RRRD032: Improving grazing management practices to enhance ground cover and reduce sediment loads

Scott N Wilkinson¹, Gary Bastin², Chris J Stokes³, Aaron A Hawdon⁴, Anne E Kinsey-Henderson⁴, D Mike Nicholas³, Vanessa Chewings², Brett N Abbott³, Karl McKellar⁵, Joseph Kemei⁴

¹ CSIRO Land and Water, Canberra
² CSIRO Ecosystem Sciences, Alice Springs
³ CSIRO Ecosystem Sciences, Townsville
⁴ CSIRO Land and Water, Townsville
⁵ Queensland DAFF, Charters Towers

Download the RRRD032 Research Outcomes Report3.48 MB

Executive Summary

Fine sediment, and attached nitrogen and phosphorus are impairing the ecological condition of the Great Barrier Reef (GBR). Three-quarters of fine sediment and large proportions of the total nitrogen and total phosphorus delivered to the GBR coast are derived from grazing lands, which occupy approximately 80% of GBR catchments. The principles of grazing management to minimise sediment loss are well known and extension packages are widely available. Intensive research at several sites has developed deep process understanding of the productivity and sediment loss outcomes of particular grazing practices. However, the practices which are appropriate for any given property across a region, and the effectiveness of those practices for improving water quality and enterprise profitability, are not well quantified or widely understood. This uncertainty can limit adoption of grazing practices which are sustainable in the long term. Spatial prioritisation is needed to determine where improved grazing land management is most needed and would be most effective.

It is recognised that historical grazing practices on individual properties have been influenced by a wide range of social and economic factors. This report investigates the long-term outcomes of a range of approaches to forage management, to help identify practice changes that can reduce the offsite impacts of the grazing industry and improve long term profitability.

This research used the Landsat-derived Ground Cover Index (GCI) of historical cover as a real-life, long-term experiment of the range of grazing practices which have been used on different properties in recent decades. A conceptual framework was developed which connects the environmental setting, grazing practices, forage productivity and utilisation, vegetation cover and sediment lost to water courses. The links within this framework were quantified using GCI data, hillslope plot measurements and process modelling. A key feature of this research was to relate the process understanding from previous site-based research to ground cover remote sensing data. The purpose of this approach was to deliver a regional-scale, ground-truthed assessment of the potential for improved grazing practices to increase forage productivity and reduce erosion.

Focus properties were selected having high, medium and low levels of historical ground cover. The existence and robustness of relationships between cattle stocking rates and water quality outcomes was tested on focus properties by assessing land condition, infiltration and soil movement at representative sites, and modelling erosion rates. This approach was replicated across three defined soil type regions between Charters Towers and Clermont, being Chromosol (red goldfields soil), Sodosol (Duplex 'spewy' soils with sodic subsoils), and Vertosol (dark cracking clay soil). The assessments at focus properties were also related to those at sites where previous research has developed detailed process understanding (e.g. Bartley et al., 2014).

Demonstrating productivity benefits from the grazing practices which can also reduce soil loss is important for achieving win-win outcomes in the context of a co-investment approach to facilitating grazing practice change in GBR catchments. Relationships between long-term stocking rates and forage productivity were tested by field measurement and grass production modelling. Broad scale assessments of the potential for practice change to improve pasture productivity were undertaken by modelling ground cover scenarios based on the focus properties.

The recently developed Dynamic Reference Cover Method was tested as a way to assess the relative differences in historical cover levels and grazing impact between properties across the soil type regions. This technique maps the deficit in historical grazed cover relative to nearby reference areas, isolating the effect of grazing from rainfall gradients. The robustness and limitations of cover deficit for property assessment across regions were established, and guidelines documented for ongoing application.

Long-term cover is broadly indicative of grazing land processes and function
This project has quantified how grazing land practices, processes and function can be related to cover remote sensing, including stocking rates, forage productivity, nitrogen availability, soil infiltration and soil stability. There was a large range in the long-term average of dry season ground cover between properties in each soil-type region, exceeding 20% over the 25 years of landsat-derived ground cover index data used in the project (1986–2010). During the three prolonged droughts in this period, the ranges in cover between properties exceeded 40%, due to lower coverage of perennial grasses including 3P species (perennial, productive and palatable) at low-cover properties. Contrasts in cover were evident between many adjacent properties, indicating that they were caused by contrasts in grazing practices.

Animal equivalent stocking rates were inversely correlated with historical cover levels, with low-cover properties having typically 2–4 times the stocking rates of high-cover properties. A High cover sites grew >2 times the forage biomass of low cover properties over the wetter than average 2011/12 wet season, and >3 times more over the drier 2012/13 wet season, in a grazing exclosure experiment. Standing dry matter within the paddocks (forage yield net of grazing) was up to 9 times higher at high cover sites. Medium cover properties had intermediate levels of forage production and standing biomass, between those on high and low-cover properties.

