Morphogenic, structural characteristics and population stability index of ryegrass tillers submitted to stocking methods

This study aimed to evaluate the morphogenic and structural characteristics and the population stability index of ryegrass ( Lolium multiflorum Lam.) tillers when the pasture was submitted to two stocking methods during grazing cycles. The experimental design was a complete randomized block, with repeated measurements over time (n=6 cycles), two treatments (stocking methods) and three area replicates. In the continuous stocking method, the highest population density of tillers was observed. The highest tiller weight occurred in the rotational stocking method. The morphogenic variables and the other structural variables were not altered by the stocking methods and showed differences during the grazing cycles. The tiller population stability index was similar in the two pasture management strategies, and both can be used for ryegrass management considering this parameter.



The livestock activity in Brazil is based on pasture utilization as the main food resource and, in the Southern states, the most utilized species is Italian ryegrass (Lolium multiflorum Lam.) due to its productive potential and good adaptation to environmental conditions in that region.The adoption of pasture management strategies aimed at maximizing both plant and animal production are extremely important.The adoption of different ways to conduct stocking, continuously or rotationally, are among these strategies.
In continuous stocking, animals have unlimited and uninterrupted access to the entire area to be grazed during the entire grazing season and in rotational stocking, alternation between defoliation and rest periods occurs.Due to the intercalation of rest and grazing periods, in rotational stocking, the regrowth process occurs in isolation from the grazing process.In this context, it is recommended that defoliation management be carried out using degree-days as it allows a direct association with the morphogenic characteristics of forages plants (Ongaratto et al., 2020).On the other hand, continuous stocking is characterized by milder changes in the sward condition over the period.
Both stockings affect the adaptive responses of forage plants differently.The persistence strategies of the plants under different defoliation methods can be accessed by studying individual tillers in the plant community.This information allows the identification of management tools that result in canopies with a favorable structure for plants and animals (Carvalho et al., 2001).
In this context, the study of the morphogenic variables associated with the tiller stability index becomes important in the comparison of stocking methods.In winter grasses, published data evaluating stocking methods are limited to information on morphogenic characteristics and other pasture variables (Cauduro et al., 2006(Cauduro et al., , 2007;;Barth Neto et al., 2013).Considering tiller population stability, research has been carried out with a focus on different grazing practices (Graminho et al., 2014;Stivanin et al., 2014;Duchini et al., 2017).
However, it is necessary to jointly evaluate pasture growth through its morphogenic characteristics when associated with plant population stability in different grazing methods.This favors the estimate of the effect of environment and management factors on the pasture, allowing a better understanding and manipulation of the processes involved (Bahmani et al., 2003;Caminha et al., 2010;Sbrissia et al., 2010).The objective of this study was to evaluate the morphogenic and structural characteristics and the stability index of the tiller population of ryegrass (Lolium multiflorum Lam.) when the pasture was submitted to two stocking methods.

