. Stenberg, B. & Torstensson, L. 1999. Microbiological and chemical changes in two arable soils after long-term sludge amendments. Biology and Fertility of Soils 30, 160- 167 Nordgren, A. 1988. Apparatus for the continous, long-term monitoring of soil respiration rate in large number of samples. Soil Biology & Biochemistry 20, 955-957. Pell, M., Stenberg, B., Stenström, J. & Torstensson, L. 1996. Potential denitrification activity assay in soil - with or without chloramphenicol? Soil Biology and Biochemistry 28, 393- 398. Sjöqvist, T. 1993. A method to test phosphatase activity in soil. In: Torstensson, L. (ed.) Soil Biological Variables in Environmental Hazard Assessment - Guidlines. report No 4262, Swedish EPA, Stockholm, pp. 134-140. Stenberg, B. 1998. Soil attributes as predictors of crop production under standardized conditions. Biology and Fertility of Soils 27, 104-112. Stenberg, B., Johansson, M., Pell, M., Sjödahl-Svensson, K., Stenström, J. & Torstensson, L. 1998a. Microbial biomass and activities in soil as affected by frozen and cold storage. Soil Biology and Biochemistry 30, 393-402. Stenberg, B., Pell, M. & Torstensson, L. 1998b. Integrated evaluation of variation in biological, chemical and physical soil properties. Ambio 27, 9-15. Stenberg, M., Stenberg, B. & Rydberg, T. 2000. Effects of reduced tillage and liming on microbial activity and soil properties in a weakly-structured soil. Applied Soil Ecology (in press), Stenström, J., Hansen, A. & Svensson, B. 1991. Kinetics of microbial growth associated product formation. Swedish Journal of Agricultural Research 21, 55-62. Torstensson, L., Stenberg, B. & Stenström, J. 1992. Determination of ammonium oxidation, a rapid method to test chemical influence on nitrification in soil. In: Anderson, J. P. E., Arnold, D. J., Lewis, F. & Torstensson, L. (eds.) The international symposium on environmental aspects of pesticide microbiology. SLU, Uppsala, Sigtuna, Sweden, pp. 49- 54. 72 Denitrification, a soil quality indicator Mikael Pell1, Kalle Svensson1 and Ewa Bringmark2 1Department of Microbiology, Swedish University of Agricultural Sciences (SLU), Box 7025, SE-750 07, Uppsala, SWEDEN E-mail: [email protected] 2Department of Environmental Assessment, Swedish University of Agricultural Sciences (SLU), Box 7050, SE-750 07, Uppsala, SWEDEN Summary The aim of the present paper is to describe the use and experience of a soil microbial test method based on the measurement of potential denitrification activity. The results and the usefulness of potential denitrification activity as a soil quality indicator is discussed in relation to other biological, chemical and physical properties. Examples are given from (1) a long-term field experiment aiming at comparing various farming practices, and (2) a single arable field sampled in a grid pattern to evaluate spatial variability by geostatistics. The potential denitrification activity (PDA) assay, used in our studies, is a rapid method for assessing the amounts/activity of denitrifying enzymes in the soil. In addition, the specific growth potential of denitrifying bacteria (µPDA) can be estimated in the same test. Compared to PDA, the µPDA was better in discriminating between a cropping system receiving farmyard manure and a cropping system receiving only inorganic fertiliser. Both parameters, however, displayed a low variability in a single arable field, especially when compared to those reported for field measurements of denitrification losses. The spatial pattern of PDA, as revealed by ordinary block kriging was distinct and resembled that of various aerobic respiration parameters, such as substrate induced respiration and basal respiration. A general conclusion is that PDA and µPDA describes the soil carbon content and quality rather than a gaseous loss of nitrogen. The PDA assay, therefor, is suitable to be included as a soil quality indicator in a test package for integrated evaluation of physical, chemical and biological parameters. Keywords: denitrification, field experiment, geostatistics, microbiology, soil microbiological tests, potential denitrification activity, soil quality Introduction Soil quality is a complex concept. Due to the increasing awareness of our dependence on the productivity of the soil ecosystem many researcher have tried to define soil quality (Stenberg, 1999). In a more narrow point of view, soil quality can be described by its physical, chemical and biological compartments. Thus, quality would be the ability of these three compartments to effectively interact and perform in the ecosystem. Methods to describe the physical and 73 chemical components have since long been well established, while standardised methods to describe the biological component are less developed (Torstensson et al., 1998). Due to their small size, hence being in intimate contact with the soil mineral and organic particles, microorganisms ought to be ideal indicators of soil quality. A logical and generally adapted strategy in soil testing seems to be to chose microorganisms and activities from the nutrient cycles of carbon (e.g. biomass and respiration patterns), nitrogen (e.g. mineralisation and nitrification) and phosphorus (e.g. acidic and alkaline phosphatase activity). Biological denitrification is commonly defined as the process where nitrogenous oxides, mainly nitrate and nitrite are reduced to the gases di-nitrogen oxide and di-nitrogen (Zumft, 1992). Denitrification is carried out by respiring bacteria that gain energy by coupling the reduction of nitrogenous oxides to electron transport phosphorylation. Almost all denitrifiers prefer oxygen as the terminal electron acceptor and therefore reduce nitrogenous oxides only under anaerobic conditions. Denitrification is not just a process where nitrogen is lost to the atmosphere but might also be an indicator of easy available organic carbon, since most denitrifiers are organotrophic and mineralise organic matter anaerobically. Moreover, the denitrifying steps in the denitrifying pathway have different sensitivity to various kinds of disturbances. Denitrifying capacity is a widely spread feature among soil bacteria. Representatives are found within most taxonomical and physical groups. The genera Pseudomonas, Alcaligenes and Bacillus are thought to be the most frequently found denitrifiers in the soil, however, the list of denitrifiers is changing with the introduction of molecular biology techniques for identification and determination of relationships. The aim of the present paper is to describe the use of a soil microbial test method based on the measurement of potential denitrification activity. Examples are given from (1) a long-term field experiment aiming at comparing various farming practices, and (2) a single arable field sampled in a grid pattern and evaluated for spatial variability by geostatistics. The results and the usefulness of potential denitrification activity as a soil quality indicator are discussed in relation to other biological, chemical and physical properties analysed. Material and Methods Example 1. Long-term field experiment Three sites (Örja, Ekebo and Orup) of a fertility experiment located in southern Sweden were sampled. The experiment, which was started in 1957, has been described by Jansson (1975), Kirchmann and Ericsson (1993), and Kirchmann et al. (1999). The experiments are designed to compare two cropping systems (system I receiving farmyard manure and inorganic fertiliser, and system II receiving only inorganic fertiliser), PK-fertiliser regime and N fertiliser regime. In our study both cropping systems, all four PK-fertiliser regimes (0 kg (A); compensation for harvested P and K (B); compensation plus 15 kg P and 40 kg K (C); compensation plus 30 kg P and 80 kg K (D) ha-1 year -1) and two N fertiliser regimes (0 kg; 100 kg N ha -1 year -1) were 74 sampled and analysed. The SAS procedure GLM was used to detect statistically significant differences (p<0.05) between treatments. Example 2. Single arable field The field chosen for the study was plot number one at the Ekhaga experimental farm (10 km E Uppsala, Sweden) (Moeller and Salomonsson, 1992). The farm is organically managed with cereals and leys in the rotation, but without farmyard manure. Soil management is conventional with mouldboard plowing between crops. The clay content ranged between 49 and 53%. In the area, the field is regarded as very fertile. Visual inspections during growing season suggested a subjective uniform field. The field was sampled in a rectangular grid of 66 x 220 m. Grid nodes of 18 x 15 m were marked, making a total of 52 grid sampling points. Along one diagonal of the fiels a transect with sampling points at 2 m interdistance was identified and sampled, making a total of 108 transect sampling points. Transect and grid data were used together to calculate variograms (GS+, ver 2.3, Geostatistics for environmental sciences, Gamma design Software, MI). Data were detrended before geostatistical calculations. Ordinary block kriging was performed on grid data and an interpolation map drawn (Goovaerts, 1998). Soil treatment Soil samples were sieved (4 mm) and stored in portions of 150 g in polyethylene bags at - 20°C. All microbiological analyses were performed within one year to avoid influence of storage (Stenberg et al., 1997). Potential denitrifying activity (PDA) Potential denitrification activity (PDA) was assayed according to the C2H2 inhibition method (Pell et al., 1996). At the start of the assay, 25 ml of a substrate containing 1 mM glucose and 1 mM KNO3 was added to each flask containing 25 g of soil. A manifold system was used to evacuate (-100 kPa) and flush (+100 kPa) the flasks with N2 four times. After venting the headspace to atmospheric pressure, C2H2 was injected to a partial pressure of 10 kPa. Slurries were incubated at +25°C on a rotary shaker at 225 rpm for 120 min. During incubation, headspace samples of 0.5 mL were withdrawn at the start and then every 30 min. They were transferred to 10 mL gas-tight vials and stored until analysis of N2O. The N2O was analysed on a gas chromatograph. Concentrations of N2O were corrected for the amount dissolved in the liquid using a Bunsen coefficient of 0.571. The rate of N2O formation increased over time and data were fitted by non-linear regression to a product equation that takes exponential growth into consideration. By using this equation the initial rates of product formation (PDA) and specific growth rate constant (µPDA) could be calculated (Stenström et al. 1991; Johansson et al., 1998). 75 Results and Discussion The PDA method Nitrous oxide emissions as well as actual denitrification rates from soil ecosystems have proven to be extremely temporal and spatial variable, and thereby hard to estimate (Christensen et al., 1990; Pennock et al., 1992). The PDA in soil could be expected to be more constant since it is not affected directly by the, sometimes, rapid changes in environmental abiotic variables. Potential denitrification activity (PDA) is a common technique to characterise denitrification in soil. However, several techniques exist for assessing PDA. The general idea of PDA is to optimise the conditions so that only the amount of denitrifying enzymes will be rate-limiting for the process. This can be accomplished in an anaerobically incubated soil slurry with additions of optimal amounts of a terminal electron acceptor, usually nitrate, and an easily- available carbon and energy source, such as glucose. The technique is rapid and has been developed for routine use in the laboratory. Chloramphenicol (CAP) was originally suggested to be used in the assay medium to inhibit synthesis of new enzymes during the incubation. Today, it is generally accepted that (CAP) not only affects the synthesis but also lower the activity of existing enzymes (Pell et al., 1996). At our laboratory we have therefor, by omitting CAP, developed a kinetic approach where not only enzyme activity but also the specific cell growth is obtained simultaneously in the same assay (Stenström et al., 1991). Specific cell growth might even be a more sensitive parameter compared with enzyme activity towards various disturbances (Johansson et al., 1998; Pell et al., 1998). The long-term field experiment In the long-term field experiment the mean value of PDA of all treatments was highest at Örja -1 -1 -1 -1 (4.65 ng N2O-N g d.s. min ), followed by Ekebo (3.11 ng N2O-N g d.s. min ), and lowest -1 -1 for Orup (2.37 ng N2O-N g d.s. min ). The mean specific growth rates (µPDA) ranged in the same field-site order, from 0.0062 to 0.0057 min-1. The enzyme activity, PDA could significantly discriminate between the two cropping systems at Orup and Örja, however, being lower in system I at Orup, while being higher in the same system at Örja, compared to system II (Fig. 1). At Ekebo no significant effect could be observed. At Orup and Örja nitrogen fertilisation increased PDA, while it decreased PDA at Ekebo. These contradictory observations might be explained by the response of denitrification to a number of abiotic factors such as carbon substrate, nitrate supply, oxygen status, water content, and pH, of which organic carbon is generally thought to be the most important. When evaluating PDA together with other soil microbial and chemical data by principal component analysis (data not shown), it could be seen that PDA was positioned in-between substrate induced respiration (SIR) and potential ammonium oxidation (PAO) in component 1. This means that PDA had a less discriminatory power compared to the other microbial parameters. 76 PDA Orup Örja 7 7 6 6 -1 5 -1 5 4 4 DM min DM min -1 3 -1 3 2 2 O-N g O-N g 2 2 1 1 ng N ng N 0 0 P_RATE B D P_RATE B D P_RATE B D P_RATE B D A C A C A C A C N_RATE: 0 kg N_RATE: 100 kg N_RATE: 0 kg N_RATE: 100 kg µPDA Orup Örja 0.008 0,008 0.007 0,007 0.006 0,006 -1 -1 0.005 0,005 min min 0.004 0,004 0.003 0,003 P_RATE B D P_RATE B D P_RATE B D P_RATE B D A C A C A C A C N_RATE: 0 kg N_RATE: 100 kg N_RATE: 0 kg N_RATE: 100 kg Figure 1. Potential denitrification activity (PDA) rates and specific growth rates of denitrifiers (µPDA) in two arable fields (Orup and Örja) included in the Swedish long-term fertility experiment. (Treatments: Cropping system I (circles) and cropping system II (squares); PK fertilisation rates (A, B, C, and D); N fertilisation rates (0 and 100). Details are given in Materials and Methods.) The specific growth rate, µPDA, was significantly higher in cropping system II at all sites, but indifferent to PK-rate. In accordance to our results Stenberg et al. (1998) reported a negative correlation between µPDA and SOM. A small but significant effect by N-rate was observed at both Orup and Örja, however, negative for the former and positive for tha latter. In general, µPDA, seemed to be a good indicator of carbon input to the cropping system, as no statistical interactions could be observed between the treatments. This is an important aspect of a soil test since interactions limit the possibility of drawing conclusions on the discriminatory power of individual treatments. The single arable field -1 -1 The mean value of PDA of the investigated arable field was 19.6 ng N2O-N g d.s. min , thus being 4-8 times higher than the soils at Ekebo, Örja and Orup. The range and coefficient of -1 -1 variation (CV) were 13.9-25.7 ng N2O-N g d.s. min and 15.4%, respectively. -1 -1 Corresponding values for µPDA were 0.0060 min , 0.0040-0.0080 min and 18.8%, respectively. Expressing PDA per gram soil carbon resulted in a mean rate of 361 ng N2O-N -1 -1 -1 -1 g C min with a range and CV of 215-561 ng N2O-N g C min and 23.8%, respectively. When measuring rates of denitrification in the field after acetylene blockage very high spatial variability with a CV ranging between 74 and 383% have been reported. The high variability 77 observed is due to hot spots i.e. patches with activities more than five standard deviations above the overall mean value (Christensen et al., 1990). A well standardised method such as PDA will, indeed, result in less variation, thus, probably measuring a basic soil microbial property. In our case the CV’s of PDA were within the range of 16-107% as reported by Parsons et al. (1991). Moreover, the CV’s for the two denitrification variables were low, not only in comparison with most other microbial variables measured, but also compared to many of the chemical variables. Strong correlations (p<0.0001) were found between µPDA and pH (r = 0.79), Mg (r = 0.79) and Ca (r = 0.76). Such good correlations were also reported between the above abiotic variables and PDA (Stenberg et al., 1998). One explanation could be a relation between soil structural ions, such as Mg, and clay resulting in a positive influence of water retention. In our study PDA also correlated well with pH (r = 0.67). However, the strongest relation was found between PDA and substrate induced respiration (r = 0.80), but also the correlation to basal respiration was high (r = 0.67). In exploring the general functional structure of 26 soils, collected all over Sweden, Stenberg et al. (1998) found a close correlation between PDA and substrate induced respiration. Again, this indicates that PDA has a strong relation to the organic component of the soil. PDA and µPDA, respectively, displayed well developed variograms as shown by the squared correlation coefficients of curve fitting (r2 = 0.92 - 0.94). The ratio nugget to sill variance for PDA was somewhat high resulting in a flat variogram compared to those of chemical and physical properties. However, when PDA was expressed per gram carbon data improved as shown by an increase in squared correlations of crossvalidation, from 0.17 to 0.54 (p<0.001). The autocorrelation ranges, i.e. range of spatial variability or distance of independent sampling were 46 m for PDA and 89 m for µPDA, respectively. When PDA was expressed per gram carbon the range increased somewhat to 52 m. Parkin et al. (1987) found a negligible spatial structure of denitrification enzyme activity within 5 m, although a weak structure was revealed when the spacing decreased to 1.5 m (Parkin et al., 1987). Interpolation maps, derived by ordinary block kriging on the denitrification variables were constructed. PDA expressed per gram carbon showed a distinct pattern with high kriged values in one end of the field that gradually decreased toward the top left of the field. This pattern was similar to that of substrate induced respiration. Conclusions The potential denitrification activity (PDA) assay is a rapid method for assessing the amounts/activity of denitrifying enzymes in the soil. In the assay, the specific growth potential of denitrifying bacteria (µPDA) will also be estimated. Compared to PDA, the µPDA was better in discriminating between cropping systems receiving farmyard manure and cropping systems receiving only inorganic fertilisation. Both parameters displayed a low 78 variability in a single arable field, especially compared to those reported for field measurements of denitrification losses. The spatial pattern of PDA was distinct and resembled that of various aerobic respiration parameters, such as substrate induced respiration and basal respiration. A general conclusion is that PDA and µPDA describes the carbon content and quality of a soil rather than a gaseous loss of nitrogen. The PDA assay, therefor, is suitable to be included as a soil quality indicator in a test package for integrated evaluation of physical, chemical and biological parameters. References Christensen, S., Simkins, S. & Tiedje, J.M., 1990. Spatial variation in denitrification: Dependency of activity centres on the soil environment. Soil Sci. Soc. Am. J. 54, 1608- 1613. Goovaerts, P., 1998. Geostatistical tools for characterizing the spatial variability of microbiological and physico-chemical soil properties. Biol. Fertil. Soils. 27, 315-334. Jansson, S.L., 1975. Long-term soil fertility studies. Experiments in Malmöhus County [South Sweden] 1957-74. J. Royal Swedish Acad. Agric. And Forestry. Supplement 10, Kungliga Skogs- och Lantbruksakademien, Stockholm (Sweden). Johansson, M., Pell, M. & Stenström, J., 1998. Kinetics of substrate-induced respiration (SIR) and denitrification: Application to a soil amended with silver. Ambio 27, 40-44. Kirchmann, H. & Eriksson J., 1993. Properties and classification of soils of the Swedish long- term fertility experiments, 2: Sites at Örja and Orup. Acta Agric. Scand., Sect. B, Soil and Plant Sci. 43, 193-205. Kirchmann, H., Eriksson, J. & Snäll, S., 1999. Properties and classification of soils of the Swedish long-term fertilitu experiments, 4: Sites at Ekebo and Fjärdingslöv. Acta Agric. Scand., Sect. B, Soil and Plant Sci. 49, 25-38. Moeller, N. & Salomonsson L., 1992. Experimental farm - ecological village. A basis for discussion. In: Salomonsson, L., Nilsson E. & Jones T.B. (eds.) Agroecosystems and ecological settlements. Swedish University of Agricultural Sciences, Uppsala, Sweden. 8- 14. Parkin, T.B., Starr, J.L. & Meisinger, J.J., 1987. Influence of sample size on measurement of soil denitrification. Soil Sci. Soc. Am. J. 51, 1492-1501. Parsons, L.L., Murray, R.E. & Smith, M.S., 1991. Soil denitrification dynamics: Spatial and temporal variations of enzyme activity. Populations, and nitrogen gas loss. Soil Sci. Soc. Am. J. 55, 90-95. Pell, M., Stenberg, B. & Torstensson, L., 1998. Potential denitrification and nitrification tests for evaluation of pesticide effects in soil. Ambio 27, 24-28. Pell, M., Stenberg, B., Stenberg, J. & Torstensson, L., 1996. Potential denitrification activity assay in soil - with or without chloramphenicol? Soil Biol. Biochem. 28, 393-398. Pennock, D.J., Kessel, C. van, Farrell, R.E. & Sutherland, R.A., 1992. Landscape-scale variation in denitrification. Soil Sci. Soc. Am. J. 56, 770-776. 79 Stenberg, B., Johansson, M., Pell, M., Sjödahl-Svensson, K., Stenström, J. & Torstensson, L., 1997. Microbial biomass and activities in soil as affected by frozen and cold storage. Soil Biol. Biochem. 30, 393-402. Stenberg, B., Pell, M. & Torstensson, L., 1998. Integrated evaluation of variation in biological, chemical and physical soil properties. Ambio 27, 9-15. Stenberg, B., 1999. Monitoring soil quality of arable land: Microbiological indicators. Acta Agric. Scand., Sect. B, Soil and Plant Sci. 49, 1-24. Stenström, J., Hansen, A. & Svensson, B., 1991. Kinetics of microbial growth associated product formation. Swedish J. Agric. Res. 21, 55-62. Torstensson, M., Pell, M. & Stenberg, B., 1998. Need of a strategy for evaluation of arable land soil quality. Ambio 27, 4-8. Zumft, W.G., 1992. The denitrifying procaryotes. Chapter 23. In: Balows, A., Trüper, H.G., Dworkin, M., Harder, W. & Schlefer, K.-H. (eds.) The Prokaryotes. A Handbook on the biology of bacteria: Ecophysiology, isolation, identification, application, Vol. 1, 2nd edn., Springer-Verlag, New York. 554-582 Acknowledgements The support for this research was provided by the Swedish Environmental Protection Agency (NV) and the Swedish Council for Forestry and Agriculture Research (SJFR). 80 Changes in redox potential and Fe mobilization due to waterlogging in cultivated and non-cultivated soils at Alta, Northern Norway Christian Uhlig1, Gunter Wriedt2, Thomas Baumgartl2 and Rainer Horn2 1Holt Research Centre, The Norwegian Crop Research Institute, 9292 Tromsø, NORWAY E-mail: [email protected] 2Institute for Plant Nutrition and Soil Science, Christian Albrechts University, Olshausenstr. 40, D-24118 Kiel, GERMANY Summary This study investigated changes in soil physical and chemical parameters in cultivated and non-cultivated soils due to waterlogging in a laboratory experiment. Mineral soil material from a birch forest and a grass-dominated meadow was taken at Alta, northern Norway. Cultivated soil material showed considerably higher concentrations of organic-C, Kjeldahl-N, and plant available P, K, and Ca. Soil material from the non-cultivated site had a higher proportion of crystallized Fe, while the cultivated soil contained more amorphous and organic-bound Fe. In a laboratory experiment, platinum redox electrodes and porous ceramic cups were placed in the homogenized and waterlogged soil material, and changes in redox potential and Fe concentration were measured over a period of 51 days. Redox potential decreased steadily with increasing time of waterlogging in soils from both sites. In the non- cultivated soil redox potential decreased from +500 to +150 mV, while in cultivated soil it declined from +400 to below 0 mV. The critical redox potential for Fe3+ reduction was between +200 and +150 mV in the cultivated soil, while no mobilization of Fe could be found in the non-cultivated soil. Results indicate that 60 years of cultivation had led to significant changes in physical and chemical responses of the investigated soils due to waterlogging. Keywords: soil cultivation, iron mobilization, redox potential, waterlogging Introduction Winter damage of meadows is a serious problem for agriculture in Norway (Årsvoll, 1972). Unfavourable physical conditions for plant survival includes freezing, desiccation, and ice/ water cover. Excess of soil water, particularly in winter and spring, is regarded as one of the most important factors causing heavy yield reduction and poor persistence, especially in older grasslands (Fossbakken, 1967, Årsvoll, 1972). However, persistence of grasslands may also depend on other factors, such cultivation practices, plant diseases and pests, and characteristics of the cultivated plant species. Excess of soil water is frequently observed in soils of northern ecosystems (Zaidelman & Bannikov, 1996). The development of a continuous frozen layer during the period of annual frost in the winter is considered as one of the major causes. During the spring melting period, the annual frost table gradually dissipates and causes water from melting snow and thawing 81 soil to accumulate in overlaying soil layers (Magnussen, 1992; Zaidelman & Bannikov 1996; Clark and Ping, 1997). Once the annual frost has dissipated, well-drained conditions may prevail and continue throughout the remainder of the growing season. These seasonal waterlogging conditions can be retained for a period of several weeks (Zaidelman & Bannikov 1996). The most important chemical change that takes place when a soil is waterlogged is the reduction of Fe and the accompanying increase in its solubility (Ponnamperuma, 1972). The reduction of Fe is a consequence of the anaerobic metabolism of bacteria and appears to be chiefly a chemical reduction by bacterial metabolites. As a consequence, the solubility of Fe increases and can accumulate to concentrations that can have a damaging effect to plants in solution culture (Drew, 1981). The single electrochemical property that serves to distinguish submerged soil from a well- drained soil is its redox potential. The low potentials (+200 to -400 mV) of submerged soils reflect this reduced state, the high potentials (+800 to +300 mV) of aerobic media, their oxidized condition (Ponnamperuma, 1972). Redox potential below +200 mV, were observed in Alaskan soils during spring, indicating the reduced state of the soil and suggesting the presence of Fe2+ (Clark and Ping, 1997). It can be expected that plants of the original northern vegetation are sufficiently adapted to periods with reduced soil conditions, while introduced cultivated plants are necessarily not. Cultivation can lead to profound changes in soil properties. Little is known about the impact of soil cultivation on changes in response to waterlogging. The present study investigated whether the transformation of a birch forest into a grass-dominated meadow caused changes in soil physical and chemical responses to waterlogging. Material and Methods Sample site description Soils from two different soil profiles in Tverrelvdalen (UTM 34WEC924613), in the vicinity of Alta, Finnmark County, northern Norway, were investigated. Both soils developed from marine primary deposits (NGU, 1979) that sedimented in the Altafjord after last glaciation, and today are located 50 m above the sea level. The distance between the non-cultivated site and the cultivated soils was approximately 100 m. The soils were considered being comparable except for the present vegetation and land-use. Both soils normally remain frozen for 6 to 8 months of the year (Høstmælingen and Sveistrup, 1999). 82 A detailed soil description was carried out in profile pits at both locations: 1) Non-cultivated soil profile: L (5.5 to 5.0 cm), Of (5 to 0 cm), Ah (0 to 5 cm), Bv (5 to 18 cm), Sg1 (18 to 23 cm), Sg2 (23 to 44 cm), C (44 to 80 cm). 2) Cultivated soil profile: Ap (0 to 25 cm), Sg (25 to 40 cm), C (40 to 80 cm). The vegetation of the non-cultivated site, which was also the previous vegetation of the whole area before its cultivation started in 1930`s, is dominated by birch (Betula pubescens) in the tree layer and the dwarfshrubs crowberry (Empetrum hermaphroditum), bilberry (Vaccinium myrtillus), and lingnonberry (V. vitis-idaea) in the bottom layer. The forest site had been and is still used for extensive grazing by sheep. At the time of the study the cultivated soil was for the fourth year under meadow, mainly consisting of timothy (Phleum pratense) and smooth meadow-grass (Poa pratensis), and was without any particular agronomic problems. Sampling of soil material After removing the upper organic layer from an area of approximately 20 m2, a pit of 1m x 1m x 0,5 m was excavated at each site. Samples of 3 times 30 l of mineral soils from between 0 and 15 cm depth were taken horizontally by using a spade from each pit. The samples were then transported to Tromsø in six 31-l plastic boxes, air-dried (20 ° C) and homogenised. Soil samples for comparison of different forms of Fe and equilibrium extract between non- cultivated (Bv-horizon) and cultivated (Ap-horizon) soils were taken from the depth of 10-15 cm. Chemical and physical analyses For particle size analysis of soil, a pipette method was used (Elonen, 1971). The soil textural were given according to the Soil Survey Staff (1951). The analysis of exchangeable cations was done according to Ogner et al. (1975, 1977). The KCl extraction was carried out in the same way as the NH4Oac-extraction. Organic carbon was determined with Leco IR 212 Carbon system. These analyses were carried out at the Holt Research Centre, Tromsø, Norway. The dithionite-citrate method (Mehra and Jackson, 1960) was used to determine free Fe. Active Fe was determined following the methods described in Reeuwijk (1995) and Schlichting et al. (1995). Organic bound Fe was measured using the pyrophosphate method (e.g. Schlichting et al., 1995). These analyses were, as the analysis for equilibrium extract, carried out at the Christian-Albrecht-University in Kiel, Germany. Experimental lay-out In the laboratory 17 kg of homogenized soil material was placed in 31-l plastic boxes, and properly mixed with 10 l of distilled water. Three individual units of platinum redox electrodes and three units of ceramic cup devices were placed in the waterlogged soil material. Changes in redox potential and Fe concentration were measured over a period of 51 days at 20º C. 83 Platinum electrodes (Pfisterer and Gribbohm, 1989) were used in conjunction with an Ag/ AgCl reference electrode (Metrohm 6.0724.140) and meter values were standardized to the H electrode potential (Eh) by adding 199 mV to the meter value (Patrick et al., 1996). Due to several uncertainties, no attempts were made to correct the potentials for the effect of pH (Bohn, 1971; Ponnamperuma, 1972; Erber and Felix-Hennigsen, 1997). Before installation all Pt-electrodes were cleaned with 5% HNO3 and rinsed with distilled water. All Pt-electrodes used were tested against an industrial standard redox solution (Metrohm 6.2306.020) and the deviation was never greater than 1 mV. For the measurements of the redox potentials a pH meter (pH 340, Wissenschaftlich-Technische Werkstätten GmbH) was used. For sampling of soil water, porous ceramic cups from the manufacturer HiPerCeramics GmbH (Mullit M3-1350, length 66 mm, outer diameter 40 mm, average pore diameter 5 µm) were used. Before use, the ceramic cups were cleaned with 1 l of 0.1 n HCl and thereafter cleaned with 2 l of distilled water. For individual conditioning, the ceramic cups were then treated with 300 ml 1:10 soil solution, and 100 ml equilibrium soil solution, from either non- cultivated or cultivated soils. To avoid oxidation of reduced Fe by oxygen inside ceramic cups due to sampling, soil solutions from inside ceramic cups were sampled by replacement with O2-free N2 gas. Sampled soil solutions were immediately filtered through a 0.45 µm membrane filer (GN-6 Metricel, Gelman Sciences) and stored in acid washed glass bottles at 0.5 °C until analyses by an AAS (Model 2380, Perkin Elmer). The soluble Fe was considered to be in Fe2+ forms (Patrick and Jugsujinda, 1992). Although homogeneous soil samples make it possible to obtain better reproducibility of redox potential measurements, conditions are different compared to physically undisturbed soils. However, it must be remembered that in this study the primary interest was in the redox potential as a possible index of potential changes in soil properties due to cultivation, and not in the absolute values of redox potentials. 