• farming systems analysis;
  • livestock;
  • sustainable land management;
  • less favoured areas;
  • common agricultural policy


  1. Top of page

Since the late 1980s, sustainable land management is one of the objectives of the European Commission in Less Favoured Areas. In this paper, we investigate the economic and environmental sustainability of farming systems in two less favoured areas in Centro and Alentejo areas of Portugal. The specific objectives were the following: (i) to characterise the farming systems; (ii) to analyse their development over a 20-year period (1989–2009); and (iii) to investigate to what extent these farming systems contribute to sustainable land management. The diversity of the farming systems was identified through a survey and cluster analysis and compared with the Farm Accountancy Data Network classification on types of farming. Indicators on the economic and environmental sustainability were estimated, namely, farm net income, return to labour and rotation management, on the basis of a survey, Farm Accountancy Data Network database and Landsat imagery, respectively. Results indicate an increased focus on livestock in the past 20 years (1989–2009). In Centro, rotation management was not affected. The small ruminant farms have been able to retain a positive farm net income but that was only possible with a below average return to labour. In Alentejo, the increased focus on livestock, cattle in particular, led to an intensification of fodder production on certain plots. Mixed crop–livestock farms show a negative farm net income since 1995 and depend heavily on subsidies to remain viable. As other studies in southern Europe have shown, farm strategies have often been directed towards lowering labour inputs, lowering forage deficits through on-farm produced resources and acquiring subsidies. Copyright © 2013 John Wiley & Sons, Ltd.


  1. Top of page

The European Commission would like to maintain environmentally friendly farming in its less favoured areas (LFAs), because that could help to preserve habitat diversity, enhance soil fertility and allow for the maintenance of firebreaks (EC, 1997). LFAs include mountain areas, areas in danger of abandonment and areas affected by specific natural handicaps (EC, 1999). However, the farming systems in LFAs are under constant pressure to improve their productivity in order to be able to compete with more intensive systems in areas with fewer limitations (de Graaff et al., 2011). It is therefore important to assess the extent to which these systems remain sustainable. In the context of sustainable land management (SLM), sustainability has been conceptualised as a combination of technologies, policies and activities integrating socio-economic and environmental concerns in order to reach simultaneously the productivity, security, protection, economic viability and social acceptability objectives (Smyth & Dumanski, 1993; Hurni, 2000).

Until now, few studies have analysed the sustainability of Portuguese farming systems. Some have tackled the classification of farming systems (Baptista et al., 1991), whereas others have looked into the profitability of specific farm enterprises (Pearson et al., 1987; Fleskens et al., 2009). But very few have analysed specific farm types, with their integrated crop and livestock enterprises and their development over time, in order to capture the dynamics of farming systems with a focus on sustainability. In this paper, a farming systems approach has been used to characterise the past and present combinations and main features of the crop and livestock enterprises on certain farm types and to assess their productivity, economic viability and environmental sustainability.

Historically, Farming System Research has evolved from an approach focused on production economics towards a holistic approach that considers the farm as a system integrated within a broader hierarchy of systems (Ruthenberg, 1980; Byerlee et al., 1982; Norman, 2002). The striking evidence, around the 1980s, of the different degrees of success of particular innovations (e.g. mechanisation and fertilisers) in different socio-economic and biophysical settings led development practitioners to this conceptual revision (Simmonds, 1985) and to cater for the need for ‘bottom-up’ approaches and implementation, one also referred to Farming Systems Research and Development. In this paper, we focus on the analysis of the role of farm managers with their farming systems on SLM.

In the past two decades, a significant part of agricultural land in Portugal has been converted to open forest land, which includes shrubby vegetation resulting from land abandonment and post-fire forest regeneration and new forest areas resulting from afforestation. The outcome of this conversion seems to have led to an increase of land degradation in some LFAs (Jones et al., 2011). This leads to several questions: What land use developments at the farm level have led to this situation and which farm types have mostly contributed to this? Could the increased land degradation be due to a higher stocking rate of cattle and/or small ruminants and as a result to shorter fallow periods?

Ruthenberg (1980) classified farming systems on the basis of such criteria as, among others: proportion of inputs produced inside the farm system (e.g. own produced animal feed), type of rotation (e.g. natural fallow systems including ley systems), and intensity of rotation, showing the extent of cropping versus fallow over the years. The rainfed crop–livestock systems are largely based on some arable and permanent cropping, some (agro) forestry, and most importantly on a combination of fodder crops and intensive and extensive grazing systems. The contribution of natural forage to the total feed consumption on a farm was used by Porqueddu (2007) to classify low-input farming systems in southern Europe. The change of these systems, with its impact on the environment, has been brought about by several strategies, ranging from pure abandonment to intensification (MacDonald et al., 2000; Caballero et al., 2007; Abu Hammad & Tumeizi, 2012; Thapa & Yila, 2012). Intensification has quite well-known effects, and abandonment can hinder the sustainability of extensive livestock systems, with socio-economic (Beaufoy et al., 1994; Caballero et al., 2007; Porqueddu, 2007) and environmental impacts (Beaufoy et al., 1994; Moreira et al., 2005; Bento-Gonçalves et al., 2012). In this paper, we hypothesised on one hand that higher stocking rates could well have reduced fallow periods and thereby increased land degradation for some farm types; whereas for others, low farm net income may soon lead to abandonment.

The objectives of this paper are therefore as follows: (i) to characterise the present farming systems as practised by specific farm types in two LFAs in Portugal; (ii) to characterise their development over time in the past two decades; and (iii) to investigate their contribution to SLM. Broader scale analysis of land-use changes across the world (Foley et al., 2005) and Europe (Bouma et al., 1998; Frost et al., 2007) have pointed out the need for small-scale detailed exploratory studies that might support sustainable land-use policy design. Ultimately, our goal is to illustrate the implications of the differential policy support of farming systems for the SLM in LFA.


  1. Top of page

Research Areas

For this research, two areas are selected, which are part of the LFAs and are both prone to desertification risk (Jones et al., 2011). In the rest of the paper, the names Centro and Alentejo will be used to indicate these research areas. Centro includes three municipalities: Mação, Proença-a-Nova and Vila Velha de Rodão, and Alentejo includes the municipality of Mértola.

Biophysical features

Centro research area (112,000 ha) lies on the border of the subhumid climatic zone. Average annual rainfall ranges between 700 and 1400 mm and is distributed over 50–75 days mainly between September and March (Figure 1). The average temperature lies between 12·5 and 15 °C. The most common soil types are eutric Lithosols and hortic Luvisols (CNA/SROA, 1978).


Figure 1. Research areas location: (A) Centro research area and (B) Alentejo research area. Mean precipitation, mean temperature and potential evapotranspiration (ETP). This figure is available in colour online at

Download figure to PowerPoint

Alentejo research area (128 Mha) lies in the semiarid climatic zone. The average annual rainfall ranges between 400 and 600 mm, distributed over 50–75 days also mainly in winter, and the average mean temperature is between 15 and 17·5 °C. The most represented soil types are eutric Lithosols and ferric Luvisols (CNA/SROA, 1978).

Socio-economic features

Centro population is about 19,000, agricultural employment is 3% and 95% of farm labour is provided by the family. Alentejo population is about 7,000, agricultural employment is 21% and 60% of farm labour is provided by the family (INE, 2011a, 2011b). Concerning farm income, only 2% of the farms in Centro and 29% in Alentejo provide the farm household with a main source of income. In fact, most of the farms in both areas have a total output lower than €4,000/year (INE, 2010). Still, 96% of the farmers in Centro and 89% in Alentejo depend mainly on farming activities as their main source of income, next to the income from pensions and salaries. Although secondary sources of income of farm households are difficult to grasp directly from one single statistical source, more than 50% of the jobs are provided by the service sector in both areas (INE, 2011a, 2011b). On average, only one annual work unit (AWU – work of one person full time) in Centro and 1·1 AWU in Alentejo are employed on the farm around the year (INE, 2010).

