Green House Gas Migration - 6 - Potential management practices and technologies for reducing carbon dioxide emissions from agriculture

Summary
6.1 - Introduction
6.2 - Overview of currently operational options
6.3 - Mitigation options relevant to New Zealand pastoral systems
6.4 - Conclusion

6 - Potential management practices and technologies for reducing carbon dioxide emissions from agriculture

Summary

The main findings on carbon dioxide mitigation options are:

Carbon dioxide emissions are currently not reported from the agriculture but are included within the energy sector and land use change and forestry national inventory calculations. However, agricultural management practices (e.g. re-seeding, grazing management) have implications for soil C sequestration.
Managing grazing land to increase carbon storage requires a larger proportion of the carbon fixed in photosynthesis to be returned to the soil. This is not economically viable as it means reduced product output relative to inputs.
Increased productivity per unit area (intensification) can be achieved without any cost in C sequestered as long as either the inputs (fertiliser, irrigation) on that area are also increased or the increased utilization of one area is matched by reduced utilization on other areas of the farm or region.
Biofuels have the greatest potential for using agricultural land to mitigate greenhouse gas emissions because they can sequester soil C and substitute for fossil fuels.
Energy crops have been investigated in New Zealand in the past but new advances in cellulose conversion technology and the current emphasis on greenhouse gas emissions means that a in depth re-evaluation is required.
Using 12% of our pastoral land to grow herbaceous feedstocks for bioethanol production would provide equivalent energy to that derived from our current total petrol usage.
The current cost of imported petrol is 67 c/l (excluding tax and levies). The costs of bioethanol production are currently 70 c/l but will fall with improvements in feedstocks and cellulose conversion technology.

6.1 - Introduction

Three main options have been identified by IPCC for the mitigation of CO₂ emissions from agriculture (Watson et al 1996):

Reduction of agricultural-related CO₂ emissions
Production of biofuels to replace fossil fuels
Creation and strengthening of carbon (C) sinks in the soil

While reduction of CO₂ emissions is not technically relevant to New Zealand agriculture as no CO₂ emissions are currently reported as occurring from this sector (fuel emissions from agricultural machinery being considered under the transport sector) (MfE 2000), issues of intensification e.g. increased use of irrigation or increased re-seeding may modify fuel use. Currently agriculture contributes only 8% of CO₂ emissions from liquid fuels (Ministry of Economic Development 2001).

Biofuels, the substitution of fossil fuels by fuels from renewable sources, have considerable potential for reducing emissions (e.g. Smith et al. 2000) because of both fuel substitution and the introduction of a perennial crop with potential gains in soil C sequestration. The IPCC report on Land Use, Land-Use Change and Forestry' (Watson et al 2000) concludes that "From a policy perspective, the potential for biofuel displacement of fossil fuel is an order of magnitude greater than any other land-use change. It may also impact atmospheric carbon levels earlier and at a lower cost than any other energy sector measures.". Current research in New Zealand is centred on woody biofuel, but international research also includes herbaceous species (perennial grasses).

International research on the management of agricultural sinks has concentrated heavily on arable crops, in particular tillage and fertiliser options. While this of considerable interest in areas such as the USA (see US Department of Energy C sequestration programme http://csite.esd.ornl.gov ), the potential for C sequestration in cropland in New Zealand is limited by the size of the cropped area (0.3 M ha compared to pasture at 13.2 M ha) and the resulting small size of the total C stock (28 Tg C cf. 1785 Tg C for pasture) (Baisden et al. 2001). Consequently, the sequestration options considered in this report will concentrate on grazed pasture.

6.2 - Overview of currently operational options

Bioenergy crops have potential for C sequestration (particularly if they replace annual crops or are planted on degraded land) but more importantly, they provide a renewable fuel source that can directly substitute for fossil fuels.

Bioenergy crops can be used in two ways:

as a solid fuel being combusted alone or in a co-combustion process with coal or,
after conversion processes, as a source of liquid fuel.

Almost any kind of agricultural waste can be used for combustion. Research in New Zealand has concentrated on wood products (short rotation coppicing as well as use of wood products from plantation forestry such as thinnings) but overseas, in particular the US and Europe, herbaceous species are being developed for this purpose. Liquid fuel production has been confined to production of ethanol from starch. The problem with this technology is that it is energy-intensive and the resulting fuel is only cost-effective in relation to petroleum when heavily subsidised. Recent developments in cellulosic conversion have the potential to change the efficiency of ethanol production from biomass and produce fuel that can compete directly with petrol. If the additional benefits in reduction of greenhouse gas emissions are considered it is evident why the US Department of Energy is investing heavily in this technology. As New Zealand data for woody biofuels is well documented this report will consider the potential of herbaceous species and in particular their potential for bioethanol production.

