Introduction

The current Chilean forest development model poses serious barriers to forest producers, both small and medium owners (SMOs) and small and medium enterprises (SMEs), with important socio-economic implications; indeed, with the exception of the three biggest forest companies, this sector involves over 19,000 SMOs and SMEs, and more than 810,000 hectares of plantations. An analysis performed by Loewe et al. (2015a) characterized SMOs socially and economically, finding them to be of low educational and low-income levels (66% earned less than US$ 330/month) and highlighting the need of SMOs and SMEs to incorporate innovations to the forest activity to improve their economic performance.

Besides a socioeconomic analysis, world trends in forestry indicate the importance of diversifying the forest activity (Pretzsch et al. 2016), considering species composition and geographical distribution; this is especially true for Chile, given the wide range of environments occurring across its territory. Diversification also helps to limit biotic and abiotic risks, withstand economic risks and market fluctuations, and maximize site use (Seidel et al. 2013). Agroforestry allows agriculture and forestry diversification, and can be implemented with different species and in different settings according to specific conditions (Sotomayor and Barros 2016).

As stated by the director of FAO’s Forestry Division (BBC 2014), agroforestry systems play a crucial role in the livelihoods of rural people by providing employment, energy, nutritious foods and a wide range of other goods and ecosystem services. Mixed systems provide rural sustainable development and enhance biodiversity while preserving landscapes; the use of conifers in these systems may expand the potential applications of agroforestry (Eichhorn et al. 2006). Reisner et al. (2007) identified regions in 32 European countries that are suitable for silvo-arable agroforestry, involving walnut (Juglans regia), cherry (Prunus avium), poplar (Populus sp.), stone pine (Pinus pinea) and Quercus ilex, confirming the species potential in combined productive systems.

The use of stone pine, a species native to the Mediterranean basin, has multiple environmental benefits, such as food source for fauna, watershed protection, dune and soil recovery, and erosion control. Stone pine is one of the nine most important species producing dried fruits in the world (Loewe and Delard 2016), its edible seeds, the pine nuts, which are the most expensive ones and have been used for millennia (Prada et al. 1997; Gil 1999; Badal 2001).

Stone pine was introduced in Chile more than a century ago by European immigrants, who used it for dune and soil improvement and livestock shading (Loewe and González 2003). The species provides an attractive option to Chilean producers, especially for the high value of its shelled nuts (Loewe and González 2007; Soto et al. 2008), with prices that have been increasing worldwide from €25 kg−1 in 2010 to €45 kg−1 in 2013, and also for the highly demanded biomass generated as by-product from its processing (Cruz 2008). Furthermore, the species is characterized by adaptation ability and relatively fast growth rate, which facilitate its management. Hence, it is also considered to be an option for small and medium landowners because it provides an annual income since the fruit productive age and it adapts to poor eroded areas (Loewe and Delard 2012).

The species is also easily integrated into agroforestry systems due to its plasticity and beauty, and the highly valued fruit in international markets. In fact, stone pine has been intercropped with agricultural crops in combined systems in Spain (Mutke et al. 2007). Furthermore, the use of grafted stone pine trees with high productivity clones becomes interesting in high intensity management systems, because of their increased pine nut production.

INFOR has investigated the species behavior in Chile for more than 25 years, aiming at developing techniques for its cultivation in this country, where conditions are suitable for stone pine growth. To assess and promote its cultivation in agroforestry settings, two experimental plots were established and implemented as mixed systems that combine pine nut production, agriculture and sheep rearing at different stages of crop development under a semi-intensive management system, using simple and affordable productive techniques. We hypothesized that agroforestry systems including stone pine would be feasible for SMO/SME and more profitable than the pure stone pine plantation.

We specifically aimed to compare two agroforestry systems that include stone pine in terms of: (1) growth (height, crown and trunk diameter) and vigor and stem form, (2) crop and sheep yields, and (3) technical and economic feasibility of mixed systems.

Materials and methods

Study site

Two plots that included stone pine nut plantations, agricultural crops and sheep grazing (Fig. 1) were established in El Carmen, Biobio region (36°56′S, 71°49′W), at 533 m a.s.l. The site is characterized by an annual rainfall of 1025 mm and average temperature of 12.4 °C; it is located in the stone pine Chilean distribution area, particularly in the South macrozone, which presents the best conditions for its growth and fruiting (Loewe et al. 2015b, 2016).

