Assessment of abandoned agricultural land resource for bio-energy production in Estonia

Abstract

The current study locates and quantifies abandoned agricultural areas using the Geographic Information System (GIS) and evaluates the suitability of abandoned fields for bio-energy production in Tartumaa (Tartu County) in Estonia. Soils of abandoned areas are generally of low quality and thereby limited suitability for crop production; as a result soil–crop suitability analyses could form the basis of knowledge-based bio-energy planning. The study estimated suitable areas for bio-energy production using willow (Salix sp), grey alder [Alnus incana (L.) Moench], hybrid aspen (Populus tremuloides Michx.×Populus tremula L.), reed canary grass (Phalaris arundinacea L.), and Caucasian goat's rue (Galega orientalis Lam.) in separate plantations. A combined land-use strategy is also presented as these crops are partially suitable to the same areas. Reed canary grass and grey alder have the highest energy potentials and each would re-use more than 80% of the available abandoned agricultural land. Energy grasses and short-rotation forestry in combined land-use strategy represents the opportunity of covering approximately a quarter of county's annual energy demand. The study estimates only agronomic potential, so further bio-energy analysis should take into account technical and economic limitations. Developed framework supports knowledge-based decision-making processes from field to regional scale to achieve sustainable bio-energy production.

Keywords:

• Abandoned fields
• biomass energy potential
• GIS
• location-specific analysis
• soil–crop suitability

Introduction

Since the last decades of the 20th century interest in the production and utilization of bio-energy has increased. The main reasons are a combination of diminishing fossil fuel reserves and increasing environmental protection awareness. European Union (EU) energy policy priorities are, as per Directive 2006/32/EC (European Union, 1995), greenhouse-gas-emission limitation and energy-supply efficiency. Increased use of renewable energy sources enables these targets to be met. During recent decades studies in Europe have examined renewable energy opportunities at the national level and provided detailed information about available renewable energy sources (Voivontas et al., 2001; Batzias & Sidiras, 2005) and about bio-energy policy-promotion tools (Streimikiene & Klevas, 2007). Effective energy utilization has maintained a central role in Estonia's public debates since 1991. The widely held conjecture is that the reserves of the most important natural resource in Estonia, oil shale, will estimably last only for a further 60 years (Valgma, 2003). Increasing the share of renewable resources in energy production is, therefore, important. A rapid decline in agricultural land use has occurred in Estonia since the restoration of independence in 1991. The aggregate area of arable land in 2006 is half of that in use fifteen years earlier (Astover et al., 2006). The scale of this decrease in arable land was the most drastic in the whole of Europe and was, by a factor of 3.9, higher than other post-Soviet European countries (Astover et al., 2006). There is necessity in Estonia for the planning of abandoned agricultural land, and the field-specific soil databases developed in this study provide a solid basis for a knowledge-based allocation of bio-energy production. The databases are sufficiently flexible to allow the results from field level to be scaled up for regional-level analysis. Increasing bio-fuel production, as a result of increasing energy demands and mindful of policies on climate change, takes up a significant area of land in many scenarios and prevents substantial abandonment of agricultural land (Bush, 2006). The use of abandoned agricultural areas is one potential way of increasing bio-energy production. Environmental awareness has forced scientific research to estimate the impacts related to bio-energy production. Studies have referred to some positive influences (McLaughlin & Walsh, 1998) but also negative ones during the process of bio-energy production (Ledin, 1998). Also, a number of farm-related factors influencing bio-energy adoption have been indicated (Roos et al., 2000). The importance of making precise estimates of the environment at both regional and local levels is therefore important. The decline in arable-land use in Estonia was regionally variable and especially high in marginal districts with low soil quality (Astover et al., 2006). Therefore the planning of bio-energy production on abandoned areas requires precise location-specific analysis. The aim of the present study was to locate, quantify, and estimate the suitability of abandoned agricultural areas for bio-energy production using Tartumaa (Tartu County) in Estonia as an example.

