The selected biodiversity indicators focus on three aspects: species richness, diversity of production and reforestation. These indicators are closely related to two of the major environmental problems associated with mono-cultural patterns of agriculture identified by the Cuban government, i.e. loss of biodiversity and deforestation (CITMA 1997).
The converted farms were characterised by the presence of large numbers of plant and animal species, i.e., about six times those at the beginning of the study (Table 2). Grain crops, root and tuber crops, vegetables, tree species, and new pasture and forage species were introduced in the design of the mixed farms. This allowed adaptation of the animal ration in the course of the year in response to seasonal climate patterns, especially rainfall, and the associated fluctuations in pasture production, one of the major problems in tropical livestock production systems (Funes 1979).
The Margalef index, as a measure of species richness, combines the total number of species in the system and the total number of individuals and reached values of 9.1 and 10.4 on the converted farms, thanks to the large number of species present (44 and 52, respectively), compared to only eight pasture species in year 0 and a corresponding index of 1.6 (Table 2). This index provides a more meaningful measure of the diversity at farm level than one only accounting for the total number of species. The large number of plant and animal species was associated with a large diversity in production (17 and 23 products, respectively), compared to only milk and beef before the conversion (Table 2).
Both farms were characterised by large numbers of trees per hectare (131 and 204, respectively), due to the establishment of trees as forage sources for animals, as well as for living fences and fruit production. Trees are an important component in MFS in the tropics. Research in Cuba and the Central American region (Benavides 1998; Hernández et al. 2001) has revealed increases in milk and meat production, and improvements in animal welfare in livestock systems following introduction of trees, especially leucaena and other leguminous species. Our results indeed indicate that trees, as major components of MFS diversification, had a positive effect on farming system productivity in terms of milk yield, energy and protein output, as tree products such as leaves, were essential components of the animal ration. Moreover, due to the deeper rooting of trees, nutrients can be ‘pumped’ from the sub-soil (Breman and Kessler 1995).
The indicators of diversity of production and reforestation are both expressed in the Shannon index, which combines either the number of products or of tree species (diversity) with the yield per product or the number of individuals per species (abundance). Shannon indices tend to be higher when the distribution of species and individuals is more even, and for relatively diverse natural ecosystems may rank between 3 and 4 (Gliessman 2001). In our mixed farms, high values of the indices of diversity of production (1.7–2) and reforestation (1.5–1.7) were attained, compared to year 0, when diversity
production was 0.2 and trees were absent. They were also appreciably higher than the values (up to 0.48) calculated for hypothetical multicropping agro-ecosystems, with two or three species and high evenness (Gliessman 2001).
Application of the Shannon and Margalef diversity indices, originally developed for evaluation of natural ecosystems, for analysis of agro-ecosystem diversity might lead to increased insight in the contribution of crop and animal diversification to the improvements in productivity, efficiency and financial indicators of mixed systems.
The increase in plant diversity also affected diversification in other aspects. In our two mixed farms, 15 natural enemies controlling potential pests have been identified (Pérez-Olaya 1998). Perennial crops, such as grasses, gliricidia [Gliricidia sepium, (Jacq.) Kunth ex Walp.] and leucaena acted as alternative hosts for natural enemies of crop pests. These observations are in line with those of Vandermeer et al. (1998) and Altieri (1999), i.e. system diversification stimulates emergence of natural enemies controlling pests, contributing to sustainability of agricultural systems.
Moreover, soil fauna biodiversity and the activity of soil biota (diplopods and worms) have been shown to increase following conversion to MFS (Rodríguez 1998).
Productivity is probably the most extensively used indicator in agronomic performance analyses. This study took into account four indicators for productivity of the farm: milk production per unit farm area and per unit forage area, and total energy and protein output.
Milk yield per unit farm area was somewhat higher than before the conversion to mixed farming (Table 2), although up to 50% of the total farm area was used for arable and horticultural crops, and therefore not directly for producing animal feed. This increase was the result of the introduction of various innovations in the mixed farms; e.g. cultivation of high-yielding perennial forages, grass–legume associations and leguminous trees and use of crop residues as animal feed, resulting in more and better quality animal feed throughout the year. This also led to a high milk yield per unit forage area after conversion (Table 2).
