4.1. Araucaria angustifolia growth in mixed and pure plots
One way to analyse whether there is competition among trees in a stand is to observe the relationship between DBH and diameter increment. The absence of correlation indicates that there is no competition, whereas a positive correlation indicates competition for light (Pretzsch and Biber 2010). A. angustifolia growth was affected by the addition of C. canjerana seedlings, as indicated by the correlation between DBH and the increment in DBH of A. angustifolia in mixed plots, in contrast with pure plots (Fig. 1). This positive correlation between DBH and the increment in DBH in mixed plots showed thatgrowthwas higher in large A. angustifolia trees than in small trees. The addition of C. canjerana in the stand increased site fertility, and consequently, the asymmetry in soil resources usage (i.e., water and nutrients) among smaller and bigger trees. Therefore, competition was higher in mixed plots than in pure plots, as larger trees are most successful in capturing limited resources (Binkley et al. 2013). As expected, the correlation gradient was flat, as competition aimed for soil resources and not light (Pretzsch and Biber 2010). According to our results, other reports stated that the growth of the sun-demanding conifers Pinus pinea and Pinus halepensis was also reduced by soil competition in even-aged mixed stands (Cattaneo et al. 2018). In that experiment, light competition between species was probably negligible as the heights of the trees of each species were very different, but plant growth was limited by the characteristic droughts of the Mediterranean environment. Therefore, plants allocated more resources to the roots and increased competition for the water available in soil. Although the A. angustifolia plantations in Misiones are inserted in a subtropical climate where average rainfall is 2,000mm year-1, the evapotranspiration is high and every year there are periods with low rainfall that limit water availability in soil, as has been demonstrated both in the native forest (Carrasco et al. 2014; di Francescantonio et al. 2020) and in Pinus taeda plantations (Faustino et al. 2013; Bulfe and Fernández 2016). The interplanting of C. canjerana, increased the availability of resources for larger A. angustifolia trees rather than for smaller ones. It is important to highlight that the mean increment in DBH of A. angustifolia in the mixed plots was higher than in pure plots. Although the mechanism underlying this result was not evaluated in our experiment, some reports describing facilitation effects when different species are mixed can be useful to suggest possible reasons: (1) the effect of nitrogen-fixing species to improve soil fertility when included in a non-nitrogen-fixing plantation (Bauhus et al. 2004; Boyden et al. 2005; Marron and Epron 2019); (2) the favorable effect of the presence of vegetation below the canopy on soil stability and aggregation, which allows better infiltration as roots are essential in reducing soil erodibility (Zou et al. 2005; Bayala and Prieto 2020); (3) fungal diversity and richness in soils closely related to the presence of certain tree species (Buée et al. 2011) and the microorganisms that improve the physical properties of soil, thus enhancing root access to water (Pardos et al. 2021), and (4) sun-demanding species benefitting from relaxation of intraspecific competition in a mixture due to differences in shade tolerance (Toïgo et al. 2021). However, in our experiment, C. canjerana is not a nitrogen-fixing species and the mixed plot was accomplished by adding a second species, not by replacing it, that is, the number of trees per ha was higher in mixed plots than in pure ones. Therefore, the mechanisms underlying this beneficial effect are probably related to a positive effect of adding young trees to soil physical and biological properties. Similarly to our results, in Eucalyptus mixed stands in the Brazilian Atlantic Forest, the largest eucalyptus trees were associated with higher neighbouring indices, irrespective of whether the neighbours were N-fixing trees or not (Amazonas et al. 2021).
