The use of energy crops on metal contaminated sandy soils

In the 19th and 20th century (up to 1970s), historic atmospheric deposition of trace metals from metal refinery activities has resulted in elevated concentrations in agricultural soils in the Campine region. As a consequence, a surface area of at least 700 km2 is now contaminated with several toxic metals, including cadmium (Cd), zinc (Zn) and lead (Pb). Previous studies also reported health problems in the vicinity of the smelters with increased risks for renal dysfunction, osteoporosis, lung cancer as well as other health-related issues due to a higher exposure of Cd. Furthermore, the soils in the region are characterised by a sandy texture and relatively low pH, which entails an enhanced risk for uptake of these metals in crops or leaching to the groundwater. Conventional soil remediation approaches tend to be overly expensive considering the moderate levels of pollution and extended areas that need a treatment. Phytoremediation involves the use of plants and their associated microorganisms for the stabilisation, degradation and/or removal of pollutants from the environment. Phytoextraction in particular aims at accumulating elements into the harvestable plant parts so that the contamination can gradually be removed through harvest. The purpose of this research is to assess the feasibility of non-food crops not only as a phytoremediation option, but also forming a risk based management approach for metal contaminated agricultural soil. In this study three non-food crops were evaluated: rapeseed (Brassica napus L.) for plant oil production (biodiesel), willow (Salix) under short rotation coppice and energy maize (Zea mays L.) for biogas production by anaerobic digestion. Rapeseed reached a total shoot biomass of 5.8±2.0 ton DM ha-1 which is in compliance with expected values. However, the seed yield was very low (0.5±0.3 ton DM ha-1). Other reports, with experiments performed in the same year and region but sown on non-contaminated soils and under optimal fertilisation, showed a seed production of 3 ton DM ha-1. The low biomass productivity was attributed to suboptimal fertilisation mainly of nitrogen and sulphur. Observations from the field experiment suggested that 27±14 g Cd ha-1 and 2.0±1.0 kg Zn ha-1 could be removed from the soil by harvesting rapeseed. This means that more than 250 growing cycles of rapeseed would be needed to reduce total Cd concentrations in the top soil layer from 5.0 to 2.9 mg Cd kg-1 (resp. the current concentration in the field plot and the site-specific threshold value of Cd for remediation of the field plot). During subsequent experiments, the growth of rapeseed failed completely. The fact that rapeseed needs to be cultivated in a rotation scheme in order to avoid possible crop failures, also contributes to lower suitability of this crop as an appropriate alternative cropping system for metal contaminated agricultural soils in the Campine region. Secondly, the cultivation of short rotation coppice (SRC) was investigated. Short rotation coppice consists of densely planted, high-yielding varieties of either willow or poplar (Populus), which are harvested every two to five years. After each harvest, new shoots spontaneously re-sprout. From the biomass productivity during the first four years of growth, it could be concluded that a rotation period of four years was needed to reach maximum annual growth potential of all clones. There were strong differences observed in biomass production between clones. After four years, Zwarte Driebast had the highest biomass (~12 ton DM ha-1 year-1), followed by Loden. In contrast, Inger achieved only a biomass production of 1.5 ton DM ha-1 year-1. The metal extraction potential of willow was much higher than that found for rapeseed. The woody biomass itself was in general 72 g Cd and 2.0 kg Zn ha-1 year-1. This means that in order to reduce the concentration of Cd from 6.5 mg kg-1 (current concentration in the field plot) to 3.0±0.3 mg Cd kg-1 (site-specific threshold value of Cd for remediation of the field plot), more than 170 years will be needed. Under normal conditions SRC is harvested in winter, but as the leaves contain high amount of metals, leaf harvest could increase the extraction potential by 40%. Previous reports showed that SRC from contaminated soils can be treated safely, yet it was not further assessed in the current study. Currently, the price of the wood does not guarantee a sufficient income for farmers, compaired to their current farmer income, to make it a suitable alternative cultivation. Neverthelles when the use of wood as energy source becomes more pronounced, wood prices will very likely also stimulate the adoption of SRC in the region. A final investigated option was the use of energy maize. Under optimal fertilisation, energy maize showed an average yield of 53±10 ton FM or 20±3 ton DM ha-1 year-1. The observed productivity levels were similar to those of the same cultivars grown elsewhere in Flanders on non-contaminated sites. As maize is a metal excluder, it contains low concentrations of metals in the different plant compartments. Due to the low concentration of metals in the plant, the extraction potential of energy maize was the lowest of the investigated crops (19±6 g Cd and 4.3±0.9 kg Zn ha-1 year-1), implying that for remediation of the experimental site (from 5.0 to 2.5 mg Cd kg-1) more than 500 years would be needed under optimistic conditions. Lab scale batch results of 14 days, performed at OWS (Organic Waste Systems, Ghent), revealed a biogas production potential of 215±23 Nm3 ton-1 FM which was similar to that of energy maize grown on a non-contaminated soil (194±4 Nm3 ton-1 FM) tested under the same conditions. For obtaining optimal biogas yields, harvest must take place when the plant reaches optimal dry matter content. Although the harvest period optimises the digestibility of the biomass, it does not affect the metal extraction rates significantly. When a semicontinuous digester is fed over a period of 435 days with maize from a contaminated site, similar biogas production levels were achieved. During this semi-continuous test, concentrations of elements in the digestate were on average 3-4 times higher than those in the input material. This is similar to concentration factors observed using energy maize from a non-contaminated site. The quality and characteristics of the digestate can be estimated and evaluated in the context of legal threshold values and limitations for usage. Economic studies indicated that the average annual income of the farmers can be maintained when they convert land use from fodder maize to energy maize. None of the tested energy crops would allow an effective short term (e.g. 20 - 30 years) removal of metals from the soils. However, the presence of contamination is not expected to introduce technical constraints in converting biomass with elevated concentrations into energy. The fate of the metals during subsequent processing of the biomass can be controlled, and eventual constraints in subsequent use and treatment of the secondary products can be managed. If economic viability of energy crop production can be ascertained, it can allow for a long term safe alternative use of the contaminated agricultural land. For energy maize as a cropping alternative, this was found to be the case. Growing energy crops aims at riskreduction, and generates an alternative income for farmers, yet in the long run also generating a gradual reduction of the pollution levels. In this way, remediation is reduced to a secondary objective with sustainable risk-based land use as primary objective. This context has been adopted as ‘phytoattenuation’, setting it apart from ‘conventional’ phytoextraction