Given that the Cuban government has defined the social mandate of the dairy sector as: ‘to produce milk for children, elderly and sick people’, increasing milk production is a political priority. However, biophysical and socio-economic constraints have reduced current total milk production in Cuba to about one-third of that in the 1980s (González et al. 2004) and present-day average annual yields in specialised commercial dairy production units do not exceed 1 Mg of milk per hectare of farmland (MINAG 2006). In commercial dairy farming, based on pasture and medium levels of concentrates, under ‘outstanding management’, production up to 3 Mg per hectare is possible (García Trujillo 1983). In year 0 of this study, the original specialised system produced 1.8 Mg ha−1, while in the mixed farms, annual yields of 3.1 and 4 Mg per hectare forage area were attained (Table 2).
In terms of total production (expressed in energy and protein, the two main components in human nutrition), livestock products in the mixed farms exceeded the yields in year 0, on top of which crop products were harvested. The highest energy (27.1 GJ ha−1 year−1) and protein (191.3 kg ha−1 year−1) outputs (Table 2), achieved at farm C50, were associated with high ‘additional’ crop production.
Productivity can also be expressed in terms of the number of people that can be fed from the protein or energy output of a system. Averaged over the 6-year period, in farm C25 the energy produced was enough to adequately feed four people, with protein for up to five, while in C50 these numbers were six and eight, respectively. These numbers are about twice as high as reported in literature for medium-intensity specialised milk production systems (Spedding 1988; Beets 1990) and at least four times higher than currently achieved in the ‘Low External Input’ specialised dairy systems in Cuba.
Human labour productivity is an essential indicator in performance assessment of MFS strategies in dairy farms in Cuba, because of the scarcity of this ‘resource’ in rural areas. Although labour-intensive designs were implemented, in practice labour input gradually decreased over time on farm C25, while on farm C50 it showed a parabolic pattern with a maximum in year 3 (Fig. 2a). Concurrently, production was maintained and therefore labour productivity increased. The higher labour demand of both mixed farms in the first years can be attributed to the initially higher number of farm activities, such as sowing legumes in grazing areas, conversion of pasture into arable land, fencing, planting of trees, establishing the crop rotation system, weed control, etc. Over the 6 years, total labour input was lower in C25 than in C50, due to the smaller cropping area.
Our results are relevant for the three major segments of present livestock production in Cuba: (1) the growing sector of small producers that received land from the state in usufruct, currently about 400,000 (Granma 2006), each with up to 5 ha of land, managed with labour-intensive methods; (2) the small farmers sector, cultivating private land and producing individually or organised in cooperatives such as Credit and Services Cooperatives and Agricultural Production Cooperatives at intermediate levels of productivity, but in most cases at low levels of crop–livestock integration and (3) the Basic Units of Cooperative Production (UBPC) that started in 1993 under Law 142. This law regulated partitioning of previous state cattle holdings into smaller units, encouraging diversification and adopting a family farm model. In total, these three segments affect about 4.2 million ha of Cuba’s agricultural land. However, recent estimates set the area of abandoned land at roughly three million hectares, i.e. about half of Cuba’s agricultural area, belonging for the greater part to the UBPC and state enterprises. Two possible directions to reverse this development are promotion of either extensive or small-scale intensive livestock-crop-tree mixed farming with low environmental impact. Under both scenarios, many of the ‘Low External Input’, low labour-intensive and high-efficiency natural resource management practices implemented in the current study are applicable. However, further simplifying managerial activities continues to be a goal, considering that labour availability remains a primary constraint, as the population has moved out of the rural areas.