84 Results and Discussion Results from soil analysis (table 1) show generally similar soil properties of the two investigated sites. However, considerably higher concentrations of organic-C, Kjeldahl-N, and plant available P, K, and Ca was found in the cultivated soil, most probably due to the application of mineral fertilizer and manure. Table 1. Physical and chemical properties of non-cultivated soils at the depth of 0-15 cm. 40 35 30 25 20 % of total Fe 15 10 5 0 silicate crystallized amorphous organic Fig.1. Distribution of different forms of iron in non-cultivated ( ) and cultivated ( ) soils. Total extractable Fe was at comparable concentrations at both sites (table 1), while there were clear differences between different Fe forms (figure 1). Soil material from the non-cultivated site had a considerably higher proportion of crystallized Fe, while the cultivated soil contained more amorphous and organic Fe. This is most likely due to cultivation, which could have increased the degree of weathering of soil material. Iron concentrations in the equilibrium extract from non-cultivated soils were about 9 times higher than those of the cultivated soil (table 1). 85 Redox potential decreased steadily with increasing time of waterlogging in soils from both locations (figure 2). In the non-cultivated soil material redox potential decreased from +500 to about +150 mV, while those of the cultivated soil material declined from +400 to below 0 mV. The Eh-time curve from the cultivated soil had a steeper negative slope than those of the non-cultivated soil (figure 2), which indicates a faster oxidation-reduction process in the cultivated soil. In general, the decrease of redox potential due to waterlogging depends on the temperature, the organic matter quantity and quality, and the crystallinity of the oxides. Since temperature was identical, other factors have to be responsible for the faster decline of redox potential in the cultivated soil. The cultivated soil had a higher organic-C content and a slightly lower C/N ratio, which might result in a higher microbial activity than in the non-cultivated soil. Furthermore, organic matter from grasslands can be considered more favorable for decay than those of the birch forest. 800 80 700 70 600 60 500 50 400 40 300 30 200 20 100 10 ecnetain(mg/l) Fe concentration redox potentialredox (Eh) in mV 0 0 -100 0 7 14 21 28 35 42 49 days of waterlogging Fig. 2. Changes in redox potentials and Fe concentrations in non-cultivated and cultivated soils due to waterlogging ( Eh non-cult.; Eh cult.; Fe non-cult.; Fe cult.). In the cultivated soil Fe concentration in soil solutions increased noticeably after two to three weeks of waterlogging, and increased with concentration with further decreases in redox potential (figure 2). The critical redox potential for Fe3+ reduction was between +200 and +150 mV, which is in general agreement with earlier reported values (Gotoh and Patrick, 1974; Patrick and Henderson, 1981; Patrick and Jugsujinda, 1992). In the non-cultivated soil no mobilization of Fe could be observed, even at redox potentials below +200 mV. Amorphous and relatively weak crystallized Fe-oxides have, in comparison to good crystallized oxides, a distinct lower stability. In the beginning, reduction processes in soils will therefore primarily affect amorphous Fe(III)hydroxides. A reduction of the crystallized Fe(III)oxides will first occur at distinct lower redox potential (Scheffer and 86 Schachtschabel,1998). Results (figure 1) indicate a higher degree of more easily reducible Fe in the cultivated than in the non-cultivated soil. To a certain extent, this might help to explain observed differences in Fe mobilisation between the two soils at similar redox potentials. Results indicate that 60 years of cultivation had led to significant changes in physical and chemical response of northern soils due to waterlogging. Field experiments should investigate the possible ecological implications of these findings. References Bohn, H.I., 1971. Redox potentials. Soil Sci. 112, 39-45. Clark, M.H. & Ping, C.-L., 1997. Hydrology, morphology, and redox potentials in four soils of South Central Alaska. In: Vepraskas, M.J. & Sprecher, S.W. (eds.), Aquatic conditions and hydric soils: the problem soils. SSSA Special Publications No. 50, 113-131. Drew, M.C., 1981. Plant response to aerobic conditions in soil and solution culture. In: Smith, H. (ed.), Commentaries in Plant Science. Pergamon Press, Oxford, 209-223. Elonen, P., 1971. Particle-size analysis of soil. Acta Agralia Fenn. 122. 122 pp. Erber von, C. and Felix-Henningsen, P., 1997. Aussagekraft kontinuierlicher Redoxpotential- Messungen in tonigen Marschböden. Mittel. Dtsch. Bodenkdl. Gesellsch. 85, 233-236. Gotoh, S. & Patrick, W.H., 1974. Transformations of iron in a waterlogged soil as influenced by redox potential and pH. Soil Sci. Soc. Am. Proc. 38, 66-71. Høstmælingen, A.S. & Sveistrup, T., 1999. Monitoring of soil freezing depth, snow depth and winter damage to grassland in Northern Norway during the winter and spring 1998-99. Rapport 24/99, Planteforsk. Fossbakken, B., 1967. Overvintringsskader i eng og beite. Særtrykk av Ny Jord 4, 141-152. Magnussen, T., 1992. Studies of the soil atmosphere and related physical site characteristics in mineral forest soils. J. Soil Sci. 43, 767-790. Mehra, O.P. & Jackson, M.L., 1960. Iron oxide removal from soils and clays by dithionite- citrate system buffered with sodium bicarbonate. Clays Clay Miner. 7, 317-327. NGU, 1979. Alta - Kvartærgeologisk Kart 1834 I, M 1:50000. Norges geologiske undersøkelse. NGU-Skrifter Nr. 349. Ogner, G., Haugen, A., Opem, M. & Sørlie, B., 1975. Kjemisk analyse program ved Norsk institutt for skogforskning. Meddr. Norsk inst. skogforsk. 32, 207-232. Ogner, G., Haugen, A., Opem, M. & Sørlie, B., 1977. Kjemisk analyse program ved Norsk institutt for skogforskning. Supplement I. Meddr. Norsk inst. skogforsk. 33, 87-101. Patrick, W.E., Gambrell, R.P. & Faulkner, S P., 1996. Redox measurements of soils. In: Methods of soil analysis. Part 3. Chemical methods. SSSA Book Series no. 5, 1255-1273. Patrick, W.E. & Henderson, R.E. 1981. A method for controlling redox potential in packed soil cores. Soil Sci. Soc. Am. J. 45, 35-38. Patrick, W.H. & Jugsujinda, A., 1992. Sequential reduction and oxidation of inorganic nitrogen, manganese, and iron in flooded soil. Soil Sci. Soc. Am. J. 56, 1071-1073. 87 Pfisterer, U. & Gribbohm, S., 1989. Kurzmitteilung: Zur Herstellung von Platinelektroden für Redoxmessungen. Z. Pflanzenernähr. Bodenk. 144, 222-223. Ponnamperuma, F.N., 1972. The chemistry of submerged soils. Adv. Agron. 24, 29-69. Reeuwijk van, L., ed., 1995. Procedures for soil analysis. 5 Edition. ISRIC Wageningen. Scheffer, F. & Schachtschabel, P., 1998. Lehrbuch der Bodenkunde. 14. Edition. Ferdinand Enke Verlag Stuttgart. Schlichting, E., Blume, H.-P. & Stahr, K., 1995. Bodenkundliches Praktikum. 2. Edition, Berlin. Soil Survey Staff 1951. Soil Survey Manual. USDA Agriculture Handbook 18. 503 pp. Zaidelman, F.R. & Bannikov, M.V., 1996. Water regime and genesis of pseudofibrous and gley soils of glaciofluvial plains (poles’s landscapes). Eurasian Soil Sci. 29, 1132-1139. Årsvoll, K., 1972. Winter damage in Norwegian grasslands, 1968-1971. Norwegian Plant Protection Institute, Division of Plant Pathology, Ås-NLH, Norway. Report No. 56. Acknowledgements The work was financially supported by The Research Council of Norway. 88 Effects of crop rotation with perennial crops on macroporosity of a clay soil Laura Alakukku Agricultural Research Centre, Plant Production Research, FIN-31600 Jokioinen, FINLAND E-mail: [email protected] Summary The effects of crop rotation with perennial crops on the macroporosity of compacted soil were investigated in a preliminary field experiment on a clay soil (Vertic Cambisol). The trial plots were compacted twice (in 1981 and 1990) with a heavy loading. In 1992-94, crop rotation treatments with perennial crops grown as green fallow crops were carried out on the field experiment. According root observations, relative root density (number of roots at 0.35 m depth/number of roots at 0.10 m depth) of goat’s rue (Galega officinalis) and cocksfoot (Dactylis glomerata) was clearly less than that of melilot (Melilotus officinalis). Below 0.35 m depth, soil macroporosity (> 0.3 and > 0.03 mm) and the number of cylindrical pores ≥ 2 mm was greater in control treatment (unloaded, oats (Avena sativa) as monoculture) than in loaded crop rotation treatments. Keywords: subsoil, green fallow, biopores, root density, melilot, goat’s rue, cocksfoot, oats Introduction Soil compaction has been found to reduce the macroporosity (> 0.03 mm) of the topsoil (Eriksson, 1982; Aura, 1983) and subsoil (Aura, 1983; Alakukku, 1996a) of mineral soils. Modifications in soil macroporosity are very important since they, in turn, affect virtually all physical, chemical and biological soil properties and processes. Tillage and natural processes alleviate the effects of soil compaction in the topsoil. Deep tillage seldom completely alleviates the compacted subsoil, and may lead to worse soil structure after recompaction (among others Kooistra and Boersma, 1994). Thus, the compacted subsoil is usually left to become alleviated by natural processes such as freezing/thawing, drying/wetting and biological activity. The effects of subsoil compaction has been found to persist in field experiments at least 9 years despite cropping and deep frost (Etana and Håkansson, 1994; Alakukku, 1996b). In these experiments, as most long-term field experiments concerning the persistence of the effects of subsoil compaction, have been relied on conventional cultivation methods and crop rotations. Actively growing plant root system may, however, offer an opportunity for modify the macroporosity of compacted subsoil (biological tillage, Dexter, 1991). Elkins et al. (1977) and Henderson (1989) reported that the roots of certain plant species can perforate compact soil layers and create easily accessible pathways for the roots of succeeding crops. Lehfeldt 89 (1988) found the roots of lupin (Lupinus spp.) and lucerne (Medicago sativa) to penetrate a plough pan and create biopores in sandy and loamy soils. Cresswell and Kirkegaard (1995) reviewed the subsoil amelioration by plant roots and found that ability to penetrate dense soil may be related to plant type (perennial v. annual), root system morphology (tap-rooted v. fibrous root systems) or specific physiological or morphological characteristics of the root, which enhance penetration ability (e.g. root tip pressure, fibrous sheaths on grass roots, resistance to root buckling). Perennial species (e.g. lucerne) may be more effective at biological tillage because of the long time and wide range of water content conditions in which to establish a deep root system. Materechera et al (1991) observed that roots, which have been growing in strong soils tend to have larger diameters than roots which have been growing in weak soil. They found also that the seedling roots of dicotyledons penetrated the strong medium more than graminaeous monocotyledons. Materechera et al. (1992) reported that tap-rooted species had bigger diameter than fibrous- rooted species and that a higher proportion of thicker roots penetrated the strong layer at the interface on topsoil and subsoil than thinner roots. Cresswell and Kirkegaard (1995) pointed out that whilst we might expect that plants are able to modify subsoil pore size distribution and that subsequent crops will benefit from the improved structure, this has yet to be demonstrated. Despite the importance of matter, biological tillage of compacted subsoil by roots has received little attention. In the present paper, the effects of crop rotation with perennial crops on subsoil macroporosity were investigated in a field experiment on a clay soil. Likewise, the root density of some perennial crops was determined. The objective of this preliminary study was to examine the ability of different (dicotyledons/monocotyledons) perennial green fallow plants to modify subsoil macroporosity. Materials and Methods The field experiment was established on a clay soil (Vertic Cambisol, FAO, 1988) at Jokioinen. The soil was had 0.03 g g-1 organic carbon and 0.48 g g-1 clay (< 0.002 mm) to 0.2 m depth, below which the organic carbon content was 0.01 g g-1 and the soil had 0.65 g g-1 clay. The experiment was established in autumn 1981 when the trial plots were compacted to 0.4 m by one pass with a tandem axle load of 19 Mg (tyre inflation pressure 700 kPa, Alakukku, 1996a). In 1990, the loading was repeated by three passes with a tandem axle load of 21 Mg (800 kPa). Control plots were not loaded. For 10 years after the first loading, spring cereals were the main crops grown (Alakukku and Elonen, 1995). In the years 11 to 13 (1992-1994) after the first loading, crop rotation treatments with perennial crops were carried out on the field experiment. The treatments were: (A) oats (Avena sativa) as monoculture; (B) oats – sunflower (Helianthus annus); (C) melilot (Melilotus officinalis) one year oats; (D) goat’s rue (Galega officinalis) monoculture; 90 (E) cocksfoot (Dactylis glomerata) monoculture. There were four replications. Sunflower, melilot, goat’s rue and cocksfoot were grown as green fallow crops (no fertilization, cutting once/growing season, crop material was left on soil surface). After the crop rotations, oats was grown one year. In subsequent years, no field traffic exceeded an axle load of 5 Mg and a tyre inflation pressure of 150 kPa. The nitrogen festilization of spring cereals was 80 - 100 kg ha-1. Undisturbed soil samples (PVC pipe 0.15 m in diameter, 0.55 m long) were taken for root density observations in 1993 (melilot) and 1994 (goat’s rue and cocksfoot) during the flowering of plants with a tractor driven soil auger (Pöyhönen et al., 1997). Three replicate samples were taken from one compacted plot of each crop from the same block. When the samples were taken, the PVC pipe was installed so that the plant was in the middle of the sampled area. In laboratory the samples were cut into subsamples in 0.05 m increments. To obtain a broken surface, cut surfaces were prepared by removing any smeared of damaged soil with a knife and brush. The number of roots was counted from the soil surface looking at the sample under a magnifying glass. Likewise, the diameter of roots was measured with a slide calliper and the type of macropore in which the root grew was determined visually. In the spring 1996, similar undisturbed soil samples as above were taken for laboratory analyses. Two replicate samples per plot were taken from the compacted plots of the treatments A and C-E and from the unloaded plots of the treatment A (control) when the soil moisture content was near field capacity. In the laboratory the samples were cut into three subsamples: plough layer (0-0.20 m), middle layer (0.20 to 0.35 m) and bottom layer (0.35 to 0.55 m). The saturated hydraulic conductivity and macroporosity (minimum diameter 0.3/0.4 mm and 0.03 mm) of subsamples were determined and soil structure was described for the profile of 0.55 m as presented in Alakukku (1996b, 1997). The number and cross-sectional area of cylindrical pores ≥ 2 mm (earthworm burrows) of subsamples were determined from the prepared surface of the sample as Alakukku (1997) described. The significance of differences in macroporosity was tested by analysis of variance using the MIXED-procedure of the SAS statistical program (SAS, 1992). The experimental design was a randomized complete block. When significant F-values (at the 5% level) occurred, means were tested at the 95% probability level by using the contrast statement. 91 Melilot 1993 Melilot 1993 0.10 0.10 0.20 De 0.20 pth (m 0.30 0.30 ) 0.40 Root diameter (mm) 0.40 < 1 > 1 0.50 0.50 0 0.5 1 1.5 0 % 50 % 100 % Roots (number cm-2) Roots in macropore Goat's rue 1994 Goat's rue 1994 0.10 0.10 0.20 0.20 De pth (m 0.30 0.30 ) 0.40 0.40 0.50 0.50 0 0.5 1 1.5 0 % 50 % 100 % Roots (number cm-2) Roots in macropore Cocksfoot 1994 Cocksfoot 1994 0.10 0.10 0.20 0.20 De pth (m 0.30 0.30 ) 0.40 0.40 Earthworm b. Root c. Crack Other 0.50 0.50 0 0.5 1 1.5 0 % 50 % 100 % Roots (number cm-2) Roots in macropore Figure 1. Mean root density profiles of melilot, goat’s rue and cocksfoot and the number of roots grown in earthworm burrows, root channels, cracks or other macropores relative to the total number of roots at each depth in compacted clay soil. Other macropore meant that roots grew, for instance, between aggregates or among straw ploughed into the soil. 92 Results and Discussion Root density observations Root density of melilot, goat’s rue and cocksfoot was determined in the last year when the crop was grown (Fig. 1). Melilot was biennial crop and the samples were taken in the second year after sowing. Goat’s rue and cocksfoot were perennial crops and the samples were taken in the third year after sowing. Root density observations were made during the flowering of crop because the root system expected to be the largest at that time. In 1994, crop roots grew deeper than to a depth of 0.40 m. All samples were, however, not able to be taken from the depth of 0-0.50 m and the root densities mean of three replicates are only shown in Figure 1. Root densities determined in different years are not comparable with each other because the weather conditions were different. Some differences between the root system of crops were, however, observed. According to present results, grass crop roots were thin (diameter < 1 mm) but other crops had also thick roots (≥ 1 mm, Fig. 1). The relative root density of goat’s rue and cocksfoot reduced faster than that of melilot. At 0.35 m depth, the root density of goat’s rue, cocksfoot and melilot were 12, 14 and 30%, respectively, of the density at 0.10 m depth. Relevant to this, Hansson et al. (1991) found that the root biomass of lucerne was higher than that of meadow fescue (Festuca pratensis) below ploughing depth (0.27 m). Elkins et al (1977), Henderson (1989) and Materechera et al. (1991, 1992) found differences between plant species in ability to penetrate strong soil. Whalley and Dexter (1994) reported that better selection of existing species or the production of new species with greater ability to improve soil structural quality is an exciting prospect for the future. An interesting observation was that approximately 50% of roots grew in biopores (root channels and earthworm burrows) below ploughing depth of 0.2 m (Fig. 1). Jakobsen and Dexter (1988) reported that a few large biopores may have a dominant effect on wheat root penetration at depth. Likewise, Pfleger and Werner (1986) found that 25% of spring barley roots grew in continuous biopores in a compacted layer even though the volume of those biopores represented only 0.84% of total porosity. 93 Table 1. Description of the clay soil structure 6 years after the second loading. In the first year after the loading oats and in the second year spring oilseed rape (Brassica rapa ssp. oloifera) were grown. After that crop rotations were: (A) 4 years oats; (C) 2 years melilot, 2 years oats; (D) 3 year goat’s rue, 1 years oats; (E) 3 years cocksfoot, 1 years oats. Control treatment (Co), no loading in 1981 or 1990 and crop rotation A. Depth (m) Description of soil structure at different depths 0-0.20 Coarse angular blocky structure in each treatment 0.20-0.55 Strong, prismatic to massive, very coarse to homogenous, poorly cleaved structure with some fine cracks, earthworm burrows and root channels to depths of 0.37, 0.42, 0.40, 0.43 and 0.39 in control and treatments A, C, D and E, respectively. In the plough pan (0.20-0.35 m), the soil of loaded plots (A, C-E) was more massive and homogenous than soil in unloaded control plots. Below the strong layer: prismatic (< 5 mm), easily cleaved subsoil with earthworm burrows and root channels Table 2. Median of saturated hydraulic conductivity (Ksat, n=4) in the clay soil 6 years after the second loading. Range in parentheses. Treatments as in Table 1. Depth Saturated hydraulic conductivity (cm h-1) (m) Co A C D E 0-0.20 207 (125-534) 127 (95-245) 178 (142-1594) 239 (91-539) 147 (108-235) 0.20-0.35 24 (6-81) 89 (8-125) 33 (4-208) 28 (2-46) 70 (34-78) 0.35-0.55 38 (4-166) 84 (0.05-192) 31 (10-242) 114 (0.2-233) 92 (42-181) Soil structure Table 1 gives the results of visible structure determinations. In each treatment the soil was massive and coarse under plough layer, but the most massive and homogenous layer was thinnest in the control plots (A, unloaded). In other studies on a clay (Alakukku, 1996b) and clay loam (Alakukku, 1998) the visible structure of soils was strong and massive greater depths in loaded than in unloaded plots at least 5 years after heavy loading despite crop rotation with 2-3 years grass (mixed timothy (Phleum pratence), meadow fescue and red clover (Trifolium pratence)). In the present experiment, the strong layer in loaded plots was still measurable nine year after the second loading. According to my unpublished data, in each crop rotation treatment soil penetration resistance was clearly higher in loaded plots than in unloaded plots from 0.30 m to 0.53 m in spring 1999. -1 In every treatment, the mean Ksat of soil was good (greater than 1 cm h , Table 2). Especially in the subsoil, the measured values varied widely. The variation was partly caused by the uneven distribution of continuous earthworm burrows, as discussed by Bouma (1991). 94 >0.3 mm, p=0.009 Depth (m) Co ab A b C c 0-0.20 D E ab ac Co A C 0.20-0.35 D >0.3 mm >0.03 mm E p=0.004 p=0.007 a a Co b b A b b 0.35- C D b b > 0.3-0.4 mm 0.3/0.4-0.03 mm E b b 0 5 10 15 20 Macroporosity (m3 100m-3) Figure 2. Mean (n = 4) soil macroporosity (> 0.3 and 0.3-0.03 mm at the depths of 0-0.20 and 0.35- 0.55 m; > 0.4 and 0.4-0.03 at the depth of 0.20-0.35 m) in the clay soil 6 years after the second loading. Treatments as in Table 1. Treatment means followed by same letters are not significantly different (p<0.05). Depth (m) Co Co A A C C 0.20 D D E E Co Co A A 0.35 C C D D E E Co Co 0.55 A A C C D D E E 0 100 200 300 400 500 600 0 0,05 0,1 0,15 0,2 0,25 0,3 -2 -2 Number of pores > 2 mm m -2 Area of pores > 2 mm (m 100m ) Figure 3. Mean (n=4) number and cross-sectional area of cylindrical pores with a diameter ≥ 2 mm in the clay soil 6 years after the second loading. Treatments as in Table 1. Likewise, the high Ksat values and variation were partly due to cutting of the samples, which evidently increased the number of continuous pores relative to unbroken soil profile. In the plough layer (0-0.20 m), the mean macroporosity in crop rotation plots was greater than that in oats monoculture plots (Fig. 2). The macroporosity (0.03 mm) of each treatment was, however, higher than the critical limit of 10 m3 100m-3 given by Aura (1983). Chan and Heenan (1996) cultivated four different winter crops in rotation with wheat (Triticum aestivum) on a red earth in Australia. They found that after two cycles (4 years), topsoils (0- 0.18 m) which had been under wheat-lupin and wheat-canola (Brassica napus) rotation were more porous and had lower shear strength that those which had been under wheat-pea (Pisum sativum) or heat-barley (Hordeum vulgare) rotation. 95 Below 0.35 m depth, the macroporosity (> 0.3 and > 0.03 mm) values in loaded plots were significantly smaller than in unloaded control plots despite crop rotation (Fig. 2). Likewise, at 0.55 m depth, the number of earthworm burrows was highest in control plots in each block (Fig. 3). In my earlier study, I found that in the layer of 0.15 m below ploughing depth the Ksat and macroporosity (>0.4 mm) of clay loam and silt soils were significantly greater in crop rotation (green fallow, mixed red clover and grass) plots than in plots grown only spring cereals (Alakukku, 1998). However, in a study on clay soil, the similar crop rotation as above (crop was cut and removed with light machines) did not change the macroporosity of subsoil (Alakukku, 1996b). McKeague et al. (1987) reported that in the Ap horizon of crop rotation plots (oats, 2 years lucerne, maize (Zea mays)) clay loam physical properties differed from those in continuous maize plots. Below 0.30 m depth, comparable horizons of soils from two plots were, however, similar According to present results, the crop rotation with perennial crops grown as green fallow did not alleviate the macroporosity of compacted subsoil (below 0.35 m). The green fallow crop was grown 2-3 years which was probably too short time. However, information on the density of root channels (pores < 2 mm) is required before final conclusions can be drawn on the effects of different crop rotations on the macroporosity of compacted subsoil. In a study on clay soil, the number of cylindrical pores less than 1 mm (root channels) in crop rotation plots was clearly higher than in continuous oats plots even though there were no significant differences in macroporosity (0.3 mm, Alakukku, 1996b). The proportion of root channels from the soil volume is usually small. Ehlers et al. (1983) found that the relative area of total area of root channels (< 2 mm) was 0.1-0.2 m2 100m-2 in the 0-0.40 m layer. Thus, the macroporosity determination used in the present paper might be too coarse to find the differences due to root channels. Even though the volume of root channels is small, they may still have a dominant effect on root growth (Ehlers et al., 1983, Fig. 1). Conclusions According to the present results, crop rotation with perennial crops grown as green fallow did not alleviate the effects of subsoil compaction. Below 0.35 depth, the macroporosity (> 0.3 and 0.03 mm) was largest in unloaded control plots (oats monoculture) in the sixth year after the heavy loading. For proper evaluation of the effects crop rotation on the macroporosity of compacted soil, more research is required into macroporosity and root channels. The effect of the root system of different crop species is also in need of further study. 96 References Alakukku, L., 1996a. Persistence of soil compaction due to high axle load traffic. I. Short- term effects on the properties of clay and organic soil. Soil Tillage Res. 37, 211-222. Alakukku, L., 1996b. Persistence of soil compaction due to high axle load traffic. II. Long- term effects on the properties of fine-textured and organic soils. Soil Tillage Res. 37, 223- 238. Alakukku, L., 1997. Properties of fine-textured subsoils as affected by high axle load traffic. Acta Agric. Scandinavica. Sec. B: Soil and Plant Sci. 47, 81-88. Alakukku, L., 1998. Properties of compacted fine-textured soils as affected by crop rotation and reduced tillage. Soil Tillage Res. 47, 83-89. Alakukku, L. & Elonen, P., 1995. Long-term effects of a single compaction by heavy field traffic on yield and nitrogen uptake of annual crops. Soil Tillage Res. 36, 141-152. Aura, E., 1983. Soil compaction by the tractor in spring and its effect on soil porosity. J. Sci. Agric. Soc. of Finland 55, 91-107. Bouma, J., 1991. Influence of soil macroporosity on environmental quality. Adv. Agronomy 46, 1-37. Chan, K.Y. and Heenan, D.P., 1996. The influence of crop rotation on soil structure and physical properties under conventional tillage. Soil Tillage Res. 37, 113-125. Cresswell, H.P. & Kirkegaard, J.A., 1995. Subsoil amelioration by plant roots – the process and the evidence. Australian J. Soil Res. 33, 221-239. Dexter, A.R., 1991. Amelioration of soil by natural processes. Soil Tillage Res. 20, 87-100. Ehlers, W., Köpke, U., Hesse, F. & Böhm, W., 1983. Penetration resistance and root growth of oats in tilled and untilled loess soil. Soil Tillage Res. 3, 261-275. Elkins, C.D., Haaland, R.L. & Hoveland, C.S., 1977. Grass roots as a tool for penetrating soil hardpans and increasing crops yields. Proc. 34th Southern Pasture and Forage Improvement Conference, Auburn, Alabama, 21-26. Eriksson, J., 1982. Markpackning och rotmiljö. Swedish University of Agricultural Sciences. Reports the Division of Agricultural Engineering. Report 354, 1-82. Etana, A. & Håkansson, I., 1994. Swedish experiments on the persistence of subsoil compaction caused by vehicles with high axle load. Soil Tillage Res. 29, 167-172. FAO, 1988., FAO/Unesco Soil map of the world, revised legend. World Resources Report 60, FAO, Rome, 138 pp. Hansson, A.-C., Andren, O. & Steen, E., 1991. Root production of four arable crops in Sweden and its effect on abundance of soil organisms. Special Pub. Series of the British Ecological Society 10, 247-266. Henderson, C.W.L., 1989. Lupin as a biological plough: Evidence for and effects on wheat growth and yield. Australian J. Experimental Agric. 29, 99-102. Jakobsen, B.F. & Dexter, A.R., 1988. Influence of biopores on root growth, water uptake and grain yield of wheat (Triticum aestivum) based on predictions from a computer model. Biol. Fertil. Soils 6, 315-321. 97 Kooistra, M.J. & Boersma, O.H., 1994. Subsoil compaction in Dutch marine sandy loams: loosening practices and effects. Soil Tillage Res. 29, 237-247. Lehfeldt, J., 1988. Auswirkungen von Krumenbasisverdictungen auf die Durchwurzelbarkeit sandiger und lehmiger Bodensubstrate bei Anbau verschiedener Kulturpflanzen. Archiv für Acker- und Pflanzenbau und Bodenkunde 32, 533-539. Materechera, S.A., Alston, A.M., Kirby, J.M. & Dexter, A.R., 1992. Influence of root diameter on the penetration of seminal roots into a compacted subsoil. Pl. Soil 144, 297- 303. Materechera, S.A., Dexter, A.R. & Alston, A.M., 1991. Penetration of very strong soils by seedling roots of different plant species. Pl. Soil 135, 31-41. McKeague, J.A., Fox, C.A., Stone, J.A. & Protz, R., 1987. Effects of cropping system on structure of Brookston clay loam in long-term experimental plots at Woodslee Ontarion. Can. J. Soil Sci. 67, 571-584. Pfleger, I. & Werner, D., 1986. Enfluβ röhrenförmiger Hohlräume in Bodeneigenschften und Pflanzenertrag sowie Massnahmen zu ihrer Minderung. Archiv Aker- und Pflanzenbau und Bodenkunde 30, 259-268. Pöyhönen, A., Alakukku, L. & Pitkänen, J., 1997. Maanäytteenoton koneellistaminen ja työntutkimus (Mechanization of the sampling of large soil cores and physical stress of soil sampling). Maatalouden tutkimuskeskuksen julkaisuja. Sarja A 22, 1-39. (In Finnish with English abstract). SAS., 1992. SAS Technical report P-229, SAS/STAT Software: Changes and enhancements, Release 6.07, Cary, NC: SAS Institute Inc. 620 pp. Whalley, W.R. & Dexter, A.R., 1994. Root development and earthworm movement in relation to soil strength and structure. Arch. Acker- Pfl. Boden 38, 1-40. 