The resources required to grow forage (rain water and nutrients) were more efficiently retained on high cover sites than low cover sites, and this was consistent across soil types:

1. High cover properties had a much higher proportion of 3P grasses than the medium and low cover sites, irrespective of soil type. Land condition assessments were consistently higher and less variable on the high cover sites.
2. The soil infiltration capacity of high cover sites was measured to be 4 times that of low cover sites for both Chromosol and Sodosol soils, indicating that high cover sites could retain more water in the root zone of the soil profile for supporting forage production. Dominance of Indian Couch (Bothriochloa pertusa) in the pasture biomass, which is associated with historical over-grazing during droughts, was correlated with impaired infiltration capacity.
3. The forage Nitrogen pool available to support plant growth and produce protein for livestock consumption was measured to be almost two times larger at the high-cover Chromosol site, relative to the low-cover site. During the 'growing season' a smaller Nitrogen pool would constrain forage production even when soil water is not limited.
4. Hillslope soil movement at high cover sites was measured to be 85% lower than at low cover sites, which assists retaining particulate nutrients and organic matter to support forage production.

Cover deficit isolates grazing impacts on ground cover across regions
The difference between property ground cover measured by GCI and local reference area cover in dry years (the grazing-induced cover deficit) was used to isolate the effect of recent grazing on property ground cover. This was more reliable than GCI itself, which displayed regional gradients that were correlated with rainfall gradients. There was a large range in cover deficit between properties within each region, highlighting the potential for targeting extension to properties with high potential for improving ground cover. We found that cover deficit performs most reliably when compared between properties having the same predominant soil type and pasture species.

Cover deficit can also be used to evaluate trajectories of property grazing impact, by benchmarking against other properties of comparable pasture composition, or of historical cover deficit. An acid test of the effectiveness of practice change will be how well ground cover is retained in the next drought, compared to previous droughts and other properties.
However, remotely-sensed ground cover and derived products form just one indicator of change. The effectiveness of practice change for improving productivity and reducing erosion should also be confirmed with on-ground assessments. We found Patchkey and Landscape Function Analysis to provide reliable indicators for assessing land condition.

Grazing management is about more than just cover
While forage productivity and hydrologic function are related to historical cover levels over decades, grazing management in the shorter term must consider more than just ground cover. For example, the widespread dominance of the exotic grass Indian couch in degrading pastures can give rise to high cover but low productivity and poor soil infiltration capacity.

Reducing stocking rates across the half of properties in each soil type having medium and low cover levels would initiate a recovery cycle. Increases in forage biomass and rebuilding of perennial species pasture with high basal area will, in turn, help to improve soil stability and rebuild the nitrogen pool, and to improve soil infiltration capacity to retain rainfall onsite to support plant growth, in a self-reinforcing cycle. The potential magnitude of these responses will increase with the degree of stock reduction. Recovery can be expected to continue for one to several decades depending on the degrees of initial degradation and stocking reduction, although thresholds of state-change may inhibit the recovery process.

The improvements in forage composition, soil and nutrient properties that are required for forage productivity to recover following run-down in dry periods have not been fully accounted for in past grass production modelling, which may have over-estimated the recovery rate. We found that over-estimating the recovery rate leads to models over-estimating the safe stocking rates for sustaining land condition. If recovery occurs more slowly than previously assumed this also increases the value of conservative stocking during drought, because the long-term productivity losses associated with slow recovery will also be greater than previously assumed.

Maintaining high ground cover reduces erosion and increases productivity in the long term
There is a demonstrated capacity in the grazing industry to modify management practices when the benefits are widely understood. Many more graziers now recognise the importance of forage management, practice wet season spelling and vary stocking rates. This suggests that evidence-based reductions in stocking rates to provide productivity gains and reduce soil loss should be achievable.

Field measurement and modelling shows that forage production on properties with long-term high cover can be several times higher than that on long-term low cover properties. Beef production (kg/ha) also declines sharply above stocking rates which cause pasture composition to degrade. Stocking within the long-term carrying capacity should help maintain or improve forage production, recover perennial pastures, and increase soil water holding capacity and nitrogen availability to support animal protein production. Improving land condition in this way would reduce runoff volumes, and so reduce gully and streambank erosion rates also in the long term.

Property-scale modelling indicates that properties in the lower half of cover levels could be maintained at ground cover levels 10–20% higher during drought years then long-term hillslope soil losses would be 30–60% lower in the long term at paddock, property and regional scales. The magnitudes of erosion reduction were similar for properties with low and medium cover levels. Achieving these magnitudes of change will require reductions in stocking rates of approximately 50% for properties with low or declining historical cover, and somewhat smaller reductions for medium cover properties.

Supporting industry change
Across the grazing industry there is limited understanding of the linkages from forage utilisation to pasture composition, productivity and profitability. Quantitative approaches to forage budgeting and stocking rate decision-making are not yet commonplace. This research has provided quantitative data on the benefits of maintaining good land condition which has been of considerable interest to some graziers, and supports recent extension messages within the industry. Supporting practice change involves ongoing and focused extension activities with individual landholders, and the property cover deficit assessment developed by this project should assist in that process. Incorporating project findings into industry communication strategies, and simple but quantitative grazier tools such as photo standards for forage budgeting, should also assist practice change to improve the profile of grazing practices which both improve productivity and reduce sediment loss.

The priorities for public investment in grazing practice change to improve water quality at regional scale need to consider spatial variations in all erosion processes such as gully and streambank erosion, in addition to forage measures such as cover deficit. A framework for integrating forage management into water quality improvement programs is proposed in the Research Outcomes Report, Appendix 4.