Material and methods

Tartu County is situated in the south of Estonia between the shores of Lake Võrtsjärv (western side of the county) and Lake Peipsi (eastern side of the county) and straddles the River Emajõgi which flows between the two lakes. The county covers 308 900 ha, which is 7.1% of Estonia's land surface of 4 369 800 ha. In 2007, agricultural land in use formed 26% of the county's total area and 9% of Estonia's agricultural land. The proportion of forest land (38.9%) in Tartu County is smaller than that in the country as a whole (51.5%).

The study identified abandoned field parcels in Tartu County, using the Estonian Basic Map (1:10 000) and the field layer of the Agricultural Registers and Information Board (ARIB) and databases of Common Agricultural Policy (CAP) payments in 2007. We considered field parcels that did not have any applications for single-area payments as ‘entirely abandoned’ and field parcels where area payments covered 50–99% of total area as ‘partially abandoned’. The study used an overlay comparison of the Estonian Basic Map and the ARIB field layers to identify agricultural areas excluded from ARIB's fields (i.e. not valid for CAP subsidy schemes), which we also tagged as ‘entirely abandoned’. We then used visual and manual correction of area boundaries based on ortho-photos to eliminate any of the remaining agricultural areas on the Basic Map that did not fit the following topographical parameters and were thus unsuitable for either bio-energy production or further analysis: (i) areas less than 0.3 ha, and (ii) areas with perimeter:area ratio over 5:1. The total agricultural land in Tartu County included for analysis in the study was 111 143 ha (103 166 ha ARIB fields and 7977 ha from Basic Map) which forms 36% of the county's total land area. The study used a GIS environment, MapInfo Professional, to perform topology analysis of the field layers and the soil map polygons. We identified the soils of abandoned land using the Estonian Land Board's digital soil map (scale 1:10 000) and depending on the soil type and texture assessed the suitability of these areas for short-rotation energy forestry and energy grasses (Laas, 2004; Kõlli, 2006). Reintam et al., (2003) provide detailed overviews about the Estonian large-scale soil map and crop-specific suitabilities. We evaluated areas suitable for potential bio-energy production using willow (Salix sp), grey alder [Alnus incana (L.) Moench], hybrid aspen (Populus tremuloides Michx.×Populus tremula L.), reed canary grass (Phalaris arundinacea L.), and Caucasian goat's rue (Galega orientalis Lam.). These crops are the most studied energy cultures under Nordic conditions (Ross et al., 1996; Uri et al., 2002; Vares et al., 2003; Lillak et al., 2007; Pahkala, 2007).

Conversion of the land potential to a bio-energy potential

The estimated abandoned land was further planned by consideration of factors ranging from soil-suitability analysis to potential energy-crop production. We calculated, for each of the five selected energy crops, the energy output of the annual biomass yield for both separate plantations and combined land-use strategies. We considered in combined land-use strategy that 30% of abandoned areas remains under natural conditions [biomass yield for natural grassland 2.0 t dry matter (DM)/ha] and 70% for energy grasses and short-rotation forestry. The land partition for energy crops was based on the results of soil-suitability analysis considering relative area proportions suitable for each crop.

