Final Report for Defra Project FO0103, Comparative Life Cycle Assessment of Food Commodities Procured for UK Consumption through a Diversity of Supply Chains
Appendix 4. Semi-quantitative comparative assessment of the impacts on wider ecosystem services.
Introduction
The increased global trade in food is leading to a greater diversity of food chains supplying the UK consumer, many of which involve overseas production with long transportation distances to the UK, often under controlled-temperature regimes. However, there can be advantages, such as climate, scales or types of production in overseas production. Appendix 1 reported quantified comparative life-cycle burdens of different food supply chains for tomatoes, strawberries, potatoes, apples, lamb, poultry and beef with respect to primary energy use (PEU), global warming potential (GWP), abiotic resource use, and emissions that can adversly affect air and water quality. The study was also required to assess the extent to which some types of chains might be exporting our environmental burdens to countries outside the UK. For example, there are significant energy reductions from importing field-grown tomatoes from the Mediterranean thus reducing the need for energy consumption in heated greenhouses in the UK, even when the additional transport energy emissions are taken into account.
Consideration also needs to be given to environmental burdens other than energy intensity. For example increased production of field-grown salads and vegetables in the Mediterranean area may create extra demands for water supplies, increase soil salination and decrease water quality through eutrophication and pesticide contamination. Increased production in any location may generate damage to valuable ecosystems through change in land use or from emissions created during the production cycle. In this Appendix we report our findings on the impacts of producing the seven foods on biodiversity, soil carbon reserves, water resources and water quality in the alternative sources of produce.
Loss of habitat and biodiversity
Substantial loss of biodiversity is directly associated with habitat change, climate change, invasive alien species, over-exploitation and pollution (MEA 2005). Habitats of critical importance include tropical forests, wetlands (including peatlands) and riverine habitats (including estuaries), grasslands and savannahs. All of these habitats are under threat from expansion of agriculture. Fragmentation of habitats can also have significant impacts on biodiversity. This is of particular concern in ecosystems such as forests or wetlands.
It is difficult to assess biodiversity as there is no single biodiversity metric that can be used to assess the impacts of environmental changes. The most commonly used metric of biodiversity is species richness, or the number of species in a given area. While it does directly measure the amount of species in a given area, it weights all species equally regardless of their abundance or role within an ecosystem. It does not differentiate between native/endemic species and exotic/weedy species. Species richness may sometimes increase in the face of human degradation since it will count invasive species that could potentially have disruptive effects on key ecosystem functions. The WWF, (WWF, 2007) however, provides data on the numbers of vertebrate species in eco-regions around the world, which are relatively large (e.g. England, Wales and Scotland are covered by just three of these regions). These include the total richness (sum of vertebrate species) as well as descriptions of the degree of vulnerability of species (Table 1). The total richness and a weighted sum were correlated and showed a very close relationship (Figure 1), so that while relatively crude, the total vertebrate richness seems to be a reasonable approximation for biodiversity. It does not explicitly account for plants and invertebrate, but they are implied as part of the ecosystem needed to support the vertebrates.
Table 1. WWF descriptions of the vulnerability of species and weighting factors used in this project
Description / Weighting factorCritically Endangered / 1.00
Endangered / 0.75
Vulnerable / 0.56
Near Threatened / 0.42
Lower Risk/Near Threatened / 0.32
Lower Risk/Conservation Dependent / 0.24
Lower Risk/Least Concern / 0.18
Least concern / 0.13
Data deficient / 0.00
Figure 4.1 Relationship between total numbers of vertebrate species in an eco-region and the number weighted by rarity for the eco-regions considered in this project
Agricultural production in the UK has had relatively little impact on biodiversity since the region’s biodiversity has already been minimised or eliminated by the early 20th century. Hence the continued production of food in the UK might be deemed preferable to production in areas of greater biodiversity value. Figure 4.2 below gives an indication of the regions of the world where biodiversity is under threat of reduction.