When we analysed the increments in DBH of A. angustifolia trees concerning their competition index (NT), it was clear that a higher number of neighbours, irrespective of their size, affected A. angustifolia growth (Fig. 2). A. angustifolia trees had up to three larger neighbours (NL), but up to eight total neighbours (NT) within a 5m radius in the mixed plots. It is important to highlight that C. canjerana seedlings were also included in NT. However, the better correlation of the increment in DBH of A. angustifolia trees was with NL, so the growth was markedly reduced by the number of larger neighbours, i.e., other A. angustifolia trees (Fig. 2). The idea that the A. angustifolia growth is mainly affected by intraspecific competition is reinforced by the negative correlation between BAT and the increment in DBH, as BAT is highly sensible to the size of the neighbours. Likewise, at the beginning of the experiment, when C. canjerana was planted, the trees with the highest DBH were those without neighbours within a 5m radius, both in mixed and pure plots (Fig. S2). Furthermore, in pure plots where only A. angustifolia was present, the negative correlation between the DBH increment and NL confirms that A. angustifolia is mainly affected by intraspecific competition. These results are consistent with the demand of this species of large disturbs to regenerate in the native forest (Souza et al. 2008), as it is a sun-demanding species with a wide monolayer crown. Also, in an unthinned commercial monospecific plantation, with an initial density of 2,500 trees ha-1, 62% mortality was registered 21 years after planting (Salto and Lupi 2019), reinforcing the idea that intraspecific competition is high.
The first two years after conversion to mixed plots, the average growth at the stand level was similar in mixed and pure plots, but it was higher in mixed plots than in pure plots for the second period. Therefore, five years after the conversion from pure to mixed stands, the complementarity of niches and the difference in age between A. angustifolia and C. canjerana implied a benefit to the conifer growth due to the intercropped plantation of the broadleaf species. This is an important point considering that in mixed plots, the density of trees was higher than in pure plots, as the second species was added without removing the preexisting plants and no mortality of A. angustifolia occurred. Similar to our results, the growth of different species in mixed plantations was higher than in their corresponding monospecific plantations (Perot et al. 2010; Forrester and Smith 2012; You et al. 2018). The fact that in our experiment, the second species was planted when the first species was 14 years old is important to ensure the optimal microenvironment for both species, which have very different requirements. Likewise, the establishment of five native tree species under 12 to 15-year-old Pinus elliottii plantations did not negatively affect the growth or the volume produced 23 years after the conversion to mixed stands (Simpson and Osborne, 2006). However, experiences carried out even with a difference of three months in the establishment of a shade-intolerant and a shade-tolerant species, have shown stratification and higher productivity in mixed plots rather than in their corresponding pure plots (Bauhus et al., 2004). As was observed 4 years after mixed stand conversion, the predominant interaction among species in the stand can change over time when competition of Robinia pseudoacacia over Populus hybrids was higher than the initial facilitation (Rebola-Lichtenberg et al. 2021). In our results, during the first 5 years after interplanting C. canjerana under A. angustifolia, an incipient beneficial effect of the mixture was observed in the growth of the sun-demanding species, but this result needs to be confirmed over a longer period of time. However, it is important to note that this is the first report of a mixed plantation with two native species of the critically endangered Altlantic Forest.
4.2. Growth and acclimation of Cabralea canjerana seedlings under Araucaria angustifolia canopy
To promote new ways to produce wood without further affecting rainforests, it is essential to evaluate the production of native forest species in already deforested areas. This is the first experiment centered on the study of physiological acclimation of a native Atlantic Forest species in the conversion from monospecific to mixed stands. C. canjerana seedlings were successfully established under the canopy of a 14-year-old A. angustifolia plantation, with an incident PFDD ranging from 4 to 58% with respect to that of open areas. Our early studies indicated that C. canjerana seedlings have high phenotypic plasticity, which gives them the ability to acclimate to different coverage conditions, even at more extreme low light intensities (below 10µmol photons m2s-1)registered in the native forest(Moretti et al. 2019b; Olguin et al. 2020). The monolayer flat umbrella-shaped crown of A. angustifolia produced a sparse canopy cover and high light intensity; sunflecks reached the understory and allowed high growth rates of C. canjerana seedlings. The microenvironmental conditions generated by thecanopy of A. angustifolia were similar among C. canjerana seedlings, with different levels of competition (Fig. 4), i.e., regardless of the number of larger neighbours (NL), which are those that mostly modify the environmental conditions. Thereby, the C. canjerana morpho-physiological response was similar at different NL. The plasticity of C. canjerana allowed it to acclimate to the different microenvironments and maintain high growth rates through physiological modifications during the first years. Two years after planting, C. canjerana decreased the SLA and the chlorophyll a:b ratio. The higher concentration of chlorophyll b is related to bigger light-harvesting antennas, which makes the plants capable of using low radiation. Then, seedlings increased the capacity to intercept more light and thus the electron transport rate (ETR), with the subsequent higher carbon gain per unit of surface area (Moretti et al. 2019b). In none of the microenvironments generated by the canopy of A. angustifolia, where C. canjerana was planted, we found any physiological indicator of water stress, excess or lack of light, such as partial or total stomatal closure, reduction in the photosynthetic rate, or changes in leaf morphology as was found in open areas (Moretti et al. 2019b). Then, the canopy of A. angustifolia prevented C. canjerana stress during its establishment.