Land and livestock endowments

Land and livestock endowments are quite distinct in Centro and Alentejo farms. While in Centro, about 3,000 farms manage 11·4 Mha of agricultural land with 1,000 livestock units (LUs – equivalent to one adult dairy cow producing 3,000 kg of milk annually, without additional concentrated foodstuffs); in Alentejo, about 700 farms manage 90 Mha of agricultural land with 11,670 LU. Agricultural area is devoted mainly to permanent crops (55%), arable crops (25%) and permanent pastures (20%) in Centro and to permanent pastures (70%) and arable crops (30%) in Alentejo. In Centro, nearly all the livestock is composed of small ruminants; whereas in Alentejo, 35% consists of cattle (INE, 2010).

Over the past two decades, livestock composition has changed significantly, likely because of market prices and common agricultural policy (CAP) subsidies. While in Centro in 1989, the goat to sheep ratio was about 3 : 1; in 2009, almost all livestock is composed of sheep. In Alentejo, goats have completely disappeared and the small ruminants to cattle ratio is now 2 : 1, whereas it was 10 : 1 20 years ago (INE, 2010). Although land-use change has favoured the increase of permanent pastures at the expense of arable crops (Jones et al., 2011), the question remains whether the management of these pastures in terms of stocking rates and fallow periods is favourable or not for sustainability in terms of reduced land degradation.


For the analysis of the present farming systems, we used primary data collected through a farm survey. We interviewed 17 and 22 households in Centro and Alentejo, respectively, covering about 12% of the agricultural area in each research area. The survey was conducted by means of single contact interviews and covered the following topics: (i) characterisation of the farmers and their household; (ii) characterisation of the farm; (iii) soil management practices; (iv) animal production practices; (v) type of farm support schemes; and (vi) estimate of farm income.


A cluster analysis was performed (SPSS/PASW 18, IBM Corp., New York, USA) with the information on land and livestock endowments per farm and the type of off-farm activity. Taking into account previous farming system analysis (Pearson et al., 1987; Baptista et al., 1991), we used the following criteria for the cluster analysis: arable area (ha), permanent crops (ha), forest with less than 30 years (ha), forest with more than 30 years (ha), goats (n°), sheep (n°), cattle (n°) and farmer's off-farm type of activity. Similar cases were grouped by computing the furthest squared mean distance on the standardised z-scores of the variables. Through this procedure, we ensured a stronger link of each farm case to its assigned cluster and that variables with different units could be compared (Field, 2005). Although the sample size is small, and clusters may not be very homogeneous, we think that it is useful to make some distinction between several farm types, with different characteristics and focus. Farm economic results such as farm net income and return to labour were computed for each group. Farm net income was obtained by deducting costs from the sum of output and subsidies and return to labour by dividing the sum of farm net income and labour input by amount of labour (FADN, 2010). The information on the direct aid payments per farm obtained through the survey was combined with the official public data from Instituto de Financiamento da Agricultura e Pescas (IFAP), available at

Farming Systems Over Time (1989–2009)

For the characterisation of the development of the farming systems over time (1989–2009), we analysed the main changes in land and farm characteristics per type of farming as classified by the Farm Accountancy Data Network (FADN). The type of farming of a farm holding is determined by the relative contribution of the standard output of the different characteristics of this holding to the total standard output of this one (EC, 2008). The type of farming reflects the characteristics of structure and management. In this paper, we will use the concept of type of farming as a proxy of farming systems in order to analyse farming systems' changes over time.

Two data sources were used: the National Agricultural Census (10-year base) and the National Accountancy Network (FADN) (annual base). The first dataset reflected the agricultural area under each type of farming at the submunicipality level for the years of 1989, 1999 and 2009. Crop farms included among others farms specialised in arable crops and olives, livestock farms included among others farms specialised in small ruminants and mixed farming included mainly crop–livestock farms. In this way, we have obtained the main trend in the change of farming systems over the past two decades (1989–2009). For a detailed view, we selected small ruminants and crop–livestock farms as the most significant both in terms of agricultural area and changes over this period.

The second dataset consisted of accountancy data of ten farms continuously surveyed by FADN during the period 1989–2009 in and closely to Centro and Alentejo research areas. This analysis shows the results in terms of return to labour and farm net income over the period 1989–2009 for the two main types of farming: small ruminants and crop–livestock farms.

Sustainability of Farming Systems

As rotation and fallow are the main practices used by farmers for conserving soil fertility, understanding the rotation dynamics is important to assess farming systems' sustainability. Because of different land and livestock endowments and level of farm income, farmers may use these practices differently. Therefore, we also investigated stocking rates and feed supply based on survey data and rotation management through the computation of Normalised Difference Vegetation Index (NDVI) – an indicator of vegetation greenness. The NDVI is a robust and widely used method to measure vegetation productivity (Wang et al., 2001). The index uses the reflectance of vegetation in the red and near-infrared (NIR) channels of the light spectrum. The index is equal to (NIR − RED)/(NIR + RED). Bare soil is characterised by NDVI values between −0·1 and 0·2, whereas dense vegetated surfaces show a variation between 0·5 and 0·8 (Carlson & Ripley, 1997). In the analysis, only the arable plots of Centro and Alentejo surveyed farms were selected. NDVI average values were computed for January to April images, when maximum greenness was expected. The minimum values for that average were assessed, and in view of those results, 0·3 was judged as a good determinant to distinguish cropping from fallow years.

Landsat 4–5 Thematic Mapper (TM) and Landsat Enhanced TM + images were acquired for the two areas for the period 2001–2010. Earlier images were not available. The quality of all images was checked manually by visual interpretation, and empty data points were not taken into consideration in the analysis, that is why 2005 was not considered in Centro analysis.

Rainfall, temperature and even fallow management (e.g. Yamamoto et al., 2009) are also important for vegetation growth but not examined in great detail, as the current method was adequate to estimate the number of cropping years over the decade (2001–2010), as well as the consecutive number of fallow and cropping years. These variables were first assessed per plot, and their mode was determined over each farm group. The analysis was performed with spatial data from the land-parcel information system. Table 1 summarises the methodological set-up.

Table 1. Methodological set-up
ObjectivesMethodological steps
  1. FADN, Farm Accountancy Data Network; NDVI, Normalised Difference Vegetation Index; IFAP, Instituto de Financiamento da Agricultura e Pescas.

Analysis of the actual farming systems1. Grouping of surveyed farms based on land and livestock resources and enterprises and off-farm activities
2. Description of management characteristics per farm type
Farming systems development over the past 20 years (1989–2009)3. Characterisation of farm types based on census data for 1989, 1999 and 2009 (INE)
4. Characterisation of the change in farm management for each farm type, on the basis of accountancy data (1990–2009) (FADN);
5. Determination of FADN farm types per group of farms in the sample
Investigate farming systems contribution to sustainable land management6. Characterisation of rotation and fallow practices based on qualitative information from survey data and vegetation index (NDVI) trends informed by land use information per plot (IFAP)


  1. Top of page

Characterization of Present Farm Types

Centro farm types

The respondents were mainly men (80%) with secondary school education, who started farming in the mid-1980s. The majority of them share the management with another member of the family: usually father and son (50%) or both spouses (30%). The household comprises on average three persons, the couple and one child who considers being successor on the farm, in case no other jobs would become available. On average, only one permanent worker was employed. Concerning land endowments, the surveyed farms ranged from 3 ha up to 628 ha, whereby fodder crops occupied from 45% to 100% of the farm land. Olives are the most significant permanent crop. The 17 surveyed farms were grouped in four clusters described in Table 2.

Table 2. Farmer and farm characteristics per farm type – Centro
 C1 (N = 9)C2 (N = 4)C3 (N = 2)C4 (N = 2)
  1. Source: own survey November 2010.

  2. a

    Variation coefficients are the least for the age variable (15%) and highest for arable over permanent area ratio (>100%).

  3. b

    We have considered renewed forest stand where not less than 30% of the forest was younger than 30 years.

