Quantication of damage and initial growth of eucalyptus in integrated crop-livestock-forest systems in the Brazilian Cerrado


 Background

The integrated crop-livestock-forest (ICLF) system is a sustainable production model composed by tree species intercropped with annual crops and forage species. The choice of the tree species that will compose the ICLF is characterized as a fundamental requirement in the adoption of the system. Therefore, this study aimed to analyze the initial growth and adaptation of two eucalyptus genotypes cultivated in ICLF systems in the Brazilian Cerrado.
Methods

The eucalyptus genotypes (E. cloeziana and E. urograndis) were intercropped with Sorghum bicolor and Urochloa brizantha in two ICLF systems in Minas Gerais State, Brazil. Qualitative data expressed by occurrences of damage and weaknesses, as well as quantitative data measured by survival rates, diameter at soil height, diameter at breast height, and total height of trees were evaluated in four periods. The data were analyzed by a correlation matrix and analysis of variance, and them qualitative and quantitative data to were analyzed in relation to the rainfall distribution during the studied period.
Results

The occurrence of Trigona spinipes and termite were the main biotic factors that caused damage and weaknesses to E. cloeziana and E. urograndis, respectively, directly impacting the initial growth and survival rate of eucalyptus.. The growth of E. urograndis was higher than E. cloeziana at 450, 630, 1020, and 1320 days after transplanting. On the other hand, E. cloeziana showed the highest survival rate (80.6%) while E. urograndis presented greater tolerance to the water stress.
Conclusion

The hibrid E. urograndis proved to be more adapted to edaphoclimatic conditions studied in the Cerrado region and can be indicated to ICLF systems implanted in regions characterized by high temperatures and irregular rainfall distribution.

The choice of the tree species that will compose the ICLF system is characterized as a fundamental requirement in the adoption of this agricultural production model (Machado et al. 2013). The eucalyptus cultivated in integrated production systems can be useful to several uses, however the commercial use as wood for sawmill, is the one that has been most used by producers, due to the greater nancial return (Oliveira et al. 2015). The choice of a model or the species that best adapt to the region, or with the system itself, can minimize the costs of implementation and conduction, increasing the production and pro tability of the integrated production system ).
Knowledge about the initial growth of seedlings is essential for the progress of the activity of producing quality genotypes, which in turn is of great importance for the success of the silvicultural activities (Mota et al. 2012). One of the tools to help understand the ecological relationships between the initial seedling development and the environment is the growth analysis (Souza et al. 2018). Growth analysis can be used to evaluate crop productivity and allows investigating the ecological adaptation of these crops to new environments, competition between species, management and treatments effects, and the identi cation of the productive capacity of the different genotypes (Alvarez et al. 2005).
In this sense, the objective of this study was to evaluate the initial growth an adaptation of two Eucalyptus genotypes (Eucalyptus urograndis and Eucalyptus cloeziana) in ICLF systems in the Brazilian Cerrado. We aimed to answer two questions: i) What were the factors (biotic and, or abiotic) that interfered with the initial growth of both evaluated genotypes? and ii) Which of the genotypes was more adapted to the ICLF system under the studied edaphoclimatic conditions?

Study area
The experiment was implemented at Barra Farm (16°38'44.02′′ S and 43°42′43.77′′ W), located in the municipality of Francisco Sá, Minas Gerais, Brazil. In November 2012, soil samples were collected in the study area at depths from 0 to 20 and 0 to 40 cm, in order to characterize the soil type, thus being classi ed as Haplic Cambisol, a clayey and medium fertility soil (Table 1). The altitude of the study area is 590 m. The region's climate according to the Köppen climate classi cation is Tropical Savanna (Aw), with rainy summer and dry winter, annual rainfall of 1,009 mm and temperature of 21.6 ° C (Alvares et al. 2013). However, during the period of conducting the experiment there was a decrease in precipitation in the study region ( Figure 1)    Selection of species and arrangements of the ICLF system For this study two eucalyptus genotypes were chosen: the specie Eucalyptus cloeziana and the hybrid Eucalyptus urograndis, which were selected due to the better quality of the wood and adaptability to regions with water de cit, respectively, when compared with other genotypes. E. cloeziana has an excellent shaft shape, with good natural durability and its wood is recommended for energy and construction purposes ). E. urograndis hybrids are the most cultivated in Brazil due to forest improvement for greater productivity and adaptability to water de cit conditions (Kullan et al. 2012;Leonardi et al. 2015).
The experiment was implemented in an area of 3.2 hectares with two productive crop arrangements, being: (i) ICLF system with E. urograndis, a hybrid of E. urophylla × E. grandis intercropped with Sorghum bicolor (sorghum) and Brachiaria brizantha cv. Marandu (marandu grass); (ii) ICLF system with E. Cloeziana intercropped with Sorghum bicolor (sorghum) and Brachiaria brizantha (marandu grass). The two ICLF systems were organized in double rows of eucalyptus (3x2m), 3m between lines and 2m between plants and spaces of 14m (alley) between double lines ( Figure 2). Implementation of the ICLF systems and management of eucalyptus Pits for planting eucalyptus were opened in December 2012 and the two genotypes seedlings were planted intercropped with sorghum. In the second year, sorghum was planted again with marandu grass between the double rows of eucalyptus ( Figure 2). After the second sorghum harvest, the system was conducted as a silvopastoral system with an arrangement composed by eucalyptus and marandu grass.
The initial fertilization consisted of the application of 7.2 g of N, 36 g of P and 7.2g of K per plant with micronutrients. During the experimental phase, the ants were controlled with granular bait before implanting the area and after planting. The termites were controlled before planting by immersing eucalyptus and sorghum seedlings in Fipronil solution [500 g of commercial product / 100 L of syrup, equivalent to 10 ml of syrup / seedling (BASF SE, Ludwigshafen, Germany)] and, by directed irrigation in transplanted eucalyptus seedlings. Irrigation management was carried out during December/2012, February and March/2013, with the application of 4L of water/plant in order to supply the seedlings during the water de cit in the region ( Figure 1B