Fig. 1
figure 1

Stone pine-based agroforestry system. Potato harvest at age 2 (upper); general view at age 3.5 (bottom)

Full size image

Plots are located on gentle slopes with deep volcanic soil that was cultivated before plantation establishment. The soil is homogeneous in both areas, with average neutral pH (6.4), high organic matter content (8.9%), no salt (EC 0.05 mmho cm−1), low nitrogen (15 ppm), medium phosphorus (10 ppm) and high potassium (198 ppm) content. Micronutrient content is very low for boron and zinc (0.13 and 0.07 mg kg−1, respectively), low for sulphur (0.5 mg kg−1) and medium for copper, manganese and iron (0.65, 2.4 and 30 mg kg−1, respectively).

Agroforestry systems established in both plots are presented in Table 1. Spacing pattern of stone pine trees differed between plots (5 × 5 m and 7 × 7 m), and total area amounted to nearly 5000 and 7000 m2, respectively. Stone pine trees (1 year-old seedlings) were planted in winter 2010 and homogeneously managed using arboriculture techniques (Buresti et al. 2001), with the aim of maximizing pine nut production and timber production quality. Management included soil preparation, weed control, initial fertilization, formative pruning and irrigation.

Table 1 Description of the productive system and yield in stone pine agroforestry plots established in the piedmont of the Andes mountain range, Biobio Region, Chile
Full size table

Fertilization and irrigation during the first 5 years were calculated considering alternating agriculture crops requirements and plot density; the amount of fertilizer applied was 3020 and 2353 kg ha−1 in plots 1 and 2, respectively. Plots were irrigated 12 and 9 times, respectively (Table 1).

Data collection

Stone pine height and diameter at 1.3 m height (DBH) measurements were taken in winter during 5 years (2010–2015). To assess the fruit production entrance, we evaluated strobili and mature 3 year-old cone quantity from year 3 onward considering that in Chile reproductive maturity age of the species is reached earlier (from the age of 8 years onward) (Loewe et al. 2016) than in its native distribution area, where cone production starts at between 10 (Crawford 1995) and 20 years of age (Goor and Barney 1976). Stem form was observed at the end of the study period and was categorized according to straightness (1: straight tree; 2: partly curved; 3: strongly curved tree). Vigor was also categorized (1: vigorous; 2: medium; 3: low).

Statistical analyses

An ANOVA was performed to test differences in stone pine growth between plots at particular ages. Pearson’s χ2 test (α = 0.05) was used to test homogeneity of proportion of qualitative variables (straightness, vigor).

The statistical analyses were conducted using the software INFOSTAT (Di Rienzo et al. 2015). Agriculture crops and sheep raising were evaluated using a descriptive comparative analysis.

Economic evaluation

In order to quantify the economics of the proposed systems, we compared two productive models: stone pine only for pine nuts, and stone pine for pine nut production in agroforestry system. Both models consider the use of non-grafted plants and semi-intensive management for cone production; the latter model also includes income and costs derived from the associated crop production and sheep raising.

Management considered weed control, irrigation during the first 8 years, annual fertilization during the whole rotation, one thinning at age 25, and pruning every 5 years. Costs are detailed in Table 2 and incomes in Table 3. The analysis considers a 1-ha plantation evaluated in a 60-year horizon with cone production increasing from 95 kg ha−1 at age 8 up to 8095 kg ha−1 at age 60, considering alternate bearing years (1 year of high production followed by two years of a halved cone production), and crop production up to 8 years after the establishment in the agroforestry system. Economic indicators obtained were net present value (NPV) at 8% interest rate and internal rate of return (IRR).

Table 2 Costs and standards used in the economic evaluation
Full size table
Table 3 Income and standards used in the economic evaluation
Full size table

Results

Intercropped cultures

Obtaines crop yields (Table 1) were lower than the Biobio regional average (INE 2014b), amounting to 3910 kg/ha for forage oat and 15,450 kg/ha for potatoes; thus, in the first 3 years of these experimental agroforestry plots we achieved 60% and 66–86% of regional-level values, respectively. The tested crops had no negative impact on tree development.

Animal raising

At age 5, sheep production in plot 1, with trees spaced at 5 × 5 m, was equivalent to 3.7 sheep ha−1 year−1, whereas in plot 2, with a lower tree density, it amounted to 4.4 sheep ha−1 year−1. Interestingly, although sheep were introduced in the plantation since its establishment, no damage was recorded on trees.

Stone pine growth

Height evolution (Fig. 2) indicated a positive and sustained growth, which was similar in both plots. Annual growth was on average nearly 50 cm in both plots.

Fig. 2
figure 2

Evolution of stone pine height

Full size image

Diameter at collar height (DCH) (Fig. 3) had a sustained growth trend, and was significantly higher in plot 2, were tree density was the lowest; annual growth was 1.98 cm in the highest density plot (plot 1), and 2.24 cm in the lowest density one. DBH annual growth at age 5 reached 0.88 and 0.86 cm in plots 1 and 2, respectively.