Age of farmer (years)aOld (61)Old (59)Young (44)Young (44)
Farm activityPart-timeFull-timeFull-timeFull-time
Live on-farmYesNo
Farm size (ha)Small (59)Small (122)Medium (205)Large (339)
Agricultural area: forest area1 : 11 : 11 : 03 : 1
Arable crops area: permanent crops area3 : 11 : 11 : 59 : 1
Fodder crop area (% of arable crop area)1001007590
Permanent pasture area (% fodder crop area)51343
Livestock (LU)Goats (12)Sheep (18)Sheep (45)Sheep (170)
ForestbDegraded pineRenewed montadoEucalypts

The first group (C1) comprises part-time relatively old farmers, living on the farm, retired from other activities or with jobs not related to farming. The farm has on average 26 ha (±28) of agricultural land and 33 ha (±24) of forest (figures in brackets are standard deviations). Nearly three quarters of agricultural land is devoted to feed production (Table 2) to sustain the herd mainly composed by goats for milk production. These small goat farmers practice a semiextensive husbandry that includes grazing of shrubby vegetation in areas in between the pasture fields. They also manage forest, mostly degraded maritime pine with more than 30 years of age. The second group of farmers (C2) includes full-time, relatively old, small farmers, occasionally also involved in marketing of farm products. On average, the farm includes 62 ha (±81) of agricultural land and 60 ha (±44) of forest. The agricultural area is evenly divided into arable and permanent cropping (Table 2). In these farms, sheep rearing is combined with permanent crops (olives) and fodder crops production, which occupy the whole arable area. The forest area comprises mainly plantations of less than 30 years of age (80%), mainly installed with the support of the afforestation measure a decade ago (EC, 1992; EC, 1999). We will designate this group as small sheep farmers.

The third group (C3) integrates full-time, relatively young, medium-sized farmers, usually also involved in cheese making. On average the farm size is 205 ha (±50), all devoted to agriculture. The share of olives in the overall land use is the highest (Table 2). About 75% of arable area is devoted to feed production with triticale or oats occasionally harvested for grains or left for ley after hay cutting. In addition, fodder crops are also grown under the canopy of olives. Livestock mainly consists of sheep for milk production and is kept in a semiconfined system with paddocks near the shed. We refer to this group as extensive sheep farmers.

The fourth group of farmers (C4) includes full-time, relatively young, large farmers. On average, the farm includes 250 ha (±308) of agricultural land and 89 ha (±101) of forest. About 90% of agricultural land is used for arable cropping, with most of it used for animal feed production (Table 2). The remainder was used for industrial crops, such as tobacco and oil seeds. These farmers practise intensive sheep rearing for both milk and meat production. Eucalyptus spp. stands for pulp production constitute most of the forest area. Forest management is undertaken under contractual agreements with the pulp industry that pays a fixed rent to the owner. We refer to this group as intensive sheep farmers.

Economics of Centro farm types

Because of the relatively small farm size of the first three farm types, only intensive sheep farms (C4) seem to be highly productive with a high farm net income and a high return to labour despite the high costs (Table 3). In fact, if we consider the regional average annual wage of €12,000 (INE, 2011a, 2011b), a threshold value under which farm abandonment becomes likely, only intensive sheep farms (C4) and small sheep farms (C2) are clearly viable in financial terms (Table 3). It is mainly thanks to the subsidies that small goat farming (C1) and extensive sheep farming (C3) become financially viable (Table 3).

Table 3. Management results per farm type – Centro
 Part-time small goat farmers (C1; N = 9)Full-time small sheep farmers (C2; N = 4)Full-time extensive sheep farmers (C3; N = 2)Full-time intensive sheep farmers (C4; N = 2)
  1. LU, livestock unit; AWU, annual work unit.

  2. Source: own survey November 2010.

  3. a

    Stocking rate is defined here as the ratio between total livestock units and agricultural area per agricultural holding.

  4. b

    AWU is equivalent to the work of one person full time in an agricultural holding (Eurostat glossary).

  5. c

    On the basis of IFAP database, rural development subsidies include afforestation, agri-environmental and less favoured areas payments.

  6. d

    On the basis of conservative estimates of livestock payments per LU, €85 per small ruminant LU and €200 per cattle LU.

Agricultural land (ha)2662205251
Rented land (ha)001400
Stocking rate (LU/ha)a0·50·30·20·7
Permanent labour (AWU)b0·20·51·03·5
Total inputs (€1,000)81822156
Total labour costs (%)38394519
Feed costs (%)25335067
Total output (€1,000)133938325
Total subsidies including SFP (€1,000)c71826126
Rural development subsidies (%)578342
Livestock payments on total (%)d1581511
Farm net income (€1,000)133942298
Return to labour (€1,000/AWU)12302673
Return to labour – without support (€1,000/AWU)7181345

While for small goat and sheep farmers (C1, C2), rural development payments are the most significant subsidies; for extensive and intensive sheep farmers (C3, C4), the single farm payments (submitted to cross-compliance obligations) are the most important (Table 3). Livestock payments represent a minor part of subsidies, but these are also partly integrated in the single farm payments as a result of the midterm CAP reform in 2003, which resulted in decoupling production from payments.

Small goat and sheep farmers (C1, C2) do not use hired labour. In order to cope with peak labour needs, small goat farmers (C1) use voluntary work from relatives often rewarded with a provision of quality products. Self-provision of quality products, pensions and other business revenues are significant for sustaining farming activity. Only two of these farmers plan to expand their activity, one with permanent pastures and the other with agrotourism. Small sheep farmers (C2) would like to decrease their livestock activities and rely more on permanent crops. However, as stated by two of them, they should then also undertake the marketing of their products. For the other two, sheep production offers, as a secondary production activity, a positive contribution to farm image.

Extensive sheep farmers (C3) employ on average one permanent worker (Table 3). For these farmers, not labour but land seems to be the limiting factor. In order to cope with this, they establish an informal labour system based on renting land from old olive farmers. While obtaining extra land for pasture, they provide assistance to old olive farmers during harvest time. This also helps old olive farmers to comply with cross-compliance requirements. With regard to the future, extensive sheep farmers (C3) find themselves in a ‘deadlock’. Because they depend on milk quantity to keep the cheese making running, they find it difficult to cope with lower yields as a result of organic farming or integrated production. Differently from small goat and sheep farmers (C1, C2) who benefit from the regional breed scheme, these extensive sheep farmers (C3) have to make use of improved breeds for milk production which activity receives less support.

Intensive sheep farms (C4) employ on average three permanent workers (Table 3). Although farm size allows for the internalisation of a significant part of animal feed production (Table 2), feed costs represent 67% of total (Table 3). Similarly to the previous type of farming, the future of their farming activities will be determined by the evolution of sheep milk prices.

Livestock activities generate the highest share of output for most farm types in Centro area, except for the small sheep farms (C2) (Table 4). With olive output, the composition of their output corresponds to a mixed cropping type of farming as defined by FADN (EC, 2008). It is interesting to notice that mixed cropping allows smaller farm types (C1, C2 and C3) to obtain a sufficiently high return to labour without subsidies (Tables 3 and 4).

Table 4. Total output (in €1,000) by farm type and enterprise – Centro
Unit = €1,000Part-time small goat farmers (C1; N = 9)Full-time small sheep farmers (C2; N = 4)Full-time medium sheep farmers (C3; N = 2)Full-time intensive sheep farmers (C4; N = 2)
  1. Source: own survey November 2010.

  2. a

    Forest output was not included; therefore, the total output of part-time small goat farmers is lower than in Table 3.

Arable cropping120105
Permanent crops42634
Total outputa113938325
Type of farmingSmall ruminantsMixed croppingSmall ruminantsSmall ruminants

With regards forest revenues, only small goat farmers (C1) derived output from forest in 2009. Although owning mostly degraded pine stands, they also own eucalypts that they manage directly. Small sheep farms (C2) and intensive sheep farms (C4), who own a considerable area with eucalypts as well, have rented the area to the pulp industry and receive every 9 years the result of wood sales at a contracted price. Extensive sheep farmers (C3) do not own any forest.

Alentejo farm types

The respondents were mainly men (90%) with high school attendance who manage their farms since the mid-1980s. The household comprises also on average three persons, the couple and one child. Farm size ranged from 236 ha up to 1250 ha. Permanent cropping occupies a very limited area on all farm types and consists almost exclusively of old olive orchards with no commercial production. Fodder crops constitute from 50% up to 90% of arable land. Most of the forest area results from recent afforestation projects (less than 30 years old) (EC, 1992; EC, 1999) mainly with stone pine (Pinus pinea L.), whereas the remainder consists of already established agroforestry – montado. About half the respondents managed their farm with family members. The 22 surveyed farms were grouped in five clusters described in Table 5.

Table 5. Farmer characteristics per farm type – Alentejo
 A1 (N = 5)A2 (N = 6)A3 (N = 3)A4 (N = 2)A5 (N = 6)
  1. Source: own survey November 2010.