98 Indication of soil degradation in strawberry fields: disappearance of earthworms Sanna Kukkonen and Susanna Vesalo Agricultural Research Centre of Finland, Laukaa Research and Elite Plant Station, Antinniementie 1, FIN-41330 Vihtavuori, FINLAND E-mail: [email protected] Summary Earthworm population densities were estimated from four organically and four conventionally cultivated strawberry fields of moraine and fine sand soil types. A pot experiment was carried out to see whether Aporrectodea caliginosa Sav. could survive better in the organically than conventionally cultivated soil and to assess the possible influence of earthworms on the growth of strawberry. Soil samples for the experiment were collected from six organic and six conventional fields. The two field types differed from each other with respect to the duration of strawberry cultivation, plant protection, vegetation between rows and fertilisation. The earthworm density and biomass in the organic strawberry fields was 86 -522 worms m-2 and 10-290 g m-2. The dominating species was usually A. caliginosa or Dendrodrilus rubidus Sav. A. caliginosa seemed to favour the unmulched parts of the field. No earthworms were found from the conventional fields, but there were low densities in the field margins. In the pot experiment, adult A. caliginosa could survive three months equally well in the soil of both field types. The biomass of adult earthworms correlated with the soil organic matter content. Although there were no acutely toxic substances in the soil, the absence of earthworms in the conventional fields might be related to the repeated use of earthworm-toxic substances. Independent of the field type or soil nutrient status the growth of micropropagated strawberry was 33% better in the presence of earthworms than in their absence. Keywords: strawberry, Fragaria x ananassa, earthworms, Aporrectodea caliginosa, soil degradation, organic cultivation, conventional cultivation Introduction Earthworms are widespread in all kinds of soil habitats. They have mechanisms to survive over periods of naturally occurring hostile environmental conditions such as drought and freezing. They do not, however, have the means to escape certain agricultural operations such as pesticide treatments and tillage. That is why cultivated land usually has lower densities of earthworms than pastures or deciduous forests (Paoletti, 1999). Organic cultivation practices usually have a favourable influence on earthworm populations. This is mainly due to the increased use of organic matter, mechanical weed control and use of pesticides not harmful to earthworms (Pfiffner and Mäder, 1997, Werner, 1997). The drastic 99 reduction or even the total disappearance of earthworms in conventional cultivation has been related to the use of fungicides or pesticides (Clements et al., 1991, Raw, 1962). There is little doubt that earthworms act to the benefit of plant production in various soil types. It has been shown that advantage is gained from certain worms through improved drainage of the soil (Clements et al., 1991, Edwards et al., 1990, Pérès et al., 1998). Other mechanisms include the manipulation of soil micro-organisms in favour of the plant (Stephens et al., 1994a, Stephens et al., 1994b, Stephens and Davoren, 1997) and increased N availability (e.g. Boström, 1988a, Parkin and Berry, 1994, Bohlen and Edwards, 1995). A hypothesis concerning the disappearance of earthworm was proposed during soil sampling in the most important strawberry production area in Finland. The hypothesis was tested by systematic soil sampling, and a simple pot experiment was designed to see if earthworms could survive better in the organically than conventionally cultivated soil. A hypothesis that earthworms could improve strawberry growth in moraine and fine sand soil types was included in the experiment. Materials and Methods The strawberry fields studied (table 1) were situated at Suonenjoki, (62° 47’ N, 27° 06’ E) and Leppävirta (62° 29’ N, 27° 40’ E), Finland. On all the fields, strawberry was grown in black plastic-mulched beds. The conventionally cultivated fields differed from the organically cultivated ones with respect to the duration of strawberry cultivation, plant protection, vegetation between rows and fertilization. In conventional fields, several insecticides (methiocarb, chinomethionat, azinphos-methyl, lambda-cyhalothrin) and fungicides (iprodion, tolylfluanid) were sprayed frequently on the rows and herbicides (diquat, glufosinate- ammonium) between the rows. Fertilization was based on inorganic fertilisers, whereas in the organic fields both compost and stone meal (apatite, biotite) were used and fresh organic matter was formed from straw, weeds or grass between the beds. Earthworm populations in strawberry fields Sampling of earthworms was carried out in August 1998 in four organically and four conventionally cultivated fields of moraine and fine sand soil types. Samples were taken from plastic-mulched strawberry beds and between the beds. Three to five parallel samples were taken per field (6-63 samples ha-1). In most cases, 1-3 reference samples were taken from the field margins. For the chemical extraction of earthworms, a circular metal frame (0.38 m2) was embedded in the soil to depth of approx. 5 cm. The soil surface inside the frame was cleared from detritus. 100 Table 1. Soil properties, duration of strawberry cultivation, preceding crop and age of current strawberry culture in the study fields. Field no. Field Soil type % clay Cultivation of Preceeding crop Age of present type strawberry, y strawberry culture, y 1 conventional fine sand 4.7 4 vegetables 4 2 conventional Moraine 5.1 10 barley 3 3 conventional Moraine 4.4 37 strawberry 5 4 conventional Moraine 4.6 >50 ley 5 5 conventional Moraine 2.9 70 barley+potatoes 7 6 conventional Moraine 1.7 27 ley 6 7 organic fine sand 9.0 7 ley 3 8 organic moraine with fine sand 2.7 3 vegetables 3 9 organic moraine with fine sand 3.4 2 vegetables 2 10 organic moraine with fine sand 5.4 2 nursery plants 2 11 organic moraine with fine sand 8.8 8 ley 8 12 organic moraine with fine sand 5.7 4 vegetables 4 13 organic very fine sand 5.0 23 barley 5 14 organic very fine sand 3 pasture 3 15 organic moraine with fine sand 4 vegetables 4 The extraction solution, 150 ml of mustard powder dissolved in 9 litres of water, was poured inside the frame. Earthworms appearing during 10 minutes were picked and preserved in 70% ethanol. This was followed by a second application of mustard solution and another 10 minutes of picking. Due to the wetness and compaction of the soil the second application of the solution did not always infiltrate completely. Immediately after the chemical extraction a square frame (25*25 cm) was placed in the middle of the larger frame. The soil inside the frame was dug with a spade to a depth of 20 cm. The sample was spread on white plastic sheet and the earthworms were picked by hand and preserved in 70% ethanol. Pot experiment Soil samples for the pot experiment were collected from a total of twelve fields, six organic and six conventional, (no. 1-12 in table 1) at the end of May 1998. Ten samples of 8 litres each were taken from the strawberry beds. The sample was dug with a spade from a 25x25cm wide and 20 cm deep area and placed on a white plastic sheet where earthworms were removed by hand sorting. The sorted sample was placed into a plastic bucket (diameter 25 cm, depth 24 cm) with a perforated bottom. There was a plastic net (mesh size 1.6 mm) on the bottom to stop the worms from escaping and a layer of gravel for drainage. Samples for determining soil particle size fractions, organic C and soluble nutrient concentrations were drilled (100 drillings, 35 ml drilling –1) from the 0-20 cm soil profile. Soluble nutrients (Ca, Mg, P, K) were measured using acid ammonium acetate (pH 4.65) extraction. Seven adult A. caliginosa (corresponding to 143 m-2) collected from a field (with silt soil) at Laukaa (62° 20’ N, 25° 59’ E) were released on the surface of five pots from each field. The total fresh weight of worms (including their gut contents) per pot was measured. A plaster block was embedded in one of the five pots with and without earthworms. The pots were then arranged outside on a clear plastic sheet according to a split plot design with 6 blocks. The 101 area was covered with a clear plastic roof. Within each block, the pots were placed into two rows, each row containing the pots from one field and one shelter pot at each end of the row. The relative soil moisture was measured twice a week and pots were given 0.5 litre of water when the moisture level was below 80%. A micropropagated strawberry (Fragaria x ananassa Duch.), ´Senga Sengana`, was planted in each pot one week after the introduction of earthworms. The experiment was harvested after 12 weeks. Fresh and dry weights of strawberry shoots (runners excluded) and washed roots were determined. The soil was hand- sorted and earthworms were picked and preserved in 70% ethanol. They were identified and the fresh (preserved) and dry weights were measured including their gut contents. The change in the fresh weight of adult A. caliginosa during the experiment and the number of juvenile A. caliginosa at the end of the experiment were used as parameter to characterize the well-being of the worms. To estimate the influence of earthworms on strawberry growth, the dry weights of roots and shoots in the control and worm-treated pots were compared. The statistical analyses was based on the common mixed model for a split-plot design in the case of strawberry-related parameters and for randomised block design in the case of earthworm- related parameters. The analyses were performed by the SAS system MIXED procedure. Results and Discussion Field earthworm populations No earthworms were found from the conventionally cultivated fields (table 2). Low densities were found from the field margins, however. In the organically cultivated fields the average biomass and density of earthworms was 88 g m-2, 218 m-2, respectively. The difference between the two field types was so obvious that no statistical tests were needed. The dominant species was either A. caliginosa or D. rubidus. Juvenile Aporrectodea sp. and Lumbricus sp. were considered to belong to species A. caliginosa and Lumbricus terrestris L., respectively, since all adults from these genuses belonged to these species. A. caliginosa was more numerous in the corridors between the beds than in the beds themselves. Table 2. Total number and dry weight of earthworms in strawberry fields. Sampling area: A=strawberry beds, B=corridor between beds, C=field margin. Field type Field no. No. earthworms m-2 Dry weight of earthworms g m-2 Sampling area A B C A B C conventional 2 0.0 0.0 1.3 0.0 0.0 0.1 3 0.0 0.0 15.6 0.0 0.0 0.2 5 0.0 0.0 186.5 0.0 0.0 8.8 13 0.0 0.0 - 0.0 0.0 - organic 8 100.5 72.0 - 4.8 3.8 - 11 53.7 122.8 5.2 5.7 14.9 1.0 14 290.0 754.4 169.5 39.3 78.6 25.7 15 200.1 150.1 - 1.8 2.1 - 102 From the earthworm’s point of view a strawberry field is a convenient habitat since the soil moisture is kept at a suitable level by irrigation, soil pH is between 6 and 7 and tilling takes place only from every four to ten years. But the frequent use of pesticides may make survival difficult in a strawberry field cultivated by conventional methods. Food may also be in short supply because of chemical weed control and the lack of organic soil amendments. Total absence of earthworms has been recorded earlier from experimental grassland plots in the UK treated with phorate (Clements et al., 1991). The long-term use of copper fungicides has been related to drastic reductions in earthworm numbers in English orchards (Raw, 1962). Of the insecticides used in strawberry production, endosulfan is moderately toxic to earthworms and methiocarb very toxic (Edwards and Bohlen, 1996). Endosulfan was used annually against strawberry mite (Phytonemus pallidus Banks) until 1996 when its use in berry production was prohibited. Endosulfan and its metabolite endosulfan sulphate have been proven to be persistent in our cool climate, since signs of these could still be seen 3.5 years after the last use (Laitinen, unpublished). Although methiocarb is even more worm-toxic it is unlikely to be responsible for all the loss of earthworms, since it had been in use for only two years before the present study. However, the toxicity of all chemicals and their mixed use is not known. The organic production of strawberry is recent in Finland and few fields are older than 8 years. It seems probable that in the youngest fields the previous crop still had a considerable effect on the number and species of earthworms. For example, earthworms were most numerous in field no. 14, previously used as a pasture. Pot experiment Survival of earthworms Earthworms were removed from 17 of 60 samples collected from organic fields. No earthworms were encountered in the soil samples from two of the organic fields and six (all) of the conventional fields. The average size of A. caliginosa introduced to the pots was equal in both field types (t=1.00, p=0.334). The biomass of adult A. caliginosa decreased 58% in the worm-treated pots during the experiment. The change in biomass was, however, equal in both field types (fig 1a). The median number of adults at the end of the trial was higher in the pots containing soil from conventional fields than from organic fields (fig 1b). The average size of worms at the end did not differ significantly between the two field types (t=7.72, p=0.106). Juvenile A.caliginosa appeared in 77% of the worm-treated pots. The median number of juveniles was higher in the pots from organic fields than from conventional fields. The number of juveniles produced per introduced adult was 0.06-2.71. Because of large variations and a significant number of pots with zero values, it was not possible to prove whether the reproduction of worms was significantly better in organic fields than in conventional fields. Two pots contained also juveniles of D. rubidus. 13% of the control pots contained juvenile and/or adult A. caliginosa. 103 The pots were almost exclusively from organic field (no. 12) with a relatively high number of worms removed before the experiment. Three control pots also contained 1-2 juveniles of Lumbricus sp. and D. rubidus. Although the egg capsules left in the soil samples did not cause considerable error, it would be useful in future experiments to remove them also. 12 50 Organic fields Conventional fields Organic Adults 10 Conventional 40 Juveniles 8 30 6 20 4 10 2 a a b b 0 Control A. cal. Control A. cal. Biomass of A. caliginosa, g pot-1 caliginosa, A. of Biomass 0 -10 0 weeks 12 weeks 5 pots-1 median caliginosa, A. of No. Figure 1. a) Average weight of adult A. caliginosa in worm-treated pots at the start and at the end of the experiment. Columns containing the same letter were not significantly different from each other at the p<0.05 level. b) Median number of adult and juvenile A. caliginosa in the control and worm- treated (A. caliginosa) pots at the end of the experiment. Error bars indicate quartile deviation from the median value. The organic C of the soil samples varied between 1.6-14.6 %. The average soil org. C content was the same in both field types when the one outlying conventional field (no. 1) was excluded (t=1.42, p=0.26). Soil organic matter had positive effect on A. caliginosa: at the end of the experiment the adults were the bigger the higher the org. C content of the soil (field no. 1 excluded) (r2=0.96, p=0.0001). As stated above, at least three of the conventional fields included in the pot experiment were devoid of earthworms. During the soil sampling, no earthworms were encountered in the other three either. Yet the survival of adult A. caliginosa was not worsened in the soil from these fields. It seems that there were no substances in concentrations causing acute toxicity. The experiment might have been too short to discover the consequences of long-term exposure, i.e. diminished reproduction. Little is known about the sensitivity of different earthworm species to toxic substances. The ability of A. caliginosa to survive when reintroduced to conventionally cultivated soil does not necessarily mean that other species could survive. The number of worms recovered from the pots was relatively low. The decline in number of earthworms during the experiment can be explained either by dying or escaping (escaping was not prevented). It cannot be totally excluded that some worms had escaped the pots for example due to overpopulation. After the covering of the experimental area, the moisture level started to fluctuate and in some pots it fell below 50% before the next watering. There seemed not to be any systematic difference between soil samples. The fields studied had not 104 received manure or other organic matter inputs for 2-8 years. Short of food could explain the decline in worm number, since earthworms are known to be food- limited in agroecosystems (Boström, 1988b, Whalen et al., 1998). The positive effect of organic matter has also been documented to be gained through reduced toxicity of chemicals (Martikainen, 1996). However, this does not explain why there was similar survival of earthworms in the two field types. The number of juveniles was also relatively low. A. caliginosa reared in unlimited food supply are known to be able to produce a cocoon every 3-6 days (Evans & Guild, 1948, Lofs- Holmin, 1983). But, as the cocoons were not counted in this experiment, it cannot be said what was the cocoon production rate. Effect of earthworms on strawberry growth Earthworms had a positive effect on the growth of both roots and shoots of strawberry measured by dry weight (fig. 2). The difference was 33% in both cases. The effect was independent of the field type, but varied 0-72% between individual fields. The zero effect was detected in the organic field, where control pots also contained worms. According to the 6 Control Figure 2. Dry weight of strawberry 5 A. caliginosa shoots (g plant –1) at the end of the 4 experiment. Left columns = Control. Right columns = A. caliginosa. Error bars 3 indicate standard deviation. Columns 2 containing the same letter were not a 1 bbsignificantly different from each other at a Shoot dry weight g plant-1 weight dry Shoot 0 the p<0.05 level. organic conventional Finnish classification (Viljavuuspalvelu, 1997), the soil fertility varied between the fields from rather poor to high. The conventional fields had somewhat higher values of K, Mg and P than the organic ones, but the plant size did not correlate with the nutrient level of the soil (data not shown). Furthermore, a significant positive effect was seen even in the most fertile soils (P> 20 and K> 250 mg/l soil). The nutrient requirements of strawberry are comparatively low, but it benefits from increased nitrogen availability (Kongsrud, 1986). The nitrogen levels were not measured in this study, but as earthworms are known to release nitrogen from soil (e.g. Boström, 1988a, Pohlen & Edwards, 1995, Parkin & Berry, 1994) it seems a very probable explanation for the observed increase of strawberry size. Earthworms have also been proven to decrease the disease severity of some soil pathogenic fungi on wheat, subterranean clover and perennial ryegrass (Stephens et al., 1994a, Stephens et al., 1994b, Stephens and Davoren, 1997). It would have 105 been advantageous to observe the root health since strawberry often suffers from root rot (Hildebrand, 1941, Nemec, 1970, Parikka, 1981). Conclusions It was proven that conventional strawberry cultivation methods can have drastic, deleterious effects on earthworms in less than 10 years. The pesticides used are very likely the cause of the total disappearance of earthworms observed. Other mechanisms such as small inputs of organic matter combined with low content of fine particles of the soil may also be involved. On the basis of this study, however, it cannot be said which part of the Finnish strawberry fields are devoid of earthworms, since soil type and cultivation practices may vary. Estimations of earthworms from coarse-textured Finnish soils are scarce and comparisons of the species composition or density of earthworms with agricultural soils of similar texture in the production of other plant species cannot be made. The survival of introduced A. caliginosa was not reduced in soil from the conventionally cultivated strawberry fields in comparison with organically cultivated soil. In the conventional fields there were no substances in concentrations causing increased mortality of adult earthworms, but there were indications of potential long-term effects on reproduction. Earthworms enhanced the growth of strawberry shoots and roots, but the mechanism was not studied. References Bohlen, P.J. & Edwards, C.A., 1995. Earthworm effects on N dynamics and soil respiration in microcosms receiving organic and inorganic nutrients. Soil Biol. Biochem. 27: 341-348. Boström, U. & Lofs-Holmin A., 1988a. Earthworm population dynamics and flows of carbon and nitrogen through Aporrectodea caliginosa (Lumbricidae) in four cropping systems. In: Ecology of earthworms in arable land. Population dynamics and activity in four cropping systems. Dissertation. Swed. Univ. Agric. Sci., Dep. Ecol. Environ. Res., Report 34. Boström, U. & Lofs-Holmin A., 1988b. Growth and cocoon production of the earthworm Aporrectodea caliginosa in soil mixed with various plant materials. In: Ecology of earthworms in arable land. Population dynamics and activity in four cropping systems. Dissertation. Swed. Univ. Agric. Sci., Dep. Ecol. Environ. Res., Report 34. Uppsala 1988. Clements, R.O., Murray, P.J. & Sturdy, R.G., 1991. The impact of 20 years' absence of earthworms and three levels of N fertilizer on a grassland soil environment. Agriculture, Ecosystems and Environment 36, 75-85. Edwards, C.A. & Bohlen, P.J., 1996. Biology and ecology of earthworms. 3 rd ed., Chapman & Hall. London. 426 pp. Edwards, W.M., Shipitalo, M.J., Owens, L.B. & Norton, L.D., 1990. Effect of Lumbricus terrestris L. burrows on hydrology of continuous no-till corn fields. Geoderma 46, 73-84. 106 Evans, A.C. & Guild, W.J.McL., 1984. Studies on the relationships between earthworms and soil fertility: IV. On the life cycles of some British Lumbricidae. Ann. Appl. Biol. 35, 471- 484. Hildebrand, A.A. & West, P.M., 1941. Strawberry root rot in relation to microbiological changes induced in root rot soil by the incorporation of certain cover crops. Can. J. Res. Sec. B Bot. Sci. 19, 183-198. Kongsrud, K.L., 1986. Nitrogen fertilization to the strawberry cultivars ’Senga Sengana’ and ’Glima’. Forsk. Fors. Landbr. 37, 281-288. Lofs-Holmin, A., 1983. Reproduction and growth of common arable land and pasture species of earthworms (Lumbricidae) in laboratory cultures. Swedish J. Agric. Res. 13, 31-37. Martikainen, E., 1996. Toxicity of dimethoate to some soil animal species in different soil types. Ecotox. Environ. Safety 33, 128-136. Nemec, S., 1970. Fungi associated with strawberry root rot in Illinois. Mycopat. Mycol. Appl. 41, 331-346. Paoletti, M.G., 1999. The role of earthworms for assessment of sustainability and as bioindicators. Agric. Ecos. Environ. 74, 137-155. Parikka, P., 1981. Strawberry root rot in Finland. Ann. Agric. Fenn. 20, 192-197. Parkin, T.B. & Berry, E.C., 1994. Nitrogen transformations associated with earthworm casts. Soil Biol. Biochem. 26, 1233-1238. Pérès, G., Cluzeau, D., Curmi, P. & Hallaire, V., 1998. Earthworm activity and soil structure changes due to organic enrichments in vineyard systems. Biol. Fertil. Soils 27, 417-424. Pfiffner, L. & Mäder, P., 1997. Effects of biodynamic, organic and conventional production systems on earthworm populations. Entomological Research in Organic Agriculture. Biol. Agric. Hortic. 1998, 15, 3-10. Raw, F., 1962. Studies of earthworm populations in orchards. I. Leaf burial in apple orchards. Annals of Applied Biology 50, 389-404. Roberts, B.L. & Dorough, H.W., 1984. Relative toxicities of chemicals to the earthworm Eisenia foetida. Environ. Toxicol. Biochem. 3, 67-78. Stephens, P.M., Davoren, C.W., Doube, B.M. & Ryder, M.H., 1994a. Ability of the lumbricid earthworms Aporrectodea rosea and Aporrectodea trapezoides to reduce the severity of take-all under greenhouse and field conditions. Soil Biol. Biochem. 26, 1291-1297. Stephens, P.M., Davoren, C.W., Ryder, M.H., Doube, B.M. & Correll, R.L., 1994b. Field evidence for reduced severity of Rhizoctonia bare-patch disease of wheat, due to the presence of the earthworms Aporrectodea rosea and Aporrectodea trapezoides. Soil Biol. Biochem. 26, 1495-1500. Stephens, P.M. & Davoren, C.W., 1997. Influence of the earthworms Aporrectodea trapezoides and A. rosea on the disease severity of Rhizoctonia solani on subterranean clover and ryegrass. Soil Biol. Biochem. 29, 511-516. Viljavuuspalvelu, 1997. Viljavuustutkimuksen tulkinta avomaan puutarhaviljelyssä. Mikkeli, Viljavuuspalvelu Oy, 20 p. 107 Whalen, J.K., Parmelee, R.W. & Edwards, C.A., 1998. Population dynamics of earthworm communities in corn agroecosystems receiving organic or inorganic fertilizer amendments. Biol. Fertil. Soils 27, 400-407. Werner, M.R., 1997. Soil quality characteristics during conversion to organic orchard management. Appl. Soil Ecol. 5, 151-167. Acknowledgements We thank Dr. Visa Nuutinen from the Crops and Soil Department, Agricultural Research Centre of Finland, for his inspiring personal communications and Dr. Jari Haimi from the Biological and Environmental Sciences, University of Jyväskylä for his comments on the research scheme. We also thank Mauri Räkköläinen for providing us with the earthworms. The work was part of the study ”Screening of the causes behind the declining strawberry fields” funded by the Ministry of Agriculture and Agricultural Research Centre of Finland. 108 Soil biological, chemical and physical properties in fields under different management systems Ansa Palojärvi1, Laura Alakukku1, Esko Martikainen2, Marina Niemi3, Pekka Vanhala4, Kirsten Jörgensen4 and Martti Esala1 1Agricultural Research Centre of Finland (MTT), FIN-31600 Jokioinen, FINLAND E-mail: [email protected] 2University of Jyväskylä, Institute for Environmental Research, FIN-40351 Jyväskylä, FINLAND 3Kemira Agro Oy, Espoo Research Centre, FIN-02271 Espoo, FINLAND 4Finnish Environment Institute, FIN-00251 Helsinki, FINLAND Summary Changes in agricultural management systems may have an impact on soil quality. The aim of the study was to investigate soil chemical characteristics, soil microbial community and functions - special attention being in nitrogen cycle-, soil fauna, and soil structure in fields of long term organic farming (> 10 years) compared with conventional ones. Ten pairs of adjacent organic and conventional fields on private farms in southern Finland were selected for the study. A 30 by 30 m sampling area was selected from each field and sampled four times: autumn 1997 and 1998, and spring 1998 and 1999. A large pattern of soil microbiological analyses was carried out, soil fauna was determined, and soil chemical and physical analyses were performed. Generally, we could detect only minor differences due to management, and the differences between the locations were greater than between the agricultural management systems. The clearest difference was seen in exchangeable P and S contents. Soil biological properties had strong spatial, and some of them also strong temporal variation. Few of the properties measured were statistically significantly higher in organic farming system. Reasons for minor differences between the management systems could be that soils react fairly slowly, and that the differences between the management systems in Finland are smaller than e.g. in Central Europe. Certain specific measures like crop rotation with grass ley seemed to be of importance, regardless of the management system. In this paper, an overview of the experiment and some general main results are given. Keywords: soil quality, management systems, organic farming, conventional farming, soil microbes, soil fauna, nitrogen cycle, soil structure Introduction It has been noticed in many studies, that agricultural management practices and management systems that has continued a long time, may have an impact on soil biological, chemical, and physical characteristics (e.g. Stenberg et al. 1998). This may lead to altered soil functions, and the changes may be either positive or negative for the agriculture. Maintenance and 109 improvement of soil quality is in central role when aiming at sustainable agriculture. Organic farming has been suggested as one possibility to improve soil quality (e.g. Bouma & Droogers 1998). Organic farming is a management system, which differs from the conventional one especially in that i) artificial mineral fertilizers and pesticides are not allowed, and ii) production is strictly based on crop rotation with nitrogen fixing legumes. Finnish studies on the effects of different management systems on soil properties are practically lacking. The effects of single management practices have been studies, but mostly only very few soil properties have been concidered at a time (e.g. Alakukku 1998). The studies are mostly carried out on experimental fields, which seldom are long term ones. The practical situation on farmer’s fields has been of little concern. The aim of the study was to investigate, whether long term organic farming (over ten years) and conventional farming on private farmer’s cereal or grass fields differs from each other on soil biological, chemical or physical properties. The aim of this paper is to give an overview of the experiment, and to show some general main results. The detailed results will be published later. Materials and Methods Ten pairs of organic and conventional fields on private farms in southern Finland were selected for the study. The cereal or grass fields in each pair located directly next to each other and had similar soil texture. Organic farming had been continued 12-41 years (Table 1). A 30 by 30 m sampling area was selected from each field and sampled four times: autumn 1997 and 1998, and spring 1998 and 1999. The analyses from the ploughing layer (0-25 cm) consisted of wide spectrum of soil microbiological, zoological, chemical and physical properties. Soil microbial biomass and community structure (phospholipid fatty acid profiles), soil respiration, substrate use efficiency (bacterial community level physiological profiles by Biolog plates), net nitrogen mineralization, potential nitrification, and several enzymatic activities were measured (see Palojärvi et al. 1997; Stenberg et al. 1998). The abundance and species (earthworms; Gunn 1992) or groups (mesofauna and nematodes) of soil fauna were determined. Soil chemical analyses were as follows: pH and E.C. (1:2 soil:water), Corg (LECO), Ntot (Kjeldahl method), inorganic N (2M KCl), exchangeable Ca, K, Mg, P and S (1M ammonium acetate pH 4.65). The inorganic nitrogen content was measured under the ploughing layer (25-60 cm), as well. Soil physical analyses were carried out from spring samples (e.g. bulk density, aggregate stability, Ksat and macro and total pore space; see Alakukku 1998; Kemper & Rosenau 1986). 