Dynamics of defoliation of associated grasses

The behavior of variables related to defoliation dynamics was evaluated, and their respective oscillations in two intercropping of tropical climate grasses managed under two grazing intensities. We used pastures formed with two intercropping (BRS Zuri, Xaraés and Basilisk; BRS Quênia, Marandu and BRS Paiaguás), managed in two grazing intensities: 40 and 60% of the pre-grazing height. In the consortium between BRS Zuri, Xaraés and Basilisk, the first canonical variable explained 72.1 and 79.2% of the total variation of the defoliation pattern for the grazing intensity levels of 40 and 60%. In the consortium between BRS Quênia, Marandu and BRS Paiaguás, the first canonical variable explained 84.3 and 89.0% of the total variation of the defoliation pattern for the grazing intensity levels of 40 and 60%. The choice of forage species of tropical climate to form the consortium may be decisive in order to perpetuate plant diversity. From the population density of tillers, it is possible to observe that grasses belonging to functional groups A and B (Zuri, Xaraés, BRS Quênia and Marandu) have a higher probability of coexisting in the same area. The choice of forage species for the establishment of intercropped pastures is important to ensure the coexistence between them and the persistence of the consortium. The analysis of canonical variables assists in explaining the dynamics of defoliation of intercropping through the estimates of severity and frequency of defoliation of extended tiller and pseudostem.

germination rate of 50% was observed for the genus Brachiaria and 20% for grasses of the genus Panicum.
The pasture structure for each grazing intensity was previously molded from May 2017 to February 2018, the lowering of the pasture height was performed by Caracu cows, with approximately 24 months of age and average body weight of 450 kg belonging to the EMBRAPA herd. This period was necessary to ensure that grasses could adapt to the grazing intensities imposed.
The heights of pre-grazing of 70 cm and 60 cm for the intercropping ZXD and QMP, were respectively established. In addition, two grazing intensities, 40% (Lenient) and 60% (High) of the pre-grazing height, were established as post-grazing goals.
The goal of grazing intensity was based on the average height of the forage canopy of the genus Panicum, since this was the predominant grass in the intercropping. With the aid of a graduated ruler, 60 points were recorded per picket related to the height of the species of the genus Panicum allowing a maximum variation of 10% between the measured height and that specified for each intensity. The grazing method adopted was intermittent stocking. The adjustment of the instantaneous stocking rate was based on the forage biomass for each desired grazing intensity.

Evaluations
The data collection period began after the stabilization of the pastures, when they reached the pre-grazing height (From March to May 2018).

Forage mass and botanical composition
In the pre and post grazing the forage mass was estimated by collecting six samples Research, Society and Development, v. 9, n. 4, e181942595, 2020(CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i4.2595 7 per picket. Forage collection was performed by cutting the forage close to the soil, which was contained in the interior of 1m² squared frames, and the samples were packaged in plastic bags and identified. The samples were taken to the laboratory for botanical separation with identification and tiller count per species of each consortium. Subsequently, the morphological components were separated for each cultivar for the quantification of the leaf lamina mass, pseudostem and dead material. After the separations, the samples were pre-dried in a forced air circulation drying oven at 55 ºC until reaching constant weight.

Defoliation pattern
To detect the oscillations in the defoliation pattern during the lowering period of the pastures, the technique of marked tillers was used (Hodgson, 1966). For this, during the pregrazing, 45 vegetative tillers were marked per picket, 15 of each grass of the consortium, identified and distributed in equidistant points. The identification of each tiller was performed by means of colored ribbons on an iron rod to facilitate the identification of the marked tillers.
In the marked tillers, the measurements of extended tiller were taken, corresponding to the distance between the ground and the apex of the highest leaf when vertically positioned. The pseudostem measurement was also taken, being the distance between the soil and the lygula of the last leaf lamina completely expanded. These measurements were performed with the aid of a centimeter graduated ruler.
After the height measurements of the extended tiller and pseudostem in the pregrazing, the animals had access to the pickets. Every two days, the evaluations of the extended tiller and pseudostem of each grass were performed again, being classified as intact (the tillers that did not have size reduction by grazing), or defoliated (when the size reduction occurred compared to the measurement performed in the pre-grazing). The animals remained in the pickets for eight days, allowing to perform four evaluations of defoliation dynamics during the canopy lowering period.

Statistical analysis
8 For characterization of the forage canopy: forage mass data, population density of tillers were grouped into means. To perform the analyses, 36 observations were quantified for each consortium, according to their respective grazing intensity. Defoliation data were analyzed according to a completely randomized model, in a factorial arrangement 2 (Intercropping) x 2 (Grazing intensities). When appropriate, Tukey's mean comparison test was performed and a 5% significance level was considered. For the analyses, the ExpDes package (Ferreira et al., 2014) of the software R version 3.5.0 was used.
The information related to the extended tiller defoliation frequency (ETDF), pseudostem defoliation frequency (PDF), extended tiller defoliation severity (ETDS) and pseudostem defoliation severity (PDS) were used to estimate the first and second canonical variables, as well as the standardized canonical coefficients for each consortium within each grazing intensity. The Candisc function of the software R version 3.5.0 for this estimation was used.

Structure of the forage canopy
The pre-grazing heights for each consortium presented values close to what was established as a management criterion. For both intercropping, approximately 50% of the forage mass was composed of leaf laminas. In the intercropping managed at 60% height removal, the highest stocking rates were observed, consequently, the lowest values of leaf lamina and pseudostem mass were observed in the post-grazing. After the stabilization period of the intercropping, it was observed the coexistence of the grasses, however, not presenting the desired proportions, because the cv. Basilisk and cv. BRS Paiaguás present a less than 10% population of tillers (Table 1).  Research, Society and Development, v. 9, n. 4, e181942595, 2020 (CC BY 4.

Defoliation dynamics between intercropped
Interaction between intercropped and grazing intensities was not observed for the variables related to defoliation dynamics. The ETDS of the QMP consortium was 10% higher compared to the ZXB consortium, and PDS in the QMP consortium was 26% higher compared to the ZXB. The ETDF and PDF do not present oscillations between the intercropped. In grazing intensities, the variables related to defoliation dynamics present proportionality (Table 2). managed with 40% and 60% removal of the height of the forage canopy.

ZXB Intercropped
In the ZXB intercropped, regardless of defoliation intensity, the standardized canonical coefficients of the first canonical variable (Can1) showed that the ETDS and PDS are the variables that most interfere over the first canonical variable, in the positive and negative quadrants, respectively. This indicates that these are the variables that most interfere on the differentiation between the defoliation patterns of the forage evaluated (Table 3). Fig. 2 shows standardized canonical coefficients for defoliation dynamics in the first and second canonical variables of the intercropped of Panicum spp. and Brachiaria sp. managed with 40% and 60% of removal of the forage canopy height. In the ZXB intercropped, the first canonical variable explained 72.1 and 79.2% and the second 20.8 and 28.0% of the total variation of the defoliation pattern, for the grazing intensity levels of 40 and 60%, respectively (Fig. 2). Research, Society and Development, v. 9, n. 4, e181942595, 2020 (CC BY 4.