Annual DM productivity forms the basis of the bio-energy potential of a plant for which reason the study needed to know the relevant values for the five selected bio-energy plants. We used the results of previous studies and calculated average annual productivity of DM as well as the calorimetric value in megajoules (MJ) for each of the study's selected plants (Table I). Willow's annual DM production was considered as 4.4 t ha⁻¹ in gleyic soils and Gleysols, and 5.9 t ha⁻¹ in histic soils and Eutric Histosols. Annual willow production varies by a factor of 4–5 depending mostly on soil water regime and nutrient sufficiency (Ross et al., 1996). The potential bio-energy production for grey alder was taken as 6.4 t ha⁻¹ and that for hybrid aspen as 6 t ha⁻¹. The main parameters influencing plantation production of grey alder and hybrid aspen coppice are age and soil. Soil parameters can influence the growth of these two types of plantations by a factor of two (Vares, 2005). The annual DM biomass yields were calculated for reed canary grass at 8 t ha⁻¹ and for Caucasian goat's rue at 7 t ha⁻¹. Reed canary grass is one of the highest-yielding perennial herbaceous grasses (Wrobel et al., 2009). Under Estonian conditions the variation coefficient of reed canary grass and Caucasian goat's rue production is accordingly 44 and 24% depending on pedo-climatic conditions and also, in the case of reed canary grass, fertilization (Rand, 1981; Eilart & Reidolf, 1987; Meripõld, 2006; Viil, 2006). The higher the soil nutrient supply is, the more stable are reed canary grass yields. In the instance of soils with low humus content (2%), the use of high levels of fertilizer (N₂₀₀P₃₅K₁₃₀) can result in a DM yield of 8 tha⁻¹. In this instance the variation coefficient was 14%. Lowering the fertilization norm by a factor of 1.5–3 can result in the same DM yield (8 tha⁻¹) on histic soil and Histosols (Rand, 1981; Eilart & Reidolf, 1987).

Table I. Annual dry-matter productivities (DM, t ha⁻¹) and calorimetric values (MJ kg⁻¹) of bio-energy crops applied in calculation of bio-energy production potential.

Crop	Annual productivity (DM t ha^−1)	Calorimetric value (MJ kg^−1)
Reed canary grass	8^1	16.6^6
Caucasian goat's rue	7^2	16.6^7
Grey alder	6.4^3	18.6^8
Hybrid aspen	6^4	18.6^9
Willow
Gleyic soils and Gleysols	4.4^5	18.6^10
Histic soils and Histosols	5.9^5	18.6^10
Sources: 1 – Rand (1981); Eilart & Reidolf (1987); Kryževiciene (2006). 2 – Meripõld (2006); Viil (2006); Lillak et al. (2007). 3 – Uri (2000); Uri et al. (2002). 4 – Vares et al. (2003). 5 – Ross et al. (1996). 6 – Burvall (1997). 7 – Hovi (1995). 8 – Tullus et al. (1998). 9 – Vares et al. (2003); Vares et al. (2005). 10 – Miles et al. (1995).

Results

Abandoned agricultural land in Tartu County covers a total of 26 351 ha of which 20 741 ha is 'entirely abandoned’ and 5610 ha is 'partially abandoned’. Abandoned field parcels are distributed homogeneously all over the county, although the density varies (Figure 1). The proportion of abandoned fields is highest near the county's biggest urban area, Tartu, but also relatively high along the banks of the River Emajõgi and the shoreline of Lake Peipsi. The mean field area differs significantly between used (21 ha) and abandoned fields (2.9 ha).

Figure 1.  The location of used and abandoned agricultural areas in Tartu County. Number of field parcels shown in brackets.

Stagnic Luvisols form on 33.5% of the total analysed agricultural land and Gleysols 20.6% (Table II). Luvisols and Cambisols form altogether 25% and are distributed as 29.5% from used parcels and 16% from entirely abandoned fields. In the case of Histosols and Albeluvisols abandoned areas form more than twice the area of that in land use, and in the case of Fluvisols nearly 23-times this soil's land-use area.

Table II. Soil distribution (percentage of area) on agricultural land of Tartu County.

Soil reference group by WRB^1	Total (%)	Land in use (%)	Entirely abandoned areas (%)
Fluvisols	1.6	0.3	6.8
Eroded soils	2.0	1.4	2.9
Deluvial soils	3.0	2.5	3.4
Gleysols	20.6	23.5	12.3
Histosols	8.8	6.9	13.9
Cambisols	6.2	7.4	4.4
Luvisols	18.9	22.1	11.9
Stagnic Luvisols	33.5	31.7	34.8
Albeluvisols	5.3	4.0	9.4
Other	0.2	0.1	0.2
^1World Reference Base 1998.