Fig. 3
figure 3

Evolution of diameter at collar in stone pine

Full size image

The reported crown diameter values correspond to measurements of year 3, since 1 year later an intense formative pruning was done, which significantly reduced this parameter. Mortality was low (below 5%) in both plots.

Although stone pine trees were vigorous in both plots, we found significant differences. The highest percentage of vigorous trees was found in plot 1 (97%), being 87% in plot 2 (Table 4). In plot 1 no low vigor trees were observed, and in plot 2 that value was very low (1%).

Table 4 Stone pine growth and productivity variables in two agroforestry systems established in the piedmont of the Andes mountain range, Biobio Region, Chile
Full size table

Records of trunk straightness exhibited statistical differences between plots. The highest percentage of straight trees was found in plot 2, the one with lower density. In plot 1, most trees (54%) were curved, although higher density usually promotes a better form. Plot 2 also had the highest percentage of strongly curved trees (9%).

Cone production was evaluated from year 3, despite the short age of the plantation. At age 5, in plot 1 over 6% of trees had at least one strobilus (1 year-old conelet). Few trees started flowering in the second year, and some of those flowers developed and reached maturity at age five, given the 3.5-year cycle the fruit needs to reach completion.

Economic evaluation

The results of the economic evaluations (IRR and NPV at 8%) (Table 5) showed that the NPV was almost seven times higher in the agroforestry system (cones and crops) than in the pure stone pine plantation (US$ 3282 vs. US$470 ha−1), and IRR rose from 8.4 to 12.3% in the agroforestry system.

Table 5 Economic evaluation comparing the stone pine agroforestry system with a traditional pure stone pine plantation, both for cone production
Full size table

Discussion

This study suggests that it is possible to associate stone pine cultivation with intercropped annual cultures and also with controlled grazing, with interesting tree development values, as it has been demonstrated for other fruit forest species (Chifflot et al. 2005), in particular with walnut (Loewe and González 2006) and cherry (Balandier and Dupraz 1999).

In the first 5 years since these experimental agroforestry plots were established, crop yields were lower than the average in the region in monoculture, reaching 60% for forage oats and 66 and 86% for potatoes in plots 1 and 2, where the available space for the intercropping was close to 70 and 80% of the total area, respectively. Additionally, sheep that fed directly in the field, which were not considered in this analysis to present a conservative scenario, should be added to the estimates.

At age 5, sheep production in plot 1, with trees spaced at 5 × 5 m, was 3.7 sheep ha−1 year−1, whereas in plot 2, with a lower density, it amounted to 4.4 sheep ha−1 year−1. Considering that mean production in medium quality pasture in Chile reaches 4 (3–5) reproductive sheep/ha (Claro 2009), our results in plots 1 and 2 represent 92 and 110% of country-level values, respectively.

Although forage production in these plantations was not enough to sustain permanent grazing, livestock from the farm grazed there during a limited number of days per month during 10 months each year, contributing to mutton production. Interestingly, although sheep were introduced into the plantation since its establishment, no damage was recorded on trees, since there was a strict control on forage availability and animal pressure.

The possibility to combine stone pine and animals was already pointed out by Agrimi and Ciancio (1994), who stated that grazing in open pineries could provide some income, even though in Europe the main species associated with livestock are chestnuts, poplars, hardwood plantations, Mediterranean oaks and fruit trees, but also pines (Pardini and Nori 2011), usually in coastal areas. Forage production in these plantations is not enough to sustain grazing, but livestock from nearby farms occasionally graze there; if grazing is periodical, it has several positive impacts on pine plantations, such as reducing growth of weeds and shrubs, fire risk and the cost of periodical mechanical or chemical cleaning. Soil fragmentation and nutrient enrichment through animal defecation during grazing has been reported, accelerating litter decomposition and nitrogen incorporation, and reducing pine needle accumulation and fire risk (Mancilla-Leyton et al. 2013). This effect of grazing is important given the increasing number of forest plantation fires and that forest plantations in Chile supply 98% of the timber used by the national industry (INE 2014a). Moreover, if there is a direct continuous regulation of livestock carrying capacity, trees would not be damaged, as it was observed in this study, and as previously pointed out by Anderson et al. (1988).

Regarding stone pine growth, our results are higher than those reported for the best growth macrozone for the species by Loewe et al. (2015b), located in the South of Chile, in which growth rate across ages and sites is 0.35 m year−1 in height and 1.50 cm year−1 in DBH. The higher growth values reported here could be due to the young age of the assessed plots, which is characterized by a fast growth is, and also to the positive effect of the association with crops, which were fertilized and irrigated during the first 5 years.