110 Table 1. Some characteristics of the study fields. Code no. of the field pairs 12345671089 Texture Clay Silty clay Loam Fine sand Organic farming continued [a] 31 31 26 41 18 19 26 13 26 12 Conventional farming: -additionally org. fertilizers x x x x x -grass in crop rotation x x x x Results and Discussion According to the preliminary results, the site specific properties had the strongest effect on almost all variables measured (results not shown). Generally, we could detect only minor differences due to management, and the differences between the locations were greater than between the agricultural management systems. This is an important factor that should be taken into account when planning studies based on comparisons between fields or areas located in distinct places. The clearest differences could be seen in the concentrations of extractable nutrients (Table 2). Especially the concentrations of phosphorus and sulphur were low in organic farmed fields. In spring, the inorganic nitrogen content was significantly higher in conventionally farmed fields. This was not surprise, since the soil sampling took place only few weeks after the fertilization. No difference could be detected from autumn samples. The soil organic carbon content did not differ between the management systems (Table 2.). This probably explanes the fact that the differences in soil microbes and fauna were minor ones (Table 2). Variation within the study field (=spatial variation) and temporal variation between the sampling times was often high in biological properties. Microbial biomass nitrogen and basal respiration were one of the few properties that according to the pairwise t-test from the mean values of the study fields showed statistically significant difference between the management systems. High mean value of earthworm biomass in organig farming is explaned by one single study field very rich in earthworms. The main soil physical properties did not differ between the management systems, as shown in Table 2. There might be several reasons for only minor differences between the management systems. One reason could be that soils react fairly slowly. On the other hand, in this study organic farming had been continued over 20 years, as an average. Another factor could be that the differences between the management systems and the intensity of agricultural management in Finland are smaller than e.g. in Central Europe. Ryan (1999) concluded that organic matter inputs are of great importance in enhancing soil biological properties. As indicated in Table 1, many of the conventional farms used additionally organic fertilizers (manure) and had grass ley in their crop rotation. These single specific measures seemed to be of importance, regardless of the management system. 111 Table 2. Average chemical, biological and physical characteristics of the farming systems. The numbers are the total means of the results from all ten study fields in all four sampling times (n=40). Differences between the management systems were tested with pairwise t-test for each sampling time separately. Organic Conventional Properties [total mean] [statistical significance]a [total mean] pH 6.1 n.s.b 6.1 E.C. [10-4 S/cm] 0.67 in spring Conclusions No major differences between the management systems could be detected. Agricultural management system seems not to be the most relevant factor for soil quality. Spatial and temporal variation was high especially in soil biological quality indicators. This raises the question about the threshold values of the indicators. In the future, research should be 112 concentrated on those most important agricultural management practices that have the most pronounced effect on soil quality. The results could be applied on all management systems. There is a lack of integrated studies on soil biology, chemistry and physics. In order to be able to study and follow the development of soil quality, several long term field experiments should be established. References Alakukku, L., 1998. Properties of compacted fine-textured soils as affected by crop rotation and reduced tillage. Soil & Tillage Research 47, 83-89. Bouma, J. & Droogers, P., 1998. A procedure to derive land quality indicators for sustainable agricultural production. Geoderma 85, 103-110. Gunn, A., 1992. The use of mustard to estimate earthworm population. Pedobiologia 36, 65- 67. Kemper, W.D. & Rosenau, R.C., 1986. Aggregate Stability and Size Distribution. Methods of Soil Analysis. Part I. Physical and Mineralogical Methods, Agronomy Monograph 9. 2nd ed. American Soiciety of Agronomy-Soil Science Society of America, Madison, WI. 425- 442. Palojärvi, A., Sharma, S., Rangger, A., von Luetzow, M. & Insam, H., 1997. Comparison of biolog and phospholipid fatty acid patterns to detect changes in microbial community. In: Insam, H. & Rangger, A. (eds.). Microbial communities: functional versus structural approaches. Springer-Verlag, Heidelberg, 37-48. Ryan, M., 1999. Is an enhanced soil biological community, relative to conventional neighbours, a consistent feature of alternative (organic and biodynamic) agricultural systems? Biological Agriculture and Horticulture. 17, 131-144. Stenberg, B., Pell, M. & Torstensson, L., 1998. Integrated evaluation of variation in biological, chemical and physical soil properties. Ambio 27, 9-15. Acknowledgements The study was financed by the Ministry of Agriculture and Forestry, and Agricultural Research Centre of Finland. We would like to thank the skilled laboratory and field staff in all the research institutes participating the project. 113 Effects of management practice on soil organic matter content Tor-Gunnar Vågen Jordforsk, Centre for Soil and Environmental Research, NO-1432 Ås, NORWAY E-mail: [email protected] Summary The term sustainable agriculture is often misused and presented as being synonymous to "organic farming", implying that only low-input or extensive farming systems can be sustainable. At the same time the increased focus on soil organic matter (SOM) as an important indicator of soil quality, and thus of sustainability, is appropriate and important. Various factors control the contents and forms of SOM, including crops and management systems. In Europe recent findings indicate decreasing SOM contents in grain monocultures. The critical level for organic matter content in agricultural soils is generally regarded to be 2% to 4%. The assumption is often that grain monocultures lead to reductions in SOM contents, and that management systems with grass production have more stable organic matter contents. However, these assumptions are not very well documented. This paper presents results from a study of the variations in SOM contents between different agricultural management systems under Norwegian climatic conditions. The management systems studied are farms with grain monocultures and no animal production, and farms with grain monocultures and/or grass production and animal production. The main objective of the study was to document whether there are differences in SOM contents in farming systems where only chemical fertilisers are applied as compared to systems where animal manure is also applied. The study was conducted for two different mineral soil types, sandy soil and medium clay soils. The results presented here are for medium clay soils where samples with higher SOM contents than 15% where excluded from the analysis to exclude organic soils from the study. The results show that soil organic carbon (SOC) contents are significantly higher in areas where animal manure is applied. The data show a lower C/N ratio in samples from farms where animal manure is applied, indicating that manure application not only leads to higher SOM contents, but also affects the composition of the organic matter in the soil due to a lower C/N ratio in the manure as compared to straw which is the main source of SOM in farming systems where no animal manure is applied. Keywords: animal manure application, Norway, soil organic C, soil quality, tillage 115 Introduction Soil Organic Carbon (SOC) is the soil attribute most often monitored in long term fertility studies, an is often considered as the most important indicator of soil quality due to its impact on other physical, chemical, and biological indicators of soil quality (Reeves, 1997). The results of long-term studies, as those in Rothamsted and Woburn (Jenkinson, 1991; Powlson and Johnston, 1994), generally show that manure, appropriate fertilisation, and cropping pattern are important for maintaining soil quality through increased C inputs to the soil. Larson and Pierce (1991) proposed a minimum dataset (MDS) of soil quality indicators which included; (i) nutrient availability, (ii) total organic C, (iii) labile organic C, (iv) texture, (v) plant available water capacity, (vi) soil structure (bulk density, Ksat), (vii) soil strength (bulk density of penetration resistance), (viii) maximum rooting depth, (ix) pH, and (x) electrical conductivity. Doran and Parkin (1994) later proposed to add biological properties to this list. In Europe recent findings indicate decreasing SOC contents in grain monoculture farming systems. For good soil quality the critical level of organic carbon content in agricultural soils is generally regarded to be 1% to 2.5%. Increasing areas of agricultural land in Europe fall within this category. The assumption is often that grain monocultures lead to reductions in SOM contents, and that management systems with grass production have a more stable organic matter content. However, these assumptions are not very well documented. There are few long-term experiments on the effects of management systems on organic matter contents in agricultural soils in Norway. However, rotation experiments with six-year rotations conducted in Ås from 1953 showed that rotations significantly affected the concentration of soil C and N (Uhlen, 1994). Experiments under similar climatic conditions in Sweden (Gerzabek et al., 1994) have also shown that aggregate stability, which was a function of SOC content, increased in the order: Farmyard manure > peat > green manure > no-N with cropping > no-N no cropping. Agricultural production in Southeastern Norway has undergone profound changes during the last 30 to 40 years, with a general intensification of production and increase in areas under grain mono cropping. In some areas intensive pig breeding has increased, with consequent increases in pig slurry applications to agricultural lands, but the dominant cropping system is generally grain mono-cropping with no farmyard manure application. The main objective of the study was to utilise available data (collected by farmers) from the Soil Database at the Jordforsk Agricultural Service Laboratory to study the impacts of management practice and farmyard manure application on SOC contents in agricultural soils of South-eastern Norway. The study formed a basis for further and more detailed field studies. 116 Materials and Methods This study is based on a comprehensive, soil sample data set from the Jordforsk Agricultural Service Laboratory. This database consists of soil samples collected by farmers themselves for routine soil analysis through a 10-year period. In addition to the soil data, detailed information on management systems, fertiliser application rates, and the area distribution of different cropping systems in each farm were acquired from Statistics Norway and cover the period from 1982 to 1996. The study did not include field measurements, and was designed to form a basis for future studies in terms of focus and design of future on-farm field experiments. Two different farming systems were studied, farms with grain monocultures where only inorganic fertilisers had been applied and farms with animal husbandry where organic (farmyard manure) and/or inorganic fertilisers had been used. From these soil samples, 207 samples from medium clay soils were analysed for SOC. Another set of soil samples consisting of 6357 samples from areas with sandy soils will be presented at a later stage. Results and Discussion As shown in Table 1, the farms included in the study are predominantly grain producing, with some vegetable and grass production. Grass production is minimal even in farms where manure is applied. The farms are small to medium in size ranging from at total cultivated area of 14.5 to 99.9 ha as a mean for the 14 year period from 1982 to 1996. None of the farms applying animal manure are organic farms, since a combination of organic and chemical fertilisers are used. Some of these farms do not apply animal manure every year, since there may be years in between where they have no pig production. It is interesting to note the very large difference in mean grain yield between the two farming systems (Table 1). The main reason for this large difference is that some of the farms with animal production have extremely low grain yields in the range of 1750 to 1800 kg/ha, while the minimum for farms with no manure application is 3160 kg/ha. The largest farm in the study, with a cultivated area of 99.9 ha has the lowest mean grain yield, the largest area under grain production, and application of farmyard manure. The relationship between farm size and area under grain production is also very clear from the data (Figure 1). 117 Table 1. Baseline data for the studied farms. Management system Manure application No manure application Number of samples 47 160 Mean grain yield (kg/ha) 2720 4166 Field crops (%) 84 98 Grain crops (%) 80 96 Grass production (%) 1.4 0.1 . 80 R2 = 0.96 70 60 50 40 30 Area under grain production (ha) 20 10 10 20 30 40 50 60 70 80 90 100 110 Cultivated area (ha) Figure 1. Correlation between farm size (cultivated area) and area under grain production. The contents of SOC in the samples show significantly higher values from farms where animal manure is applied than where no animal manure is applied. The overall difference between the mean values is not large (Table 2), but there is skewness in the data and the results from areas with manure application have an asymmetric tail extending more towards positive values than the data from areas where no manure is applied. However, the variability is very high for both the farming systems (Figure 2). 118 Table 2. Summary statistics for SOC (%) contents in soil samples from areas with and without manure application. Management system Manure application No manure application Mean 2.78 2.42 Median 2.70 2.35 Range 4.30 4.60 Skewness 2.26 1.51 Kurtosis 6.83 5.61 Figure 2. Soil organic carbon contents in farms with manure application and farms where no manure is applied. The mean diamonds show mean values and the 95% confidence limits for each of the management systems. The SOC contents in samples from farms where animal manure is applied are higher than in farms where no animal manure is applied despite the higher yields in the latter farms. The data show no correlation between yield size and SOC contents, or any effects of percentage grass production versus grain production, and the difference between the management systems thus seems to be attributed to higher inputs of organic matter where animal manure is applied. However, the variability within the data is very large, and there is actually no clear relationship or correlation between animal manure application (kg/ha) and SOC content (R2 = 0.03). In addition to the quantity of organic matter in the soil, the quality of the organic matter is important. The C/N-ratio for the studied soil samples was therefore calculated and looked at in more detail. The C/N-ratio is important, e.g. for crop response to fertiliser N, and an important indicator for the quality of the soil organic matter. As Table 3 indicates, C/N ratios 119 are higher in management systems where no manure is applied. An analysis of variance shows that the difference between the management systems is statistically significant (Figure 3). The lower C/N-ratio in farms where animal manure is applied is probably due to the lower C/N- ratio in animal manure as compared to straw, which is the source of organic matter in farming systems with no manure applications. However, as for SOC contents, the variability in C/N- ratios is also very high, and it is therefore very difficult to actually explain the differences between the management systems from the data. For example, an analysis of the relationship between grain yield and C/N-ratio shows a slight upward trend, but this is not statistically significant, and does not explain the difference between the management systems. Table 3. Summary statistics for the C/N-ratio calculations in soil samples from areas with and without manure application. Management system Manure application No manure application Mean 11.7 13.0 Median 11.5 13.0 Range 8.0 15.1 Skewness 0.17 0.54 Kurtosis 0.07 0.63 Figure 3. C/N-ratios in farms with manure application and farms where no manure is applied. The mean diamonds show mean values and the 95% confidence limits for each of the management systems. 120 Conclusions The data show that there are significantly higher contents of SOC in soil samples from farms with manure application than from farms where no manure is applied. However, the variability is very high within both ”treatments”, as there are a number of other factors that determine SOC contents (e.g. yields, cropping systems, tillage practices, etc.), and there is a need for more detailed studies to explain this variability better. The quality of SOM is as important as the quantity, and the C/N-ratio is therefore an important indicator of soil quality. The C/N-ratio is higher in soil samples from farms where no animal manure is applied than from areas where manure is applied, indicating that the lower C/N-ratio in animal manure than in straw is important for the overall quality of the SOM. The C/N-ratio seems to be more influenced by factors like grain yield than the SOC contents, but these trends are not statistically significant. The data presented are based on a limited number of samples, and need to be supplemented by more samples. Additional field studies are also necessary to better understand the dynamics of SOC contents, including temporal trends, in different management systems under Norwegian conditions. More detailed field studies are therefore currently being implemented to document the effects of management systems on SOC contents and quality not only between farms, but also with regard to the temporal trends under different management systems. References Doran, J.W. Parkin, T.B., 1994. Defining and assessing soil quality. In: Doran, J.W., Coleman, D.C., Bezdicek, D.F., Stewart, B.A. (Eds.), Defining Soil Quality for a Sustainable Environment, SSSA Special Publication No. 35, Soil Sci. Soc. Amer., Amer. Soc. Agron., Madison, WI, pp. 3-21. Gerzabek, M.H., Kirchmann, H., Pichlmayer, F., 1994. Response of soil aggregate stability to manure amendments in the Ultuna long-term soil organic matter experiment. Z. Pflanzenernähr. Bodenk. 158, 257-260. Jenkinson, D.S., 1991. The Rothamsted long-term experiments: Are they still of use? Agron. J. 83, 2-10. Larson, W.E., Pierce, F.J., 1991. Conservation and enhancement of soil quality. In: Evaluation for Sustainable Land Management in the Developing World (Elliott, C.R. ed.) Vol. 2. IBSRAM Proc. 12, 2 Technical Papers, International Board for Soil Research and Management, Bangkok, Thailand, pp. 175-203. Powlson, D.S., Johnston, A.E., 1994. Long-term field experiments: their importance in understanding sustainable land use. In: Greenland, D.J., Szabolcs, I. (Eds.), Soil Resilience and Sustainable Land Use. Proc. Symposium, Budapest, Hungary 28 Sept. – 2 Oct. 1992, CAB International, Wallingford, UK, pp. 367-394. 121 Reeves, D.W., 1997. The role of soil organic matter in maintaining soil quality in continuous cropping systems. Soil and Tillage Research. 43, 131-167. Uhlen G., Kolnes, Anne-G., Torbjoernsen, B., 1994. Effects of long-term crop rotations, fertilizer, farm manure and straw on soil productivity: I. Experimental design and yields of grain, hay and row crops. Norwegian J. of Agric. Sci. 8, 243-258. 122 Organic wastes and soil quality Søren O. Petersen1, Kasia Debosz1 and Frank Laturnus2 1Danish Inst. Agricultural Sciences, Dept. Crop Physiology and Soil Science, Tjele, DENMARK E-mail: [email protected] 2 Dept. Plant Biology and Biogeochemistry, Risø National Laboratory, Roskilde, DENMARK Summary Organic waste products may have both beneficial and adverse effects on soil quality. The input of nutrients and organic matter generally stimulate mechanisms that promote soil structural properties, but pollutants and pathogens associated with the waste represent risks to consumers. Results are presented from field and laboratory experiments designed to quantify effects of organic waste products on soil characteristics, as well as the turnover of waste constituents. The response of physical, chemical and microbiological parameters to waste amendment were transient and in the same order of magnitude as seasonal fluctuations. The nutrient availability from different waste materials varied significantly, as reflected in the plant uptake from two sludge types. Distribution of the waste material will influence the potential for interaction between waste and soil, and short-term effects on the soil microbial biomass around sludge particles spiked with bis(diethylhexyl)phthalate (DEHP) were described in the laboratory, while long-term (5 months) microbial dynamics and DEHP disappearance inside sludge particles were studied under realistic field conditions during growth of an oats crop. The presence of plants may significantly alter the direction and dynamics of organic waste decomposition due to aeration and colonization of the sludge by penetrating roots. Introduction Recycling of nutrients in urban wastes, such as sewage sludge or household composts, can support strategies towards sustainability in agriculture. In Denmark, ca. 70% of the sewage sludge produced was recycled to arable soil in 1996, while practically no household compost was applied (Petersen, 1999). Organic waste materials are a source of energy and nutrients and, even though the chemical composition is often not optimal, organic waste amendment has generally been found to be beneficial for soil properties like bulk density, aggregation, water retention characteristics, nutrient status, N and C mineralization, and soil biological activity (e.g., Kladivko and Nelson, 1979; Levi-Minzi et al., 1985; Dar, 1997; Krogh et al., 1996; Elsgaard et al., submitted). 123 However, organic waste can also be a source of pathogens and pollutants. In the past, toxic effects have mainly been associated with heavy metals, less frequently with organic micro- pollutants such as plasticizers and detergents. The increasing awareness of these trace components has made farmers, as well as consumers, concerned about the consequences of waste amendment for product and soil quality. In response to this concern, the Danish government in 1997 initiated the research programme ’Centre for Sustainable Land Use and Management of Contaminants, Carbon and Nitrogen’ (http://www.landuse.dk/) involving 12 Danish research institutes and universities. The main aim is to improve our capacity for risk assessment, and for prediction of fertilizer and soil improvement value of organic waste products. This contribution presents data from field and laboratory studies to describe basic properties of soil-waste interactions. Waste application versus climate Laboratory incubations have frequently been employed to document an effect of waste application on soil parameters. However, soil characteristics typically have a seasonal variability as well, and waste-derived effects must be held up against these natural fluctuations (Domsch et al., 1983). In order to quantify separate and interactive effects of waste and climate on various soil physical, chemical and microbiological parameters, samples with soil alone, or soil amended with anaerobically stabilized sludge or household compost were incubated for up to 11 months under constant conditions (10° C, dark) or in the field. Soil and waste was fully mixed, and at sampling analyzed for aggregate stability, dispersible clay, inorganic N, polysaccharides, CO2 production rates, fluorescein diacetate (FDA) hydrolysis and β- glucosidase activity, biomass C, and phospholipid fatty acid (PLFA) composition (unamended soil only). Selected results are presented in Fig. 1 showing the fraction of soil in wet-stable aggregates >250 µm (top), inorganic N (centre), and FDA hydrolysis activity (bottom). All three parameters were initially stimulated by both waste types, although the anaerobic sludge was apparently more reactive than compost. With sludge, aggregation was immediately increased, probably reflecting a physical interaction between sludge constituents and soil particles, but all treatments eventually reached a level of ca. 90%. We assume that soil handling (storage, sieving, repacking) had reduced the stability of soil aggregates prior to the experiment, and that the soil gradually recovered from this disturbance to a level determined by inherent soil properties. Compost and sludge differed with respect to the initial N dynamics, with net N mineralization in the compost treatment and net immobilization with sludge. Gross N dynamics, quantified in a separate field study using 15N, demonstrated much higher N turnover rates with sludge than with compost (P. Ambus, pers.comm.). All field samples were depleted of N during fall and winter due to leaching, but the lab incubations revealed a net N mineralization from sludge (difference between sludge-amended and unamended soil at the 124 last sampling) which would correspond to ca. 50 kg N ha-1. In contrast, the N accumulation with compost was not significantly above the control. FDA hydrolysis activity, a broad index of hydrolytic enzymes, exhibited a brief response to waste amendment, then gradually declined. At the last sampling the following spring an increase was observed in all field samples relative to the lab incubations, possibly due to substrates released during spring thaw under field conditions. 100 Compost Sludge Control 90 m (%) µ 80 70 60 aggregates >250 Soil fraction in wet-stable in wet-stable Soil fraction 50 100 Compost Sludge Control 75 dry wt. soil dry -1 50 mg N kg mg N 25 0 80 -1 Compost Sludge Control soil h -1 60 40 20 mg fluorescein wt. kg dry 0 10 Jun 98 10 Jun 98 Jun 10 10 Jun 98 10 Jun 21 98 Apr 21 Apr 98 Apr 21 21 98 Apr 23 Mar 99 Mar 23 26 Aug 9826 Aug 11 98 Nov 23 Mar 99 23 Mar 23 Mar 99 Mar 23 26 Aug 98 Aug 26 98 Nov 11 26 Aug 9826 Aug 11 98 Nov 13 98 May 13 May 98 May 13 13 98 May Figure 1. Variations of wet-stable macro-aggregates (top), inorganic N (centre) and FDA hydrolysis activity (bottom) in soil incubated under constant conditions (white symbols) or in the field (grey symbols) for up to 11 months (vertical bars indicate S.D., n=2). 125 The temporal dynamics of the microbial community was examined, but in unamended soil only, using the composition of PLFA, that is, long-chain fatty acids derived from membrane lipids, as a fingerprint (Petersen and Klug, 1994). The concentration of PLFA varied only moderately, though the concentrations in lab incubations had dropped by 30-40% at the last sampling after 11 months (Fig. 2). There were only minor differences between the PLFA composition of lab and field incubated soil, even at the last sampling (data not shown), indicating that microbial turnover had been limited. It is well known that microbes will degrade cell constituents, including membrane lipids, to sustain basic metabolism during starvation (Thomas and Batt, 1969; Oliver and Stringer, 1984), and the observed difference in PLFA concentration at the last sampling may thus reflect a depletion of degradable substrates in soil incubated under constant conditions in the laboratory, as also suggested by the FDA hydrolysis measurements. 90 60 dry wt. soil) wt. dry -1 30 PLFA (nmol g 0 0 100 200 300 Time (days) Figure 2. The concentrations of phospholipid fatty acids (PLFA), an index of microbial biomass, in soil incubated under constant conditions (white bars) or in the field (grey bars) (vertical bars indicate S.D., n=2). Nutrient value of organic wastes Danish farmers receiving organic wastes must take the nutrient value into account when calculating the supplementary fertilizer they can apply (sludge: 30% of total N; household compost: 10% of total N). In the seasonal study described above, N mineralization differed between waste types, which could then make a significant difference to crop yields. We subsequently compared the fertilizer value of two sludge types under field conditions. The field study was carried out in the spring of 1999 to study N and P dynamics (P will not be discussed here) during growth of an oats crop. The treatments included i) anaerobically stabilized sewage sludge, ii) activated sewage sludge (from the aeration tank) and iii) + inorganic fertilizer, and application rates per hectare were i) ca. 125 kg Ntot (ca. 15 kg NH4 - + N) , ii) ca. 200 kg Ntot (ca. 30 kg NH4 -N), and iii) 80 kg N calcium ammonium nitrate; the 126 sludge application rates were chosen to give similar levels of C input. Soil samples were ana- lyzed for inorganic N and biomass N, and above-ground plant parts were analyzed for total N. The application of activated sludge, but not anaerobically stabilized sludge, significantly increased pools of biomass N in the soil (Fig. 3A). Between 3 May and 28 June the concentrations of biomass gradually declined in all three treatments, corresponding to a net release of 40-50 kg N. It supports the hypothesis that the microbial biomass can serve as an 140 ) -1 120 A 100 80 60 Figure 3. Concentrations of N in 40 20 the microbial biomass (A), in soluble Biomass N (kg N ha (kg N N Biomass 0 forms (B), and in above-ground parts ) -1 80 B of an oats crop fertilized with 60 inorganic N (white bars), 40 anaerobically stabilized sludge (grey bars) or activated sludge (black bars) 20 Inorganic N (kg N ha (vertical bars indicate S.D., n=2). 0 160 Note: In B and C one of duplicate ) C -1 140 field plots receiving inorganic 120 100 fertilizer were disregarded, since by 80 60 mistake it had received too much N. 40 Plant N (kgPlant N ha 20 0 9 9 9 9 9 r 9 9 9 9 9 p ay un u n u n A M J J J 1 5 3 7 17 28 important temporary storage of nutrients (Smith, 1994), and that biomass dynamics influence the nutrient supply of arable crops. Inorganic N clearly increased following the fertilization (Fig. 3B); the levels were not closely linked to the application of inorganic N, presumably due to turnover of soil and fertilizer N between the first sampling and fertilization, respectively. The immediate availability of nutrients in plots receiving inorganic fertilizer was reflected in an earlier plant uptake (Fig. 3C). But by 28 June the highest concentrations were found in plots with application of activated sludge. In this treatment the N uptake, in above-ground plant material alone, exceeded the net release of N from the microbial biomass and inorganic pools, showing that N was also supplied from other N pools, although it may have passed through the microbial biomass prior to plant uptake. Effects of waste distribution on turnover processes The absence of hazardous chemicals and pathogens are essential properties of arable soils, but such pollutants can be introduced with application of organic waste products. Models of risk assessment typically assume that the subsequent degradation of waste components will 127 proceed in accordance with the characteristics of the soil environment, for example that conditions are predominantly aerobic, and that pollutants are fully mixed with the soil matrix (as reflected in the application of EC [effect concentration] values and octanol:water partition coeffecients). However, in practice the waste will be distributed in discrete particles, and this may significantly influence the temporal dynamics of decomposition processes, as well as the exposure of soil organisms to the pollutants. Observations by Petersen et al. (1999) have indicated that the microflora of sewage sludge partly survived for several months after field application in November, as revealed by PLFA fingerprints, and argued that this was due to maintenance of the sludge environment inside sludge particles. Henriksen et al. (1999) found that a clumpwise distribution of sludge initially delayed the decomposition of bis(diethylhexyl)phthalate (DEHP) compared to a completely mixed system, but after several weeks the picture was reversed, presumably due to an enrichment of DEHP degraders near the sludge-soil interface. The interaction between sludge and soil was examined in a laboratory experiment with an intact soil column which was amended with well-defined cylindrical sludge particles spiked with ca. 50 mg DEHP kg-1 dry wt. sludge. After a week, sub-samples were removed from five distance intervals relative to the sludge and analyzed for PLFA composition. Figure 4 presents the sampling intervals, as well as the results of a principal component analysis based on the mol percentage distribution of 31 PLFA. Contamination of zone 3 (soil next to the sludge) during sampling can not be excluded, and so the effect of sludge on soil microbes at 0-2 mm distance from the sludge can not be evaluated. At 2-6 mm distance from the sludge particles, changes in PLFA composition were very limited compared to the undisturbed soil, suggesting that the effects of sludge on the soil microflora were highly local. The volume of soil affected by a sludge particle will likely depend on the chemical properties of any given component (mobility, degradability, toxicity), but in fact a similar distribution of C turnover was indicated around artificial manure clumps, in which microbial growth was concentrated within 2 mm distance from the soil-manure interface (Frostegård et al., 1997), while soluble C was largely metabolized within 4 mm distance from the manure (Petersen et al., 1996). The responses of soil quality parameters described in the previous section (cf. Fig. 1 and 2) could be modified in practice due to the discrete distribution of organic wastes. This has been addressed in an incubation study with two soil types, three waste materials, and full mixing vs. a clumpwise application (data analysis not completed yet). 