C sequestration in ecosystems occurs when C entering the system through gross primary production (photosynthesis) is greater than the C leaving the system through plant and heterotrophic respiration, lateral transfers, leaching and harvest. While pasture can achieve rates of gross photosynthesis and net primary productivity (NPP) equivalent to dense forests (Goudriaan 1992), the ephemeral nature of leaves in pastures means that the fixed above ground C is turned over rapidly rather than accumulating as standing biomass as in forests. A consequence of this rapid turnover is a large flux of C is directed to the soil sink, and, indeed, soils under grasslands contain large amounts of C in relation to other ecosystems on a per unit area basis (Goudriaan 1992; Roy et al. 2001). In pastures a key factor controlling the amount of NPP returned to the soil is the amount of NPP removed by the grazing animal. Consequently, management of C sequestration in grasslands is more complex than in non-grazed ecosystems but there is, perhaps, greater scope for manipulation.

6.3 - Mitigation options relevant to New Zealand pastoral systems

6.3.1 - Bioenergy

As the technology for cellulose conversion develops, the potential for biofuel crops grown specifically for this purpose increases. Rapid advances in cellulose treatment are being made due to major research programmes in the US. In Canada, Petro-Canada (the second-largest petroleum refining and marketing company in Canada) are co-funding research with the Canadian government on biofuel to ethanol conversion. In Europe a research consortium is funded from EU monies (see the European Energy Crops site at http://www.eeci.net ). Further contributions can be seen on the International Energy Authority site relating to bioenergy. Considerable attention is being paid to the biofuel potential of perennial, rhizomatous grasses such as Panicum virgatum L (switchgrass) and Miscanthus sinensis x giganteus (Japanese elephant grass).

Because of the costs of growth, conversion etc. there are energy costs to be factored in when considering the reduction in greenhouse gases associated with biofuels. A current estimate for starch-based conversion systems indicates that for each litre of petrol substituted for by a litre of ethanol there is a saving of 35-46% in CO₂ emissions. However, estimates for the mitigation potential of cellulosic conversion are far higher, in the range 85-130% Bioethanol can yield greater than 100% reduction because the co-product of the conversion process is lignin that can be burned as 'green-coal'.

Current values for herbaceous crops from Oak Ridge National Laboratory, USA suggest a crop producing 10 tonne dm per hectare will yield approximately 3.3 tonne of bioethanol per hectare. If 12% of our pastoral land was used for bioethanol production this would provide the same energy as our current total petrol consumption (i.e. 2.25 Mt petrol or 3.36 Mt of bioethanol - bioethanol having a lower energy value). The cost of production using current technology (as employed by Iogen and Petro-Canada) for conversion from herbaceous feedstocks (woody feedstock conversion is more costly) is estimated to be 70 c/l with a feedstock cost of $57 per tonne (http://www.pyr.ec.gc.ca/ep/wet/section16.html ). The current cost of petrol before taxes and levies is approximately 67 c/l (includes Singapore wholesale price plus international freight and importer margin, http://www.med.govt.nz/ers/oil_pet/prices/fuelprices.pdf ). Advances in feedstock through plant breeding and selection, and advances in cellulosic conversion are expected to bring these costs down substantially. The OECD estimated costs of ethanol production will decrease 25-50% over the next 10 years (OECD 2000) and the National Renewable Energy Laboratory calculates 50% in the next 15 years (Wooley et al 1999).

Energy crops were investigated in New Zealand in the 1970's and early 1980's. The view then, that has been recently restated by the Energy Efficiency and Conservation Authority (EECA 1997) is that bioenergy crops are not a good option in New Zealand. The basis for this conclusion is that:

energy crops compete with food and fibre crops for good arable land and there is no surplus of such land here,
the costs of production and conversion are too high to make biofuels competitive with fossil fuels.

However, the new advances in cellulosic conversion require a reassessment of this conclusion. First, it will be possible to use herbaceous species for bioethanol production opening the possibility of integrating perennial grasses into pastoral farm systems (i.e. not removing high quality arable land from food production). Second, the efficiency of conversion will ensure that bioethanol is directly competitive with fossil fuels even without the additional benefit of being a renewable fuel with the consequent greenhouse gas emission advantages.

