Version 2, 30 november 1998

Ma arten Hens & Oscar Schoumans

In the autumn of 1998 this document was developed in order to coordinate the activities of one of the research topics of Working Group 2 of COST action 832 "Methodologies for estimation the agricultural contribution of eutrophication". The major purpose of WG2 is to develop a methodology to estimate P losses at field scale. At the first COST 832 meeting in Greenmount College, Antrim, N. Ireland, June 16-17, 1998, it was decided to group the WG2 activities into 4 topics:

Topic 1 deals with soil processes and three key objectives were identified for this topic. The activities regarding this ‘soil processes’ topic should be finished by May 1999, when a workshop is organized on ‘Soil Processes and P Loss’ together with WG 1 in Cordobá, Spain, May 13-15, 1999.

In order to initiate and further organize the WG 2 activities for the ‘soil processes’ topic, a small organizing meeting was held in Brussels on November 19^th. A first version of this document was discussed at this meeting. Based on the outcome of the discussions, the document could be refined. Furthermore, a balanced list of action points was compiled. An overview of these action points is included in this ‘version 2’ of the summarizing document.

Maarten Hens

Oscar Schoumans

 2.1. Role of soil processes within the overall WG 2 framework

 The objective of WG 2 is to develop a methodology to estimate P losses from agricultural soils at the field scale based on current scientific knowledge. The activities of WG 2 were organized according to four topics:
Topic 1 soil processes
Topic 2 erosion / surface runoff all at the field scale
Topic 3 leaching / subsurface flow
Topic 4 critical load and risk assessment

At the november ‘98 meeting in Brussels, the concepts and terminology of P transfer were discussed.

Simplified, P loss or P transfer from agricultural soils to receiving waters can be seen as a two step process: first, the transfer of any P-containing compound from the immobile phase (bulk soil) to the mobile phase (water) by e.g. desorption or detachment, followed by net movement of the carrying water phase. It was agreed that topic 1 ‘soil processes’ would only deal with the study of P transfer mechanisms between the mobile and immobile phase, irrespective of hydrological pathway. Once P is ‘loaded’ into a mobile water phase, several hydrological pathways can be followed. The most general subdivision differentiates between subsurface pathways (e.g. leaching and ‘vertical’ erosion/migration) and surface pathways (e.g. runoff and ‘horizontal erosion). According to this framework, the different WG 2 topics could be (re)defined as follows:
Topic 1 P transfer mechanisms between bulk soil and a mobile water phase
Topic 2 P loss by surface pathways at field scale
Topic 3 P loss by subsurface pathways at field scale
Topic 4 critical load and risk assessment at field scale

 2.2. P transfer between bulk soil and a mobile water phase

 In the first version of this document, it was attempted to subdivide P release and retention mechanisms into two broad categories. On one side, a wide range of P-specific soil chemical, physico-chemical and biochemical reactions can be identified controlling solution P intensities. On the other hand, a series of non P-specific soil physical mechanisms controls the concentration of P-containing particles in the passing water flow.

However, it was felt that this approach ‘masks’ or ‘ignores’ the gray zone of colloidal P release mechanisms. The scheme below presents an attempt to organize P release and retention mechanisms according to (classical) dissolved-colloidal-particulate continuum. Two general terms, solubilization and erosion have been introduced to describe the release of ‘dissolved’ and ‘particulate’ compounds respectively. Again, the use of two terms is not fully able to reflect colloidal behavior.

 It is important to realize that any of these release mechanisms can occur irrespective of the hydrological pathway considered. Thus, this scheme can be used to tackle ‘subsurface transport of eroded particles’ and ‘surface transport of solubilized P compounds’.

It should be noted that e.g. the term ‘erosion’ is defined above purely as the transfer of soil colloids and particles to a carrying water flow. Usually, the term ‘erosion’ is linked to a hydrological pathway, e.g. surface transport (overland flow, ...). Similarly, it was felt that the widely used term ‘leaching’ did not fit well in the above picture, as it combines solubilization and (subsurface) transport processes. Furthermore, the term leaching usually does not include subsurface transport of particles. For this combination of process and pathway the term ‘migration’ is used sometimes.

Field-scale variability of the P release/retention mechanisms is governed by the spatial variability of the resp. controlling soil parameters (e.g. texture, organic matter, P sorption capacity, pH, temperature, ...). Some of these parameters are time-constant, others like pH, Eh and soil solution ionic strength are more temporally variable. Within topic 1, it is likely that this aspect ‘at field scale’ will not receive that much attention. On the other hand, the temporal and spatial variability at the field scale becomes important when release mechanisms are combined with hydrological pathways and P loss from field scale is considered (activities of topics 2 and 3 within WG2).