128 5 4 3 2 1 10 mm '5 '3 '3 '5 '1'2'2'1 '4 PC(6.8%) 2 '4 PC 1 (85.5%) Figure 4. The PLFA composition of samples from the five intervals indicated was analyzed by a principal component analysis. The microbial community of the sludge (1 and 2, left-hand side of the score plot) was distinctly different from the soil community. Microbial dynamics in field-applied sludge The survival of sludge organisms mentioned above was further investigated in a realistic field situation with spring application of anaerobically stabilized sludge and growth of an oats crop. Sludge was applied as a 4 × 4 cm sludge string to enable sampling of well-defined intervals relative to the sludge. Sampling in and around the sludge, as well as in undisturbed soil, took place 1 d after application (12 May) and again after harvest (10 Oct), and the sludge itself was also sampled twice during spring (30 May and 26 June). The microbial community was described by PLFA fingerprints, while concentrations of DEHP and selected heavy metals were also monitored. At this time only samples from the centre of sludge strings, i.e. sludge without physical contact with the soil, have been analyzed. Concentrations of PLFA are shown in Fig. 5; there was initially a significant increase, but at harvest the concentration had dropped several-fold. Crop roots started to colonize the sludge after a few weeks, and at the last two samplings the root mass was completely interwoven inside the sludge. Roots could not be separated from the sludge material at sampling, and the potential contribution from root cell membranes to the PLFA profiles must therefore be considered. In Fig. 4 the sum of fatty acids present in oats (Welch, 1975) is also shown; all of these fatty acids are also found among soil microorganisms. The fatty acids potentially derived from roots could only account for a minor part of the initial increase, and this fraction 129 actually dropped during June when root colonization was most extensive. Hence, there was no evidence for a large contribution from root cells to PLFA extracted from the sludge. Fig. 5 also shows the ratio between the plasticizer DEHP and PLFA; the disappearance of DEHP was clearly not as fast as the decline of the microbial biomass. 8000 PLFA, total 0,08 Figure 5. Concentrations of PLFA, oats PLFA DEHP/nmol mg total PLFA and PLFA found 6000 DEHP/PLFA in oats were quantified in 0,06 dry wt. -1 field-applied sludge during a 4000 5 months period. The 0,04 microbial biomass declined more rapidly than DEHP nmol PLFA g PLFA nmol 2000 0,02 (vertical bars indicate S.D., n=3). 0 0,00 99 99 99 99 5- 5- 6- 0- -0 -0 -0 -1 12 30 26 06 Figure 6. A principal component analysis of Soil * May12May12 May12 PLFA profiles of sludge May12 applied to an oats crop. Oct10 Oct10Oct10 Oct10 Samples were taken Oct10Oct10 between 1 and 168 days May30 Jun26Jun26 May30Jun26 after application. * The PLFA profile of the soil at this site is shown for reference; this profile Scores,2 (14.6%) PC May30 was obtained in a different experiment. Scores, PC 1 (54.1%) A principal component analysis of the PLFA composition of all samples, along with that of undisturbed soil from this site, revealed an interesting picture (Fig. 6). Between application in May and the last sampling in October there was a gradual change in the composition of PLFA towards the PLFA composition of the soil microbial community, even though there was no physical contact between the sludge material sampled and the soil. At this point it is not possible to decide whether selective survival of sludge microorganisms or colonization by soil organisms was the main mechanism behind the observed changes, but the roots may well have served as a vector for dispersal of soil microbes into the sludge. 130 Conclusions The results summarized above have described different aspects of the turnover of sludge in soil. Stimulation of both structural, chemical and (micro)biological parameters was observed, but the effects of such a one-time application were transient and in the same order of magnitude as seasonal fluctuations. Fully mixed experimental systems will tend to synchronize turnover processes, and it is likely that the heterogenous distribution under field conditions will delay and/or mitigate effects of organic waste application on soil properties. The quality of organic wastes in terms of plant available nutrients can vary dramatically, as seen by the plant uptake under field conditions. The soil microbial biomass and, possibly, the microbial biomass of sludge (but cf. Fig. 5) was a source of N. Organic waste distribution in the soil probably has a large impact on the degradation of sludge constitutents, and on the survival of sludge microorganisms. This may be of great importance for the degradation pattern of organic micro-pollutants to the extent that these processes are carried out by organisms from the sludge. Our results suggest that the soil volume exposed to pollutants from the waste may be restricted. Finally, the results demonstrated that the turnover of organic wastes can be influenced by the presence of a growing crop. Crop roots may promote aeration of the sludge and stimulate colonization by soil microorganisms. This could significantly change the turnover dynamics of organic micro-pollutant, many of which are not degradable under anaerobic conditions, compared to a laboratory set-up with constant water potential. Roots penetrating sludge particles may also become exposed to high concentrations of pollutants, and this aspect should be considered in the design of risk assessment studies. References Elsgaard, L., Petersen, S.O. and Debosz, K. Concentration effects of LAS in agricultural soil. 2. Effects on soil microbiology as influenced by sewage sludge and exposure time. Submitted to Environ. Toxicol. Chem. Frostegård, Å., Petersen, S.O., Bååth, E. and Nielsen, T.H., 1997. Dynamics of a microbial community associated with manure hot-spots as revealed by phospholipid fatty acid analysis. Appl. Environ. Microbiol. 63: 2224-2231. Henriksen, K., Roslev, P. and Møldrup, P., 1999. Biodegradation of DEHP in sludge- amended agricultural soil. In: J. Petersen and S.O. Petersen (Eds.) Use of Municipal Organic Waste, pp. 75-82. Proceedings of NJF Seminar No. 292, 23-25 November 1998 in in Jokioinen, Finland. DIAS Report No. 13, Danish Inst. Agric. Sci. Kladivko, E.J. and Nelson, D.W., 1979. Changes in soil properties from application of anaerobic sludge. J. Wat. Poll. Cont. Fed. 51, 325-332. 131 Krogh, P.K., Holmstrup, M., Jensen, J. and Petersen, S.O., 1996. Ecotoxicological assessment of sewage sludge amendment to arable land. Working Report from the Danish Environmental Protection Agency No. 43, 52 pp. (in Danish) Levi-Minzi, R., Riffaldi, R., Guidi, G. and Poggio, G., 1985. Chemical characterization of soil organic matter in a field study with sewage sludges and composts, pp. 151-160. In: J.H. Williams et al. (Eds.) Long.term Effects of Sewage Sludge and Farm Slurries Applications. Elsevier Applied Science Publishers, New York, NY. Oliver, J.D. and Stringer, W.F., 1984. Lipid composition of a psychrophilic marine Vibrio sp. during starvation-induced morphogenesis. Appl. Environ. Microbiol. 47, 461-466. Petersen, J., 1999. Legislation in Denmark and nutrient value of waste products. In: J. Petersen and S.O. Petersen (Eds.) Use of Municipal Organic Waste, pp. 13-18. Proceedings of NJF Seminar No. 292, 23-25 November 1998 in Jokioinen, Finland. DIAS Report No. 13, Danish Inst. Agric. Sci. Dar, G. H., 1997. Impact of lead and sewage sludge on soil microbial biomass and carbon and nitrogen mineralization. Bull. Environ. Contam. Toxicol. 58, 234-240. Domsch, K.H., Jagnow, G. and Anderson, T.H., 1983. An ecological concept for the assessment of side-effects of agrochemicals on soil microorganisms. Res. Rev. 86, 65-105. Petersen, S.O. and Klug, M.J., 1994. Effects of sieving, storage, and incubation temperature on the phospholipid fatty acid profile of a soil microbial community. Appl. Environ. Microbiol. 60: 2421-2430. Petersen S.O., Nielsen T.H.. Frostegård Å. and Olesen T., 1996. Oxygen uptake, carbon metabolism, and denitrification associated with manure hot-spots. Soil Biology and Biochemistry 28: 341-349. Petersen, S.O., Debosz, K., Elsgaard, L. & Krogh, P.H., 1998). Effects of organic wastes on microbiological aspects of soil quality, p59-64. In: J. Petersen and S.O. Petersen (Eds.) Use of Municipal Organic Waste, pp. 13-18. Proceedings of NJF Seminar No. 292, 23-25 November 1998 in Jokioinen, Finland. DIAS Report No. 13, Danish Inst. Agric. Sci. Smith, J.L., 1994. Cycling of nitrogen through microbial activity. In: J.L. Hatfield and B.A. Stewart (Eds.) Soil Biology: Effects on soil quality, pp. 91-120. Adv. Soil Sci., Lewis Publ. Boca Raton, FL. Thomas, T.D. and Batt, R.D., 1969. Degradation of cell constituents by starved Streptococcus lactis in relation to survival. J. Gen. Microbiol. 58, 347-362. Welch, R.W., 1975. Fatty acid composition of grain from winter and spring sown oats, barley and wheat. J. Sci. Fd. Agric. 26, 429-435. Acknowledgements This work was funded by ’Centre for Sustainable Land Use and Management of Contaminants, Carbon and Nitrogen’. 132 Levels of structural and functional complexity in soil OM turnover as revealed by physical fractionations Bent T. Christensen Dept. Crop Physiology and Soil Science, Danish Institute of Agricultural Sciences, Research Centre Foulum, PO Box 50, DK-8830 Tjele, DENMARK E-mail: [email protected] Summary Physical fractionation of soil according to particle size and density allows the separation of free and occluded uncomplexed OM (UOM) and of primary and secondary organomineral complexes (OMC). This methodological approach to soil OM turnover emphasizes the importance of interactions between organic and inorganic soil components, and recognizes that the overall regulation of decomposer activity is through soil structure which determines gas exchange, water availability and transport of solutes. Results from physical fractionations of soil suggest three levels of structural and functional complexity in soil. Primary OMC account for the primary level of complexity. The clay, silt and sand sized OMC are seen as the basic units in soil, surface reactions between substrates, organisms and minerals being main regulatory mechanisms. Secondary OMC reflect the degree of aggregation of primary OMC and refer to the secondary level of complexity. Physical protection of UOM and organisms, and the creation of gas and moisture gradients are additional features regulating OM turnover at this level of complexity. The structurally intact soil (the soil in situ) constitutes the tertiary level of complexity. This level integrates the effects of primary and secondary OMC. Additional features associated with this level are the effects of resource islands/hot spots, macropores, root distribution and mesofaunal activity, soil tillage and compaction on solute transport, gas exchange, distribution and comminution of litters and UOM. OM turnover in soil The biologically mediated transformation of plant and animal residues containing organic matter (OM) and plant nutrients, and the simultaneous formation of soil OM (SOM) are key functions of the decomposition subsystem. The net release of nutrients in plant available forms influences the composition and the productivity of the plant subsystem, causing mineralization processes to feed back on the amount and quality of the litter entering the decomposition subsystem. Through its effect on physical and chemical soil properties, the formation of SOM feeds back on the biological activity by affecting the decomposer environment and on the plant subsystem by affecting germination of seeds, plant establishment, retention of water and nutrients, and root distribution and activity. Terrestrial ecosystems accumulate OM in the decomposition subsystem, which provides a reservoir of 133 energy and nutrients that buffer the system against perturbations. The decomposing litter fractions represent a dynamic component while SOM formed during litter decomposition accounts for a slow component. The regulation of OM turnover is a main feature of the decomposition subsystem and spatial and temporal variations in substrate availability are important regulatory mechanisms. Substrate patchiness and the variability in associations between substrates and mineral soil components determine the abundance and activity of the microorganisms and soil fauna. Physical fractionation The use of physical fractionations in studies of OM turnover in soil has increased steadily over the past two decades. This development arises from an increasing awareness of OM turnover being due to biological processes under the overall regulation of soil structure. Moreover it is recognized that the availability of substrates to decomposers depends not only on the intrinsic chemical nature of the substrates but also, and perhaps more importantly, on the nature of the association of OM with soil mineral components. This association is manifested in the formation of organomineral complexes. Thus mechanisms responsible for the retention of OM in soil include inherent chemical recalcitrance of the organic component, chemical stabilization of the OM by chemical reactions with mineral surfaces, and physical protection of substrates through the creation of physical barriers between substrates and decomposer organisms. Physical fractionation of soil according to particle size and density is based on the application of various degrees of soil dispersion to break physical bonds between the elements of soil structure, and it allows the separation of uncomplexed OM (UOM) and of differently sized organomineral complexes (OMC) (Fig. 1). The fractionation procedures attempt to minimize in-process chemical changes in the OM but realizes a partial kill-off of soil microorganisms. Uncomplexed OM (UOM) Uncomplexed OM (UOM) is defined as the fraction of OM which is neither present as readily recognizable litter components (typically >2 mm) nor incorporated into primary OMC. UOM is recovered by density and size fractionation procedures and combinations thereof. Density separation based on heavy liquids (~1.2-2.0 g ml-1) has been widely used. The yield of UOM is very sensitive to fluctuations in liquid density, indicating that choice of density, exact density adjustment, and temperature control during processing all are critical factors for the outcome of density fractionations (Christensen, 1992). UOM consists mainly of particulate, partly decomposed plant and animal residues but may also encompass fungal hyghae, spores, faecal pellets, faunal skeletons, root fragments, and seeds (Christensen, 1992; Gregorich and Janzen, 1996; Golchin et al., 1997) 134 Figure 1. Separates isolated by physical fractionation procedure Uncomplexed organic matter • free – UOM • occluded – UOM (in aggregates) Primary organomineral complexes (primary OMC)’ • < 2 µm clay-sized • 2-20 µm silt-sized • 20-2000 µm sand-sized Secondary organomineral complexes (secondary OMC) • < 250 µm microaggregates • > 250 µm macroaggregates For soils regularly exposed to burning of vegetation, charcoal may account for a significant proportion of the UOM (Skjemstad et al., 1990; Cadish et al., 1996). UOM is considered a transitory pool between litter and mineral-associated OM, its mean rate of C turnover being slower than that of recently shed litter but higher than that of clay and silt-bound OM. The proportion of SOM recovered as UOM varies widely. UOM is affected by land use, vegetation type, and other factors that influence the balance between litter input and decomposition (e.g. climate, soil type, faunal activity). Accumulation of UOM is favoured in cold and dry climates, in soils low in pH, and in continuously vegetated soils with a high return of plant litter such as forests and grasslands (Christensen, 1992). While UOM in soils with permanent vegetation may account for 15-40% of the OM in surface horizons, UOM in long continued arable soils usually make up less than 10% of the OM in the tilled layer. Due to the dominant influence of plant litter inputs, UOM shows significant seasonal and spatial variability, especially in forest soils. Seasonal variations are less prominent in arable soils, but UOM has been found to reflect differences in cropping sequences and soil tillage systems (Christensen, 2000). UOM tends to be readily depleted when native soils with permanent vegetation are brought under cultivation (Christensen, 1992; Besnard et al., 1996; Cadish et al., 1996; Gregorich et al., 1997; Balesdent et al., 1998; Six et al., 1998), and UOM shows an increase when arable soil is reverted to native grassland (Jastrow, 1996). Often the drop in UOM accounts for a major part of the initial, cultivation-induced loss of SOM. UOM is neither a readily available nor major source of mineral N in soil. A high CO2 evolution potential indicates that UOM may be a source of readily available C and energy to the decomposers, and that its turnover 135 may well be coupled with N immobilization. Thus in some situations UOM may act as a net- sink rather than a net-source of plant available mineral N. UOM in soil occurs as “free” and “occluded” OM. Free-UOM includes free OM particles in the soil and particulate OM adhering to the exterior of secondary OMC. Free-UOM is also termed inter-aggregate OM. Occluded-UOM is the fraction of UOM which is trapped and physically protected within secondary OMC (intra-aggregate OM). The distinction between free and occluded fractions of UOM is based experimentally on a stepwise dispersion of soil combined with size and/or density separation. Free-UOM is recovered from minimally dispersed samples in which aggregates remain intact, while occluded-UOM subsequently is isolated after dispersion of aggregates. Golchin et al. (1997) concluded that free-UOM consists mainly of partly decomposed litter residues whereas occluded-UOM has undergone more extensive decomposition during its physical protection within aggregates. Occluded-UOM is less decomposable than free-UOM as reflected by differences in 13C signatures of UOM fractions from soils where C3 vegetation has been replaced by a vegetation with C4-type photosynthetic pathway (Besnard et al., 1996; Gregorich et al., 1997). However, Six et al. (1998) suggested that free-UOM represented a mixture of recently deposited crop residues and older UOM previously occluded within aggregates but released from degraded aggregates following the depletion of available substrates in the occluded-UOM. Free and occluded UOM originate from different soil locations with different exposure to microbial decomposition, differ in chemical composition, in 13C signature, and exhibit different lability towards microbial turnover. These experimentally verifiable UOM fractions are therefore considered to play different roles in organic matter turnover in the decomposition subsystem and free and occluded UOM have been suggested as candidates for structurally and functionally meaningful compartments in dynamic simulation models of C turnover (Christensen, 1996b; Elliott et al., 1996). Recent developments in concepts linking free and occluded UOM pools to aggregate dynamics and accumulation of soil OM have been presented by Elliott et al. (1996), Golchin et al. (1997), Six et al. (1998), and Christensen (2000). Primary organomineral complexes The separation of soil into differently sized primary OMC implies that OM associated with mineral particles of different size, and therefore also of different mineralogical composition (e.g. see Schjønning et al., 1999), differs in structure and function. Further, it is anticipated that the differences in OM characteristics between primary OMC size classes are more significant than the variation found within individual OMC size classes. The differential influence of individual mineral species and particle size on microbial activity and on the retention of metabolic products in soil is well documented (e.g. Huang and Schnitzer, 1986). 136 A complete dispersion of soil is essential for the isolation of primary OMC. Ultrasonic vibrations in water appears to be the most feasible method for dispersing soils prior to isolation of primary OMC from soils low in UOM. Significant contents of UOM should be removed from soil samples prior to ultrasonic dispersion. Fractionation procedures should be optimized for the soil types under study and the resulting particle size distribution calibrated against conventional textural analyses. For temperate arable soils, clay sized OMC (<2 µm) shows the highest concentrations of OM. Silt sized (2-20 µm) OMC is less enriched and size separates >20 µm are usually very low in OM. For permanently vegetated soils, UOM may raise OM contents in sand size separates significantly. However, the amount of OM engaged in true organomineral associations as organic coatings of sand grains remains low also in such soils. To compare concentrations of OM in various OMC classes isolated from different soils, -1 -1 element enrichment factors (e.g. EC = mg C g OMC/mg C g whole soil) can be calculated for each OMC size class. Effects of differences in SOM levels of whole soils are thereby eliminated. Clay sized OMC isolated from Danish arable soils shows EC between 2 and 15, while EC values for silt OMC typically range from 1 to 5. Excluding soils rich in UOM, sand sized separates show an EC <0.1. EC and EN for clay and silt sized OMC were found to be inversely related to the proportion of these separates in whole soils (Christensen, 1996a), whereas E values for separates >20 µm were unrelated to whole soil sand contents. For similar yields of OMC dry matter, higher E values are found for clay than for silt sized OMC, indicating that clay may be quantitatively more effective in SOM sequestration than silt. Clay sized OMC tends to be more enriched in N than in C while the opposite is true for silt sized OMC. The different patterns of EC and EN for clay and silt OMC are reflected in their C/N ratios, OM in clay having a lower C/N ratio than OM associated with silt. The higher OM enrichment of clay and silt sized OMC in soils low in these particles infers that the OM binding capacity of clay and silt is more saturated in these soils than in soils richer in fine particles. This observation raises the question whether clay and silt sized minerals have a finite capacity to store OM. Relating the amount of OM in clay and silt sized OMC classes to their specific surface area shows that the area specific C load increases with increasing particle size (Christensen, 1996b). Values for clay and silt typically range from 0.5-1.5 and 1-5 mg C m-2, respectively. For fine textured marine sediments, the area specific C load usually falls within the interval of 0.5-1.0 mg C m-2, corresponding to what has been termed a “monolayer-equivalent” C loading of the particles (Mayer, 1994; Hedges and Oades, 1997). However, these data provide no information on the actual distribution of the OM on mineral surfaces. Most likely a continuum exists in the composition of clay and silt OMC, the ratio between organic and inorganic components being a continuous variable rather than a fixed parameter. Individual soil particles are not envisaged to have a finite capacity for accommodating OM. However, it is hypothesized that the strength with which recently 137 formed organic compounds become attached to primary OMC declines as the distance between the mineral surface and the new SOM increases. Thus OM already associated with soil minerals is considered to impede chemical stabilization of new OM. In temperate arable soils, 50 to 75% of the SOM is present in clay sized OMC while silt accounts for another 20 to 40% and sand sized separates for <10%. The proportion isolated along with sand may be more significant (up to 40%) in soils enriched in UOM. It is important to note, however, that the actual level of OM in whole soils is not directly related to contents of fine particles (clay + silt), and that the proportion of SOM associated with clay and silt cannot generally be related to the SOM level in the soil (Christensen, 1992, 1996a; Hassink, 1997). These observations underline the importance of secondary and tertiary soil structural features in determining SOM levels in the decomposition subsystem, including the effects of free- and occluded-UOM pools. The composition of OM in a given OMC is influenced by plant as well as microbially derived products and their successive transformations. However, the extent of decomposition of plant derived compounds increases from sand to silt to clay sized OMC (Guggenberger et al., 1995). Clay OMC is dominated by products of microbial synthesis while silt appears to be rich in plant derived aromatic compounds. A number of features of 13C NMR spectra acquired for sand sized separates are analogous to spectra of plant litter. During decomposition microbial products such as diaminopimelic acids, amino sugars, teichoic acids and microbially derived sugars are formed, providing a greater microbial input to clay than to coarser OMC (Christensen, 1996a, 2000). Solid-state 13C NMR spectroscopy has revealed significant differences in the chemical composition of OM in clay, silt, and sand sized separates (Christensen, 1992, 1995, 1996a; Guggenberger et al., 1995; Randall et al., 1995). Aromatic structures show the smallest relative concentrations in clay OMC, the OM in clay being dominated by alkyl-C. The high alkyl-C content of clay cannot only be ascribed to a gradual accumulation of plant derived wax-like biopolymers including cutin, suberin and tannin (Preston, 1996) but most probably also involves in situ synthesis of alkyl substances by microorganisms associated with clay OMC surfaces (Baldock et al., 1992). Very small amounts of aromatic structures have been found to be synthesized during decomposition of 13C-labelled glucose (Baldock et al., 1990; Golchin et al., 1996) and it is considered that the abundance of aromatic structures in silt sized OMC is due to a selective preservation of plant derived lignin moities and not to de novo synthesis by microorganisms (Baldock et al., 1992). The potential bioavailability of OM in primary OMC increases as OMC size decreases regardless of aeration status, mineralization of SOM being dominated by contributions from clay sized OMC. Mineralization rates for UOM in sand are high for C but low for N, the high CO2 evolution indicating a high microbial activity which probably is coupled with 138 immobilization of N. Sand sized UOM usually has a high C/N ratio and shows little or no net N mineralization. Thus seasonal and annual fluctuations in UOM may well have a controlling influence on soil net N mineralization. The greater biological activity associated with clay sized OMC is reflected in the higher content of microbially derived products in clay than in >2 µm OMC. Clay OMC also maintains the highest concentrations of OM suggesting that a high microbial activity provides more microbial metabolites and residues (including cell wall materials) that become stabilized and accumulate in clay OMC, e.g. by “irreversible” or protective sorption (Hedges and Oades, 1997; Kaiser et al., 1998). The high microbial activity thereby links to the accumulation of SOM in clay OMC. Land use and cultivation may affect the quality of OM associated with primary OMC, but the main effects are on the amount of SOM and its distribution across OMC size classes. The differences between differently sized OMC in the chemical nature of OM dominate over differences induced in individual OMC size separates by contrasting land uses (Christensen, 2000). The chemical composition of UOM isolated along with sand size separates is more readily affected by cultivation than OM in clay and silt OMC. Clay and silt associated OM is dominated by the interaction between the soil microbial biomass, its decomposition products and the mineral soil matrix and the end products of the decomposition process appear to be similar for contrasting inputs (Oades et al., 1988; Randall et al., 1995). Changes in management procedures applied to arable soils (e.g. changes in crop rotation, fertilization regime, crop residue disposal, soil tillage system) generally introduce less drastic changes than those induced by the initial cultivation of native soils and other soils with permanent vegetation. Results from various studies suggest that the chemical composition of OM in clay and silt sized primary OMC is little influenced by changes in soil management, while some changes can be reflected in sand sized UOM. Similarly, the bioavailability of OM in clay and silt appeared remarkably uniform regardless of straw management, UOM in sand being an exception. It is concluded that primary OMC delineates structural and functional entities of SOM, and that primary OMC and UOM represent basic and biologically meaningful OM pools in the studies of OM turnover in the decomposition subsystem. Figure 2 summarizes some characteristics of primary OMC isolated from temperate sandy soils under long-term arable use: 139 Figure 2. Characteristics of primary OMC from temperate sandy soils under long-term arable use Sand associated OM (= UOM) • enriched in plant polymers • high C/N ratio • • high C and low N mineralization low proportion of total soil OM (< 10%) • CEC: 10-150 mmol/kg • most OM not firmly associated with minerals 2 • surface area (N2-BET): <10 m /g Silt associated OM • enriched in plant derived aromatics • holds 20-40% of total soil OM • • low OM decomposition rate CEC: 60-350 mmol/kg • surface area (N -BET): 10-50 m2/g • medium C/N ratio 2 • 1-5 mg C/m2 surface area • medium OM concentration (1 – 5 x) Clay associated OM • high OM concentration (2 – 15 x) • low C/N ratio • • enriched in microbial products contains 50-75% of total soil OM • CEC: 300-900 mmol/kg • depleted in plant residue components 2 • surface area (N2-BET): 25-100 m /g • high C and N mineralization • 0.5-1.5 mg C/m2 surface area Secondary organomineral complexes Individual primary OMC and UOM particles may occur as discrete structural units in the soil, but in most soils the majority of the primary OMC is incorporated into differently sized secondary OMC. The degree of aggregation of primary OMC is reflected in the size distribution of the secondary OMC and in their stability towards disruption. Secondary OMC may encompass mineral-associated and uncomplexed OM, microorganisms (bacterial colonies, fungal hyphae, microfauna) and fine roots, whereby secondary OMC represent a higher level of structural and functional complexity than individual size classes of primary OMC. The potential of a soil to form aggregates depends on the size distribution of the primary OMC and their characteristics which in turn relate to basic soil properties, in particular the amount and types of clay minerals, polyvalent cations and sesquioxides. Secondary OMC are also referred to as aggregates in accordance with common terminology. The formation and dynamics of soil aggregates have been reviewed intensively in recent years (e.g. Oades, 1993; Haynes and Beare, 1996; Golchin et al., 1997). The formation, stabilization and degradation of secondary OMC involves abiotic processes such as wet/dry and freeze/thaw cycles, physical impacts of traffic and tillage, and biotic factors including extracellular microbial biopolymers, faunal activity, plant roots and fungal hyphae. 