The increment in height and collar diameter in C. canjerana seedlings was affected by the number of total neighbours (NT) and, specifically, by the number of larger neighbours (NL) (Fig. 3), similarly to the pattern observed in A. angustifolia. BAT correlated with collar diameter increment, but there was no correlation with height increment; therefore, higher BAT diminished stem thickness. As higher BAT implied more A. angustifolia trees near the C. canjerana seedlings, the growth in height and collar diameter of C. canjerana seedlings was affected mainly by the competition of the largest neighbours. Consistently, NL was the competition index that best explained the relationship between the increase in height and collar diameter and competition. In other words, C. canjerana seedlings growth was lower when there were more A. angustifolia neighbours within a 5m radius. This result was repeated in the two A. angustifolia mixed stands with similar basal area (17 and 18m2ha-1),where C. canjerana growth in height was lower when NL was higher (Fig. 6). However, in the A. angustifolia stand with higher basal area (BA=28m2ha-1), C. canjerana growth in height was low and separate from NL, as the site was high stocking, then any microenvironment was good enough for C. canjerana seedlings. Unlike A. angustifolia trees in which intraspecific competition is higher than interspecific competition, in C. canjerana seedlings competition with the big trees would prevail over intraspecific competition. Despite these competitive relationships, C. canjerana seedlings survival and growth rates were very high (Fig. 5 and Fig. 6), higher than the growth registered in plantations in gaps in the native rainforest (Moretti et al. 2019a; Olguin et al. 2020). The initial growth registered is consistent with previous experiments that reported plants with a 22cm DBH and 12m height, 22 years after planting (Paniagua et al. 2006). Thus, it can be stated that although further measuring until harvest is required to be able to assess its yield, the initial C. canjerana establishment under the canopy of A. angustifolia was successful, and conversion from an even-aged monospecific A. angustifolia stand to a mixed uneven-aged stand is possible. In mixed uneven-aged plantations, the differences in planting moments of the species produced microenvironmental heterogeneity, which is advantageous in the establishment of both species. The species planted first has the advantage of establishing and growing without interspecific competition. The species planted later, in this case C. canjerana, benefits from being protected by a canopy during the first years, as the canopy buffers high and low temperatures, wind speed, and reduces soil drying compared to deforested open areas. Then, there is a trade-off between facilitation and competition that allows species to coexist. In this sense, the facilitation five years after planting C. canjerana below the canopy of A. angustifolia would have a higher incidence than the incipient competition for resources, with high growth rates observed for both species when mixed. However, in more advanced stages, sustained competition could affect growth rates (Ledo et al. 2014).
It is important to highlight that the microenvironmental conditions were not significantly affected by the number of larger neighbours in each seedling (Fig. 4), so all the seedlings were similarly acclimated to those microenvironments (Table 1). In this sense, the selection of the planting site in the stand should not seek the protective effect of the proximity of many neighbours, neither systematically nor randomly. It is important to highlight that during the first 24 months, C. canjerana growth was indifferent to the number of A. angustifolia neighbours within a 5m radius. This time lapse coincides with that reported for tropical species as the most critical to ensure establishment (Campoe et al. 2014). After two years of planting, the growth rate markedly increased in C. canjerana seedlings with only one or two neighbours. Therefore, within the A. angustifolia stand, the plantation of C. canjerana should be carried out in positions that imply the seedling will have one or two neighbours within a 5m radius, and positions with a higher number of neighbours should be avoided.