Detailed analysis of herbaceous bioenergy options is outside the scope of this report but further investigation is proposed. This should include the identification of desirable attributes for bioenergy species and examination of some potential candidates - including a range of native species; assessment of current research in cellulosic conversion, the role that New Zealand expertise in rumen function might play in this area, and an analysis of how bioenergy cropping might be included in a pastoral farming enterprise.

6.3.2 - Increased C sequestration by grazed pasture

International literature on C sequestration in grasslands (see review of Conant et al. 2001 and Watson et al 2000) is heavily weighted towards studies involving management of unimproved grasslands. In this situation, it is frequently the case that improvements, such as reseeding with more productive species, fertiliser, irrigation and appropriate stocking rates, result in increased soil C. In New Zealand, management effects have to be considered against a background of largely improved pasture. In this situation there is a fine balance between C fixed and C removed or utilized, the outcome depending on a number of complex interacting processes involving plants, animals and decomposers. Ecosystem models are an essential tool in unravelling this complexity and providing a theoretical foundation on which to base mitigation strategies.

Figure 6.1 The relationship between a) gross and net aboveground primary productivity and b) the consequences of herbivory for the return of C (for more detailed description see Parsons & Chapman 2000).

A great deal is known of the impact of many 'management' variables on the flows of C in grassland, including flows to the soil (e.g. Thornley 1998; Parsons et al 2000). Management can have a substantial impact on C fluxes in that it controls biomass and so the source of C in photosynthesis, as well as the subsequent fate of C, as outlined below.

The first principle to recognize is that:

Any factors that increase the specific rate of growth, or allow a greater biomass to be sustained, increase the flow of C to soil.

The grassland canopy contains a high proportion of leaf (photosynthetic) tissues. This and its high leaf area index (and leaf area duration, LAI integrated over time) allow rates of canopy gross and net photosynthesis equivalent to dense forest (Goudriaan 1992). However, the leaves are highly ephemeral (lifespan c. 30 days) so that the high rates of photosynthetic C flux are not reflected in a sustained large biomass but constitute a substantial flow of C to soils. In the absence of harvesting the C gained in net primary productivity is available for return. How much of this C becomes stored in the ecosystem then depends on decomposition processes. As NPP (gross productivity minus the C lost in respiration) increases then the amount of C returned also increases (Figure 6.1a). We can imagine such an effect occurring as a result of increased use of fertilizer or irrigation.

The second important principle recognizes the part played by grazing in the system:

Any factors that aim to increase the intensity of harvesting by animals ('utilization') of the plant material will act to decrease the flow of C in two ways:

First, harvesting removes photosynthetic material during its brief ephemeral life thus reducing the standing biomass and the rate of biomass production. This is the case as we move to the left of the graph in Figure 6.1b. Second, not all the C removed by the animal is then returned to the ecosystem. At intermediate levels of grazing intensity variable amounts of C are removed by the animals (Figure 6.1b). In percentage terms, approximately 30% of the C ingested is returned in dung and urine. Less than 10% is removed as animal product, while 60% or more is lost to the atmosphere either as methane (5%) or as respired CO2 (55%) (data in Crush et al. 1992). As well as having a profound effect on the quantity of C returned, animals also strongly modify the C/N ratio of the material entering the decomposer pathway as 60% or more of the C in herbage ingested is lost to the atmosphere whereas almost none of the nitrogen is lost.

These principles can be seen in operation using an ecosystem model to study the consequences of specific management practices. Fig 6.2a shows increased C sequestration in response to fertiliser in accord with principle 1 and Fig 6.2b shows a reduction of C stored as stocking rate increases and a greater proportion of NPP is removed by animals as expected from principle 2.

It should be noted that well-validated, mechanistic models are the only tool to explore potential management combinations for their effects on C sequestration. Empirical relationships or field data do not have the generality to enable us to make robust predictions when we want to consider a range of potential management options, some of which might lie outside our current experience. Although models such as the Hurley Pasture Model and Century have been extensively evaluated and are based on sound, well-tested processes, it is reassuring to compare model behaviour with field data. In fact, there are few data that have measured management effects on C sequestration. Although models such as the Hurley Pasture Model and Century have been extensively evaluated and are based on sound, well-tested processes, it is reassuring to compare model behaviour with field data. In fact, there are few data that have measured management effects on C sequestration. However, recent data from Lambert et al. (2000) can be reanalysed to extract this information (Fig 6.3). The data show the change in C in the soil organic matter over a 15 year period on hill country pasture in the North Island. Two fertiliser treatments were compared and stocking rates were adjusted to utilise the increased herbage production that developed over time. The data show that greater fertiliser inputs result in increased C sequestered cf. low and high fertiliser at any particular stocking rate.