140 Current models of the formation, stabilization and breakdown of secondary OMC are essentially modifications of the concept of aggregate hierarchy proposed by Tisdall and Oades (1982) and further developed by Oades and Water (1991), Oades (1993) and Golchin et al. (1997). This concept is thought to be applicable to soils containing appreciable contents of clay sized OMC and where OM is the dominant stabilizing agent. The concept recognizes three basic levels of structural organization of secondary OMC (small microaggregates <20 µm, large microaggregates 20-250 µm and macroaggregates >250 µm) and classifies the organic binding agents into transient (mainly polysaccharides), temporary (roots and hyphae) and persistent (strongly sorbed organic polymers and polyvalent cations). The concept links mechanisms of aggregate formation and stabilization with the spatial and temporal distribution of OM and thus links processes with scale expressed in the size of the secondary OMC. Small microaggregates (<20 µm) contain little if any occluded-UOM and the main mechanisms in the stabilization of these aggregates are microbial products, root exudates, polyvalent cations and other persistent binding agents. The small microaggregates are considered to be the most stable units in the secondary structure of soil. Larger microaggregates (20-250 µm) are composed of small microaggregates, primary OMC and occluded-UOM and their stabilization involves both persistent and transient binding agents. It is also suggested that root-hairs and fungal hyphae can be active in the stabilization and formation of larger microaggregates. The level of microaggregates in a soil, in particular those smaller than 20 µm, is relatively stable over long periods of time and hence not significantly influenced by changes in soil management. Macroaggregates (>250 µm) are made up of microaggregates, primary OMC and UOM particles held together by transient and temporary binding agents. It is indicated that temporary binding mechanisms are more important for large macroaggregates than transient ones. The level of macroaggregates exhibits a dynamic nature and their size distribution is susceptible to seasonal fluctuations in climate and to changes in land use and soil management. So permanently vegetated soils generally contain more and larger macroaggregates than frequently tilled arable soils with short season crops. It follows that soil sampling and fractionation procedures are decisive for the resulting size distribution of macroaggregates and for the interpretation of results acquired in subsequent analyses of aggregate size classes (Beare and Bruce, 1993). All classes of organic binding agents are subject to a simultaneous loss and gain, but their rate of turnover is considered to be widely different. Persistent agents may remain for decades while temporary agents may last for days only. The concept of aggregate hierarchy does not per se imply a sequential formation of micro and macroaggregates. New microaggregates in the size range 20-250 µm may be generated within the more stable fraction of the macroaggregates as a result of the microbial activity responsible for the gradual 141 transformation of UOM occluded within macroaggregates (Jastrow et al., 1996; Golchin et al., 1997; Six et al., 1998). One consequence of the concept of aggregate hierarchy is the porosity exclusion principle (Oades, 1993). According to this principle smaller aggregates have smaller pores, greater contact between sub-units, and a higher density than larger aggregates which also include the larger pores that exist between the constituent smaller aggregates. Thereby the size of aggregates relates to conditions for microbial and microfaunal growth and activity by affecting potentials for air exchange, flow/diffusion of nutrients and substrates and the retention of microbially available water. The role of UOM particles in the formation of macroaggregates and as nucleation point for ~50-250 µm microaggregates has been increasingly emphasized in recent developments of the aggregate hierarchy concept (Oades and Waters, 1991; Golchin et al., 1997; Jastrow and Miller, 1997; Six et al., 1998). A continuous input of plant litter (the main source of UOM) to the decomposition subsystem may be crucial for the formation of macroaggregates and large microaggregates, while the general level of SOM may be of less importance. Any determination of the size distribution of aggregates incorporates a determination of aggregate stability, because some level of disintegrating force is required to break the tertiary soil structure into secondary OMC. The aggregate size distribution will therefore reflect both the forces that stabilize the aggregate against breakdown in situ and the abrasiveness of the fractionation procedure applied (Beare and Bruce, 1993). Traditionally aggregates are isolated by dry or wet sieving procedures (Kemper and Rosenau, 1986). Dry sieving reflects the stability of aggregates against mechanical breakdown when dry soil is exposed to relatively short periods of rotary sieving in air. In contrast, wet sieving procedures also include effects due to the wetting process. The mechanical forces imposed on aggregates during wet sieving are considered less abrasive than those exerted by dry sieving. Abrasive forces operating during wet sieving cause slaking and dispersion. Slaking is the break-up of larger aggregates into smaller ones while dispersion is the release of primary OMC from aggregates (Emerson and Greenland, 1990). For studies targeting the relationship between secondary soil structure and SOM turnover, the traditional methods of aggregate fractionation are often modified and supplemented according to the specific purpose of the individual study. Soil samples may be freed of free-UOM before being subjected to wet sieving. Density separations may be applied to aggregate size fractions before and after their dispersion in order to isolate occluded-UOM. The incorporation of primary OMC and UOM particles into secondary OMC is considered to increase the protection of SOM against decomposition. A main protection mechanism is the creation of physical barriers between the decomposer community and otherwise labile substances, but reduced oxygen and water availability may also be involved. When occluded within aggregates the surfaces of UOM become less accessible to microbial attack and adsorbed OM may have reduced exposure to microbial colonization in points of contact 142 between individual primary particles. Further, the pore structure established during aggregate formation and stabilization may include a significant proportion of small and unconnected pores (Schjønning et al., 1999). It follows that the breakdown of aggregates should expose previously protected OM to decomposition. The simultaneous kill-off of microorganisms by the aggregate dispersion process contributes to the pool of easily available substrate and thereby confounds with the response due to release of protected substrates. For soils to which the concept of aggregate hierarchy is considered to apply it is generally found that crushing of macroaggregates increases the release of CO2 upon incubation in the laboratory (Christensen, 1996a). Microaggregates release more CO2 upon incubation than macroaggregates and it is assumed that smaller aggregates represent larger aggregates that have been subject to slaking during the fractionation. The concept of aggregate hierarchy was developed from studies of soils rich in clay + silt and with relatively high soil OM contents. Results from light-textured Danish arable soil appear to lend little support for the existence of aggregate hierarchy. In these soils the decomposition of microbial biomass killed by the fractionation processes may contribute significantly to the mineralization response recorded during incubation of aggregates. The contribution from non-biomass OM previously protected within macroaggregates appears quantitatively less important. Although somewhat preliminary, this conclusion is in accordance with mineralization studies on primary OMC (Christensen, 1987; Christensen and Olesen, 1998) showing that pooled contributions from individually incubated primary OMC size classes agreed with C and N mineralization from (sieved) whole soil samples. However, in the more clayey soil pooled contributions exceeded that of whole soil. From this evidence it may be inferred that abiotic and other stochastic processes play a dominant role in the formation of secondary OMC in sandy arable soils. This would explain why bulk soil and differently sized macroaggregates appear similar in textural composition and generally also in microbial biomass and OM content (Christensen, 2000). Consequently, UOM and microbial processes may play inferior roles in aggregate formation. Still the concept allows for a reduced lability of OM in primary OMC when these are incorporated into secondary OMC because OM located on the surfaces of particles will be less available in zones of contact between particles. To some extent the concept allows biotic processes to add to the stabilization of secondary OMC although analyses of the microbial community do not suggest that specific groups of organisms are involved in aggregate stabilization. It is suggested that the effect of aggregation of primary OMC on soil OM turnover is mainly through effects on soil pore structure and physical soil processes which in turn affect the general microbial activity through regulations of air exchange and water availability. Thus the effect of formation of secondary OMC on OM turnover is indirect. 143 Levels of structural and functional complexity The spatial organisation of the decomposition subsystem is dynamic. Exogenous forces (e.g. freeze/thaw and dry/wet cycles and tillage induced physical disturbance), internal properties (soil microbial and faunal activity), and the patchiness of the plant subsystem (including root distribution and activity) are important in the simultaneous creation, stabilization and deterioration of the physical structure of the decomposition subsystem (Oades, 1993; Beare et al., 1995; Lavelle et al., 1997). Some perturbations have only transient influences on decomposer activity (e.g. fluctuations in moisture and temperature) while others (e.g. soil tillage) markedly alter the mosaic of microsites and the distribution and availability of substrates, and thereby exert a more permanent effect on soil structural elements and microbial activity. Both exogenous perturbations and bioturbation operate across a wide range of spatiotemporal scales. Various levels of complexity in the decomposition subsystem can be distinguished and classified according to size, function, structure and timescale involved. Although the levels defined may differ according to the purpose of individual studies, it is generally recognized that the significance of processes occurring at one level of resolution is expressed at the next higher level of complexity while the mechanisms behind the processes have to be sought at the next level down or sometimes at even finer scales of resolution (Wu and Loucks, 1995). Three levels of structural and functional organization have been recognized in the work referred here (Fig. 3). Primary OMC refer to the primary structure of soils as defined by soil texture. These complexes are seen as the basic structural and functional unit and are isolated following complete dispersion of soils (Christensen, 1985, 1992). The distribution of primary OMC into clay, silt and sand-sized separates matches the size distribution obtained by conventional textural analysis. In organic matter studies, organomineral size separates may be considered as a functional analogue to soil texture. The main mechanisms responsible for reducing the biological accessibility of OM in primary OMC are the stabilization of otherwise degradable substrate through electrostatic forces, sorption and other chemical reactions with mineral surfaces. Secondary OMC represents the next higher level of organization, the secondary soil structure being determined by the degree of aggregation of primary OMC. 144 Figure 3. Levels of structural and functional complexity in OM in soil Structural entities Functional features Primary structure • clay, silt, and sand sized OMC • chemical OM stabilization • UOM • modification of micro-environment • scale in µm-mm • surface reactivity • adhesion of microbes Secondary structure • aggregated primary OMC • physical OM protection • UOM • soil porosity/aeration • fine roots • microfaunal habitat • fungal hyphae • water retention • scale in mm-cm • slaking/erodibility Tertiary structure • the intact soil • bioturbation by macrofauna • resource islands • pore system continuity • macropores • particle transport • macroroots • preferential flow • scale in cm-m • compaction, drainage • crop productivity While the size distribution of primary complexes represents a static property of a given soil, the distribution of secondary organomineral complexes is a dynamic property affected by bioturbation and exogenous perturbations. In addition, soil sampling and fractionation procedures are decisive for the size distribution of secondary complexes. With regard to air exchange, water and resource availability and thus the general conditions for microbial activity, the secondary OMC represents a higher level of complexity than primary OMC. The topmost level of complexity is expressed in the tertiary soil structure representing the structurally intact soil in situ. At this level of organization, macroporosity, resource islands, mesofaunal and plant root activity, soil tillage and compaction become important phenomena for organic matter turnover. In this way the tertiary soil structure exhibits a dominant control over microbially mediated processes in terrestrial decomposition subsystems (e.g. van Veen and Kuikman, 1990; Golchin et al., 1997). The tertiary soil structure is very dynamic, however, and responds readily to physical disturbances whether these are natural (e.g. mediated by abiotic factors or by ecosystem engineers (Lavelle et al., 1997)) or anthropogenic (cultivation, soil tillage). The impact of cultivation and soil management on the spatial organization of the decomposition subsystem and thus on organic matter turnover may be analysed and 145 interpreted in the conceptual framework outlined above. Cultivation and soil tillage can be viewed as triggering a cascade or successional series of changes that move down through the levels of soil organization. The direct and immediate effect of physical disturbances is on the tertiary soil structure which in turn triggers effects on the secondary structural level. Changes in secondary OMC subsequently affect primary OMC. The cascade of events introduces simultaneous changes in the complex three-dimensional soil matrix of solids, liquid and gas phases and eventually affect all three levels of organization. It is indicated, however, that while effects of soil disturbance on the tertiary level occur immediately, the derived effects on secondary OMC may lag behind. The slowness with which primary OMC responds quantitatively is reasonably well documented. It is suggested that this view, principally based on physical soil fractionation studies, represents a viable perspective for research on soil organic matter turnover and may contribute to an improved management of agricultural soils. One major challenge in this respect is to incorporate the time and intensity of soil tillage as a tool in improving the synchrony between soil net N mineralization and crop N uptake potentials. A greater recognition of the hierarchical nature of soil structure will assist in identifying important gaps in our knowledge and facilitate increasingly interdisciplinary research activity. References Baldock, J.A., Oades, J.M., Vassallo, A.M. & Wilson, M.A., 1990. Solid-state CP/MAS 13C NMR analysis of particle size and density fractions of a soil incubated with uniformly labelled 13C-glucose. Australian Journal of Soil Research 28, 193-212. Baldock, J.A., Oades, J.M., Waters, A.G., Peng, X., Vassallo, A.M. & Wilson, M.A., 1992. Aspects of the chemical structure of soil organic materials as revealed by solid state 13C NMR spectroscopy. Biogeochemistry 16, 1-42. Balesdent, J., Besnard, E., Arrouays, D. & Chenu, C., 1998. The dynamics of carbon in particle-size fractions of soil in a forest-cultivation sequence. Plant and Soil 201, 49-57. Beare, M.H. & Bruce, R.R., 1993. A comparison of methods for measuring water-stable aggregates: implications for determining environmental effects on soil structure. Geoderma 56, 87-104. Beare, M.H., Coleman, D.C., Crossley Jr., D.A., Hendrix, P.F. & Odum, E.P., 1995. A hierarchical approach to evaluating the significance of soil biodiversity to biogeochemical cycling. In: H.P. Collins et al. (Eds.). The Significance and Regulation of Soil Biodiversity. Kluwer Academic Publ., Dordrecht, The Netherlands, 5-22. Besnard, E., Chenu, C., Balesdent, J., Puget, P. & Arrouays, D., 1996. Fate of particulate organic matter in soil aggregates during cultivation. European Journal of Soil Science 47, 495-503. Cadish, G., Imhof, H., Urquiaga, S., Boddey, R.M. & Giller, K.E., 1996. Carbon turnover (δ13C) and nitrogen mineralization potential of particulate light soil organic matter after rainforest clearing. Soil Biology and Biochemistry 28, 1555-1567. 146 Christensen, B.T. & Olesen, J.E., 1998. Nitrogen mineralization potential of organo-mineral size separates from soils with annual straw incorporation. European Journal of Soil Science 49, 25-36. Christensen, B.T. 1985. Carbon and nitrogen in particle size fractions isolated from Danish arable soils by ultrasonic dispersion and gravity-sedimentation. Acta Agriculturae Scandinavica 35, 175-187. Christensen, B.T., 1987. Decomposability of organic matter in particle size fractions of soils with straw incorporation. Soil Biol. Biochem. 19, 429-435. Christensen, B.T., 1992. Physical fractionation of soil and organic matter in primary particle size and density separates. Advances in Soil Science 20, 1-90. Christensen, B.T., 1995. Land use and fertilization effects on the chemical nature of soil organic matter in primary organomineral complexes. Nordic Humus Newsletter 2, 7-16. Christensen, B.T., 1996a. Carbon in primary and secondary organomineral complexes. In M.R. Carter & B.A. Stewart (Eds.): Advances in Soil Science - Structure and Organic Matter Storage in Agricultural Soils. CRC Lewis Publ., Boca Raton, Florida, 97-165. Christensen, B.T., 1996b. Matching measurable soil organic matter fractions with conceptual pools in simulation models of carbon turnover: Revision of model structure. In D.S. Powlson, P. Smith & J.V. Smith (Eds.) Evaluation of Soil Organic Matter Models. NATO ASI Series, Volume I 38, Springer-Verlag, Berlin, 143-159. Christensen, B.T., 2000. Organic matter in soil – structure, function and turnover. DIAS Report No. 30, Plant Production. Danish Institute of Agricultural Sciences, Research Centre Foulum, Tjele, Denmark. Elliott, E.T., Paustian, K. & Frey, S.D., 1996. Modelling the measurable or measuring the modelable: A hierarchical approach to isolating meaningful soil organic matter fractions. In: D.S. Powlson et al. (Eds.). Evaluation of Soil Organic Matter Models. Springer-Verlag, Berlin, Germany, 161-179. Emerson, W.W. & Greenland, D.J., 1990. Soil aggregates-formation and stability. In M.F. De Boodt et al. (Eds.). Soil Colloids and Their Associations in Aggregates. Plenum Press, New York, 485-511. Golchin, A., Baldock, J.A. & Oades, J.M., 1997. A model linking organic matter decomposition, chemistry, and aggregate dynamics. In: R. Lal et al. (Eds.). Soil Processes and the Carbon Cycle. CRC Press, Boca Raton, FL, USA, 245-266. Golchin, A., Clarke, P. & Oades, J.M., 1996. The heterogeneous nature of microbial products as shown by solid-state 13C CP/MAS NMR spectroscopy. Biogeochemistry 34, 71-97. Gregorich, E.G. & Janzen, H.H., 1996. Storage of soil carbon in the light fraction and macroorganic matter. In: M.R. Carter & B.A. Stewart (Eds.). Structure and Organic Matter Storage in Agricultural Soils. CRC Lewis Publishers, Boca Raton, FL, USA, 167-190. Gregorich, E.G., Drury, C.F., Ellert, B.H. & Liang, B.C., 1997. Fertilization effects on physically protected light fraction organic matter. Soil Science Society of America Journal 61, 482-484. 147 Guggenberger, G., Zech, W., Haumaier, L. & Christensen, B.T., 1995. Land use effects on the composition of organic matter in particle size separates of soil II. CPMAS and solution 13C NMR analysis. European Journal of Soil Science 46, 147-158. Hassink, J., 1997. The capacity of soils to preserve organic C and N by their association with clay and silt particles. Plant and Soil 191, 77-87. Haynes, R.J. & Beare, M.H., 1996. Aggregation and organic matter storage in meso-thermal, humid soils. In: M.R. Carter & B.A. Stewart (Eds.). Structure and Organic Matter Storage in Agricultural Soils. CRC Lewis Publishers, Boca Raton, FL, 213-262. Hedges, J.I. & Oades, J.M., 1997. Comparative organic geochemistries of soils and marine sediments. Organic Geochemistry 27, 319-361. Huang, P.M. & Schnitzer, M., Eds., 1986. Interactions of Soil Minerals with Natural Organics and Microbes. SSSA Special Publication No. 17, Soil Science Society of America, Inc., Madison, WI, USA. Jastrow, J.D. & Miller, R.M., 1997. Soil aggregate stabilization and carbon sequestration: Feedbacks through organomineral associations. In: R. Lal et al. (Eds.). Soil Processes and the Carbon Cycle CRC Press, Boca Raton, FL, 207-223. Jastrow, J.D., 1996. Soil aggregate formation and the accrual of particulate and mineral- associated organic matter. Soil Biology and Biochemistry 28, 665-676. Jastrow, J.D., Boutton, T.W. & Miller, R.M., 1996. Carbon dynamics of aggregate-associated organic matter estimated by carbon-13 natural abundance. Soil Science Society of America Journal 60, 801-807. Kaiser, K., Guggenberger, G. & Zech, W. (1998) Formation of refractory soil organic matter by sorption. Mitteilungen der Deutschen Bodenkundlichen Gesellschaft 87, 211-224. Kemper, W.D. & Rosenau, R.C., 1986. Aggregate stability and size distribution. In: A. Klute et al. (Eds.). Methods of Soil Analysis, Part 1, Physical and Mineralogical Methods, Second Edition. ASA and SSSA Publisher, Madison, WI, 425-442. Lavelle, P., Bignell, D., Lepage, M., Wolters, V., Roger, P., Ineson, P., Heal, O.W. & Dhillion, S., 1997. Soil function in a changing world: the role of invertebrate ecosystem engineers. European Journal of Soil Biology 33, 159-193. Mayer, L.M., 1994. Relationships between mineral surfaces and organic carbon concentrations in soils and sediments. Chemical Geology 114, 347-363. Oades, J.M. & Waters, A.G., 1991. Aggregate hierarchy in soils. Australian Journal of Soil Research 29, 815-828. Oades, J.M., 1993. The role of biology in the formation, stabilization and degradation of soil structure. Geoderma 56, 377-400. Oades, J.M., Waters, A.G., Vasallo, A.M., Wilson, M.A. & Jones, G.P., 1988. Influence of management on the composition of organic matter in a Red-brown earth as shown by the 13C nuclear magnetic resonance. Austr. J. Soil Res. 26, 289-299. Preston, C.M., 1996. Applications of NMR to soil organic matter analysis: History and prospects. Soil Science 161, 144-166. 148 Randall, E.W., Mathieu, N., Powlson, D.S. & Christensen, B.T., 1995. Fertilization effects on organic matter in physically fractionated soils as studied by 13C NMR: Results from two long-term field experiments. European Journal of Soil Science 46, 557-565. Schjønning, P., Thomsen, I.K., Møberg, J.P., de Jonge, H., Kristensen, K. & Christensen, B.T., 1999. Turnover of organic matter in differently textured soils I. Physical characteristics of structurally disturbed and intact soils. Geoderma 89, 177-198. Six, J., Elliott, E.T., Paustian, K. & Doran, J.W., 1998. Aggregation and soil organic matter accumulation in cultivated and native grassland soils. Soil Science Society of America Journal 62, 1367-1377. Skjemstad, J.O., Le Feuvre, R.P. & Prebble, R.E., 1990. Turnover of soil organic matter under pasture as determined by 13C natural abundance. Australian Journal of Soil Research 28, 267-276. Tisdall, J.M & Oades, J.M., 1982. Organic matter and water-stable aggregates in soils. Journal of Soil Science 33, 141-163. van Veen, J.A. & Kuikman, P., 1990. Soil structural aspects of decomposition of organic matter by microorganisms. Biogeochemistry 11, 213-233. Wu, J. & Loucks, O.L., 1995. From balance of nature to hierarchical patch dynamics: a paradigm shift in ecology. The Quarterly Review of Biology 70, 439-466. 149 Management of biodiversity in arable soil by field inoculation - an example using deep burrowing earthworms Visa Nuutinen1 and Jyrki Pitkänen2 1Agricultural Research Centre, Crops and Soil, FIN-31600 Jokioinen, FINLAND E-mail: [email protected] 2Employment and Economic Development Centre for South-Eastern Finland, Box 1041, 45101 Kouvola, FINLAND Summary We investigated the possibilities to encourage by field inoculation the population growth of the deep burrowing earthworm Lumbricus terrestris L.. Inoculation was done in southern Finland in a clayey field and its margins where no L. terrestris had earlier been found. The species is present in many nearby soils. In half of the field autumn mouldboard ploughing had been replaced with autumn stubble cultivation four years before the inoculation, a shift which is known to be beneficial for L. terrestris. Two years after the inoculation L. terrestris had locally got a foothold at the permanent grass strips in the field margins. There the population had also started to spread from some of the inoculation points. Inside the field, no L. terrestris were found from ploughed or non-ploughed areas. The absence of L. terrestris from the site may thus have been partly due to limited dispersal but the tillage method was obviously not the key factor limiting the settlement. It is likely that inside the field the very high clay content in the subsoil with the accompanying high water table level and compacted soil structure was inhibiting the establishment of L. terrestris population. The inoculation was evidently done in relatively harsh conditions and generalizations on the merits of field inoculation are unwarranted. The results nevertheless suggest that earthworm inoculations into fields, which are clearly within the species’ current geographical range may be of limited value. Inoculations done in more benign conditions, in fields cleared recently into habitats with low natural population densities and/or outside the extant species range could be more well-founded and successful. Keywords: soil management, biodiversity, earthworms, Lumbricus terrestris, field inoculation Introduction Abundance of deep burrowing, surface residue feeding earthworms often rises after replacing mouldboard ploughing with reduced tillage (Edwards & Bohlen 1996). The consequences of the change are regarded as beneficial for the incorporation of surface residue and cycling of plant nutrients (Blair et al. 1995) and also for soil structural and hydrological properties (Edwards & Shipitalo 1998; Pitkänen & Nuutinen 1998). There is considerable variation in 151 the response of deep burrowers to implementation of reduced tillage and at times they remain absent from a field years after ploughing has ceased (Nuutinen 1992). Earthworms are slow dispersers and it is possible that the lack of population increase is sometimes due to limited dispersal. We wanted therefore to study if active field inoculation of deep burrowing earthworms could enhance the increase in their numbers after introduction of reduced tillage. Here we report a case of an experimental inoculation of the earthworm Lumbricus terrestris L. done in 1996 in south-western Finland Material and Methods Inoculation site The place of inoculation was an experimental drainage field which consists of four parallel plots, each 33 m x 140 m (Table 1; Turtola 1999). From 1992 on, autumn mouldboard ploughing was replaced with autumn stubble cultivation to a depth of 8 cm in two of the plots. In the two other plots autumn mouldboard ploughing to a depth of 20-25 cm was continued. Before the earthworm inoculation, spring cereals had been grown in the field for seven years except for an intervening year of timothy in 1993. In autumn 1988 a comprehensive earthworm sampling using formalin extraction had been carried out in the field (Nuutinen, unpublished). No L. terrestris were then found. Neither had recent soil sampling and excavation in the area revealed any signs of L. terrestris activity. The study area is within the geographical range of L. terrestris in Finland (Terhivuo 1988) and the species is found in many field, forest and garden soils in the nearby areas. Inoculation units During spring 1996 individuals of L. terrestris were collected in large quantities at a park lawn in Jokioinen. Using these worms, one hundred earthworm inoculation units were prepared in May 1996 applying the technique developed by Butt et al. (1995). The units consisted of a plastic bag containing eight litres of clayey soil with horse manure added on the soil surface for worm food. Five large L. terrestris individuals, at least two of them sexually mature, were introduced in each unit and the bags were closed. The units were provided with aeration and were maintained in +16ºC at 12:12 hrs light cycle. Soil moisture content was kept at approximately 22% and food was added when necessary. Before the inoculation ten units were randomly chosen and their contents were wet sieved on 2 mm mesh to estimate earthworm survival and cocoon production in the units. A sample of 70 cocoons was incubated on moist filter paper in small vials with perforated lids to evaluate the hatching success of the cocoons. The vials were kept in constant darkness at +12ºC. 152 Table 1. Characteristics of the inoculation site. ______ LOCATION: Experimental drainage field, Jokioinen, south-west Finland, 60º49' N, 23º30' E. SOIL (data: Turtola 1999): Vertic Cambisol (FAO 1988), very fine Typic Cryaquept (Soil Survey Staff 1992). ______ 0-25 cm 25-40 cm 40-80 cm ______ clay% 61 83 90 Org. C% 2.7 0.6 0.4 pHwater 5.9 6.3 6.9 -1 KSAT (cm h ) 62 0.63 0.005 ______ SIZE & SLOPE: 132 m x 140 m, 2% (1-4%). DRAINAGE: Tile drainage 1962, improved with additional plastic pipes 1991. MANAGEMENT: The area has been under cultivation for approximately 50 years, recently spring cereals and grass for silage; NPK, no organic fertilizers; autumn ploughing omitted in half of the area in 1992. EARTHWORMS (1988): Aporrectodea caliginosa (Sav.) and Lumbricus rubellus (Hoffm.) in low numbers. ______ Prior sampling at the inoculation site In October 1996, 1.5 weeks before the inoculation, L. terrestris sampling was carried out using mustard extraction (Gunn 1992) to check once more that L. terrestris was absent from the inoculation site. Sampling was done at the upper margin of the field covering half of the planned inoculation area. Six samples were taken from both of the non-ploughed plots as they were the ones most likely to maintain L. terrestris. In addition, six samples were taken at the permanent grass strips at the upper margins of the two plots. At the grass strips the sampling unit size was on average 0.23 m2, inside the field on average 0.47 m2. Narrowness of the border strips made it necessary to use smaller sized samples at the field margins. Emerging earthworms were rinsed in fresh water and placed in vials with moist water towels. They were allowed to empty their guts for 24 hours, anesthetized using diethyl ether and stored in 70% alcohol. Species were determined following Sims & Gerard (1985). Inoculation Inoculation of the L. terrestris units was done at the upper end of the field in mid October 1996 after the autumn tillage. The majority of the inoculations were in 5.5 m intervals along three transects, each approximately 135 m long (Fig. 1). The first transect was outside the cultivated area, at the grass strip between the field margin and an open ditch. The remaining two transects were inside the field, at 6 and 8 m distances from the field margin. They ran 153 through all four experimental plots, inoculations being made also at the grass strips in the plot boundaries. For the inoculation, a 15-20 cm deep pit was dug and the contents of an unit were carefully emptied into it. Altogether 82 units were inoculated, 40 inside the field and 42 at the grass strips. The pits were covered with horse manure and an ample pile of straw to provide food and insulation. At the time of the inoculation soil moisture content at 0-10 cm and 30-40 cm depths were 25%. During a few nights following the inoculation, air temperatures fell below 0ºC. Soil temperature was followed in one of the inoculation pits at the depth of 10-15 cm. The temperature was +6ºC at the inoculation. The first below 0ºC temperatures in the pit were measured 4 weeks after the inoculation. Figure 1. The inoculation area in the upper end of the experimental field. Circles indicate the places where inoculation was done in autumn 1996. Crosses mark the points where L. terrestris were sampled in autumn 1998. Arrows point at the inoculation pits around which additional L. terrestris samples were taken at 1.5 m’s distance in autumn 1998. Field observations and sampling In May 1997 the top few centimeters at 13 inoculation pits (5 at the grass strips, 8 inside the field) were investigated for the presence of L. terrestris individuals. In May 1998, before the spring cultivations, observations on L. terrestris middens were made at the 40 inoculation points inside the field. In September 1998 comprehensive sampling of L. terrestris was carried out at the inoculation area. Altogether 39 samples were taken at the points of inoculation, 16 samples inside the field and 23 samples at the grass strips (Fig. 1). Additional samples were taken 1.5 m away from five inoculation points to find out if L. terrestris individuals had started to disperse (Fig. 1). The sampling procedure was the same as in the prior sampling in 1996. At the time of sampling topsoil (0-5 cm) temperature was +13ºC and moisture content 37%. 