MATERIALS AND METHODS
The experiment was approved by the Ethics Committee for Animal Experimentation of Universidade Federal de Santa Maria (UFSM), Protocol 9708210518.The experiment was performed from July to November 2019 at the Universidade Federal de Santa Maria (UFSM), Santa Maria, RS, Brazil.Climate, according to Köppen´s classification, is humid subtropical.The meteorological data for the months that comprised the experimental period were obtained from the UFSM Meteorological Station (Fig. 1).Al -3 = 0.7cmolc L -1 ; Ca +2 = 3.6cmolc L -1 ; Mg +2 = 1.6cmolcL -1 ; CTC pH7 = 11.4.The fertilizer consisted of 200kg ha -1 of the formula 05-20-20 (N-P-K) and 115kg of N ha -1 in urea form split into three applications.
We evaluated two grazing management strategies: continuous stocking -keeping forage mass between 1.200-1.600kg DM ha -1 and rotational stocking -with post-grazing target sward height of 10±1 cm.Criterion utilized to determine the rest period for rotational stocking was the thermal sum (TS) of 187.5 degree-days (DD), equivalent to 1.5 Italian ryegrass phyllocron value (Confortin et al., 2010).Thermal sum (TS) was calculated by the equation: TS= ∑(Tmd) -5, where: Tmd is the daily average temperature of the stocking cycle; the value of 5 grade is the minimum temperature required for growth of cool season forage species.
Experimental animals were Angus heifers with initial age and body weight (BW) of eight months and 157±3.94kg, respectively.Three test animals (permanent animals during all stocking period) were used per experimental unit.For the maintenance of post-grazing canopy height in rotational stocking and forage mass in continuous stocking, a variable number of putand-take animals were used.The pasture utilization comprised six stocking cycles (1-Jul./07 to Aug./13; 2-Aug./14 to Sep./02; 3-Sep./03 to Sep./17; 4-Sep./18 to Oct./02; 5-Oct./03 to Oct./17; 6-Oct./18 to Nov./01.Morphogenic measurements were performed in cycles 1 to 4 and stability index of the tiller population from 1 to 6.The duration of stocking cycles, in days, was similar for both stocking methods. For every stocking cycle 20 measurements of canopy height per experimental unit (pre-and post-grazing in rotational stocking) were taken.In continuous stocking, forage mass (FM, kg DM ha -1 ) was determined by visual estimation technique with double sampling.For rotational stocking, FM was determined by two herbage mass samples taken at ground level every preand post-grazing using frames of 0.25m 2 .These places were representative of the paddock average canopy height.The forage from cuttings were dried at 55 ºC for 72 h.The forage allowance (FA; kg DM kg BW -1 ) was calculated in each stocking cycle according to the methodology of Sollenberger et al. (2005).The stocking rate (kg BW ha -1 ), was calculated by adding the average weight of the test animals to the average weight of each put and take animal, multiplied by the number of days that it remained on the paddock divided by the number of days of the stocking cycle.The instantaneous stocking rate, in the rotational stocking, was calculated by the quotient between the heifer's live weight and the area of the plot occupied by them.
Morphogenetic and structural characteristics were evaluated in 18 tillers per experimental unit, by of marked tillers.Every beginning of a new stocking cycle, a new group of tillers was selected for evaluation.In continuous stocking and in the rest periods in rotational stocking measurements were made twice weekly.In rotational stocking the measurements were made daily when the plot was being grazed.In these occasions, the length of fully expanded, expanding, and senescent leaf blades, canopy height, and height of pseudostem were measured in cm.From these measurements, the following variables were calculated: leaf appearance rate (degree-days), leaf expansion rate (cm degreedays tiller -1 ), leaf senescence rate (cm degreedays tiller -1 ), leaf elongation duration (degreedays), leaf lifespan (degree-days), phyllochron (inverse of the rate of appearance of the leaves (degree-days) and number of live leaves (Lemaire and Chapman, 1996).
The tiller population density and tiller population stability index (IS) variables were evaluated by monitoring tiller generations.The first generation of tillers was tagged with plastic wires of same color, in two fixed frames (0.0078 m 2 ) in each experimental unit, at the starting of first stocking cycle.In this occasion, the number of tillers of this generation was similar (P>0.10) in all paddocks.Each new grazing cycle the living tillers tagged in the previous generation were counted again and untagged tillers (new tillers) were tagged with plastic wires of a different color.The dynamics of tillering was carried out from the identification and counting of the remaining living tillers and the appearance of new tillers.With the sum of the number of tillers belonging to each generation, it was possible to calculate the density of tillers in each generation (tillers m -2 ).We calculated rates of tiller appearance (RTA), mortality (RTM) and survival (RTS; tiller tiller -1 m -2 ) and population density of tillers (PDT; tillers m -2 ).The population stability index (SI) of tillers was calculated according Bahmani et al. (2003), in which: SI = RTS*(1+RTA).To determine weight per tiller (g DM tiller -1 ), cuts were made in two areas (0.0625 m²).The number of tillers in these areas was quantified and, subsequently, these tillers were dried in a circulating air oven at 55°C for 72 h and after weighed.The dry mass value was divided by the number of tillers in the sample.
A randomized complete block following a repeated measures arrangement (n= 4 stocking cycles for morphogenetic variables or n= 6 for tillering variables) was used, with two stocking management strategies (continuous and rotational) and three area replications (paddocks).To compare the stocking systems, the variables with normal distribution were evaluated considering the fixed effects of stocking systems, stocking periods and their interactions and the random effects of blocks, residuals and paddocks nested in stocking systems using the Mixed.We performed a structure selection test, following the Bayesian Information Criterion (BIC) to determine the model that best fit the data.When not fitted to regression models, the mean values were compared using the 'lsmeans' procedure.The interaction between stocking systems and stocking cycles was broken down, when significant at 5% probability.The variables were also analyzed using Pearson correlation analyses.