  2. a

    Variation coefficients are the least for farm size (20%) and highest for permanent pasture over fodder area ratio (>100%).

  3. b

    We have considered renewed forest stand where not less than 30% of the forest was younger than 30 years.

Age of farmer (years)aYoung (52)Young (53)Young (51)Old (61)Young (52)
Farm activityPart-timeFull-timeFull-timeFull-timeFull-time
Live on-farmYesNoNo
Farm size (ha)Medium (515)Medium (437)Medium (696)Small (287)Large (963)
Agricultural area: forest area13 : 16 : 12 : 11 : 94 : 1
Arable crops area: permanent crops area57 : 11 : 01 : 06 : 111 : 1
Arable crops area: fallow area1 : 21 : 31 : 11 : 1
Fodder crop area (% of arable crops area)65509070
Permanent pasture area (% of fodder crop area)14415151
Livestock (LU)Cattle (108)Sheep (94)Cattle (171)Sheep (23)Mixed sheep + cattle (283)
ForestbRenewed montadoRenewed montadoOld montadoNew pineRenewed montado

The first farm type (A1) comprises mainly part-time farmers with an off-farm activity in the domain of agrotourism, often living on the farm. The average farm size is 515 ha (±243), being almost exclusively devoted to arable cropping and evenly distributed between grains (triticale), fodder crops (oats and ryegrass) and area left to ley. Ley is the regrowth of fodder crops, very often the same species cultivated for grain, but cut for hay instead. The ley area is accounted for under fodder crops, which occupies about 65% of arable area (Table 5). With fertilisation, ley area is kept from 2 up to 5 years, being in this case accounted as permanent pasture. Although herds are usually mixed with cattle and sheep, cattle are predominant. These part-time cattle farmers own the smallest share of forest that is mainly renewed montado.

The second group (A2) consists of full-time farmers with medium-sized farms (437 ± 112 ha), of which about 85% is agricultural area evenly devoted to grain and fodder crops. About 40% of fodder area is permanent pasture. The main activity of these farms is sheep rearing for meat production combined with feed production in open oak stands (montado). Oak acorn offers also an alternative source of feed from November to February, when stubble or ley has been completely grazed and pastures installed in October are still emerging. Contrary to the traditional system, animals are kept semiconfined in paddocks. This allows a rotational management; where for each hectare used, three are left fallow. The share of fallow is the highest share among all farm types. Forest consists mainly of renewed montado. We designate this group as sheep farmers.

The third group (A3) includes full-time farmers who established their farming activity with the support of a young farmer project. Most live outside the farm. On average, the farm has 696 ha (±56) that includes about one third of forest (Table 5). About 90% of agricultural land consists of fodder crops and about half of this is permanent pasture. Fallow and cropped areas are about the same size. Although herds are usually mixed with cattle and sheep, cattle dominates. Forest area combines new and traditional oak plantations, where old montado (Quercus rotundifolia Lam.) is predominant. We designate these farmers as full-time cattle farmers.

The fourth group (A4) comprises relatively old farmers who have converted most of their farm land (287 ± 71 ha) to forest with the support of the European Union (EU) afforestation measure (EC, 1992; EC, 1999). They have no arable crops at the moment, and sheep are kept in the old montado area. They pay a forest contractor to conduct forest maintenance, an obligatory requirement in order to receive the subsidy. New oak area largely surpasses old montado area; still, stone pine is the predominant species in afforested areas. We designate these farmers as retired farmers with new forest.

The last group (A5) includes full-time farmers, occasionally involved in the marketing of meat products. Most of them do not live on the farm. They manage large farms (963 ± 207 ha), which include about one fifth of forest area. About 70% of agricultural land is devoted to fodder crops from which about half is permanent pasture. For each hectare of cropped area, another hectare is left to fallow. Herds are usually mixed with sheep and cattle. The forest area is dominated by renewed montado, where new oak stands represent nearly 60% of montado area. We designate these farmers as large mixed livestock farmers.

Economics of the Alentejo farm types

All farm types, except full-time cattle farmers (A3), manage to obtain a sustainable farm net income and return to labour, partly thanks to subsidies. Mixed livestock farms (A5) have by far the best economic results. Compared with the regional average annual wage of €10,000 (INE, 2011a, 2011b), only part-time cattle farmers (A1) and mixed livestock farmers (A5) manage to obtain a good return to labour (Table 6). It is interesting to notice that these farms contract the highest amount of permanent labour.

Table 6. Management results per farm type – Alentejo
 Part-time cattle farmers (A1; N = 5)Sheep farmers (A2; N = 6)Full-time cattle farmers (A3; N = 3)Forest retired farmers (A4; N = 2)Mixed livestock farmers (A5; N = 6)
  1. LU, livestock unit; AWU, annual working unit.

  2. Source: own survey November 2010.

  3. a

    Stocking rate is defined here as the ratio between total livestock units and agricultural area per agricultural holding.

  4. b

    AWU is equivalent to the work of one person full time in an agricultural holding (Eurostat glossary).

  5. c

    On the bass of IFAP database, rural development subsidies include afforestation, agri-environmental and less favoured areas payments.

  6. d

    On the basis of conservative estimates of livestock payments per LU, €85 per small ruminant LU and €200 per cattle LU.

Agricultural land (ha)47737144929759
Rented land (ha)1011500284
Stocking rate (LU/ha)a0·20·30·30·20·3
Permanent labourers (AWU)b0·80·20·70·54·5
Total inputs (€1,000)372918813141
Labour costs (%)2474735
Feed costs (%)32558643
Total output (€1,000)76371256269
Total subsidies (€1,000)c46356642132
Rural development subsidies (%)7202710033
Livestock payments on total (%)d37284426
Farm net income (€1,000)8543334260
Return to labour (€1,000/AWU)523972956
Return to labour – without support (€1,000/AWU)269−33132

While for part-time cattle farms (A1), rural development subsidies constitute only 7% of total subsidies; for mixed livestock farmers (A5), they represent 33%, and for forest retired farmers, even 100%. Sheep farmers (A2) and forest retired farmers (A4) depend on subsidies to make their operations viable. In fact, for sheep farmers (A2), the amount of subsidies received is about the same as total output, and forest retired farmers (A4) subsidies are almost the sole contributions to farm income. Full-time cattle farms' (A3) return to labour is just positive even with subsidies and without those it would be negative. This is due to very high feed purchases, and they cannot compensate that enough with their output. For these farms, livestock payments constitute the highest share of direct payments.

In a characterisation of extensive livestock systems in Europe by Moreira & Coelho (2010), an extensive grazing system in medium to large private farms is defined as a type of extensive grazing existing in south Iberian Peninsula that relies on an increasing substitution of labour by costly investments in fences and automatic water points. To some extent, we identify this trend in full-time cattle farms (A3), where input costs exceed largely the output, and feed costs represent 86% of the total input (Table 6). Despite the restricted area left to fallow (one for each cropped hectare) and the high proportion of grazing area in arable land (90%) (Table 5), livestock feed requirements have to be supplemented with a high amount of purchased feed. This is in part due to the higher number of cattle. Unlike sheep, cattle cannot be allowed to graze in sown pastures in late spring because of the soil disturbance they would cause; so, they have to be provided with feed. Although this is more costly in terms of feed, it avoids hiring labour to manage the herd from paddock to paddock. Beyond labour and feed requirements, livestock subsidies and market prices play a significant role in farmers' decisions to favour cattle over sheep; as a farmer explained quite clearly: ‘where a cow eats, five sheep could eat instead; however, the subsidy is at least ten times more for the cow, and lamb meat prices are roughly the same as 10 years ago’. These results are in line with the research of Pearson et al. (1987), Coelho & Reis (2009) and Madeira (2008).

With regard to labour, both part-time and full-time cattle farmers (A1, A3) hire one permanent worker. Part-time cattle farmers (A1) tend to have the help of relatives in cropping activities. This gives them extra time to run the agrotourism business. Three out of five part-time cattle farmers (A1) were sheep farmers before 2008. In two cases, this change involved the conversion of grain crop area into sown pasture area. The third farmer reduced labour and land rented from others. In the present system in which high feed requirements are met with external feed sources cattle rearing is less land and labour demanding than sheep rearing. Also, two out of three full-time cattle farmers (A3) reduced sheep herds to buy cattle. They have a business attitude and do not see themselves as ‘nature managers’ as they claim the actual policy framework wants to make of them.