The following analyses represent six different concept strategies for producing biomass from abandoned agricultural land in Tartu County; three of them concern short-rotation forestry, two energy grasses (Table III), and one combined bio-energy land-use strategy (Table IV).

Table III. Potential annual biomass and energy production from suitable abandoned areas in Tartu County using separate willow, grey alder, hybrid aspen, reed canary grass, or Caucasian goat's rue plantations.

	Entirely abandoned agricultural land			Partially abandoned agricultural land
Crop	Area (ha)	Total yield^1 (t)	Total energy^2 (GWh)	Area (ha)	Total yield^1 (t)	Total energy^2 (GWh)
Reed canary grass	17 433	139 462	643	4883	39 060	180
Caucasian goat's rue	14 411	100 877	465	4486	31 403	145
Grey alder	15 914	101 850	526	4876	31 205	161
Hybrid aspen	13 140	78 842	407	4211	25 265	131
Willow
Gleyic soils and Gleysols	8283	36 445	188	2076	9133	47
Histic soils and Eutric Histosols	3668	21 643	112	682	4022	21
^1Total potential yield from suitable abandoned areas in tonnes. ^2 Total potential energy production from suitable abandoned areas in gigawatt hours.

Table IV. Potential annual biomass and energy production in Tartu County in the case of combined land-use strategy.

	Entirely abandoned agricultural land			Partially abandoned agricultural land
Crop	Area (ha)	Total yield^1 ( t)	Total energy^2 (GWh)	Area (ha)	Total yield^1 (t)	Total energy^2 (GWh)
Reed canary grass	3484	27 876	129	903	7226	33
Caucasian goat's rue	2904	20 326	94	825	5773	27
Grey alder	3194	20 442	106	903	5781	30
Hybrid aspen	2613	15 680	81	785	4713	24
Willow
Gleyic soils and Gleysols	1603	7053	36	383	1685	9
Histic soils and Eutric Histosols	720	4249	22	128	753	4
Natural grassland	222	12 445	57	1683	3366	16
Total	20 741	108 070	524	5610	29 296	142
^1 Total potential yield from suitable abandoned areas in tonnes. ^2 Total potential energy production from suitable abandoned areas in gigawatt hours.

Short-rotation forestry

Gleyic soils, histic soils, Gleysols, and Eutric Histosols are most suitable for growing willow. There are 11 951 hectares of entirely abandoned agricultural land in Tartu County which are suitable for growing willow. Gleyic soils and Gleysols form 69% and histic soils and Eutric Histosols 31% of these areas. The total quantity of willow dry wood, if grown in all of the entirely abandoned areas, would weigh 58 087 metric tonnes with an energy value of 300 GWh. There are also 2757 hectares of partially abandoned field parcels which are suitable for growing willow. The total production of willow dry wood in the aggregate of the partially abandoned areas would be 13 155 metric tonnes with an energy value of 68 GWh.

Grey alder is most productive on high-fertility Cambisols and Luvisols but satisfactory productivity also occurs on Stagnic Luvisols and Albeluvisols. Grey alder could be cultivated on 15 914 hectares of entirely abandoned agricultural land. The potential bio-energy production of these fields is as high as 526 GWh. There are 4876 hectares of partly abandoned areas where grey alder cultivation could potentially produce an energy value of 161 GWh. Grey alder biomass production represents the highest re-use potential of abandoned parcels (80% in total) by the three energy forests.

Hybrid aspen's soils requirement coincides mostly with that of grey alder; the most suitable soils are moderately moist soils, gleyic loamy sand, and loam soils. There are 13 140 hectares of entirely abandoned and 4211 hectares areas of partially abandoned agricultural land that are suitable for growing hybrid aspen as energy forest. The potential energy values of these areas are accordingly 407 GWh and 131 GWh.

