Crop and pasture productivity
Maize productivity found in this study (2.4 t ha− 1) was lower than the Brazilian average for the season (5.6 t ha− 1), independent of the treatment. This was an effect of competition with Crotalaria juncea planted alongside the maize (Conab, 2021). The productivity of C. juncea (12 t ha− 1 of dry mass) was higher than that reported in other studies (Collier et al., 2020; Pissinati et al., 2018). Chieza et al. (2017) found reduction of approximately 50% in maize productivity when planted in consortium with C. juncea in the summer season due to the fast initial growth of the C. juncea. Thus, we can assume that the productivity of the crops planted simultaneously are limited by their spatial and temporal adequation, being affected by the growth architecture, habits, and physiology of each crop.
We found no effect of tree species on crop productivity during the first year (Table 2). This was because of the smaller size of trees in the first year of the agroforestry system. The triple tree row arrangement reduced the productivity of the maize and Crotalaria juncea planted in the first season by approximately 40%. This was due to the reduction in the occupation percentage of these crops. This also resulted in lower silage production in the second year.
In the second year, the tree species affected silage productivity. Greater silage production was observed in E. cloeziana, which has a lower growth rate. A greater tree growth rate increased competition for growth resources, especially light and water. Shading caused by the trees can also decrease maize productivity (Pezzopane et al., 2019).
A reduction in the productivity of crops in agroforestry systems, when compared with monospecific plantations, is often observed (Adhikari et al., 2020; Pardon et al., 2020). However, the low productivity of a single crop is compensated by the higher production for other crops or forests, increasing land use efficiency. Other ecosystem services include carbon sequestration, soil quality, biodiversity, and nutrient cycling (Adhikari et al., 2020; Pardon et al., 2020).
Pardon et al. (2020), working with the agroforestry system of Juglans aged 75 years growing alongside wheat, triticale, and barley, found an increase in the crop protein level when compared with monospecific cultivation.
Pasture productivity
As in the first two years with crops, the triple row arrangement also reduced pasture productivity because of the reduction in the proportion of the area with pasture. Similar results were found by Cabral et al. (2017) in an agroforestry system of Tectona grandis with pasture. These authors found pasture productivity of 0.258 kg m² in the agroforestry and 0.327 kg m² (reduction of 27%) in monospecific pasture cultivation.
The reduction in pasture productivity at 2.5 m from the tree rows found in our study was stronger in the E. urophylla because of the large growth rate of this species. A larger growth rate means greater resource consumption and increased light interception, which decreases resource availability for the pasture. Pacciullo et al. (2011) found that in an agroforestry system, the shading caused by trees affects the intensity and quality of the radiation reaching the understory, and above a certain level, shading affects the productivity of tropical grass. Owing to the east-west orientation of the tree rows and the low latitude of the experimental site (16 °S), the pasture at 5 and 7.5 m from the tree rows was almost unaffected by tree shading. In contrast, the pasture 2.5 m from the tree rows was under shading, especially in winter. Furthermore, close to the trees, there is greater competition for water and nutrients.
Some authors, such as Pent (2020), reported higher productivity of pastures under an agroforestry system. This is mostly because the implementation of agroforestry systems, demands higher technology and care with soil chemistry, especially pH correction and fertilization, resulting in higher productivity.
It is clear that trees affect pasture productivity, especially next to the tree rows, as found in this, and other studies (Gamarra et al., 2017; Paciullo et al., 2011). Thus, the tree population and distance between rows should be carefully planned according to the primary goal of the agroforestry system.
Tree growth dynamics
E. urophylla planted in triple rows at 1 m among trees resulted in the highest wood productivity. The wood volume produced by this treatment at 55 months of age was approximately 230 m3 ha− 1, resulting in a mean annual increment (MAI) of 50 m3 ha− 1 ano− 1. This productivity is considered high even for monospecific Eucalyptus plantations (Binkley et al., 2020). The higher productivity of E. urophylla in comparison with other tree species was expected because of its genetic characteristics, tree breeding level, and plasticity of cutting (Binkley et al., 2020). Despite their low growth rate, C. citriodora and E. cloeziana have a higher wood density; thus, their productivity difference in terms of biomass tends to be lower (Segundinho et al., 2019).