Statistical analysis
Qualitative variables were analyzed using Pearson's Correlation Matrix. For quantitative variables the analysis of variance (ANOVA) was performed and subsequently the means were compared by the F test (p <0.05), using R statistical software (R Core Team, 2019). Additionally, both quantitative and qualitative data were analyzed in relation to the rainfall distribution during the studied period.

Qualitative data
In the evaluation carried out at 450 DAT (April 2014), the occurrence of competition with weeds (42.4%), climbing plants (39.6%), attack of Trigona spinipes (31.9%), bifurcation (21.9%), damage to the stem (20.5%) and branching (10.1%) was observed in E. cloeziana (Table 2). In this evaluation, a correlation (p < 0.05) was observed between the attack of T. spinipes and the presence of branching and stem damage in E. cloeziana, and a high correlation between the occurrence of weeds and climbing plants (Table 3). For the E. urograndis, there was a higher occurrence of interference with weeds (24.7%), climbing plants (20.1%) and termite attack (9.0%) ( Table 2). There was also a high correlation between the occurrence of weeds and climbing plants (Table 3). After the rst sorghum harvest, in May 2013, there was a severe attack of termites in the ICLF lines planted with E. urograndis, which resulted in serious damage, reducing the root area and resulting in high plant mortality.   The second evaluation (630 DAT) was carried out at the end of the dry season (October 2014), a period characterized by low rainfall (Fig. 1D). In plants of E. cloeziana there was an increase in the occurrence of competition with weeds (42.4%) and climbing plants (42.4%), followed by the attack of Trigona spinipes (39.2%), damage to the stem (39.2%), tree branching (30.6%) and bifurcation (21.9%) ( Table 2). There was a correlation between the occurrence of weed and climbing plants competition (Table 4). The plants of the E. urograndis had a better response, with a reduction in the occurrence of tree branching (7.6%), water stress (7.3%), weeds (5.6%) and climbing plants (5.6%). In addition, there was a correlation between the water stress and bifurcation variables and between the occurrence of weeds and climbing plants (Table 4). In the third evaluation, at 1020 DAT (November 2015), E. cloeziana continued to damage the stem (39.2%), as well as branching (25.7%), bifurcation (19.4%) and dieback (9.0%). In this evaluation, water stress symptoms (19.4%) were also found in this species, which was strongly correlated (p < 0.05) with dieback (Table 5). It was observed the occurrence of water stress (5.2%) and the presence of weeds (13.9%) in trees of E. urograndis evaluated (Table 2), with a correlation (p < 0.05) also between the occurrence of water stress and dieback (Table 5). In the fourth evaluation carried out at 1320 DAT (September 2016) there was an increase in damage to the stem in E. cloeziana, which occurred in 55.9% of the evaluated plants (Table 2), followed by branching (64.6%), weed competition (42.4%), bifurcated plants (31.9%), water stress (24%) and climbing plants (15.3%). It was found that, over the evaluated period, the damages and weaknesses observed increased in this species. The correlation matrix for the parameters evaluated in 1320 DAT showed a direct relationship between weed competition and stem damage, and between climbing plants and bifurcation for E. cloeziana (Table 6). E. Urograndis area showed an increase in weed interference (39.9%), branching (32.6%), bifurcation (27.4%), nutritional de cit (25.3%) and termite occurrence (5.6%). Correlations (p < 0.05) were found between weed competition and branching, and between nutritional de cit and branching (Table 6). The correlation analysis of qualitative data, including all observation periods (450, 630, 1020 and 1320 DAT) showed a signi cant correlation (p < 0.05) between stem damage and branching for E. cloeziana and, between branching and bifurcation for E. urograndis (Table 7). Quantitative data E. cloeziana and E. urograndis showed differences (p < 0.05) at 450, 630, 1020 and 1320 DAT for survival rate (Table 8), DSH, DBH and Ht (Table 9). However, no differences were found between the two genotypes in relation to DSH and Ht at 90 DAT (Table 9).
E. cloeziana had a survival rate of 94.1% at 90 DAT, and this percentage decreased at 450 DAT, reaching 86.8% (Table 8). During the evaluations (450 to 1320) a 6.2% decline in the survival rate of this species was observed. Whereas, E. urograndis had a survival rate of 90.1% at 90 DAT (March 2013), decreasing abruptly and reaching 67.4% at 450 days (April 2014). For evaluations between 450 to 1320 DAT, the survival rate of E. urograndis was practically unchanged, with a reduction of only 0.7% (Table 8). Thus, the survival rate of E. cloeziana was 13.9% higher than that of E. urograndis in 1320 DAT (September 2016). Legend: DAT: days after transplanting.
As for the growth parameters, at 450 DAT, the DSH was greater in E. urograndis in relation to the E. cloeziana trees. For the DBH, growth was greater in E. urograndis at 450, 630, 1020 and 1320 DAT ( Table 9). The period between 450 and 630 DAT was characterized by the dry season (Fig. 1D). The increase in DBH in the E. cloeziana and E. urograndis was 0.30 and 0.78 cm, respectively. E. urograndis also showed a higher Ht at 450, 630, 1020 and 1320 DAT. At 450, 630 and 1020 DAT, the Ht of this genotype was, respectively, 1.17, 1.43 and 3.93 m higher than that of E. cloeziana, con rming the growth of the E. urograndis hybrid also during dry periods.