Mixed livestock farmers (A5) hire the highest number of workers, about four persons, whereas sheep farmers (A2) and forest retired farmers (A4) hire the lowest number of workers. In the case of mixed livestock farmers, more labour is needed because of large farm size and more variety of activities. Mixed livestock farmers (A5) have invested in more permanent cropping and improved permanent pastures with clover spp. However, in order to cope with high maintenance costs, two of them have partially substituted permanent pastures by sown pastures with mixed grain crops, which can be harvested as concentrates for own use or for the market when prices are profitable.

Four out of six sheep farmers (A2) have reduced their herds substantially after 2003. The only milk producer, who also owns a cheese making unit, wants to sell the farm. Part of the herd reduction is not yet reflected in direct payments.

Forest retired farmers (A4) hire forest contractors, who provide labour and technical equipment, to conduct forest maintenance operations. They have given up sheep rearing and count on the collaboration of relatives to supervise forest operations. The majority of afforestation projects are reaching maturity within 5 years and payments will then stop, which may cause problems in complying with minimum forest cleaning operations.

For most farm types, arable crops and livestock activities contribute equally to output that results in a crop livestock type of farming (e.g. A1, A2, A3) (Table 7). In large mixed livestock farms (A5), permanent crops contribute with a significant share to total output that results in a mixed cropping farming type. Farms with herds almost exclusively composed of cattle or sheep such as those of cattle farmers (A1, A3) and sheep farmers (A2) seem to have a lower return to labour than mixed herds such as those of mixed livestock farms (A5) (Tables 6 and 7).

Table 7. Total farm output (in €1,000) by farm type and enterprise – Alentejo
Unit = €1,000Part-time cattle farmers (A1; N = 5)Sheep farmers (A2; N = 6)Full-time cattle farmers (A3; N = 3)Forest retired farmers (A4; N = 2)Mixed livestock farmers (A5; N = 6)
  1. Source: own survey November 2010.

Arable cropping311565045
Permanent crops0000136
Total output76371256269
Type of farmingCrop–livestockCrop–livestockCrop–livestockN/AMixed cropping

Farm Types Over Time (1989–2009)

In this section, we will use the type of farming accounted in national statistics as a generalisation of farming systems. In last section, we have characterised the present farm types and we have determined the types of farming to which they were more closely associated with (Tables 4 and 7). In Centro, we identified three farm types as small ruminant farms (C1, C3 and C4) and one farm type as mixed cropping farm (C2). In Alentejo, three farm types were identified as crop–livestock farms (A1, A2, A3) and one as mixed cropping (A5). For retired forest farms (A4), the concept type of farming was not applicable.

Overall, during the past two decades and in both research areas, farms specialised in livestock activities (e.g. small ruminant farms) have increased their business at the expense of farms with a mixed production pattern (e.g. crop–livestock farms)

In both research areas, small ruminant farms are the most represented among livestock farms. In Centro, they occupy 26% of agricultural area (11,363 ha) and own 78% of total LUs (1,006 LU); whereas in Alentejo, these shares are 30% (90,018 ha) and 39% (11,674 LU), respectively (INE, 2010). Since 1989, this represents more than a sevenfold increase in Centro and twofold in Alentejo in terms of area; whereas in terms of total LUs, it represents more than a twofold increase in both research areas (INE, 1990).

In contrast, crop–livestock farms that occupy 18% of the area in Centro and 15% in Alentejo; registered during the same period, a drop of more than one third in Centro and two thirds in Alentejo. In terms of total LUs, the drop was 88% and 65%, respectively (INE, 1990; INE, 2010).

With regard to grazing, not only the share of pastures increased but also those with a permanent character increased both in small ruminant and crop–livestock farms in the two research areas (Table 8). Overall, this has led to a decrease of stocking rates in both areas. For example in small ruminant farms, stocking rates decreased from 0·9 to 0·3 LU/ha in Centro and from 0·4 to 0·2 LU/ha in Alentejo (Table 8). These trends favour SLM. The increased allocation of arable land to permanent pasture contributes to decreased soil erosion by reducing the number of tillage operations and by providing vegetation cover throughout the year. Decreased stocking favours a more balanced management of pasture area, avoiding overgrazing and soil compaction.

Table 8. Livestock and grazing area characteristics per type of farming – 1989–2009
  1. Source: INE, 1990; INE, 2000; INE, 2010.

  2. a

    Grazing area includes permanent pastures, temporary pastures and fodder crops; agricultural area includes arable crops, permanent crops and grazing area.

Small ruminants/total livestock (%)      
Small ruminant farms1009310010010094
Crop–livestock farms729493887571
Grazing area/agricultural area (%)a      
Small ruminant farms755177603982
Crop–livestock farms182525453351
Permanent pastures/grazing area (%)      
Small ruminant farms333256868993
Crop–livestock farms52959859196
Stocking rate (LU/ha)      
Small ruminant farms0·91·40·30·40·60·2
Crop–livestock farms1·21·30·20·30·50·3

It is important to notice that, although not shown in the table, the majority of small ruminants owned by crop–livestock farms in Centro are sheep, which to a large extent have substituted goats, particularly in the last decade. This is also a reason for the increase of permanent pasture, because sheep only graze, whereas goats also browse (on shrubs).

In Alentejo, crop–livestock farms have been replacing small ruminants for cattle. The share of small ruminants (almost exclusively sheep) has decreased from 88% in 1989 to 71% in 2009 (Table 8), whereas the share of cattle increased from 12% to 29%. The research of Madeira (2008) on the two livestock systems for meat production in Alentejo area shows that as result of the Agenda 2000 CAP reform, support for cattle increased from €20/ha in 1992 to €136/ha in 2004; whereas for sheep, the increase was far more modest, from €69/ha in 1992 to €72/ha in 2004. Unlike the increase in the share of permanent pasture in the total grazing area and the decrease of stocking rates, the increase of sheep over goats in Centro and the increase in cattle over small ruminants in Alentejo, although to some extent with the use of regional breeds, has a rather detrimental effect on SLM as will be shown further on this paper.

Economics of small ruminant and crop–livestock farms over time (1989–2009)

The assessment of the economics of the two main types of farming over time focused on the analysis of the return to labour and farm net income.

Over the past two decades (1989–2009), small ruminant farms in Centro had in most years a return to labour (even with support) below the national minimum wage (Figure 2). This has been partially sustained by labour with low opportunity cost (e.g. part-time workers and retired persons) and farming rationales other than market orientation. Nevertheless, recently, it seems that return to labour is slightly increasing, mainly as a result of a decrease in variable costs and an increase of farm income. Although the total level of support has remained nearly the same since 1992, the share of rural development payments in total payments has increased, which seems to have had a positive effect on farm net income. Recently, with the exception of 2005, which was a particularly dry year (IM, 2012), farm net income without subsidies shows an increasing trend.


Figure 2. Main farm results of small ruminant farms in Centro [four farms in Farm Accountancy Data Network (FADN) sample] over the period 1991–2009 (data source: GPP/FADN database Portugal).This figure is available in colour online at

Download figure to PowerPoint

In Alentejo, crop–livestock farms had on average a return to labour (with support) slightly above the national minimum wage, despite the decreasing trend in farm net income (Figure 3). In fact, since 1996, farm net income without subsidies has been negative. Thanks to the direct payments, these farms can continue their activities. Rural development subsidies form a minor share of direct payments for these farms, except between 2000 and 2002 when they benefited from specific agri-environmental measures mainly targeting permanent pastures. Variable costs, which during the whole period seem to have been correlated with the amount of aid, are now apparently stabilising.


Figure 3. Main farm results in crop–livestock farms in Alentejo [six farms Farm Accountancy Data Network (FADN) sample] over the period 1991–2009 (data source: GPP/FADN database Portugal). This figure is available in colour online at

Download figure to PowerPoint

Farming Systems and Sustainable Land Management: Rotation and Fallow Practices

We have characterised earlier the different farm types in terms of their present resource endowments and management results. Small ruminants for milk production and mixed herds of sheep and cattle for meat production are the main activities of the specific farm types distinguished, respectively four in Centro and five in Alentejo. In the previous section, we characterised the change over time of these two main types of farming. In the present section, we will try to understand the rotation dynamics of the installation of fodder areas, which is a key issue for the sustainability of farming systems on the basis of livestock production in both environmental and economic terms. The analysis will focus on the farm types identified earlier.