Increasing the efficiency of input use was identified as an important objective in the management of the prototype mixed farms. The small sizes of the two experimental farms allowed use of animal traction and intensive human labour, instead of mechanised operations. Human labour was the largest component in energy inputs on both mixed farms that were designed as labour-intensive management systems, with the other components (i.e. diesel and feedstuffs) accounting for about 20% of the total (Fig. 3). Energy input linearly decreased with time since establishment on farm C25, while on farm C50 it showed a parabolic pattern with a maximum in year 3, in parallel to the labour inputs (Fig. 2b), and was lower on farm C25, due to the smaller area devoted to crop production. Realizing high levels of production, at the lowest possible level of inputs (Hilhorst et al. 2001) would indeed be an advantage under the conditions of scarcity and uncertain supply of inputs prevailing in Cuba. This is a strong argument in favour of continuation of MFS, even when the economic situation improves.
Higher energy efficiencies on the mixed farms were primarily the result of transformation of part of the pasture area into arable crops, leading to an increase in total energy output and a reduction in total energy input (Table 2). Energy efficiency shows an increasing trend with time after conversion on both farms, associated with decreases in total energy input, mostly in the form of human labour, while energy output was stable (Fig. 2a–d).
In energy terms, protein was produced more efficiently in the mixed systems (i.e. lower energy costs of protein production than in the specialised system. Moreover, although energy efficiencies in animal and crop production systems have a different biological basis (Spedding 1988; Stout 1990), our results indicate that higher production of animal protein per unit forage area can be attained using MFS strategies. This type of farm-scale energy efficiency analyses is consistent with studies of Pimentel (2004) and Giampietro et al. (1994) who in sustainability analyses, focused on energy flows in food production at system level. Energy conversion analyses should not be considered as an alternative to financial analyses, but rather as a complement to better cover the complex web of interrelationships between finances and the environment in which food systems operate (Giampietro et al.
In countries where fossil energy is abundantly available or where the use of high energy inputs is subsidised, energy-intensive farming systems do not face many technical constraints. However, for countries such as Cuba, where energy and/or capital are scarce resources, energy efficiency is a critical issue for national food security (Funes-Monzote and Monzote 2001). Furthermore, economic considerations such as high oil prices on the international market and environmental issues such as global warming associated with CO2 emissions, and the pollution of water and air, are leading societies worldwide to demand more responsible use of fossil energy. High dependence on fossil fuels is generally considered an indicator of low sustainability. Renewable energy alternatives such as biogas, wind power, solar energy, biomass and biofuels, have high potential applications for the development of energy self-sufficient agricultural systems (Pimentel et al. 2002).
Our two mixed farms achieved higher gross margins and higher benefit-cost ratios than the specialised farm (Table 3). This is the result of the inclusion of arable crops, the high productivity per unit farm area, and the higher prices for crop products than for milk and meat (Appendix 1). Therefore, increasing whole farm income by selling crop products in regions where arable farming is possible, might be a suitable strategy for supporting cattle operations and making dairy farming more attractive. This is in line with the results presented by De Koeijer et al. (1995) and Thomson et al. (1995) who have indicated financial advantages of MFS, as a result of a more intensive use of natural resources and beneficial interactions between crop and livestock production.
The total value of production was higher in the two mixed systems than in the specialised dairy system in year 0, but the total costs of production were also higher, associated with the higher labour costs and the capital demand to establish the crop production activities (Table 3). Economic incentives are important to sustain or to increase the population in rural areas. However, lack of incentives and centralised top–down decisions constrain development of the dairy sector. The price of milk for the consumer in the vulnerable sectors of the population is set to 0.25 CUP/litre by the government, the only official milk processor and retailer, while the producer price is set to about 1.00 CUP/litre, which is low, considering the costs of production. Therefore, milk production is a low-income activity, not economically attractive for producers. While a reduction in the cost price of milk is difficult to realise in low external input DFS, in MFS, milk production tends to become more feasible when combined with other, highly profitable activities such as cash crop and fruit production.
The results of this study are not in contradiction with the national policy of prioritisation of the dairy sector. To be politically acceptable, any diversification strategy should first demonstrate that it does not negatively affect the ‘main goal’ of producing milk, associated with the ‘social mandate’ given to livestock enterprises. Hence, any MFS strategy should be able to produce milk with ‘minimal environmental damage’ and at low costs in external inputs.