154 Results In the prior sampling no L. terrestris were found in the field or its margins supporting the conception that the field was devoid of this species. Inside the field Aporrectodea caliginosa (Sav.) was present. In the grass strip at the field margin Lumbricus rubellus (Hoffm.) and Dendrobaena octaedra (Sav.) were found. At the time of the inoculation, there were on average 23 (SE=6.2, N=10) egg capsules of L. terrestris per inoculation unit. Their laboratory hatching percent was 46%, which is a relatively low figure compared with the hatchability of 80-90% reported by Holmstrup et al. (1996). Only a couple of newly hatched L. terrestris were found in the units. On average three (SE=0.5, N=10) of the original five earthworm individuals had survived till the time of inoculation. One can then estimate that roughly a total of 250 large L. terrestris individuals and 1900 cocoons were inoculated in the 82 units. In May 1997 newly hatched L. terrestris individuals were observed in two pits out of the thirteen investigated. Both pits were inside the field, one in ploughed, the other one in a stubble cultivated plot. In May 1998, no L. terrestris middens were found at the inoculation points inside the field suggesting that the settlement in the cultivated area had failed. In September 1998 L. terrestris was present at 10 out of the 23 points sampled at the grass strips (Fig. 2). Maximum estimate of population density was 28 individuals m-2. The majority of the individuals were juveniles hatched from the inoculated cocoons (Fig. 3). L. terrestris had not got established inside the field: no L. terrestris were found in either ploughed or non- ploughed plots. At two grass strips L. terrestris individuals were found 1.5 m down the strip from the inoculation point in densities of 5 and 8 individuals m-2, respectively. Figure 2. Distribution and abundance of L. terrestris in the field in autumn 1998. 155 Figure 3. The age distribution of L. terrestris individuals sampled at the inoculation points in autumn 1998. Discussion The successful establishment of a L. terrestris population in the field margins indicates that the absence of the species from the site may have partly been due to limited dispersal. Future monitoring at the site is, however, needed to confirm that a truly permanent settlement took place. Inside the field the tillage method was obviously not the key factor limiting the settlement of L. terrestris. Obstacles for population growth in the cultivated area are likely to include the very high subsoil clay content, periodically high water table level and the accompanying compact soil structure. There are namely earlier reports of low L. terrestris numbers in soils with high clay content (Guild 1948, Nuutinen 1992) and soil compaction is known to adversely affect L. terrestris (Rushton 1986). Further, our recent observations on relatively high population densities of L. terrestris at subsurface drain pipe lines in another clayey field indicate that high water table level may be a factor limiting population densities (Nuutinen et al., in preparation). Compared with the cultivated area the living conditions at the grass strips were likely to be better for L. terrestris in a number of ways. Roots and other biological activity may have worked the soil more suitable for burrowing, and soil compaction due to field traffic is absent at the strips. Excessive soil moisture may also have been less of a problem at the grass strips due to their slightly elevated position and, at the field margin, owing to the drainage by an open ditch next to the upper margin of the field. 156 One factor, which may generally have impaired the success of the inoculation were the low temperatures during the nights following the inoculation. This could have caused mortality of worms and cocoons or at least interfered with the feeding and burrowing of worms. However, according to our own observations L. terrestris digs burrows and feeds actively in temperatures at least as low as +4ºC. The temperatures inside the pits remained at that level during the first weeks after the inoculation. Therefore it seems unlikely, that immediate mass mortality caused by the first night frosts would have been the major factor affecting the success of the inoculation Conclusions The inoculation was obviously done in conditions, which were very demanding for L. terrestris. Therefore generalizations on the merits of active earthworm inoculation based on the present results are unwarranted. However, the results suggest that this type of L. terrestris inoculation into fields, which are clearly situated within the species’ geographical range may be of questionable value. One can assume that in such instances the absence of L. terrestris is often due to the prevailing soil conditions. If the conditions in the area are generally suitable for the species, it is likely to be present at the vicinity of the field. It may then be sufficient to count on natural migration into the field after a beneficial change in the soil tillage has occurred. This is particularly so considering the notable labour effort which the active inoculation requires. There are examples of successful earthworm introductions and the practice is without doubt reasonable when carried out in well-considered manner (Edwards & Bohlen 1996). For instance, thinking of L. terrestris introductions in northern Scandinavia, inoculations done in more benign soil conditions, in fields cleared very recently into habitats with low natural population densities and/or outside the extant species range could be well- founded and beneficial. In Finland the distribution of L. terrestris displays northward, long- distance dispersal ”jumps” into human impacted soils which most probably result from anthropochorous dispersal (Terhivuo 1988). This may indicate scope for actively introducing this species to agricultural soils outside its present, predominantly southern range in Finland. References Blair, J.M., Parmelee, R.W. & Lavelle, P., 1995. Influences of earthworms on biogeochemistry. In: Hendrix, P.F. (ed.). Earthworm Ecology and Biogeography in North America. Lewis Publishers, Boca Raton, 127-158. Butt, K.R., Frederickson, J. & Morris, R.M., 1995. An earthworm cultivation and soil- inoculation technique for land restoration. Ecological Engineering 4: 1-9. Edwards, C.A. & Bohlen, P.J., 1996. Biology and Ecology of Earthworms. 3rd Edition. Chapman & Hall, London. 426 pp. Edwards, W.M. & Shipitalo, M.J., 1998. Consequences of earthworms in agricultural soils: aggregation and porosity. In: Edwards, C.A. (ed.). Earthworm ecology. St. Lucie Press, Boca Raton, 147-161 157 FAO, 1988. FAO/Unesco Soil Map of the World. World Resources Report 60 FAO, Rome. Reprinted as Technical Paper 20. ISRIC, Wageningen, 140 pp. Guild, W.J. McL., 1948. Studies on the relationships between earthworms and soil fertility III. The effect of soil type on the structure of earthworm populations. Annals of Applied Biology 35: 181-192. Gunn, A., 1992. The use of mustard to estimate earthworm populations. Pedobiologia 36:65- 67. Holmstrup, M., Østergaard, I.K., Nielsen, A. & Hansen, B.T., 1996. Note on the incubation of earthworm cocoons at three constant temperatures. Pedobiologia 40: 477-478. Nuutinen, V., 1992. Earthworm community response to tillage and residue management on different soil types in southern Finland. Soil & Tillage Research 23: 221-239. Pitkänen, J. & Nuutinen, V., 1998. Earthworm contribution to infiltration and surface runoff after 15 years of different soil management. Applied Soil Ecology 9: 411-415. Rushton, S.P., 1986. The effects of soil compaction on Lumbricus terrestris and its possible implications for populations on land reclaimed from open-cast coal mining. Pedobiologia 29: 85-90. Sims, R.W. & Gerard, B.M., 1985. Earthworms. Synopses of the British Fauna. No. 31. The Linnean Society of London and The Estuarine and Brackish-Water Sciences Association. E.J. Brill / Dr. W. Backhuys, London. 171 pp. Soil Survey Staff ,1992. Keys to Soil Taxonomy, 5th edition. SMSS Technical Monograph No. 19. Blacksbury, Virginia. 541 pp. Terhivuo, J., 1988. The Finnish Lumbricidae (Oligochaeta) fauna and its formation. Annales Zoologici Fennici 25: 229-247. Turtola E., 1999. Phosphorus in surface runoff and drainage water affected by cultivation practices. Dissertation (Ph.D.), University of Helsinki. 108 pp Acknowledgements We thank Risto Seppälä, Taisto Sirén, Carolien Helmich and Fredau Miedema for their help in the field and laboratory. Asko Hannukkala and the Plant Protection unit of the Agricultural Research Centre are acknowledged for generous logistic support. For co-operation and helpful comments our thanks are due to Eila Turtola. Participants of the NJF Seminar No. 310 ”Soil stresses, quality and care”, Douglas Karlen and Susanne Elmholt in particular, are thanked for useful comments. The study was partly funded by the Academy of Finland (Project No. 34038). VN acknowledges additional support by the Finnish Drainage Foundation. 158 Land use changes and degradation of forest and soil in watersheds of Nepal – A review B.K. Sitaula1, K.D. Awasthi2, N.R. Chapagain 2, G. S. Paudel3, R.P. Neupane3, P.L. Sankhayan4, B.R.Singh1 and O. Hofstad4 1Department of Soil and Water Sciences, Agricultural University of Norway, N-1430 Ås, NORWAY E-mail: [email protected] 2Institute of Forestry, Tribhuvan University, P.O. Box 43, Pokhara, NEPAL 3The School of Environment, Resources and Development, Asian Institute of Technology, Pathumthani 12120, Bangkok, THAILAND 4Department of Forest Sciences, Agricultural University of Norway, N-1430 Ås, NORWAY Summary The watersheds in the hills of Nepal are facing severe stresses from deforestation, soil erosion, and depletion of plant nutrition resulting in the loss of soil quality. These watersheds have inherently fragile ecology with unstable geology, steep slopes, intense monsoon rains, and subsistence farming systems. This paper presents an overview on watershed degradation, deforestation/forest degradation, land use changes and other important factors and processes affecting soil quality, such as soil erosion and depletion of plant nutrients (soil fertility decline) in Nepal. Soil degradation in the form of soil erosion and fertility decline appears a major cause of farmland degradation. Introduction Watersheds located in hills of Nepal are undergoing gradual degradation (Mahat, 1987; Ahmed, 1989; Applegate and Gilmore, 1987; Mishra and Bista, 1998; Ives and Messerli, 1989). Several interlinked factors, such as, increased demand for food, wood fuel and timber of rapidly increasing human and livestock populations, uncontrolled and excessive grazing, poor soil and forest management practices, slash and burn practices and other environmentally unsound infra-structural activities are contributing to the land degradation processes in the watershed (Partap and Watson, 1994). The overall watershed degradation in Nepal is related to the land use changes, high rate of deforestation and degradation of forests and soils. In order to undertake countermeasures, or simply to diagnose problems of watershed degradation, a broader and more thorough understanding of the process of overall land use changes is essential. Estimating land use changes in a given watershed would provide useful information on the magnitude of the deforestation and land degradation over a period of time. Locating and estimating land use changes in the watershed serves as the first step towards answering various questions, such as, 159 how forest resources in the watershed are changing, how much of forest or soil is degraded, what are the causes of deforestation, what becomes of the deforested land, and, what role do the ecological and socio-economic conditions play in this transformation. In this paper, we have reviewed studies that focus upon deforestation/forest degradation, land use changes, and different types of soil degradation (such as soil erosion and decline in soil fertility) at the watershed level. All of these have relevance for soil quality in broader sense. Watershed Degradation There is a lack of updated information on the extent and severity of watershed degradation covering whole of Nepal. A 20 years old study undertaken by Nelson et al. (1980) indicated that 13 percent land was severely degraded and another 36 percent was degrading in Nepal. Table 1. Classification of watersheds according to degradation status in Nepal. Rank Name of District Value Rank Name of Value Rank Name of District Value District Very Poor Good Excellent 1 Surkhet 5,118 26 Ilam 1,468 51 Achham 471 2 Kavrepalanchok 4,958 27 Baitaidi 1,449 52 Kapilvastu 462 3 Dang 4,944 28 Nawalparasi 1,392 53 Solokhumbu 389 4 Okhaldhunga 4,941 29 Makwanpur 1,370 54 Sankhuwasabha 263 5 Mustang 4,824 30 Sindhuli 1,325 55 Chitwan 261 6 Shyangja 4,725 31 Sallyan 1,294 56 Dhanusha 257 7 Nuwakot 4,646 32 Panchthar 1,277 57 Bardiya 246 Sindhupalan 1,268 58 Dadhaldhura 219 Poor chok 8 Kathmandu 3,627 33 Dhankuta 1,254 59 Morang 213 9 Gulmi 3,477 34 Bajhang 1,159 60 Rauswa 194 10 Arghakhanchi 3,476 35 Rolpa 1,144 61 Kailali 170 11 Bhaktpur 3,450 36 Jajarkot 1,036 62 Taplejung 157 12 Ramechhap 3,299 37 Tehrathum 1,000 63 Mugu 138 Fair Kaski 982 64 Dolkha 94 13 Parbat 2,822 38 Doti 944 65 Sunsari 93 14 Manag 2,486 39 Gorkha 934 66 Saptari 92 15 Piuthan 2,341 40 Bhojpur 929 67 Sirah 42 16 Khotang 2,261 41 Darchula 823 68 Jhapa 10 17 Myagdi 2,126 42 Kalikot 812 69 Mahottari 0 18 Palpa 2,032 43 Jumla 636 70 Sarlahi 0 19 Dolpa 1,990 44 Banke 627 71 Rautahat 0 20 Udiyapur 1,945 45 Bajura 623 72 Bara 0 21 Rukum 1,854 46 Rupandehi 575 73 Parsa 0 22 Dhading 1,765 47 Lamjung 575 74 Humla 0 23 Tanahun 1,744 48 Baglung 569 75 Kanchanpur 0 24 Lalitpur 1,695 49 25 Dailekh 1,544 50 Source: Summarized from Shrestha et al. (1983) and Nelson et al. (1980) 160 Using the results of Nelson (1980), Shrestha et al. (1983) classified 75 districts of Nepal into five categories of watersheds (very poor, poor, moderate, good and excellent) by computing their value and rank. This was done by estimating an index relating the prevailing state of watershed condition in the area compared with watershed condition for that area under natural or well-managed conditions. Low rank and high value indicate poor watershed condition, and low value and high rank indicate better watershed condition (Table 1). The definitions of five major groups based on the extent of degradation are as follows: Class 1. Excellent: The watersheds were near to pristine. Natural erosion processes, including landslides may be present. Class 2. Good: Minor amount of disturbances caused by land use may be present. Productivity of land is not impaired and any negative effect of the disturbances can be mitigated through normal management practices. Class 3. Fair: There are significant disturbances present due to accelerated soil erosion. Productivity of land shows a diminishing trend. A combination of education and structural measures (e.g. check dams) are needed for rehabilitating such watersheds. Class 4. Poor: Disturbances by accelerated soil erosion are serious, and extension, structural and land use changes are required to upgrade the land to a productive condition. Class 5. Very Poor: Accelerated soil erosion has broken down the soil structure and productivity is significantly reduced. Rehabilitation requires structural protection and high investment to upgrade the watersheds. Based on the watershed condition defined and ranked by Nelson et al. (1980), Shrestha et al. (1983) calculated the percentage area under each category (Table 2). Table 2. Summary of watershed condition in Nepal. Class Definition Area in sq.km* % No. of districts** Very poor Badly eroded, difficult to rehabilitate 1410 1 7 Poor Eroded but can be rehabilitated 1410 1 5 Moderate High erosion need immediate rehabilitation 15360 11 13 Good Leading to degradation 50432 36 25 Excellent Near pristine 71400 51 25 Total 140 012 100 75 Source: Summarized from *Nelson et al. (1980) and ** Shrestha et al. (1983) This showed extending watershed degradation in larger areas during 1980’s. Rapidly growing human and livestock populations since 1983, may have had an important influence in accelerating the extent and severity of watershed degradation. The present population density in farming areas is very high, on an average five persons per ha of arable land, and almost 90% of the Nepalese population are engaged in agriculture (Shrestha, 1996). Majorities of the households own very small parcels of land. Such a high pressure on farmland accompanied by growing crops with low inputs of manures and fertilizers, are likely to lead to gradual 161 decrease in soil quality and to overall soil degradation. Investigations carried out by ICIMOD (1999) on a typical watershed located in the middle hills of Nepal (Jhikhu Khola watershed) revealed that the land degradation was related to population growth, agricultural intensification and deforestation. The population in this watershed was growing at a rate of 2.9% per year followed by increased land use intensities. Deforestation and forest degradation Studies undertaken in the mountain watersheds of Nepal have paid overwhelming attention to the problems and prospects of managing forest and rangeland resources. Different studies have reported degradation of forest resources (Enke, 1971; Eckholm, 1976; World Bank, 1979; Kawakita, 1997; Wallace, 1981; Bajracharya, 1983; Metz, 1994). Contrarily, a few micro level studies noticed improvements in forest quality and forest coverage despite growing population pressure in the hills of Nepal (Stevens, 1993; Fox, 1993; Gilmour, 1989; Carter and Gilmour, 1989). Studies carried out in the late 1970’s and 1980’s reported massive deforestation in the hills of Nepal. The descriptive analysis of early studies did not present enough empirical evidence on the extent of appreciation or depreciation of the forest coverage. However, the national forest inventory done by the government of Nepal indicates that the extent of deforestation reported by earlier studies was somewhat overestimated (World Bank, 1979; Ruddle and Rondinelli, 1983). Two indicative studies undertaken in the mid hill watersheds indicated both increasing and decreasing trends in forest coverage (Table 3). Table 3. Land use changes in the mountain watersheds of Nepal. Watersheds Land use Base Final Base year Final year % of % of % year year area in area in area in area in change hectares hectares base final in land year year use Upper Forest 1957 1988 7534 7522 37.0 36.98 -0.2 Pokhara Pasture and shrub ” ” 2993 2785 14.7 13.7 -6.9 Valley1 Agricultural land ” ” 9387 9607 46.2 47.2 2.3 Other land 424 424 2.1 2.1 0 Total ” ” 20338 20338 100 100 - Jhikku Forest 1972 1991 2181 3359 19.5 30 54.0 Khola Pasture and shrub ” ” 3041 1383 27.0 12.4 -54.5 watershed2 Agricultural land ” ” 5497 6073 49.5 54.6 10.5 Other land ” ” 422 306 4.0 3.0 -27.5 Total ” ” 11121 11121 100 100 - Source 1Thapa and Weber (1993), 2Shrestha and Brown (1995) Higher encroachment over shrub and pasture land than in forest was noticed. Forest coverage had slightly decreased in Upper Pokhara Valley (Thapa and Weber 1993), and significantly increased in Jhikku Khola watershed (Shrestha and Brown 1995). Decreasing forest cover was not so problematic as depreciation in forest quality (forest degradation). Even the studies, 162 which found increasing forest cover, noticed degradation of forests due to logging in the interior parts and associated thinning process. The community forestry programs initiated since 1978 have progressively increased forest cover near the village settlements. At the same time, it has increased pressure over open access forests, though the coverage remains the same as before (Paudel, 1997). An empirical study on forest thinning process (forest degradation) undertaken in Andhikhola Watershed (Western hills of Nepal) indicated that the ground cover, tree density and forest bio- diversity declined significantly from geographical center to outer boarder of the forest due to rampant cutting (Table 4). Table 4. Forest thinning process (in 102 m plots) in Andhikhola Watershed (Paudel, 1997). At geographical At central point between outer boarder At the outer boarder of mean center of the forest and geographical mean center the forest (4 km inside the forest) (2 km inside the forest) (100 meter inside the forest) Tree class No of trees No of No of trees No of species No of trees No of species species Canopy 15 2 16 2 8 1 Secondary486103 808 Primary 157 16 79 13 - - Sub-total 220 24 95 18 88 9 Bushes 1,110 110 110 25 - - Total 1,330 134 205 43 88 9 Ground 80% - 40% - 20% - cover Findings of some recent work show deforestation as a continuous process in many parts of Nepal. A study undertaken by Koirala (1999) in the Kathmandu valley observed a significant decline in forest area, increase in the urban area, slight decrease in agriculture area, and continuous depletion of forest resource. Sah et al. (1997) reported an average rate of forest depletion of 0.57% per year in the forest area in one of the watersheds in eastern Nepal. They have also observed that the deforestation rate varied from 0 to 3% of the forested area. Awasthi and Balla (2000) have estimated the annual deforestation rate between 0.6% and 2.1% in their study of two watersheds located in mid hills of Nepal. A Multi-temporal Land Cover Assessment was carried out by ICIMOD (1999) using National Oceanographic and Atmospheric Administration’s (NOAA) Advanced High Resolution Radiometer (AVHRR) data for the year 1992-93. It suggested that while the forest cover of Nepal declined from 39 % in 1980 to 28% in 1992-93, the agriculture land has increased from 20.9% to 31.4% and shrub land decreased from 4.7% to 1.5% during the same time period. 163 Soil erosion Nepal is estimated to loose over 240 million m3 of fertile top soil annually, and about 10,000 hectares of land in the mountainous area has shown desert-like conditions (Ghimire and Upreti, 1997). The soil erosion estimates from dominant land uses in Nepal showed large variation within and between the land use categories (Table 5), most probably due to variations in both biophysical and socio-economic factors. These estimates provide little information on the effects of land management on soil losses. Moreover, some of these work estimated soil loss using universal soil loss equation (USLE) developed for North American conditions, which does not account for socio-economic causes of soil erosion and soil conservation strategies adopted by farmers of Nepal. Table 5. Soil loss estimates from different land uses in watersheds of Nepal. Watersheds Land Uses Soil Loss Source (t ha-1 yr-1) Trijuga Agriculture 35-41 Sah (1996) Andhikhola Agriculture 70 Pahari, (1993) Kulekhani Maize/Bean/Millet 1.5-2.25 FAO (1994) Likhu Khola Maize/Milllet 6-56 Shrestha and Rice 0.1-0.8 Zinck (1999) Trijuga Protected forest 1.7-2 Sah (1996) Degraded forest 6-11 Andhikhola Primary forest 2.2 Pahari (1993) Secondary forest 15.3 Nakkhu khola Forest 5 Tiwari (1990) Likhu Khola Dense forest 0.1-0.4 Shrestha and Degraded forest 0.1-8.6 Zinck, (1999) Trijuga Shrub land 123-180 Sah (1996) Andhikhola Shrub land 29.8 Pahari (1993) Nakkhu khola Shrub land 42.49 Tiwari (1990) Likhu Khola Grazing lands 1.6-19.8 Shrestha and Zinck (1999) Andhikhola Grazing lands 213 Pahari (1993) Nakkhu khola Grazing lands 173 Tiwari (1990) Soil nutrient losses and fertility decline Several studies are available on nutrient losses and decline in soil fertility and crop productivity in the farmlands of Nepal. Most often the studies have confined in monitoring crop yields, performed cost benefit analysis and described soil properties. The majority of these reported declining soil fertility in private farmlands (Carson, 1992; Vaidya et al. 1995; Pandey et al., 1995; Schreier et al., 1995; Shah and Schreier, 1995; Keating et al., 1999, Tripathi, 1996; Shah, 1996). Though soil erosion and intensification of cropping are well- recognized causes of declining soil fertility in the hills of Nepal, effects of soil erosion on fertility decline are still insufficiently understood. Soil erosion rates in agricultural land would 164 vary significantly from one area to another depending on slope gradient, land use, types of farmland and soil management practices employed by farmers. Moreover, these studies (ibid) did not distinguish the land type and portrayed only a general picture. Gharbari (homestead upland), bari (rain-fed uplands normally terraced) and khet (lowland areas used for rice cultivation) are three major types of cultivated areas in the Hills. These lands have distinct characteristics and different management priorities, which largely affect the soil loss and fertility decline. One study shows that the bari, which are sloping terraces or used for shifting cultivation, are more prone to loss of soil and its fertility than khet lands which are level terraces (Table 6). Table 6. Estimated soil and nutrient losses by rainfall erosion under different types of farm lands. Descriptions Level Terraces Sloping terraces Shifting Cultivation Soil loss (tons ha-1 yr-1) 5 20 100 Organic matter loss (kg ha-1 yr-1) 150 600 3000 Nitrogen loss (kg ha-1 yr-1) 7 30 150 Phosphorus loss (kg ha-1 yr-1) 5 20 100 Potassium loss (kg ha-1 yr-1)1040200 Source: Land Resources Mapping Project (1986) The nutrient loss from leaching and surface run-off has been a problem in maintaining agricultural production in many hill watersheds of Nepal. Soils are heavily leached during the short periods of intense monsoon rains, more than 80% of which are concentrated in July- September. This problem of excess moisture moving down the soil profile is more severe in bari than in khet soils. The perennial tree crops grown in bari, which are rain fed and vulnerable to accelerated soil erosion, help in maintaining soil moisture for longer duration and control soil erosion. In warm and dry months, when crop cultivation in bari becomes difficult due to increased soil temperature, plant cover and shade help to control soil temperature fluctuations. Table 7. Nutrient loss from farmland by leaching and surface run-off. Land type Estimated for Nutrient loss from leaching and surface run-off (kg ha-1 yr-1) OM N P K Rainfed bari land 1 (Eastern hills) 664 55.3 2.5 7.9 Rainfed bench terrace (Tarikhet)2 Nepal 75 3.8 5.0 10 Rainfed marginal land (Bari land)2 Nepal 300 15 20 40 Bari land3 Landruk (Western mid hill) - 85.2 39.6 158.5 Bari land3 Bandipur (Western mid hill) - 13.8 75.5 28.5 Bari land3 Nayatola (Western mid hill) - 46.3 19.2 44.5 Source: 1Serchan and Gurung (1996); 2Carson (1992); 3Tripathi et al. (1999) Results of studies outlined in table 7 showed the loss of organic matter to be at 75 kg ha-1 yr-1 as minimum in rain fed bench terraces (tarikhet) to 664 kg ha-1 yr-1 as maximum in bari lands (up-land non-irrigated maize – millet field). Similarly, nitrogen loss was estimated to be 165 4 kg ha-1 yr-1 as minimum in tarikhet and 85 kg ha-1 yr-1 as maximum in outward facing bari terraces. Even the phosphorous loss is estimated higher compared to its inputs. The loss is estimated to be 3 kg ha-1 yr-1 as minimum to 75 kg ha-1 yr-1 as maximum. Loss of potassium is recorded ever highest, and it ranges from 7.9 kg ha-1 yr-1 to 158 kg ha-1 yr-1 (Table 7). Besides leaching and run off losses, the nutrient uptake by crops is estimated higher than farmers’ inputs (Tripathi et al., 1999). Hence, the low inputs by farmers, higher uptake by crops and losses from the natural processes have made mid hill farmlands vulnerable to infertility. Conclusions Watershed degradation in Nepal is a serious problem. It is primarily attributed to deforestation, forest degradation, soil erosion and fertility decline. The degradation processes may have resulted in significant land use changes. Soil degradation in the form of soil erosion and fertility decline appears a major cause of farmland degradation. The processes contributing to land use changes, deforestation, and forest and soil degradation in Nepal is inadequately understood. Forest degradation as manifested through decline in forest cover, and the resulting soil erosion losses and fertility decline is caused by a complex coupling of biophysical, socio-economic and technological factors in the Nepalese watersheds. Therefore greater understanding of degradation processes will require a systems analysis by considering the problem in its entirety rather than the various components in isolation. References Ahmed, A. E., 1989. Land restoration and revegetation, role of forestry in combating desertification. FAO Conservation Guide 21. FAO, Rome. Applegate, G. E. & D. A. Gilmore, 1987. Operational expenses in forest management development in the hills of Nepal. ICIMOD Occasional Paper No 6. ICIMOD, Kathmandu. Awasthi, K.D. & Balla M.K., 2000. Degraded lands in mid hills of central Nepal: A GIS appraisal in quantifying and planning for sustainable rehabilitation. Institute of Forestry, Pokhara, Nepal. Bajracharya, D., 1983. Fuel, food or forest? Dilemma in a Nepali village. World Development 11, 1057-1073. Carson B., 1992. The land, the farmers and the future. A soil fertility management strategy for Nepal. ICIMOD Occasional Paper No.21, Kathmandu, Nepal. Carter, A. S. & Gilmour D.A., 1989. Increase in tree cover on private farmland in central Nepal. Mount. Res. Develop. 