Although the trend over the past 20 years indicates that the overall share of permanent pastures in grazing area is increasing, the different farm types are adopting distinct strategies to cope with the feeding needs of livestock. Table 9 shows the amount of money spent on feed purchases and the value of feed produced on-farm by the different farm types identified in Centro area.

Table 9. Sustainability indicators regarding feed consumption and fodder area per livestock unit – Centro
 Part-time small goat farmers (C1; N = 9)Full-time small sheep farmers (C2; N = 4)Full-time extensive sheep farmers (C3; N = 2)Full-time intensive sheep farmers (C4; N = 2)
  1. Sources: own survey November 2010; Landsat images 2001–2011.

  2. a

    The value of the feed produced on-farm does not include the value of the grass consumed fresh (grazing).

  3. b

    Includes montado area and arable crops without fallow area.

  4. c

    It was not possible to obtain an image for 2005.

Feed produced on-farma (€1,000 per farm)12054
Feed purchased (€1,000 per farm)2611104
Feed self-sufficiency (% in value)3423234
Average fallow years3333
No. cultivation years (in 9)c455

Small goat farms (C1) and intensive sheep farms (C4) have the highest share of feed produced on-farm (34% of feed expenditure), yet they have different strategies regarding feed provision (Table 9). While the first manage to be highly self-sufficient by assigning a large area to each LU (4 ha/LU), the second uses land the most intensively, assigning only 1 ha/LU. With regard to fallow duration, the different intensity in land utilisation seems not to result in shorter fallow periods, as all farm types manage to leave land to fallow 3 years in a row. This is also confirmed by the analysis of satellite images through the estimates of NDVI, which identify a similar number of cultivation years for all farm types within the decade, usually 4–5 years of cultivation (Table 9).

On extensive sheep farms (C3), the value of feed produced on-farm represents the lowest share among Centro farm types, only 2% (Table 10) that matches also with the lowest share of arable land devoted to fodder crops (75%) (Table 2). Small sheep farmers (C2) manage to produce on-farm about 20% of feed requirements, half of them by organic and integrated production methods. These farmers devote the largest share of fodder crops to permanent pastures (13%, Table 2).

Table 10. Sustainability indicators regarding feed consumption and grazing area per livestock unit – Alentejo
 Part-time cattle farmers (A1; N = 5)Sheep farmers (A2; N = 6)Full-time cattle farmers (A3; N = 3)Forest retired farmers (A4; N = 2)Mixed livestock farmers (A5; N = 6)
  1. Sources: own survey November 2010; Landsat images 2001–2011.

  2. a

    The value of the feed produced on-farm does not include the value of the grass consumed fresh (grazing).

  3. b

    Includes montado area and arable crops without fallow area.

  4. c

    One part-time cattle farm and two sheep farms were not located and therefore not considered in NDVI analysis.

Feed produced on-farma (€1,000 per farm)24156145
Feed purchased (€1,000 per farm)121616260
Feed self-sufficiency (% in value)67472743
Average fallow years3434
No. cultivation years (in 10)c4467
No. consecutive fallow years3333
No. consecutive cultivation years3133

In Alentejo, the identified farm types use the land more intensively than in Centro. In order to provide feed for one LU all farm types use less than 3 ha. Understandably, cattle farmers have the largest area per LU, 2·6 ha/LU on full-time cattle farms (A3) and 1·7 ha/LU on part-time cattle farms (A1) (Table 10). However, with the larger area per LU, full-time cattle farmers (A3) provide the lower value of feed, 27% against 67% on part-time cattle farms (A1). Although one might conclude that full-time cattle farms (A3) are less intensive than part-time cattle farms (A1), it is not the case; because within the last decade (2001–2011), the whole plots showed to have been cultivated for 6 years against only 4 years on part-time farms (Table 10). In the sheep farms (A2), the area devoted to each LU is the lowest among all farm types (1·1 ha/LU). Still, they manage to provide about half the feed expenditure with on-farm production while leaving to fallow the highest share of land (3 ha for each cropped hectare, Table 5). This seems to indicate a higher suitability of sheep to land and climate conditions.

All farm types manage to include 3 years of fallow in their rotations. In fact, the mode of the maximum number of consecutive years of fallow identified within the decade (2001–2011) is 3 years for all farm types. Within the decade, full-time cattle farmers (A3) and mixed livestock farmers (A5) seem to use land more intensively as the total number of cropping years is 6 and 7 years, respectively (Table 10).

Our research on the respective farm types in two areas in Portugal and their economic and environmental sustainability brought to light the narrow margin of their economic viability in terms of return to labour, their strong dependence on public handouts and their trend towards specialisation in detriment of mixed farming systems while preserving extensive features. Porqueddu (2007) shows similar trends in extensive grassland farming systems in southern Europe and Caballero et al. (2007) do so for LFAs in Europe. These authors consider that these systems result from traditional systems that made modifications to overcome the socio-economic and environmental constraints, mostly related to labour and forage deficit. The adopted strategies to overcome such constraints have been to lower labour inputs (Caballero, 2001), to lower forage deficit with on-farm produced resources (Porqueddu, 2007) and to acquire subsidies (Caballero et al., 2007).

Traditional grassland farming systems have been changing through the replacement of grazing by hay cutting (MacDonald et al., 2000; Caballero, 2001) and fencing (Caballero et al., 2007; Moreira & Coelho, 2010). This translates into changing grazing patterns with localised concentration of livestock around farmsteads and intensification of forage production on better quality plots, whereas others are being abandoned (MacDonald et al., 2000). Our results are not conclusive with regard to grazing patterns, but through the assessment of the number of cultivation years within the decade 2001–2011, we were able to identify localised intensification at the farm level at least for two farm types in Alentejo: full-time cattle farmers (A3) and mixed livestock farmers (A5). Although the sequence of images was not complete for Centro, the results seem to indicate that there is no localised intensification.

The distortion effect of livestock payments in extensive grassland farming systems has been widely reported (Beaufoy et al., 1994; Caballero et al., 2007; Madeira, 2008). Our results illustrate that situation quite well. On the one hand, small goat and sheep farmers (C1, A2) represent the most traditional systems that make good use of poor pasture resources, which are common on idle land in Centro and on heavily degraded soils in Alentejo. On the other hand of the intensification «spectrum» are intensive sheep farmers and full-time cattle farmers (C4, A3) who rely mainly on purchased feed. These farm types depend mainly on first pillar CAP payments and on the strategy followed by farmers for maximising these subsidies. Somewhere in between are the other farm types either adjusting feed production with more ley (C3, A1), feed requirements with less cattle (A5) or investing in supplementary activities (e.g. quality olive oil) with the support of rural development payments (C2). We have shown that these strategies have not the same value in terms of SLM.

In Centro, it seems important to maintain grazing practices of poor pasture resources for vegetation control in order to avoid wildfires; the replacement of goats for sheep and the increased focus on specialised livestock farming seems to contradict that expectation.

Extensive grazing, particularly with goats, contributes to a well-managed forest-pasture mosaic in depopulated mountainous areas where shrub encroachment is the trigger of recurrent fire events (e.g. Álvarez-Martínez et al., 2013). Álvarez-Martínez et al. (2013) found that in Spanish Cantabrian Mountains, most benefits were perceived from extensive grazing in combination with other practices (e.g. trimming and prescribed fires). Other solutions have been tried out for the immediate intervention after fire occurrence such as the application of hydromulching – a mixture of seeds, wood fibbers, a surfactant, nutrients, a natural biostimulant and a green colourant (Prats et al., 2013). In Centro, although the intervention proved effective at the plot level, it is expensive and not yet completely established for large areas (Prats et al., 2013). Such interventions might be justified for the control of flood risk that can be an issue for pine and eucalypt forests after fire when the already existing soil water repellency seems to increase (e.g. Stoof et al., 2011; Santos et al., 2013). Therefore, providing subsidies for the maintenance of extensive grazing farms can be cost-effective for safeguarding landscape production and regulation services at stake.