Qualitative data
In the rst evaluation at 450 DAT, the occurrence of weeds was the main factor that interfered in the areas cultivated with E. cloeziana and E. urograndis, and a high occurrence of climbing plants was also recorded ( Table 2). According to Serra et al. (2019), competition with weeds can cause delay in tree growth, so weed management must be planned in advance. Another factor that in uenced the growth of E. cloeziana was the attack of the T. spinipes, classi ed by Damascena et al. (2017) as a kind of common occurrence in the Brazilian Cerrado. T. spinipes jaws cut the tissues of the plants causing a release of certain secretions, resins or brous laments for the construction of their nests in the branches of the trees or, in abandoned termite mounds (Moura et al. 2017). In addition to the injuries caused by the direct attack of this insect, such as destruction of ower buds, new branches, owers and fruits (Gallo et al. 2002), the exposure of tissues can allow the entry of pathogens, causing even greater problems in production (Alves et al. 2018).
Having observed termite attack in the E. urograndis genotype at 450 DAT (Table 2), the control of these insects has become essential, as the damage caused can induce nutritional and water de cits, which can culminate in the death of the plant (Ahmad et al 2021). Another important factor that may have in uenced the result was the low rainfall observed in the months of January and February 2014 (Fig. 1D). These facts explain the damage and injuries caused mainly by the presence of insects, such as bees and termites, as well as the competition of weeds and the high number of climbing plants observed in the two evaluated systems of the ICLF, since plants with water stress have no capacity to react to the damage, presenting symptoms more easily (Katsaruware-Chapoto et al. 2017;Ramesh et al. 2017).
The second evaluation (630 DAT) was carried out in a period of low rainfall (Fig. 1D) and, probably, the development of the E. urograndis plants inhibited the occurrence of damage and weaknesses. For the E. cloeziana, the occurrence of weeds and climbing plants increased in this evaluation period (Table 2). Analyzing the effects of spacing and layout on the growth of tree species, the ICLF system contributes to the incidence of solar radiation between the double rows, increasing weed growth and infestation (Santos et al. 2012). According to Santos et al. (2010), weed interference must be controlled continuously, with direct application of herbicide or manual weeding. In relation to the season, this must be done at the beginning of the rainy season to guarantee the success of the integrated production system (Gomes, 2010).
In the evaluation at 630 DAT, the plants of E. Cloeziana continued to be attacked by the bee T. spinipes, which contributed to the appearance of lateral rami cations and bifurcation of the plants. According to a study by Chiaradia et al. (2003), plants are more susceptible to the attack of T. spinipes in the early years of planting in their apical shoots, causing damage to the young tissue to obtain secretions for the construction of their nests. Thus, when the attack of the bees causes damage to the young stems and growth points to remove the resin and exudates, it also induces stress in the plant, leading to the emission of lateral branches. In addition, other factors such as water stress, weed competition and climbing plants may have contributed to the delay in the development of E. cloeziana in the ICLF system (Ngugi et al. 2004;Muller et al. 2017).
At 1020 DAT there was a recurrence of damage to the stem, branches of bifurcations in trees of the E. cloeziana genotype (Table 2). In the ve months preceding this assessment (June to October 2015), low rainfall was recorded in the study area (Fig. 1E), which may be directly related to records of pointer drought and water stress recorded for this genotype. In the two genotypes of eucalyptus analyzed, a correlation between water stress and pointer drought was observed, and this is common in eucalyptus, due to the nutritional de cit that occurs during the dry season or under conditions of low precipitation, even with the availability of macro and micronutrients in the soil. According to Alfenas et al. (2009), up to 3 years of age, plants may present water de cit symptoms such as leaf lesions and death in parts of the main branches.
The damage and weaknesses reported at 1020 DAT in E. Cloeziana trees continued to increase at 1320 DAT. For E. Urograndis, the presence of weeds, branches and bifurcations was recorded again (Table 2). In addition, between February and September 2016, all precipitation records were very low (Fig. 1F), which directly in uenced the observed results, especially in relation to the occurrence of water stress, weed competition and nutritional de cit in the two genotypes evaluated (Hu et al. 2008;Borken and Matzner 2009;Valdés et al. 2013;Maseda and Fernández 2016;Barbosa et al. 2016).