Figure 4.2. Status of Terrestrial Ecoregions (from WWF 2006). Red = Critical/Endangered, Orange = Vulnerable, Green = Relatively Stable/Intact, Grey = No Data.
As can be seen from Figure 4.2 above, parts of Israel, Spain and Brazil relevant to this study are assessed as critical or endangered, while much of the relevant area of Brazil is vulnerable. The trend in species richness (Figure 4.3) moves from the relatively low indices of New Zealand, through the UK and increases through Spain and Israel, and up into Brazil. The range in Brazil is large, befitting a country of its vast scale, and indicates the diversity of land used for grazing. The Pampa area (defined as WWF area NT0710 - Uruguayan savannah) has the lowest index while the Cerrado (NT0704) has almost the highest index (marginally below Alto Paraná Atlantic forests), but also covers the single largest area of an eco-region. Using Cerrado for grazing has some interesting consequences relating to the biodiversity. In one farm in the agri benchmarking project (Deblitz, 2006), calf mortalities were relatively high at 6%. This was attributed to the high number of snakes and other predators in the area. In contrast, there are no predators of sheep in NZ.
Figure 4.3. Range of vertebrate species richness in the main eco-regions of interest in the this project.
Tomatoes
A converse of the greater intensity of UK production under protection is the much greater pesticide and land requirement for Spanish production by factors of *7 and *5 respectively. Figure 1 in Appendix 1 illustrates the extent to which land cover in the SE coastal strip is dominated by polythene structures producing horticultural crops. The far smaller land area required for UK production, together with the isolation of agrochemical inputs, leads to a much smaller potential adverse impact on biodiversity in the UK. Appendix 1 reports the potential for greatly reducing GHG emissions from UK greenhouse production by establishing production close to sources of industrial waste heat. There is a potentially very significant synergy here, in that such industrial sites are located in urban, or peri-urban areas, where derelict or blighted land may be available for development.
Potatoes
Cultivation of uncultivated land will also have negative impacts on biodiversity and may increase human-animal conflicts. Thus, although much of the potato production in Israel is on reclaimed desert it might be argued, on strict biodiversity criteria, that potato production is reducing biodiversity. A counter argument would be that, on a global scale, since the area covered by deserts is large, and increasing, some reduction in the area of desert habitat found within Israel is not a cause for concern.
Poultry and beef
Since Brazil was the source of imported poultry and beef in this study, these two commodities will be considered together. With respect to habitat loss the crucial consideration is the respect to which production of these commodities drive rainforest destruction.
The drivers of rainforest destruction are numerous and complex, and in many of the areas of concern, logging (legal or illegal) and slash and burn agriculture are also important. Beef and poultry production for export to the EU are not considered to be primary factors. The vast majority of cattle raising for the EU will be in the central Cerrado (Savannah) regions with a smaller percentage from the Pampa (grassland) region. Poultry, raised in large indoor units similar to those in Europe, are generally found near to the urban centres and ports and, occupy little space.
While the relative amounts of biological diversity contained in grasslands and savannahs is less than other habitats of concern, Conservation International’s Biodiversity Hotspots programme includes the Cerrado region of Brazil. The Cerrado region comprises 21% of the country and contains the most extensive woodland-savannah in South America (see Figure 4.2). Only 1.5% of the Cerrado lies within Federal reserves. It is estimated that only 35% of the Cerrado retains large proportions of its native vegetation (Casson 2003). It features a pronounced dry season and supports an array of plant species adapted to drought and fire and a large number of endemic birds. Large mammals such as the giant anteater, giant armadillo, jaguar and maned wolf also survive here, but are competing with the rapid expansion of Brazilian agriculture. These mammals are key indicator species for the ecosystem of the region.