An overview of nomenclature regarding hydrological pathways and P loss mechanisms was presented by Phil Haygarth at the WG 2 Antrim meeting and can be found at http://www.ab.dlo.nl/eu/cost832/welcome.html (click on: Meetings, past Cost Meetings, WG2, Haygarth)

Comparisons between soil P mobility studies are severely hampered by differences in analytical methodology and terminology for solution P fractions. On the Antrim WG 2 meeting the need for a standardized analytical methodology and nomenclature was formulated. Reviews of operationally defined P fractions can be found in Broberg and Persson (1988) and Robards (1994). On the P Workshop in Wexford (september 1995), A. Sharpley and H. Pionke also proposed a list of "unified" P terminology (see Johnston et al., 1997). Here, we briefly summarize fractionation procedures for solution P and propose a scheme for fraction terminology.

P fractionation procedures for natural water samples are built up along two fractionation axes: a size fractionation axis and a chemical fractionation axis.

Physical or size fractionation reflects the particulate - colloidal - dissolved concept. As a standard procedure in both limnology,oceanogrpahy and soil science, 0.45-µm membrane filtration is used to distinguish between particulate and dissolved material. If colloid-sized material is present, part of it will in general pass 0.45-µm filters and be present in the "dissolved" phase. The occurrence of P containing colloids in soil solutions and soil leachates was demonstrated by a.o. Gerke (1992), Haygarth et al. (1997) and Sinaj et al. (1998).

Chemical fractionation reflects the different P-bonds occurring in solution. As a standard procedure, molybdate reactive phosphorus and total phosphorus are determined on (filtered) soil waters. Further chemical fractionation may involve the use of UV photo-oxidation or acid hydrolysis to distinguish between organic phosphates (P-O-C and P-C bonds) and condensed phosphates (P-O-P bonds).

The terminology given below is based on "conventional" size fractionation with 0.45 µm filters. Hence, a terminology for the truly "dissolved" species (e.g. determined after ultrafiltration at 10 kD) is not included.

Therefore, it would be more appropriate to mention explicitely the size fractionation used, rather than to keep the letter D(issolved) open for interpretation¼ In such a scheme, MRP (molybdate reactive P) is a general term used to indicate a chemical fraction only, without size specification. One would refer to the MRP content of a 0.45-µm filtrate simply as MRP (<0.45µm).

The difference between total P and molybdate reactive P is often termed "organic P" instead of "unreactive" P. The term organic neglects the potential presence of condensed phosphate species in solution. "True" organic phosphorus can be determined after UV photo-oxidation of the sample as the difference of MRP before and after UV-photooxidation.

Simplified, P loss rates are P concentration (mass volume^-1) ´ flow (volume time^-1). Our topic "Soil processes", as defined in the previous chapter, only deals with prediction of concentrations which can be mobilised and transported by a given hydrological flow. The terms 'solubilization' and 'erosion' were introduced to cover the wide range of P release mechanisms.

The main objective with regard to the joint WG 1 and WG 2 meeting in may 1999 (Cordoba, 13-15 May, Spain) is to review the state-of-the-art of both P release mechanisms. This review exercise consists of

In general, prediction of solution P concentrations should take into account the entire soil P cycle. For recent reviews of the soil P cycle, see e.g. Frossard et al. (1995), Magid et al. (1995). Studies of subsurface transport of P have mainly focused on orthophosphate or MRP losses from soil, while studies looking at surface transfer of P mainly deal with the total and bioavailable amounts of P exported from given area. The mobilization of other P compounds, such as organic P and P containing colloids is much less understood. However, both types of P species may be much more mobile than free orthophosphate, since the negatively charged orthophosphate ion is protected from sorption by covalent and/or electrostatic bonding to carbon or metal ions in the resp. molecules / colloids. Current knowledge gaps include:

Therefore, reviews should:

In general, prediction of solution P concentrations should take into account the entire soil P cycle. For recent reviews of the soil P cycle, see e.g. Frossard et al. (1995) and Magid et al. (1995).