DBH and height at age 5 was similar in both plots despite the differences in fertilization dose and irrigation quantity, which were 28 and 33% higher in plot 1, respectively. The fact that 6.3% of 5 year-old trees in one of the plots have strobili, which should reach maturity at age 8 years given the long fruit development cycle (42 months), is remarkable, since in its native distribution area first cone production starts in trees of 10 (Crawford 1995) and 20 years of age (Goor and Barney 1976). This result indicates the overall beneficial effects of the system on tree reproductive development, and could be an advantage for traditional plantations with seedlings, which are significantly less expensive than grafted plants.

Forest management planning provides a framework to establish priorities, set objectives and devise strategies to deal with risks (Day and Pérez 2013). From the management point of view, this technical combination turned out to be a simple scheme, easy to follow by small and medium size landowners with limited capital and knowledge base.

The amount of fuels associated with forest management is very important for forest fire occurrence; in fact, it has been reported that a reduction of fuel biomass decreases the probability and intensity of wildfires (Peña and Pedernera 2004). Preventive forestry aims at reducing the number of fires, and more importantly, their size (Haltenhoff 2006). Accordingly, USDA (2006) noted that the decrease in stand density and understory and surface fuel loads are factors that explain historical fire rates. Hence, the efforts to prevent damage and losses in plantations aim at including measures also related to decreasing horizontal and vertical fuel continuity, contributing to the reduction of wildfire likelihood, minimizing losses and supporting actions for a rapid extinction if a fire occurs. The proposed agroforestry system then turns out to be not only a productive solution but also a management practice with foreseeable positive effects on reducing fire likelihood, given the important quantity of fuel reduction and continuity defined by the lower density and by the presence of animal grazing, while providing several benefits such as improved access, extra revenue and reduced fire hazards.

From an economic point of view, we found the combined system to be significantly more profitable than the stone pine monoculture oriented to cone production. Economic results can improve if the owner adds a primary elaboration to sell in-shell pine nuts.

Our results are in line with experiences reported in New Zealand. where it was found that the combined system was more profitable than pastoral, agriculture and plantation forestry individually (Arthur-Worsop 1984), and in Australia in pine agroforestry systems (Garland et al. 1984), besides environmental benefits reported fin South America, such as the strong influence exerted by trees in the creation of favorable microclimate within silvopastoral systems (Dube et al. 2013).

Furthermore, the agriculture/animal production during the first years generates income that in a traditional stone pine plantation is not available until sexual maturity is reached. Therefore, this important additional production is attractive not only due to its economic implications, but also because land owners need continuous annual income, which is an additional benefit of the tested system.

The duration of the associated animal and crop production under the present conditions is not well known and will depend upon crown development; however, according to the present analysis, the rain-fed forage and/or grazing, and irrigated potato and oat and rye grass pasture is expected to last over 8 years. After that age, a stone pine thinning will be necessary to keep the agroforestry systems as such.

The important technical and commercial gaps that affect forest SMOs and SMEs, characterized by a lack of options partly due to their fragmentation, could benefit from agroforestry systems designed to fit different situations, considering available capital, owner’s goals, capacity of operative management, and market conditions. The preliminary results of the proposed agroforestry system are a relevant input for future policy design and implementation, since its adoption could be a way of enhancing the incorporation of SMOs and SMEs to a more sustainable forest activity, as recommended by Loewe and Venegas (2005). In particular, the associated productive system showed an economic performance that can justify a government subsidy for agroforestry system establishment and management in order to improve the rural economy.

Conclusions

Stone pine is a highly interesting species due to its high-value pine nuts, its relative fast growth rate, simple management and excellent phytosanitary performance in Chile, without any disease or pest having been recorded. It was found to be a species suitable for its establishment in combined agroforestry systems based on trees for both nuts and timber, intercropped with agricultural crops and animal grazing.

The annual income derived from crops and animals is relevant for the household economy, especially during the first years of tree development when they still are not fruiting.

Forage production in these plantations is not enough to sustain permanent grazing, but sheep grazing during certain programmed periods contributes to mutton production; in addition, periodic grazing reduces shrub growth, fire risk and the cost of periodical mechanical or chemical cleanings. No damage by animals was observed on trees.

The tested stone pine—agricultural crops—sheep grazing system showed a positive economic performance; given the socio economic limitations characteristic of this sector, these results could justify a government subsidy for agroforestry system establishment and management in order to improve the rural economy and SMOs and SMEs.