Fugure 6.2

Figure 6.3

As stocking rate increased the amount of C sequestered was reduced; however, the stocking rate effect could be ameliorated by greater fertiliser inputs. The loss of C per stock unit calculated from the Lambert data was 0.34 tonnes C per hectare for high fertility and 0.8 tonnes per hectare at low fertility. On average, C was being lost from these systems at a rate of 0.2 tonnes C per hectare per year.

Further supporting evidence for the critical role of grazing in determining C inputs and ultimately C in the soil is provided by Hoglund (1985). In a 3 year trial on ryegrass/white clover pasture near Kirwee in Canterbury, Hoglund varied grazing intensity by sheep and measured the consequences for soil N and C pools. The range of residual herbage left after grazing was from 470 to 1100 kg dry matter/ha. The change in soil C was linearly related to residual mass (Fig 6.4) with the change in soil C in the top 100 mm being 1020 kg C/ha for each 100kg change in residual dry matter. Over the full sampled depth the difference between the most severe and most lax grazing was 6.5 t C.

Figure 6.4

We can make a theoretical calculation to assess the relative importance of changes in soil C in relation to other greenhouse emissions. Assume a pasture with a stocking rate of 20 sheep/ha with a total intake of 10t organic matter (or approximately 4t C/ha), then each sheep has an intake of 4/20 or 0.2t C. Approximately 5% of intake is lost as methane (Crush et al. 1992) so the change in C (loss or gain) per sheep/ha/yr is 0.01t C. If we now assume that there is a potential rate of change in soil C of 20t spread over 100 years (i.e. 0.2t C per year, see Lambert et al. 2000) this means that one sheep can effect the C change in the soil at a rate of 20/(20*100) or 0.01tC sheep/ha/yr i.e. the change in emissions that would follow from a change in stocking rate could be of the same order of magnitude for the soil C effect as for the methane effect.

The question of steady state

Saggar et al. (2001) have argued that soil C in New Zealand pastures has reached a steady state; the proviso being that this is under conditions of constant management as 'management practices always undergo incremental evolution ... and these factors can cause a shift in the eventual steady state' (Saggar et al. 2001). These views are consistent with those expressed above as both recognise the important role of management in determining soil C levels. Saggar et al. (2001) provide data on soil C from long-term fertiliser trials and show no effect of fertiliser on soil C despite increased pasture growth. The implication here is that the soil is at steady state and therefore any increased input of C is matched by increased C mineralization - other interpretations are possible. It is important to recognise that an increase in pasture growth (as measured for example by pre-grazing cuts) does not necessarily mean greater C inputs as the extra pasture growth could be removed by animals if stocking rate is adjusted to match the higher feed supply. The experiments quoted by Saggar et al. (2001) are only a test of fertiliser effects on C storage if stocking rate - or more correctly utilization - was held constant. Saggar et al. (2001) also discuss the data of Lambert et al. (2000) and describe this as an example of an experiment in which fertiliser application resulted in a loss of soil C. Saggar has previously described this experiment as one in which 'the soils differed only in SSP (single superphosphate) fertilizer application' (Saggar et al. 2000). As we have seen in the previous section, stocking rate was doubled in the low fertility treatment and trebled in the high fertility treatment over the course of the trial. Figure 6.3 shows the loss in soil C can be explained by the increased utilization and that this deleterious effect on soil carbon can be ameliorated by increased fertiliser application. These results are consistent with the principles already stated.

Two questions need addressing:

is management of C sequestration a practical option for managing greenhouse gas emissions?
has 'intensification' since 1990 resulted in changes in the C sequestered in New Zealand's pastoral soils?