Quantitative data
Regarding survival rate, E. cloeziana showed better results in all evaluations when compared to E. urograndis, which can be referred to a favorable adaptation of E. cloeziana to the study site. According to Borges et al. (2012), the survival rate of eucalyptus plants depends directly on climatic conditions. Reis et al. (2017) also found higher survival rates in E. cloeziana when compared to E. urograndis in the Brazilian Cerrado.
The possible causes of mortality in E. cloeziana in evaluations between 90 and 450 DAT may be associated with late planting, weed competition, severe attack of Trigona spinipes and occurrence of water stress at the initial moment of plant growth, factors already reported by Joern and Mole (2006). While for E. urograndis, the high mortality rate between evaluations at 90 and 450 DAT may be associated with the severe attack of termites, since there is an increase in damage and weaknesses in the plants after the attack of these insects as mentioned in others studies (Gauchan et al. 1998;Debelo and Degaga 2017).
Is important to note that chemical control with pronil was performed in E. urograndis. However, after the sorghum harvest there was a greater attack of termites in this genotype, increasing plant mortality (Table 8). In addition, the survival rate of this species was practically unchanged between the evaluations performed at 450 and 1030 DAT, even though there was a period of low rainfall in the study region (Fig. 2). It can be inferred from this that there was resistance of this eucalyptus hybrid to the water stress (Silva et al. 2014).
E. urograndis showed greater growth, which can be attributed to the greater number and intensity of damage and weaknesses observed in E. cloeziana, due to the continuous attack of Trigona spinipes, causing branches and bifurcations and also weed interference (Table 2). According to Santos et al. (2017), it is important to use eucalyptus clones that are more adapted to local edaphoclimatic conditions and that are more productive and have quality wood for multiple uses. Pulrolnik et al. (2010) studied the growth of E. cloeziana and E. urograndis in the Cerrado of the Amazon region and reported higher DSH and Ht values than found in the present study. At 540 DAT, the authors found higher DSH values, with 1.14 and 2.42 m for E. urograndis and E. cloeziana, respectively. For Ht, the values were 1.42 and 3.06 m higher for E. urograndis and E. cloeziana, respectively. However, it is important to note that the rainfall in the study region of Pulrolnik et al. (2010) were larger and better distributed than in the region of the present study, which directly in uences the growth rates of the evaluated genotypes. Comparatively, in the present study, E. cloeziana was more affected by the frequency and distribution of rainfall than E. urograndis.

Conclusions
Trigona spinipes and termite attacks were the two main biotic factors that caused damage and weaknesses in E. cloeziana and E. urograndis, respectively, reducing growth and increasing mortality rates of the two genotypes.
In the two evaluated arrangements of ICLF systems, E. cloeziana genotype showed the highest survival rate, but E. urograndis hybrid showed greater resistance to the water stress condition observed by its growth performance evaluated by BSH, DBH and Ht.
The E. urograndis showed tolerance to water stress and adaptation to edaphoclimatic conditions in the present study, Therefore, it can be indicated do ICLF systems implanted in regions characterized by high temperatures and irregular rainfall distribution.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study was funded by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES; Finance Code 001), and the Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG; grant numbers APQ-00953-14 and PPM-00133-15). The funding agencies had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.

Authors' contributions
Conceived and designed the study: Barbosa