9, 111-134. Eckholm E.P., 1976. Loosing ground: Environmental stress and world food prospects. World Watch Institute and UNEP, Norton Company INC, New York. Enke, S., 1971. Projected cost and benefit of population control. Population and Development CIDA Seminar Paper No 2, Kathmandu, Nepal.17-22. FAO, 1994. The collection and analysis of land degradation data. Report of the expert consultation of the Asian Network for Soil Problem. Bangkok, Thailand. 166 Fox, J. M., 1993. Forest resources in a Nepali village in 1980-1990: The positive influence of population growth. Mount. Res. Developm. 13, 89-98. Ghimire & Upreti, 1997. Combating desertification: Report of the national seminar on desertification and land improvement, November 4 - 5 1997, MOPE-UNCCD Kathmandu Nepal. Gilmor, D.A., 1989. Forest resources and indigenous management in Nepal, Working Paper No.17, East West Center Honolulu, Hawaii. ICIMOD, 1999. Fragile mountain ecosystems: The Hindu Kush Himalayas. In ICIMOD home page, http://www.icimod.org.sg/hkh/mtn.htm. Ives, J. D. & B. Messerli, 1989. The Theory of Himalayan environmental degradation: What is the nature of the perceived crisis? Routeledge, New York. Kawakita, J., 1997. Nurturing in the Himalayas in purpose of an alternative development model based on participation and field science. The Institute of Himalayan Conservation Occasional Paper No.2, Yoyogi, Japan. PP 48-66. Keating, J.D.H., Qui, A., Wheeler, T.R., Subedi, M., Shah, P.B., Ellis, R.H., 1999. Annual legume species as green manures/cover crops in low-income farming systems of Nepal. Mount. Res. Developm. 119, 325-332. Koirala, H.L.1999. RS and GIS in assessing urban environment: A case study of Kathmandu Metropolitan City & Kathmandu Valley. Proceedings of the international conference Geoinformatics: Beyond 2000. Indian Institute of Remote Sensing, Dehradun, India, March 9-11, 1999. Land Resources Mapping Project, 1986. Land System Report, LRMP Report, Kentig Earth Science Ltd., Kathmandu. Mahat, T. B. S., 1987. Forestry - farming linkages in the mountains, ICIMOD Occasional Paper No 7, ICIMOD, Kathmandu. Metz, J.F., 1994. Forest product use at an upper elevation village in Nepal. J. Environ. Managem.18, 371-390. Mishra, S. B. & S. Bista, 1998. Soil erosion. Compendium on Environmental Statistics Nepal, HMG/N, National Planning Commission Secretariat, Central Bureau of Statistics, Kathmandu. Nelson, D., Labon, P., Shrestha, B.D., & Kandel, K.P., 1980. A reconnaissance inventory of major ecological land units and their watershed conditions in Nepal. IWMP/DSCWM/FAO, Project Field Document WP/17, Nep/74020, Kathmandu, Nepal. Pahari, K.J. , 1993. Soil erosion susceptibility using remote sensing and GIS: A case study of Andhikhola watershed Nepal. Dessertation (M.Sc.) NR 93-16. Asian Institute of Technology, Bangkok Thailand. Pandey S.P., Tamang, D.B. & Baidya, S.N., 1995. Soil fertility management and agricultural production issues with reference to middle mountain regions of Nepal. In: Scheier H., Shah, P.B. & Brown, S. (eds), Challenges in mountain resource management in Nepal: Processes trends and dynamics in middle mountain watersheds. Proc. workshop of the ICIMOD/IDEC/BC, Kathmandu, Nepal.41-49. 167 Paudel, G.S., 1997. Integration of forest and rangeland management for livestock development in the hills of Nepal. Dissertation (M.Sc.) NR 97-16, Asian Institute of Technology, Bangkok Thailand. Partap, T. & Watson, H.R., 1994. Sloping agricultural land technology (SALT). ICIMOD Occasional Paper No.23, Kathmandu, Nepal. Ruddle, K & Rondinelli, D.A. (1983). Transforming natural resources for human development: A resource systems framework for development policy. Resource Systems Theory and Methodology Series, No.1. United Nations University, Japan. Sah, B.P., 1996. Degradation and its scio economic impacts using RS and GIS: A case study of Trijuga watershed Nepal, Dissertation (M.Sc.) NR 96-20. Asian Institute of Technology, Bangkok, Thailand. Sah, B.P, Honda, K. & Murai, S. 1997. Sub watershed prioritisation for watershed management using remote Sensing and GIS. Proceedings, 18th ACRS conference, October 1997, Malaysia. Schreier H., Brown, S. & Shah, P. B., 1995. Identification of key resource issues: Discussions and recommendations. In: Scheier H., Shah, P.B. & Brown, S. (eds), Challenges in mountain resource management in Nepal: Processes trends and dynamics in middle mountain watersheds. Proc.wrkshop of the ICIMOD/IDEC/BC, Kathmandu, Nepal. 247-252. Serchan, D.P. & Gurung G.B., 1996. Sustainable soil management issues in eastern hills of Nepal. Proceeding of the Workshop: Fertility and Plant Nutrition Management, NARC, Lalitpur Nepal. 27-41. Shah, P.B., 1996. Soil fertility and erosion based unsustainability concerns in Nepal. Proc. of Workshop: Soil Fertility and Plant Nutrition Management, NARC, Lalitpur Nepal. 78-88. Shah, P.B. & Schreier, H., 1995. Maintaining soil fertility in agriculture and forestry. In: Scheier H., Shah, P.B. & Brown, S. (eds), Challenges in mountain resource management in Nepal: Processes trends and dynamics in middle mountain watersheds. Proc.workshop of the ICIMOD/IDEC/BC, Kathmandu, Nepal. 171-182. Shrestha, S. & Brown, S., 1995. Land use dynamics and intensification. In: Scheier H., Shah, P.B. & Brown, S. (eds), Challenges in mountain resource management in Nepal: Processes trends and dynamics in middle mountain watersheds. Proc.wrkshop of the ICIMOD/IDEC/BC, Kathmandu, Nepal. 129-140. Shrestha, S.S., 1996. Watershed planning for sustainable development of natural resources using remote sensing and GIS: A case study of Tinau watershed Palpa Nepal. Dissertation (M.Sc.) NR 96-21. Asian Institute of technology, Bangkok, Thailand. Shrestha, B.D., Ginnekh P.V., Sthapit. K.M., 1983. Watershed condition by district of Nepal, DSWC, Kathmandu Nepal. Shrestha, D.P. & Zinck, J.A., 1999. Land degradation assessment using GIS: A case study in the middle mountain region of the Nepalese Himalaya. Proceedings of the international conference:Geoinformatics-Beyond200. http://pages.hotbot.com/edu/geoinformatics/f64.htm. 168 Stevens, S.F., 1993. Tourism change and continuity in the Mount Everest Region of Nepal. J. Am. Geographical Soc. 83, 76-85. Thapa, G.B. & Weber, Karl E., 1993. Managing mountain watersheds: The Upper Pokhara Valley Nepal, HSD Monograph No. 22, AIT Bangkok, Thailand. Tiwari, D.N., 1990. Watershed modelling estimation of surface runoff and soil erosion rate: A case study of Nakkhu khola watershed Nepal. Dissertation (MSc) NR 90-1. Asian Institute of Technology, Bangkok, Thailand. Tripath, B.P., 1996. Present soil fertility research status and future research strategy in the western hills of Nepal. Proceeding of workshop: Soil Fertility and Plant Nutrition Management, NARC, Lalitpur Nepal. 42-60. Tripathi, B. P., Gardner, R., Mawdesley, K.J., Acharya, G. P., & Sah, R. P., 1999. Soil erosion and fertility losses in the western hills of Nepal: An overview. Lumle Seminar Paper No 99/9, Pokhara, Nepal. Vaidya A., Turton C., Joshi K.D. & Tuladhar J.K., 1995. A system analysis of soil fertility issues in the hills of Nepal: Implications for future research. In: Scheier H., Shah, P.B. & Brown, S. (eds), Challenges in mountain resource management in Nepal: Processes trends and dynamics in middle mountain watersheds. Proc.workshop of the CIMOD/IDEC/BC, Kathmandu, Nepal. 63-80. Wallace M.B., 1981. Solving common property resource problems: Deforestation in Nepal. Dissertation (Ph.D.) Harvard University, University Micro Film International Ann Harbour, Michigan, USA. World Bank, 1979. Nepal development performance and prospects: A World Bank Country Study, South Asia Regional office, World Bank, Washington DC. Acknowledgements This research is a part of research project ”Soil and forest degradation in marginalized agriculture of Nepal” funded by the Research Council of Norway (Project number 131692/730). Financial assistance by the Research Council is gratefully acknowledged. 169 Soil Stresses, Quality and Care: Concluding remarks from discussions in working groups and plenary sessions of NJF-Seminar no. 310 Susanne Elmholt1, Bo Stenberg2, Arne Grønlund3 and Visa Nuutinen4 1Department of Crop Physiology and Soil Science, Danish Institute of Agricultural Sciences, Research Centre Foulum, DK-8830 Tjele, DENMARK E-mail: [email protected] 2Department of Agriculture Research Skara, SLU, PO Box 234, SE-532 23 Skara, SWEDEN 3Centre for Soil and Environmental Research, Jordforsk, N-1432 Ås, NORWAY 4Agricultural Research Centre of Finland, Crops and Soil, FIN-31600 Jokioinen, FINLAND This NJF seminar on ‘Soil Stresses, Quality and Care’ brought together about 50 specialists and scientists with a wide array of specialities within agriculture sciences at large, and soil science in particular. The seminar focused on how to implement a soil quality concept into the consciousness of specialists, decision-makers and farmers in Scandinavia. Focus was also put on how to assess and evaluate the quality of soils and to identify key properties that may be used as indicators of soil health. The interaction between various factors was regarded as an important topic in order to increase our understanding about the role of soil as a link between agricultural production, human welfare and environmental concerns. Thus, the aim was to emphasise the interdisciplinarity within the soil quality concept and to address new approaches for monitoring soil quality and improving soil care. By the end of the seminar the plenary was divided into six working groups with the aim to discuss four predefined Soil Quality issues: • Soil quality for what? • Soil quality for whom? • Is there a basis for common Scandinavian approaches to soil quality? • How to assess and evaluate soil quality? In the following paper we have put together some of the main issues from the working groups and the following plenary discussions to give an impression of the views and concerns of the seminar participants regarding the concept of Soil Quality and its use. Definition of Soil Quality and Soil Degradation An often quoted definition of soil quality is ‘the capacity of a specific kind of soil to function within natural or managed ecosystem boundaries, to sustain plant and animal productivity, maintain or enhance water and air quality, and support human health and habitation’ (Karlen et al., 1997). The seminar discussions revealed that the soil quality concept - as this definition describes it - was new to many of the participants and that they rather conceived soil quality as synonymous to soil fertility. More simplified, soil quality can be defined as ‘the fitness of the soil to function’. The phrase capacity or fitness to function refers to the different functions 171 of the soil, which can be grouped in three ecological and three socio-economical and technical functions (Blum, 1998). The ecological functions are (1) production of biomass, ensuring food, fodder, renewable energy and raw materials; (2) filtering, buffering and transformation capacity between the atmosphere, the groundwater and the plant cover and (3) biological habitat and gene reserve. The socio-economical and technical functions are (4) a physical and spatial base for technical, industrial and socio-economic structures; (5) a source of raw materials, energy and water and (6) a geogenic and cultural heritage. Because of the many functions and uses of soil, there are many ideas and opinions of what a good quality soil is. For a farmer, high soil quality is associated with a high level of production and good yield quality, with low fertiliser input, especially in years with unfavourable weather conditions. From an environmental point of view, soil quality may signify its ecosystem functions, including maintenance or enhancement of biodiversity, water quality, nutrient cycling, and biomass production. Soil characteristics influencing soil quality can either be inherent (genoform) or dynamic (phenoform) (Droogers & Bouma, 1997). The inherent characteristics are determined by the basic soil forming factors (parent material, climate, time, topography, and vegetation) and are reflected in soil classification systems (soil series and soil types). The dynamic characteristics result from the long term effect of management of a specific soil type. The quality of a soil type will therefore also depend on land use and anthropogenic influences, in addition to climate conditions and position in the landscape. Thus each soil type, occurring in a certain climatic zone, has a characteristic range of production rates, risk and environmental effects. Bouma (2000) referred to this behaviour of soil as its ‘window of opportunities’. Soil degradation can be defined as a loss or reduction of the capacity of a soil to function, due to natural or human-induced processes. The impact of degradation processes on the soil quality depends on the soil’s resistance and resilience. Resistance is the ability of a soil to remain stable and withstand stress factors and prevent degradation, (e.g. erosion, compaction, acidification and nutrient depletion). Resilience is the soil’s ability to recover, to bounce or spring back after a perturbation caused by poor management or adverse weather conditions (Lal, 1998). Hence, soils of high quality should either have a high resistance or a high resilience. On a geological time-scale, the Scandinavian soils are young compared with most other soil in the world. The climatic conditions can be characterised as temperate and humid, which result in slow decomposition of organic matter. The soil organic matter content is therefore relatively high and a substantial part of the soils are classified as organic. Another consequence, however, is a slower and less complete decomposition of harmful organic substances including pesticides. The most severe processes of soil degradation are water 172 erosion, especially in hilly and undulating terrain, acidification, compaction due to heavy machines, and locally, pollution by heavy metals and pesticides. Common environmental side-effects are water pollution by nutrients and pesticide residues and greenhouse gas emission caused by drainage and mineralisation of organic soils. These facts will influence ‘the window of opportunities’ for the soils. A main challenge is to develop the concept of soil quality to ensure sustainable and productive agricultural soils with minimal detrimental impacts on the environment. Soil quality studies - for what and for whom do we need them? These two questions were posed to the working groups and especially the latter raised an intense discussion. Taken as a whole the groups and the plenary discussion had the following inputs as to why we need soil quality (SQ) studies: • SQ studies are interdisciplinary by nature. As such they enable the different and often very specialised scientific communities to elucidate interactions between different components of the soil ecosystem and they promote the communication between different groups of scientists. • SQ studies are needed as a tool in scientific studies to obtain a better understanding of soil processes. They can monitor changes in the soil ecosystem over time with the aim to improve soil management according to defined goals. • SQ studies can be used to rank soils with the aim to evaluate them for different land use. • SQ studies can be used to pinpoint vulnerable soils. • SQ studies shall help enabling farmers and decision makers • to meet the demand for sufficient supplies of high quality food and feeds • to protect the environment • to maintain or increase biodiversity in soil. • SQ studies can be used to predict consequences of uncontrolled land use politics. • SQ studies shall contribute to create sustainable management systems that guarantee productivity for future generations. The soil quality concept is global but it is important to realise that its implementation depends on a range of regional and local considerations (climate, geological setting, natural vegetation etc.). The soil quality concept can be adapted to a range of different land uses (natural resources, agriculture, forestry, recreation, habitation etc.). This implies that SQ studies will be highly goal-dependent, i.e. the approach depends on the demands of soil functioning by a 173 given land use purpose. There was a general agreement at the seminar that this is very important to bear in mind when deciding how to assess soil quality. The goal-settings will also depend very much on regional differences. As an example pointed to by several speakers at the seminar there must be different SQ goal-settings for the peat soils of northern Scandinavia and the fertile, arable soils of southern Scandinavia. The seminar participants agreed that SQ is an important issue but especially the plenary discussion revealed major frustrations regarding nature and use of the SQ concept. As one participant stated: 'The subject seems to slip out of hand all the time!'. One reason is that most scientists agree that we have to consider land use purpose in order to discuss meaningfully the SQ concept as outlined above. On the other hand – it was objected – that is exactly why the SQ concept and SQ studies are needed: To evaluate what kind of land use is appropriate in a given area. It should be borne in mind, however, that we as soil scientists are not to define land use in a given area. This is the task of decision makers. It was put forward that one reason why the subject 'slips out of hand' is that the term 'quality' needs to be defined. We need to put objective meanings into it before it can be communicated properly, both among scientists and to non-scientific communities. Soil quality studies are demanded by different 'target groups' as outlined below: • Soil scientists - who continuously explore and validate the scientific basis of the SQ concept and who seek means to communicate their understanding of soil functioning to non-scientists as well as to other scientists, outside their own, often narrow discipline. • Decision-makers - who are setting up targets for SQ in relation to different land uses. • Extension services and farmers - who receive recommendations for soil management improvement. • The general public - who needs information about the importance of SQ in relation to environmental protection. The seminar participants agreed on the target groups. However, there was a pronounced disagreement as to how to extend this knowledge to the non-scientific communities, i.e. politicians, farmers and the human society. Based on personal experiences from conflict areas in their daily work, several participants were bothered about presenting politicians with a number. Even though Karlen & Andrews (2000) tried to stress that they were illustrating a framework for assessment and not a fixed number, the use of scoring functions and SQ indices were of great concern. Well knowing (all too well!) that soil is a very complex medium, the use of a single number in the form of an index value may be misused by decision-makers. On the other hand, the scientific community is well aware that it has to 174 obtain some way to present SQ information in an operational manner to 'get the message through'. It is very important to enable decision-makers to conclude if there is something wrong, that we are e.g. approaching unwanted, irreversible changes in SQ, and in that case what to do. Recently there have been several examples that policies affecting land use (e.g. recycling of organic wastes, support for alternative farming systems) are being based on emotions rather than informed decisions. Therefore it is considered very important that the general public as well as decision-makers can obtain relevant information on SQ. It is also important to address decision-makers in relation to the whole food chain (from 'soil to table'). There was a general agreement that the SQ concept is important to farmers and to the agricultural extension service because it can be used to confer relevant messages on soil functions. However it was put forward that especially the farmers need more knowledge and inspiration on SQ topics rather than some more or less non-transparent index number. Concern was raised that scoring functions and SQ indices may simplify the information so much that the end users may actually loose the sense of their soil having a 'window of opportunities'. This window is not only linked to the genoform of the soil but very much also to its phenoform and therefore sensitive to management changes, resulting in e.g. soil erosion or compaction as discussed by Bouma (2000). The inclusion of the 'general public' as a target group for SQ information reflects the need to communicate that soil is an important part of the environment and not just something the farmers grow their crops in or something we use for habitation etc. This message should be possible to get through regarding the global environmental concerns of the public. The need to address the whole human society was also stressed by the invited speakers. They recommended to increase the focus on soil quality to the same level as long put to air and water quality, partly because soil quality is very important to both water and air quality (Bouma, 2000; Karlen & Andrews, 2000). Is there a basis for common Scandinavian approach to soil quality issues? This subject was taken up only on a very general level. An opinion was put forth that there is such a common basis owing to shared environmental, cultural and scientific background. It was further pointed out that among the Nordic soil science communities there is likely to be a relatively good agreement on the biological, physical and chemical variables important in soil quality evaluation. The themes covered by the seminar presentations also show that there are a number of common concerns relating to the sustainable management of arable soils in Scandinavia and the Baltic countries. It was suggested that some of the present long-term field experiments in the different countries could form a base for joint soil quality studies. 175 In the final discussion the participants were encouraged to address the need of a common, Nordic soil quality forum for discussions and exchange of methodological experiences and data. Also, the need for common research initiatives was discussed. About half of the participants expressed that in principle they were interested in common Nordic activities. During the discussion no concrete suggestions of joint research initiatives emerged. Doubts were also raised whether there would be resources and funding available for such activities. It was suggested that Section I of NJF could actively encourage general discussions on soil quality. Later workshops, dedicated to e.g. different soil quality indicators and practical implementation of the soil quality concept, could be organised. It was not regarded as necessary to initiate Nordic, e-mail discussion groups or homepages dedicated to soil quality issues. The availability of information on the characteristics of Nordic soils, their responses to different management and the location of most ‘vulnerable’ soils were shortly discussed. Although not covered by the seminar presentations, these matters have been addressed in a number of earlier activities. These include co-operation in Nordic soil mapping (Rasmussen et al., 1991) and a project on soil compaction (Håkansson et al., 1987) just to mention two. Important research has also been carried out under the Nordic council of ministers concerning e.g. an initiative of soil monitoring in Nordic countries and regionalisation of erosion and nutrient loss risks from agricultural land (Rekolainen & Leek 1996). The continuation of these types of activities is very valuable, as illustrated for instance by one common, important goal, which is shared by many member countries of NJF, namely the protection of the Baltic Sea. How to assess and evaluate soil quality? This question is the key issue towards a strategy with the potential to promote interdisciplinary communication about soil, to identify risk areas, to monitor changes for better soil management and to promote sustainable land use. It addresses a series of fundamental problems in soil science and soil monitoring, for example: • The complexity of the soil system. • The variability in time and space at all scales. • The influence of external forces such as the climate. In the presentations, in the working groups and in the plenary discussion it was generally recognised that - due to the complexity of soil - the development of a working strategy and a framework to assess soil quality requires an interdisciplinary approach. Although it was generally agreed that this approach should consider both the chemical, physical and biological components of soil, only a few of the presentations actually had this interdisciplinary focus. Another aspect of the complexity, which was put forward, is the need of integrating indicators of soil quality. That is, indicators which collate influences from several soil functions through their interactive properties, thus reducing the need for numerous specific measures. In the presentations dealing with methodological approaches to soil organic matter turnover, more specific microbiological processes, and soil structure and aggregation, it was shown that such 176 indicators exist. However, it was also recognised that such integrative indicators will conceal the actual characteristics of the quality of a given soil, thus limiting the possibilities to use the results of a study for recommendations towards a more sustainable management. This contradiction between economy and transparency of soil quality assessment could be overcome by a strategy including an additional step of analyses, targeted to resolve problems indicated by an initial, orienting round of analyses. In this first round indicators should be few, they have to be integrative but still dedicated to specific soil functions so that they can direct additional analyses to the relevant functions. The variability in time and space raised two questions. Firstly, can two soils be compared? And secondly, how large changes in a soil quality indicator could be regarded as natural fluctuations? These questions can be collated into the one question: what references should we use for the indication of a high or low soil quality? This is probably the key obstacle that has to be overcome before a working strategy to assess soil quality can be formulated. The problem includes the evaluation as well as the presentation of results. Is it possible to find universal optimum or threshold values for single variables? Or do they differ between different soil types, or even within a given soil type? In such case, how should the soil quality be presented without confusion? Could the variability be handled by using 'internal references' in an integrative evaluation of some sort, putting all indicators in relation to each other to reveal a functional structure of the soil? Suggestions in this direction were given by a few presentations that had used multivariate methods for simultaneous evaluation of numerous indicators. Another way suggested by some of the working groups could be to use a reference in time by monitoring i) specific sites over time to give an overview of the direction soil quality is taking, and ii) long term field experiments to study the effects of certain management practices. In both cases the number of sites has to be large enough to allow for interpolation over e.g. a region or a country. In this context the sampling strategy was also stressed as an area that needs a lot of consideration to give reliable data. The climate - as an example of an external force - has large influences on the soil processes as such, as they are often temperature and water dependent. These influences fall under the discussion above. However, at the same time the climate is an integrative part of many other soil quality indicators, such as texture and structure indicators as they regulate the water regime. The climate under which a certain soil is influenced will in part be decisive for the quality of the soil. That is, different climates suit different soils differently. The 'window of opportunities' changes with the climate. To conclude, there are many question marks which urge for interdisciplinary research before a reliable soil quality assessment strategy could be formed. However, it appears that there is no lack of potential indicators but rather a lack of knowledge in what these indicators really say and how they interact with each other. Therefore, field studies dedicated to the soil quality concept as a whole are required. Both traditional field experiments - preferably on long term - 177 to study the influences of management, and inventory screenings to increase our knowledge on the existing natural variation, are needed. Final comments Although soil quality in its broad and interdisciplinary sense was a new way of thinking to many participants, we found that the seminar discussions were vivid and came up with interesting ideas and thoughtful remarks. We would like to express our gratitude towards the keynote speakers, Johan Bouma and Douglas Karlen, for setting the scene and for their enthusiastic participation in the discussions. And we would like to thank the seminar participants for their interest and hard work during the workshop. It is our hope that the seminar experiences of the participants may contribute to or even initiate further soil quality work in Scandinavia. References Blum, W.E.H, 1998. Soil functions and key impacts on soil – an overiew. In: Kraemer, R.A., Hollerbuhl, S. & Labes, G. (eds), Soil protection policies within the European Union. Workshop 9-11 December 1998, Bonn, Germany. Federal Ministry for the Environment. Nature Conservation and Nuclear Safety. 63-70. Bouma, J., 2000. The land quality concept as a means to improve communications about soils. In: Elmholt, S., Stenberg, B., Grønlund, A. & Nuutinen, V. (eds), Soil Stresses, Quality and Care. Proceedings from NJF Seminar 310, 10-12 April 2000, Ås. DIAS Report 38, Danish Institute of Agricultural Sciences, Foulum. 1-14. Droogers, P. & Bouma, J., 1997. Soil survey input in exploratory modelling of sustainable soil management practices. Soil Sci. Soc. Amer. J. 61, 1704-1710. Håkansson, I., Voorhees, W.B., Elonen, P., Raghavan, G.S.V., Lowery, B., van Wijk, A.L.M., Rasmussen, K. & Riley, H. 1987. Effect of high axle load traffic on subsoil compaction and crop yield in humid regions with annual freezting. Soil & Tillage Res. 10, 259-268. Karlen, D.L. & Andrews, S.S., 2000. The soil quality concept. A tool for evaluating sustainability. In: Elmholt, S., Stenberg, B., Grønlund, A. & Nuutinen, V. (eds), Soil Stresses, Quality and Care. Proceedings from NJF Seminar 310, 10-12 April 2000, Ås. DIAS Report 38, Danish Institute of Agricultural Sciences, Foulum. 15-26. Karlen, D.L., Mausbach, M.J., Doran, J.W., Cline, R.G., Harris, R.F. & Schuman, G.E., 1997. Soil quality: A concept, definition, and framework for evaluation. Soil Sci. Soc. Am. J. 61, 4-10. Lal, R., 1998. Soil quality and sustainability. In: Lal, R., Blum, W.H., Valentine, S. & Steward, B.A. (eds), Methods for assessment of soil degradation. Advances in Soil Science, CRC Press, Boca Raton. 17-30. Rasmussen, K., Sippola, J., Urvas, L., Låg, J., Troedsson, T. & Wiberg, M., 1991. Soil map of Denmark, Finland, Norway and Sweden. Scale 1: 2 000 000. Landbruksforlaget, Oslo. 178 Rekolainen, S. & Leek, R. (eds.), 1996. Regionalisation of erosion and nitrate losses from agricultural land in Nordic countries. Nordic Council of Ministers. TemaNord No. 615. Copenhagen. 68 pp. Acknowledgements Douglas Karlen and Johan Bouma are gratefully acknowledged for valuable comments on the manuscript. Furthermore we would like to thank Mette K. Bjørnlund and Rita Bundgaard for technical assistance during the preparation of the proceedings. 179