In Alentejo, although the maintenance of permanent pasture is positive to protect the already degraded soils, its association with high feed requirements such as those of cattle farms might cause an over-intensification on other plots of the farm if the feed value of these pastures is not properly targeted.

For crop–livestock farms in Nigeria, Thapa & Yila (2012) also showed that farmers tended to choose management practices that they perceived as bringing the highest return to labour. Although it could be desirable to have more livestock in order to have a larger pool of manure to integrate into the soils, the lack of labour might be a constraint to do so. An alternative intensification path is needed for these traditional systems that are still operating.

The need to redirect subsidies to the support of sensible management alternatives that might render these systems more sustainable in the future is also a recurrent recommendation (Beaufoy et al., 1994; Caballero et al., 2007). A good example is the improvement of self-sown legume-based pastures, which constitutes a long-term lowering cost strategy (Porqueddu, 2007) and an opportunity for the rehabilitation of degraded land (Porqueddu et al., 2013). This recommendation could be of use for farm types where pasture efficiency was not at its best, for example, in the case of extensive sheep farmers in Centro (C3) and full-time cattle farmers in Alentejo (A3). At the EU level, a number of agri-environmental measures were designed to support beneficial traditional practices (e.g. the upkeep of permanent pastures), which have been abandoned over the years in the course of intensification biassed policies (Barbayiannis et al., 2011; Calatrava et al., 2011; Prager et al., 2011; Prosperi et al., 2011). Although the design of these incentive-based measures potentially delivers the desired benefits, a better target at the farm level management needs to be put into practice in order to deliver the intended benefits (Louwagie et al., 2011; Posthumus et al., 2011). A measure for the improvement of permanent pastures would have a high initial cost and would have to be well articulated with livestock payments and cross-compliance measures. This recommendation stems also from the research of Kutter et al. (2011) on the EU-27 policy measures with a soil conservation focus.


  1. Top of page

This paper focused on the analysis of land management at the farm level in order to find out whether land degradation was due to higher stocking rates and shorter fallow periods.

The analysis shows that in spite of an increased focus on livestock activities at the expense of mixed farming stocking rates decreased and the share of permanent pastures increased. Livestock payments in particular for cattle seem to have encouraged high expenditures on external inputs (e.g. full-time cattle farmers, A3), whereas rural development payments seem to have encouraged more sustainable strategies such as the improvement of yields of mixed farming systems (e.g. small sheep farmers, C2, and mixed livestock farmers, A5). This is also perceived from the trend of farm net income composition over time (1989–2009) constructed with the national FADN data.

In Centro, goat and extensive sheep farms (C1, C3) showed to have lower returns to labour and goat farms (C1) were in fact unsustainable without subsidies. Intensive sheep rearing (C4) showed to be viable on large farms managing to internalise feed production costs. Permanent cropping with olives contributes to the higher returns to labour and farm net income of mixed cropping farms (C2). In these farms, about 20% of the total amount spent on feed is met with on-farm production. For each LU, 3 ha is available and about 13% of the area devoted to fodder crops is maintained with permanent pastures.

In Alentejo, all farm types manage to remunerate labour above regional average annual wage, except full-time cattle farms (A3) that present a negative return to labour (€−33,000/AWU) depending on subsidies to stay viable. Mixed herds of sheep and cattle were responsible for the higher returns to labour and farm net income of mixed cropping farms (A5). In these farms, about 40% of the amount spent on feed is produced on-farm. About 1·5 ha is devoted to each LU, and about 51% of fodder crops area is maintained with permanent pastures.

While in Centro, the change from perennial to arable crops (for animal feed) seems not to have affected the rotation management; in Alentejo, the increased focus on cattle did lead to a higher number of cultivation years and hence seems to indicate a trend towards less sustainable land use. In order to support economically viable and environmentally sustainable farming systems, future policy design should take into consideration the suitability of livestock and pasture associations. In the past 20-year period (1989–2009), the most economically viable and environmentally sustainable systems in Centro and Alentejo seem to have been associated with mixed cropping farming types.


  1. Top of page

We would like to thank Engineer Fernando Delgado, Dr José Ferragolo da Veiga, Engineer Ciel Cileno, Engineer Teresa Laia and Engineer Patricia Guerreiro for facilitating the contacts with farmers. We are also grateful to Engineer Rita Araújo, Engineer Luis Ramos, and Engineer Maria da Luz Mendes, for providing secondary data. This work has been financially supported by the Fundação da Ciência e Tecnologia (SFRH/BD/43214/2008) and EU-project Desire (EU-36087-FP6-SUSTDEV) (website: We also wish to thank the Editors and two anonymous reviewers for their valuable help in improving this manuscript.