Moreover, if economic or political changes lead to price increases for milk and meat, other goals, related to environmental protection and sustainable rural development will be sufficiently important to retain mixed farming on Cuba’s future agricultural agenda.
Farms in the UBPCs are increasingly turning towards prioritising diversification for self-sufficiency (feeding workers and their families at low costs and selling possible surpluses in local or external markets to improve their financial sustainability), which makes these results even more relevant. Other emergent activities that might be combined in diversified MFS such as agro-tourism, nature conservation and education are also attractive options and need to be seriously considered. However, as indicated before, structural changes and economic incentives are necessary to stimulate the return of people to the rural areas and make economic use of available land. Our results show that the importance of the financial impact of adopting MFS to promote changes in Cuban agriculture should not be underestimated.
Soil fertility of the Ferralsols in year 0 was classified as medium. According to DNSF (1982), the content of SOM was low and pH moderately to slightly acid. Levels of available P and exchangeable K+ were medium, while the sum of exchangeable cations (SEC = base saturation) was half the ‘typical’ values for this type of soil (around 20).
After conversion to mixed farming, SOM contents tended to increase. Although in some fields this increase was statistically significant, these data should be interpreted with caution. In the Walkley and Black analytical method it has been assumed for the calculation of SOM that 77% of the organic carbon is oxidised and that SOM contains 58% carbon. Since these are average values that may vary widely, depending on soil type and management practices, respectively, the results in terms of changes in SOM over time after adaptations in soil management, are highly uncertain.
Soil pH increased slightly and remained moderately to slightly acid, except in the cash crop (C1) and the diversified garden (C2), where it increased significantly. Available P decreased to low in A1 and A3, remained medium in A2 and B2 and increased to high in B1, C1 and C2; however, the differences were not statistically significant. Exchangeable K+ changed very little, except in sugar cane (Saccharum officinarum, L.) (B1) and in king grass (B2) where it declined. SEC hardly changed, and remained low for all land use types (Table 4).
The application of on-farm produced compost and vermi-compost at annual doses of between 4 and 6 Mg ha−1 in the crop sub-system, and other soil-restoring practices such as planting legumes and trees, and mulching, might allow maintaining or even slightly increasing SOM in the arable land (De Ridder and Van Keulen 1990). However, roughly 40 Mg of compost per ha should have been added annually during 5 years to increase SOM by 1% (B. H. Janssen, Group Plant Production Systems, Wageningen University, pers. comm.). Such quantities were certainly not incorporated in the mixed systems, confirming the uncertainties associated with the Walkley and Black method.
The slight decrease in available P in the grazing sub-system may be attributed to the continuous phosphorus export through sales of milk and meat, and manure collected in the stable (about 3.6 Mg annually). Increases in SOM, pH and available P have been reported in a silvo-pastoral system in Cuba (Crespo and Rodríguez 2000). Hence, there was no reason to expect P depletion in the silvo-pastoral sub-system. However, studies in Australia and New Zealand have shown acidification effects as a consequence of biological N-fixation of legumes, leading to a reduction in availability of some nutrients such as P (Haynes 1983; Helyar and Porter 1989; Ledgard and Steele 1992). In the king grass sub-system, apparently K is being depleted and needs to be restored. This process has been extensively documented (Herrera 1990) and maintenance of a favourable soil K-status in high-yielding forage areas should be a goal in any MFS.
The overall picture arising from these data is that as a result of nutrient exports from the farm in the form of products, and the redistribution of nutrients via organic transfers, nutrients accumulate in some of the arable fields, while some other fields (particularly pastoral) are ‘mined’ (Hiernaux et al. 1998; Archard and Banoin 2003). This is especially true for P and K. The information on carbon dynamics is inconclusive, as there is doubt about the quality of the analytical data. However, accumulation seems to take place in the arable sub-systems, especially the annual crops and the sugar cane. Medium-term rotation (5–7 years) of the crop and livestock sub-systems might be a solution to this problem. However, longer-term research is necessary to establish the long-term effects of rotations and in general of agro-ecological management on soil fertility at farming system level.