Energy grasses

Reed canary grass could be grown in 17 433 hectares of entirely abandoned areas, which forms almost 86% of these fields. Biomass production from these areas could reach as high as 139 462 metric tonnes with an energy value of 643 GWh. There are also 4883 hectares of partly abandoned fields which are suitable for growing reed canary grass. The potential energy production in these fields is as high as 180 GWh (Table III).

The potential bio-energy production of cultivating Caucasian goat's rue on both the entirely and partially abandoned agricultural land (18 897 hectares) is 132 280 metric tonnes of dry biomass with an energy value of 610 GWh.

Combined land-use strategy

Soil-suitability analysis for each of the selected crops indicated there were overlaps in soil suitability between the five crops. We decided therefore to compile a combined land-use strategy for evaluating potential bio-energy production from total abandoned areas using all five available crops. Since we had previously declared that 30% of abandoned areas would remain as natural grasslands the data for this category of abandoned land are included. The biomass of energy forests and grasses grown on abandoned fields in Tartu County would weigh 121 555 tons, of which 95 625 tons would come from entirely abandoned land and 25 930 tons from partially abandoned land (Table IV). The total bio-energy production from these fields could be as high as 594 GWh, which in relation to separate plantations is lower than the energy production from reed canary grass, Caucasian goat's rue, or grey alder but higher than that potentially from hybrid aspen or willow. Biomass production from natural grasslands would form 15 811 tons with an energy value of 73 GWh. Potential bio-energy production from total abandoned areas in Tartu County is as high as 667 GWh.

Discussion

Several studies have estimated agricultural land resource potential for bio-energy at global, EU, and national scales (Voivontas et al., 2001; Edwards et al., 2005; Hoogwijk et al., 2005) but investigations at a more detailed spatial level are few (Förster et al., 2008). Sustainable bio-energy planning recommendations and related land-use decisions must be made on as detailed a scale as possible. Our field-scale GIS approach contributes knowledge and methodology which can be easily applied nationwide as the required input data for analysis are available. Owing to this we were able to avoid the main reason for the lack of many spatial-decision support systems that input data for models are unavailable, expensive, or difficult to collect. The study used GIS to store, modify, and analyse geographically distributed data. GIS has also been used for spatial-distribution estimation of biomass including available forest and agricultural crops residuals (Panichelli & Gnansounou, 2008; Shi et al., 2008) and cultivated energy crops (Förster et al., 2008). Since Mitchell (2000) analysed different computer models of bio-energy systems and suggested decision-support systems’ development should be aimed towards bio-energy application, development of more complex biomass-management tools has occurred (Batzias & Sidiras, 2005). Any decisions about the cultivation of appropriate energy crops for a given area are best taken at a regional or local level (Fischer et al., 2005), which the results of our paper support. The application of a large-scale soil map with the combination of soil–crop suitability models provided the framework for spatial bio-energy planning.