The greatest association between meat production and rainforest loss is the indirect impact of the cultivation of soya to feed to livestock. Soya is a crop whose agricultural intensification and expansion comes at the expense of important regions of biological diversity. Soybean production has been identified as one of the leading causes of deforestation in Brazil’s forests, particularly the Mato Grosso seasonal forest ecoregion (Grau et al., 2005; Casson, 2003; Morton et al., 2006). These authors have identified links between increases in the area under soybean cultivation and decreases in uncultivated land and biodiversity in the Cerrado and (more recently) rainforest. Evidence of these changes was obtained from statistical data from the Brazilian Government and FAO (USDA, 2005). The USDA has provided an analysis of historical soybean production in Brazil, which links soybean expansion firmly to the price commanded for soybean on the international market and increased profitability for Brazilian farmers (USDA, 2005). Some soybean producers clear forests themselves. Other buy the land from small producers, often colonists, who have already cleared it. These same small producers then move further into the frontier and clear more land. In addition to direct habitat conversion, soybean production in pristine areas also requires the construction of massive transportation and other infrastructure projects. Moreover, the infrastructure developments for soya production unleash a number of indirect consequences associated with opening up large, previously isolated environments to population migration and to other land uses. This infrastructure contributes directly and indirectly to habitat conversion. Casson (2003) also reports the displacement of smallholders by soybean plantations, causing them to migrate into the Amazon region where they clear forest for agriculture or cattle ranching (for domestic consumption).
Nevertheless, it must be recognised that the expansion of soya production is mainly due to demand from Europe for animal feed. Moreover, since poultry raised in the UK are fed amounts of soya similar to those fed to poultry in Brazil, and much of the soya imported into the UK originates in Brazil, there is little or no reason to suppose that poultry production in Brazil is any greater stimulant to soya cultivation than poultry raised in the UK. The key driver here is not the country of origin of poultry and other meat products but the growing global demand for meat.
Loss of soil C
Potatoes
A positive feature of arable cropping in the Negev Desert is that soil organic carbon (SOC) has increased - starting from virtually zero. This is in marked contrast to the UK where potato cultivation is regarded as one of the most damaging to soil and leading to particularly large losses of soil C. In addition to ploughing prior to potato cultivation, increasing use is made of stone separation and this is a particularly vigorous form of tillage causing more soil disturbance than ploughing. In addition harvest of potatoes leads to further disturbance of the soil. Hence in comparison with cereals and other combinable crops which may be ploughed very year, soils on which potatoes are grown are vigorously cultivated at least three times in one year. Moreover, carbon losses increase with increasing organic matter contents. On degraded lands, carbon stocks can be increased after restoration to productive use. It is recognised that carbon storage in degraded soils can be partly restored by practices that reclaim productivity including: re-vegetation (e.g., planting grasses); improving fertility by nutrient amendments; applying organic substrates such as manures, biosolids, and composts; reducing tillage and retaining crop residues; and conserving water (Lal, 2001b; 2004b; Bruce et al., 1999; Olsson and Ardö, 2002; Paustian et al., 2004) and most, if not all of these activities have been introduced to the Negev by the cultivation of potatoes. In contrast, where potatoes are grown on soils with initially large concentrations of SOC, such as the East Anglian Fens, the introduction of the intense cultivation associated with potato growing will inevitably lead to very large reductions in SOC.
The RothC model (Coleman and Jenkinson, 1996) of soil C turnover was applied to the unusual circumstances of Israeli potato production. Simulations were run from 1948 until the model reached equilibrium. Meteorological data came from the Israel Meteorological Service station at Beer Sheva. Rotations were simulated both with and without potatoes, using fallow as the alternative to potatoes. The soil C returns from potato residues were derived from the IPCC (2006) default coefficients. Over the last 60 years, SOC increases attributable to potatoes totalled about 6 t C/ha. Considering the next 20 years (as defined in PAS 2050, Final draft), the storage of C is estimated to be equivalent to kg CO2 Equiv. per t potatoes per year and represents about 1% of the total GWP that is incurred during the production of potatoes.