The increase in wood volume produced with an increase in the number of trees per hectare (Fig. 4) is usually found in monospecific forest plantations (Stape and Binkley, 2010; Binkley et al., 2020). The rate of this increment, as well as the inflection found in C. citriodora (Fig. 4), was an effect of the assessment age and growth rate of each species. The inflection found in C. citriodora is a result of intraspecific competition for growth resources (Binkley et al., 2020).
Interaction among the system compounds
When the amount of growth resources (light, water, and nutrients) is limited, system compounds (trees, crops, and grass) compete in an intra-and interspecific manner. Intraspecific competition, especially among trees, has been previously discussed, therefore, we investigated interspecific competition and, furthermore, the interaction between these compounds.
We found a negative relationship between tree growth and pasture productivity, starting in the second year (Fig. 5). This effect was even stronger for pastures distanced 2.5 m from the tree line (Fig. 3). With E. urophylla, the species with the larger growth rate, the pasture productivity was 40% lower at 2.5 m from the trees, when compared with the productivity at 5 and 7.5 m from the tree line. Many authors have reported deleterious effects of excessive shading on pasture productivity, especially in tropical African grass (Feldhake et al., 2015). In addition to shading, water competition, especially due to the long dry season in winter (Fig. 1), can also be responsible for the reduction in pasture productivity, as reported by Vieira et al. (2021).
In the triple row arrangement, owing to the aforementioned reason, grass was not found between the three tree lines, resulting in a 6 m band of bare soil. This reduced the overall pasture productivity under this treatment. Furthermore, this bare soil band could increase the risk of soil erosion, especially in this region, due to the large volumes of rain in the summer.
Despite competition among the system compounds, the area also exhibits complementarity interactions. Complementarity consists of the different niche occupations by species, which increases the total amount of available resources. Despite the lack of direct evaluation, we believe that the exploration of different soil layers by crops and trees is the main complementarity relationship in agroforest systems. Eucalyptus trees can access deep soil layers and uptake water and nutrients below the root zone of crops (Germon et al., 2018; Christina et al., 2017; Christina et al., 2011; Silva et al., 2011).
Management considerations
Integration systems, especially with trees, increase the overall productivity, particularly in tropical areas, owing to their complementarity. This results in different niches being occupied by trees, crops, and pastures, which increases the amount of available resources. However, there is a negative relationship between the productivity of system compounds. As a result, there is no ideal arrangement, or tree density, for an agroforest system, because it depends on the main goal of the system. If wood production is prioritized, planting more trees per hectare reduces pasture productivity. However, even when the goal is prioritizing wood production, triple rows are not recommended in regions with strong climatic seasonality. This is because the triple row results in a band of bare soil, increasing the risk of soil erosion (field observations). Some studies suggest that simple tree rows increase the exposure of the trees to wind, improving the growth tensions of the wood and reducing the wood quality, especially for sawlogs (West and Smith, 2019), however this was not assessed in this study.
In this study, we found that up to 333 trees per hectare of C. citriodora or E. cloeziana planted in simple rows did not affect pasture productivity (Table 2). Only a slight reduction in pasture growth was observed next to tree lines. This reduction in pasture growth can be compensated for by an increase in pasture protein content (Xavier et al., 2014; Lopes et al., 2017); plants next to the trees had larger access to N from the nutrient cycling provided by the trees, alongside an increase in the leaf proportion of the pasture. This alteration ca resulted in an increase in the protein concentration of pastures next to the trees.
Beyond the complementarity and overall productivity increase found in agroforestry systems, secondary effects were also noted. The main related effect is the improvement of animal well-being. Tree shading promotes better thermal comfort during grazing and rumination, resulting in better productive and reproductive zootechnical indices (Reis et al., 2021; Giro et al., 2019; Martins et al., 2020).