The density of abandoned fields in Tartu County is higher near the county's biggest urban area partly because of urban sprawl. Limited accessibility and unsuitable soils for traditionally cultivated crops could be the reasons for the high proportion of abandoned areas along the banks of the River Emajõgi and the shoreline of Lake Peipsi. The compositions of soils in currently used and in abandoned agricultural areas are remarkably different (Table II). The proportion of soils with low quality and limited suitability is higher in abandoned areas compared with used fields. Whereas Astover et al. (2006) verified a higher abandonment rate in regions with lower soil quality (at the level of municipalities), our research provides, for the first time, evidence of this phenomenon at a detailed spatial scale (the level of mapped soil polygons 1:10 000). This peculiarity indicates the necessity to consider site-specific soil information. In Estonia, the large-scale digital soil map is available for the entire land surface, but is still rather rarely used in the decision-making process because of the complexities of the decision-support systems and the limited knowledge of the decision makers. The development of GIS-based decision-support systems where specific soil criteria will be converted for stakeholders to more understandable format (i.e., to functional suitability maps) can contribute to overcoming this shortcoming. For instance, Förster et al. (2008) developed a site-optimized suitability model where a medium-scale soil map was the basis for biomass-production analysis. We used suitability models to estimate the agronomical and biological potential of energy grasses and short-rotation forestry on various soils. We could, thereafter, calculate the possible energetic value of the biomass potential from abandoned areas. We calculated the energy production of the selected energy grasses and energy forestry using the average yields but analysing bio-energy production in detail, yielding variation based on soil texture, water content, nutrient sufficiency, climatic conditions, and plantation age, should be taken into account. In the case of combined land-use strategy potential bio-energy production could cover approximately 20–25% of total energy consumption in the study area whereas reed canary grass could provide 24%, grey alder 20%, Caucasian goat's rue 18%, hybrid aspen 16%, willow 11%, and natural grassland 11% to the energy grid. The relative significances of these different crops must be handled provisionally because they depend on some fixed assumptions (i.e., 30% of abandoned land will remain as natural grassland) and on soil–crop suitability. However, for more complex analysis several additional criteria should be included to the suitability analysis – technical, economical barriers, policy impulses, environmental restrictions, ethical values, etc. As these parameters influence bio-energy implementation, EU structural funds could be used to relieve some, for example, economic barriers (Streimikiene & Klevas, 2007).

Structural and communion funds considerably prioritize subsidizing energy production from renewable (including biomass) and alternative energy sources. Estonia's government has also adopted legislative measures (for example, excise tax exemptions for biofuels and biomass, CO₂ tax for combustion installations, etc.) for supporting bio-energy implementation. Psychological barriers include public awareness about bio-energy cultures and farmers’ hesitancy towards bio-energy profitability.

The average size of abandoned field parcels in the study area is relatively small (by a factor of about seven compared with used fields), they are often fragmented and dispersed around the county. Hence the re-use of abandoned areas could have several technical and economical limitations and must be included in further analysis. Increased interest in bio-energy production indicates the necessity of evaluating production costs. Since Noon & Daly (1996) pointed out the importance of choosing power-plant location, further studies include distance calculations (Voivontas et al., 2001; Panichelli & Gnansounou, 2008; Shi et al., 2008). The maximum cost-effective transportation distance in Finland in the case of reed canary grass is estimated at 60 km (Pahkala, 2007). Our research does not include any distance calculations of potential plantations to available power plants. Therefore the further planning process of bio-energy production on abandoned areas in Estonia should be supplemented with distance calculations and economic criteria.

The precise spatial determination of abandoned agricultural land resource forms a solid basis for further bio-energy suitability analysis without reducing existing food and fibre production. This is especially important for Estonia and also for other Eastern European countries with a high proportion of abandoned land. The current study concerning the allocation of suitable areas for bio-energy crops considers only the use of abandoned fields, because use of these areas does not have any negative effect on Estonia's level of food self-sufficiency. This is an important consideration since Estonia's agricultural self-sufficiency became negative in 1997 (Rask & Rask, 2004) and is currently caused also by low productivity. Increased land-use efficiency and crop productivity from utilized fields would be the basis for decreasing negative self-sufficiency. The use of abandoned agricultural areas for bio-energy production can help to improve the overall profitability of the agricultural sector and promote the economic stabilization of rural regions. Re-using abandoned agricultural areas could also ease the level of unemployment in rural areas (Streimikiene & Klevas, 2007) since the employment rate between 1990 and 2005 decreased in Estonia's agriculture and hunting sectors from 16.6 to 3.9%.

Soil–crop suitability analyses serve the basis for knowledge-based allocation of bio-energy production. As the soils on abandoned fields have a lower quality with limited suitability, the consideration of local pedological conditions is crucial. Our study provides a basis and framework to develop GIS-based decision-support systems for multi-criteria site-optimized bio-energy planning.

Acknowledgements

We acknowledge the input of Estonian University of Life Science organized BOUA Summer School, How to write a Scientific article. We also thank Marcus Denton for the correction of language.