  1. Top of page
  • Abu Hammad A, Tumeizi A. 2012. Land degradation: socioeconomic and environmental causes and consequences in the Eastern Mediterranean. Land Degradation & Development 23: 216226. DOI: 10.1002/ldr.1069.
  • Álvarez-Martínez J, Gómez-Villar A, Lasanta T. 2013. The use of goats grazing to restore pastures invaded by shrubs and avoid desertification: a preliminary case study in the Spanish Cantabrian Mountains. Land Degradation & Development. DOI: 10.1002/ldr.2230.
  • Baptista FO, Baptista MP, Caldas JC, Coelho IS, Lourenço F, Moreira MB, Novais A, Rodrigo I. 1991. Zonagem e caracterização dos principais tipos de agricultura no continente – sistemas de produção dos concelhos de castelo branco, idanha a nova, penamacor e vila velha de ródão. Centro de Economia Agrária e Sociologia Rural da UTL: Lisboa, Portugal.
  • Barbayiannis N, Panayotopoulos K, Psaltopoulos D, Skuras D. 2011. The influence of policy on soil conservation: a case study from Greece. Land Degradation & Development 22: 4757. DOI: 10.1002/ldr.1053.
  • Beaufoy G, Baldock D, Dark J. 1994. The nature of farming: low intensity farming systems in nine European countries. Institute of European Environmental Policy: London, UK.
  • Bento-Gonçalves A, Vieira A, Úbeda X, Martin D. 2012. Fire and soils: key concepts and recent advances. Geoderma 191: 312.
  • Bouma J, Varallyay G, Batjes NH. 1998. Principal land use changes anticipated in Europe. Agriculture, Ecosystems & Environment 67: 103119.
  • Byerlee D, Harrington L, Winkelmann DL. 1982. Farming systems research – issues in research strategy and technology design. American Journal of Agricultural Economics 64: 897904.
  • Caballero R. 2001. Typology of cereal-sheep farming systems in Castile–La Mancha (south-central Spain). Agricultural Systems 68: 215232.
  • Caballero R, Riseth JA, Labba N, Tyran E, Musial W, Molik E, Boltshauser A, Hofstetter P, Gueydon A, Roeder N, Hoffmann H, Moreira MB, Coelho IS, Brito O, Gil A. 2007. Comparative typology in six European low-intensity systems of grassland management. Advances in Agronomy 96: 351419. DOI: 10.1016/S0065-2113(07)96001-0.
  • Calatrava J, Barberá GG, Castillo VM. 2011. Farming practices and policy measures for agricultural soil conservation in semi-arid Mediterranean areas: the case of the Guadalentín basin in southeast Spain. Land Degradation & Development 22: 5869. DOI: 10.1002/ldr.1013.
  • Carlson TN, Ripley DA. 1997. On the relation between NDVI, fractional vegetation cover, and leaf area index. Remote Sensing of Environment 62: 241-252.
  • CNA/SROA. 1978. Carta dos Solos (CARTA III.1), 1: 1,000,000. Accessible at. Atlas do Ambiente, Agência do Ambiente - Ministério do Ambiente, do Ordenamento do Território e do Desenvolvimento Regional on March 2010
  • Coelho IS, Reis P. 2009. Pastoralismo Mediterrâneo: competitividade, sustentabilidade dos territórios e diversificação da economia rural. Pastagens e Forragens 29/30: 99117.
  • EC. 1992. Council Regulation (EEC) No. 2080/92 of 30 June 1992 instituting a community aid scheme for forestry measures in agriculture. Official Journal of the European Communities L 215: 96-99.
  • EC. 1997. Fact sheets environment: towards a greening of the common agricultural policy.
  • EC. 1999. Council Regulation (EC) No 1257/1999 of 17 May 1999 on support for rural development from the European Agricultural Guidance and Guarantee Fund (EAGGF) and amending and repealing certain regulations. Official Journal of the European Communities L 160: 80102.
  • EC. 2008. Commission Regulation (EC) No 1242/2008 of 8 December 2008 establishing a community typology for agricultural holdings. Official Journal of the European Union L 335: 324.
  • FADN. 2010. Farm Accountancy Data Network: an A to Z methodology. FADN: EC DG Agriculture, Brussels; Belgium.
  • Field A. 2005. Discovering statistics using SPSS (and sex, drugs and rock'n' roll). SAGE Publishers: London, Thousand Oaks, New Delhi.
  • Fleskens L, Duarte F, Eicher I. 2009. A conceptual framework for the assessment of multiple functions of agro-ecosystems: a case study of Tras-os-Montes olive groves. Journal of Rural Studies 25: 141155. DOI: 10.1016/j.jrurstud.2008.08.003.
  • Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Stuart Chapin F, Coe MT, Daily GC, Gibbs HK, Helkowski JH, Holloway T, Howard EA, Kucharik CJ, Monfreda C, Patz JA, Colin Prentice I, Ramankutty N, Snyder PK. 2005. Global consequences of land Use. Science 309: 570574. DOI: 10.1126/science.1111772.
  • Frost LC, Willems E, Lathuy C, Calvo-Iglesias MS. 2007. An assessment of landscape heterogeneity in the European Union using CORINE land cover 2000 and LUCAS survey data XI Congreso Internacional de Ingenieria de Proyectos. Lugo: Spain.
  • de Graaff J, Kessler A, Duarte F. 2011. Financial consequences of cross-compliance and flat-rate-per-ha subsidies: the case of olive farmers on sloping land. Land Use Policy 28: 388394. DOI: 10.1016/j.landusepol.2010.08.001.
  • Hurni H. 2000. Assessing sustainable land management (SLM). Agriculture, Ecosystems & Environment 81: 8392. DOI: 10.1016/s0167-8809(00)00182-1.
  • IM. 2012. Situação de seca meteorológica, 15 de março 2012., accessed in 30-03-2012
  • INE. 1990. National Agricultural Census 1989, data selection from GPP for the research areas – Ministry of Agriculture and Decenial statistics Portugal, Agriculture Census historical series http: accessed on 22/09/2011
  • INE. 2000. National Agricultural Census 1999, data selection from GPP for the research areas - Ministry of Agriculture and Decenial statistics Portugal, Agriculture Census historical series accessed on 22/09/2011.
  • INE. 2010. National Agricultural Census 2009, data selection from GPP for the research areas – Ministry of Agriculture and Decenial statistics Portugal, Agriculture Census historical series accessed on 22/09/2011
  • INE. 2011a. National Population Census 2011, dados preliminares accessed on 16/10/2011
  • INE. 2011b. Regional Statistics Yearbook 2010, decenial statistics Portugal, Agriculture Census historical series accessed on 06/10/2011
  • Jones N, de Graaff J, Rodrigo I, Duarte F. 2011. Historical review of land use changes in Portugal (before and after EU integration in 1986) and their implications for land degradation and conservation, with a focus on Centro and Alentejo regions. Applied Geography 31: 10361048. DOI: 10.1016/j.apgeog.2011.01.024.
  • Kutter T, Louwagie G, Schuler J, Zander P, Helming K, Hecker J-M. 2011. Policy measures for agricultural soil conservation in the European Union and its member states: policy review and classification. Land Degradation & Development 22: 1831. DOI: 10.1002/ldr.1015.
  • Louwagie G, Gay SH, Sammeth F, Ratinger T. 2011. The potential of European policies to address soil degradation in agriculture. Land Degradation & Development 22: 517. DOI: 10.1002/ldr.1028.
  • MacDonald D, Crabtree JR, Wiesinger G, Dax T, Stamou N, Fleury P, Gutierrez Lazpita J, Gibon A. 2000. Agricultural abandonment in mountain areas of Europe: environmental consequences and policy response. Journal of Environmental Management 59: 4769. DOI: 10.1006/jema.1999.0335.
  • Madeira JP. 2008. A política agrícola comum e o percurso dos sistemas de agricultura de sequeiro no sul do Baixo Alentejo. Master degree thesis, Universidade Técnica de Lisboa. Lisboa.
  • Moreira MB, Coelho IS. 2010. Determinants of change on extensive livestock systems: a theoretical framework. Rivista di Economia Agraria 65: 487499.
  • Moreira F, Pinto MJ, Henriques I, Marques T. 2005. The importance of low-intensity farming systems for fauna, flora and habitats protected under the European ‘birds’ and ‘habitats’ directives: is agriculture essential for preserving biodiversity in the Mediterranean region? In Trends in biodiversity research, Burk AR (ed.). Nova Science Publishers: New York, USA.
  • Norman DW. 2002. The farming systems approach: a historical perspective 17th Symposium of the International Farming Systems Association. IFSA: Lake Buena Vista, Florida.
  • Pearson SR, Avillez F, Bentley JW, Finan TJ, Fox R, Josling T, Langworthy M, Monke E, Tangermann S. 1987. Portuguese agriculture in transition. Cornell University Press: Ithaca.
  • Porqueddu C. 2007. Low-input farming systems in southern Europe: the role of grasslands for sustainable livestock production Low-Input Farming Systems: an opportunity to develop sustainable agriculture JRC scientific and technical reports.
  • Porqueddu C, Antonio Re G, Sanna F, Piluzza G, Sulas L, Franca A, Bullitta S. 2013. Exploitation of annual and perennial herbaceous species for the rehabilitation of a sand quarry in a Mediterranean environment. Land Degradation & Development DOI: 10.1002/ldr.2235.
  • Posthumus H, Deeks LK, Fenn I, Rickson RJ. 2011. Soil conservation in two English catchments: linking soil management with policies. Land Degradation & Development 22: 97110. DOI: 10.1002/ldr.987.
  • Prager K, Schuler J, Helming K, Zander P, Ratinger T, Hagedorn K. 2011. Soil degradation, farming practices, institutions and policy responses: an analytical framework. Land Degradation & Development 22: 3246. DOI: 10.1002/ldr.979.
  • Prats SA, Malvar MC, Vieira DCS, MacDonald L, Keizer JJ. 2013. Effectiveness of hydromulching to reduce runoff and erosion in a recently burnt pine plantation in central Portugal. Land Degradation & Development DOI: 10.1002/ldr.2236.
  • Prosperi P, Terres JM, Doublet S, Pointereau P. 2011. Conservation agriculture effects and policy support to mitigate soil degradation in Midi-Pyrennes (France). Land Degradation & Development 22: 7083. DOI: 10.1002/ldr.1021.
  • Ruthenberg H. 1980. Farming systems in the tropics. Clarendon Press: Oxford.
  • Santos JM, Verheijen FGA, Tavares-Wahren F, Wahren A, Feger K-H, Bernard-Jannin L, Rial-Rivas ME, Keizer JJ, Nunes JP. 2013. Soil water repellency dynamics in pine and eucalypt plantations in Portugal – a high-resolution time series. Land Degradation & Development DOI: 10.1002/ldr.2251.
  • Simmonds NW. 1985. Farming systems research. A review. World Bank: Washington, D.C.
  • Smyth AJ, Dumanski J. 1993. FESLM: an international framework for evaluating sustainable land management. FAO: Rome, Italy.
  • Stoof CR, Moore D, Ritsema CJ, Dekker LW. 2011. Natural and fire-induced soil water repellency in a Portuguese shrubland. Soil Science Society of America Journal 75: 22832295. DOI: 10.2136/sssaj2011.0046.
  • Thapa GB, Yila OM. 2012. Farmers' Land management practices and status of agricultural land in the Jos Plateau, Nigeria. Land Degradation & Development 23: 263277. DOI: 10.1002/ldr.1079.
  • Wang J, Price KP, Rich PM. 2001. Spatial patterns of NDVI in response to precipitation and temperature in the central Great Plains. International Journal of Remote Sensing 22: 38273844. DOI: 10.1080/01431160010007033.
  • Yamamoto Y, Oberthür T, Lefroy R. 2009. Spatial identification by satellite imagery of the crop-fallow rotation cycle in northern Laos. Environment, Development & Sustainability 11: 639654. DOI: 10.1007/s10668-007-9134-z.