Much effort has been put on developing simplified relationships between soil solution P concentrations and soil P status (see also "environmental soil P tests"). In general, these models (only) deal with sorption and desorption of orthophosphate in a soil-water system. A good example of this approach is the Dutch protocol on P-saturated soils (Van der Zee et al. 1987), in which the concepts "P sorption capacity" and "degree of P saturation" were defined for acid sandy soils. This protocol has been used extensively in a.o. the Netherlands, Belgium, Northern Germany to map the P saturation parameters from field to regional scales.

Mechanisms of both organic P and colloidal P release are not well understood. Most authors stress the role of microbial activity, while other papers have focussed on sorption/desorption behaviour of natural organic matter and/or selected organic P compounds.

Reviews should summarize current concepts and process descriptions of soil chemical, physico-chemical and biochemical reactions ('solubilization') as well (P-related?) mechanisms for erosion of (colloidal) soil particles. Moreover, given the knowledge gaps with regard to mobilization/erosion of non-orthophosphate P-forms, they should also address:

The need for an environmental soil P test is intimately linked with the establishment of upper critical limits for P in soils. These limits should take into account the soil P concentration for optimal crop production and the soil P concentration which "produces" environmentally sound (sustainable) P losses. Sibbesen and Sharpley (1997) give a good overview of the philosophy behind environmental soil P testing.

An environmental soil P test only looks at the soil compartment. It should be kept in mind that the risk for P loss is a combination of the soil P risk and the hydrological risk (topic 4 of WG 2; risk assessment).

In agreement with the availability concept (available P = all P that CAN be taken up be plant roots), one can attempt to define mobile P as all P that CAN be lost from the soil. The environmental impact will be determined by the loss rate, (combination of P concentration and composition of mobile phase and water flow leaving the soil or field compartment by different pathways). The usefulness of this concept is clear for leaching, much less clear for erosion.

Environmental soil P test values can be derived from (generalised) intensity-quantity relationships. The intensity corresponds to the activity of P in soil solution and is often determined by a water or CaCl₂ solution extract. The quantity factor refers to the plant-available P reserve and is determined by chemical extraction methods, by anion exchange resin extractions, or by isotopic dilution (³²P) techniques. The capacity factor is generally related to the phosphate buffering power and phosphate fixing capacity of the soil. The mechanism by which the intensity, quantity , and capacity factors are regulated is particularly influenced by soil properties such as amorphous Al and Fe oxides, soil pH, clay or organic matter contents, and by the prevailing form of soil P.

The most straightforward way of establishing environmental soil P test value is to study quantity - intensity relationships on individual soil samples. However, the P intensity (P concentration and composition in the mobile phase) of a topsoil is not directly related to the concentrations and/or amounts of P lost from the pedon or field (cfr. sorption in subsoil, ¼ ). Therefore, attempts have been made to correlate "conventional" soil P quantities (Olsen-P, Mehlich-P, ¼ ) of the topsoil to P concentrations in runoff and leachate waters (e.g. Sibbesen and Sharpley, 1997).

Again, most of the work done only considers sorption/desorption of orthophosphate and "neglects" other P forms and release mechanisms. The above description reflects some of the conceptual problems with environmental soil P testing:

Should we include organic and colloidal P forms in the intensity measurements?

Point-approach (mechanistic - relating intensity and quantity for individual soil samples) or spatial approach (correlation - relating intensities in water flows leaving the system to an averaged soil P quantity)

Does a similar concept exists for erosion? Are environmental soil P tests useful anyway?

Actions for the may 1999 workshop include:

At the small organizing meeting in Brussels, a list of "action co-ordinators" was composed to elaborate on the different action points. The task list was organized according to the "solubulization" and "erosion" mechanism categories.

1. Concepts, definitions and terminology M. Hens, O. Schoumans

2.1 Review experimental data (field/lab studies) S. Sinaj, B. Turner, M. Hens
relative importance of different P-forms
influence of soil type and land-use

2.2. Review concepts and process descriptions

2.2.1. "chemical" mechanisms W. Chardon, O. Schoumans

2.2.2. "biological" mechanisms F. Gil-Sotres, B. Turner

2.3. Environmental soil P test J. Vanderdeelen, M. Hens, O. Schoumans

3.1. Review experimental data (field/lab studies) E. Barberis, P. Withers

3.2. Review concepts and process descriptions P. Withers/J. Quinton

3.3. Environmental soil P test P. Withers/J. Quinton,

H. Hartikainen

Development of a methodology to estimate P losses O. Schoumans/M. Hens

at field scale: I) the influence of soil P processes

General reviews of P losses from agricultural soils to aquatic ecosystems include Sharpley et al. (1995) and a series of articles in Tunney et al. (1997).