Beef and poultry
Soya production in the Cerrado is estimated to lead to average soil losses of 8 tonnes/hectare/year (Fritsche et al., 2006); loss of SOC is a serious problem in the soya-producing areas of Brazil due to the warm climate and dry winters.
Land use changes and the effects on SOC have been studied in Brazil, especially in the Brazilian Amazon (e.g. Cerri et al., 2003, 2004, 2007 cite over 50 papers). Cerri et al. (2003, 2004, 2007) studied conversion of forest to pasture for cattle ranching using simulation models, including RothC and Century, and experimental measurements. They showed how conversion to pasture led to initial falls in SOC stocks (0-20 cm depth), but in the majority of cases this was followed by a slow rise to levels exceeding those under native forest. One exception to this pattern was a degraded pasture. They conclude overall that well-managed pastures can be useful in increasing SOC stocks after deforestation. In one study (Cerri et al., 2004) reported how pasture SOC recovered, after about 10 years decline, to the original SOC concentration in the top 0-20 cm layer, although loss of forest SOC continued, it was exceeded by the accumulation of pasture SOC. It cannot, however, be assumed that all conversion will be as well managed. It also needs to be remembered that these measurements of SOC take no account of the loss of C sequestered in forest vegetation.
Calegari et al. (2008) reported how SOC in a once-forested part of Paranà, decreased for about 10 years when cultivated for arable crops. This was followed by increases in SOC at differing rates according to the management practices used over a 19 year period, particularly at different depths of soil. No tillage (NT) cultivation sequestered more SOC in the upper soil layer than conventional plough-based tillage (CT). Using winter cover crops with NT was the most successful method of increasing SOC storage and was the only method that approached the original SOC content of forest soil and increased C storage at a rate of abut 1.2tC ha-1 year1. Again, the possibility of restoring some lost SOC was demonstrated, but not all management practices necessarily yet operate at the same high level. The potential for sequestration also varies with soil types, as some are more capable of protecting SOC from degradation than others. Moreover, such estimates of the impact of NT on SOC fail to recognise that increases in SOC concentrations take place only near the soil surface and that the SOC content of lower horizons may decrease (Gál et al., 2007)
The impact of grazing on SOC in the Cerrado is less than on forest. Maquere et al. (2008) reported that total C stocks to 1 m depth under pastures that had been established for 20 and 80 years were numerically greater than, albeit not significantly different to, total SOC to 1 m of native Cerrado. The total SOC estimated, at 84 t/ha, was greater than the default value of 66 t/ha for Brazilian savannah cited by IPCC (2006). This finding was consistent with the results of earlier studies cited in the paper. However, when evaluating fluxes of CO2 to the atmosphere, it needs to be remembered that only considering changes to SOC takes no account of changes in above-ground carbon stocks which may be greatly reduced by land use change. The current estimate of total carbon storage, both SOC and above-ground, is c. 360 t/ha for Brazilian rainforest (IPCC, 2006).
It is clear that land use changes contribute to the loss of stored C, either from soils or from wood that may be burned during deforestation. There is also the potential for some recovery from the use of pasture or better arable cropping practices. It is a non-trivial task to estimate this for all the Brazilian crops and pastures that contribute to domestic Brazilian production for export to the UK or indeed for soya for export to the UK. In the same way, it is non-trivial to estimate the loss of SOC caused in the UK by the forest clearances that occurred many years (indeed centuries in some cases) ago. It is beyond the scope of this project to undertake these calculations.
It is worth observing that UK soils generally lost SOC following the great increase in arable cultivation both during and after the Second World War. These changes should be approaching a new equilibrium about now, although the asymptote may take another 100 years or so to be reached. Somewhat greater rates of loss of soil C than might be expected in England and Wales were reported by Bellamy et al. (2005). One hypothesis was that climate change has led to warmer winters so that SOC loss was accelerated during the part of the year when C supply from growing crops and grass was minimal.