First, intensification, defined as increased productivity per unit area of land, does not necessarily result in reduced C storage. This is only the outcome if utilization increases i.e. the increased productivity runs ahead of increases in inputs. Even if increased utilization occurs on part of an enterprise this is not to say that the total C balance will alter because the concentration on production in a part of the system might have resulted in de-intensification on other parts. This argument can be applied at a regional and national scale. Managing C storage in the face of intensification will require careful analysis of the spatial distribution of changes and the tradeoffs between greater productivity on some areas and reduced productivity on other areas. Such an analysis would also allow for inclusion of erosion effects on C fluxes. A recent OECD study (OECD 2001) notes that at least 26% of New Zealand's land area is at risk from soil slippage and 39% is affected by wind and water erosion. There are potentially significant losses of C on eroded land and changes in stocking intensity on 'at risk' areas need to be factored into the intensification/de-intensification tradeoff.

Second is the question of whether national agricultural soil C stocks have changed over the past decade as New Zealand's agriculture has become more 'intensified'. As we have seen in the previous sections, we need to carefully define intensification, particularly in respect of increased productivity on a per unit area basis being matched by an increase in inputs. It would be impracticable to investigate this balance of inputs and removals over the last decade on a farm-by-farm basis. One approach to this question would be to consider the balance of inputs and outputs at the national scale using an estimate of products removed as the outputs (taking account of efficiencies derived from livestock improvements) and fertiliser use (and irrigation) as the inputs. Energy could be used as a common currency for this purpose and this would have the additional advantage of matching the currency used for methane inventory purposes.

6.3.3 - Botanical composition

In this section we consider the potential options in relation to grassland species rather than comparisons with species suitable for different land uses such as trees. Introduction of legumes and improved grasses into unimproved pastures can markedly increase C in the soil (Conant et al. 2001) but differences between cultivars and species in improved pastures in their C sequestering capacity have not been researched extensively.

Greater C sequestration should result from selecting a species that can produce large C inputs of a low degradability. Low degradability i.e. production of recalcitrant C, can occur under cereal crops and in some C4 pastures in which the C/N ratio of the litter is wide. For example in some tropical pastures the C/N can be four to five times higher than in C3 pasture (Wilson et al 1986). Unfortunately, one consequence of a higher C/N can be immobilisation of N (see modelled consequences in Thornley and Verbene 1989) resulting in a negative feedback on plant growth and subsequent reduction in the mass of litter produced. Legumes might be capable of reducing the negative N feedback but, in general, legume litter is of high quality and therefore does not possess the desirable low degradability. One possibility is to evaluate the C sequestration potential of legumes that are high in tannins as tannins can protect litter from rapid decomposition. Very high values for root mass have been recorded for Lotus uliginosus stands e.g. a fine root mass in the 0-5 mm layer of greater than 19 t dm was measured by A. Nordemeyer (personal communication) 14 years after planting into a felled Pinus contorta stand but little quantitative data are available for other situations. High tannin containing species (Lotus, Sulla) are being evaluated as possible alternatives to reduce ruminant methane production. In the absence of relevant data it would be desirable to measure elements of C sequestration in experiments that are currently employing tannin-containing species as a methane mitigation option.

6.4 - Conclusion

Management options for increasing C storage on grazed land appear to be limited.

Increased storage would require maintained or increased inputs and reduced products or outputs - a policy that would be undesirable from an economic standpoint.

Increased productivity per unit area (intensification) can be achieved without any cost in C sequestered as long as either the inputs (fertiliser, irrigation) on that area are also increased or the increased utilization of one area is matched by reduced utilization on other areas. This outcome offers the possibility of concentrating production on smaller areas and releasing more marginal areas to be used as C sinks with potential benefits also for control of C losses through erosion. Potential gains in C sequestration need to be calculated for different intensification options at a range of spatial scales from the farm to the regional level.
The use of pasture species specifically to sequester C has little merit in itself (see reseeding arguments in methane section) but is certainly an aspect of plant performance that should be considered in plant-based strategies to reduce methane emissions.

Biofuel options appear to have a great potential ('an order of magnitude greater than any other land-use change', Watson et al 2000) for using agricultural land to mitigate greenhouse gas emissions.

They have the double benefit of a perennial crop capable of sequestering C and the production of fuel to directly replace fossil fuel use.
Rapid advances in ethanol production technology overseas, our own understanding of the breakdown of plant fractions through our expertise in rumen function and the climate and experience to produce high yields of herbaceous species in New Zealand present an exciting opportunity to develop biofuel options for this country.
The potential of biofuel crops, initial tests of appropriate species and their integration into farming systems should be investigated.