RESULTS
Meteorological data of the experimental period showed that the average values of temperature, rainfall and insolation were 11.4%, 15.6% and 10.9% higher in relation to the historical averages (Fig. 1).
There was no interaction (P>0.05) between stocking methods × stocking cycles for forage mass, stocking rate and forage allowance variables.The forage mass was similar in paddocks used to evaluate stocking methods (Table 1).The forage mass differed between stocking cycles (P<0.10)being higher in the sixth (1654.6±117.0kgDM ha -1 ), intermediate in second and fifth cycles (1512.2±117.0kg DM ha -1 ) and lower for the other cycles (1414.0±117.0kgDM ha -1 ).The post-grazing canopy height mean value in the rotational stocking was 11.5±0.7cm.The average canopy height in continuous stocking was 14.9±2.0cm.The grazing heifers, in both stockings, were submitted to the same forage allowance (1.34±0.15kgDM kg BW -1 ; P = 0.4680).The stocking rate was 18.5% higher in rotational compared to the continuous stocking (Table 1) and did not differ between stocking cycles (P>0.10;1188.4±149.9kgha -1 of BW).The instantaneous stocking, in rotational stocking, presented an average value of 7819.5±1323.8kgha -1 of BW.
There was no interaction (P>0.05) between stocking method × stocking cycles for the variables leaf appearance rate, phyllochron, leaf lifespan, leaf senescence rate, leaf expansion rate and number of live leaves.These variables did not differ between stocking methods (Table 1).There was a difference between stocking cycles (P<0.10) for leaf appearance rate, phyllochron, leaf senescence rate, leaf expansion rate and number of live leaves.The leaf appearance rate Arq.Bras.Med.Vet.Zootec., v.74, n.6, p.1134-1142, 2022 was higher and similar in grazing cycles 1 and 2 (0.008±0.003 leaf degree-days -1 ), and lower in 3 and 4, similar to each other (0.007±0.003 leaf degree-days -1 ).The phyllochron was smaller in grazing cycles 1 and 2 and similar to each other (137.6±5.8 degrees-day) and greater (154.55±5.86degrees-day) in 3 and 4 that did not differ between themselves.
The leaf senescence rate was higher and similar in grazing cycles 1 and 2 (0.043±0.005cm degree-days -1 ), lower and similar to each other in cycles 3 and 4 (0.025±0.005 cm degree-days -1 ).The leaf expansion rate was higher and similar in grazing cycles 1 and 2 (0.069±0.007cm degreedays -1 ), lower in cycles 3 and 4 (0.033±0.007 cm degree-days -1 ).There was no difference between generations of tillers for leaf lifespan (504.8±19.7 degree-days; P>0.10).The number of live leaves was adjusted to the decreasing linear regression model in relation to the thermal sum (Ŷ = 4.3725-0.0016x;r² = 0.57; P = <0001; CV = 8.20).There was interaction (P<0.05) between stocking methods × stocking cycles for pseudo-stem height.In the stocking cycle 4, there was a difference between the stocking methods for pseudo-stem height, being greater in continuous stocking method (6.2±0.32 cm; P = 0.0221) in relation to rotational (5.2±0.32cm).In the other stocking cycles the height of the pseudo-stem was similar (P> 0.05) for both methods, being 4.1±0.32cm in cycle 1, 4.34±0.32cm in cycle 2 and 4.8±0.32cm in cycle 3.
There was no interaction (P>0.05) between stocking methods × stocking cycles for tiller population density and tiller weight.There was a difference between stocking methods (P=0.0238), with tiller population density being 13 % higher in the continuous (2176.1±74.8tillers m²) than in the rotational stocking (1919.0±74.8tillers m -²).Tiller weight was 14% higher in the rotational stocking (P<0.10;0.033±0.002g)than in the continuous stocking (0.029±0.002 g).
There was no interaction (P>0.05) between stocking methods × stocking cycles for the tiller population stability index.There was a difference between tiller generations (P<0.10) for the tiller population stability index.The stability index of tillers (Figure 2) was higher and equal (P>0.05) in G2 and G3 generations (1.66±0.1), the G5 generation (1.13±0.1)did not differ from the generation G4 (1.28±0.1)and G1 and G6 generations were smaller and similar to each other (0.96±0.1).

DISCUSSION
The stocking practices alter the pasture structure and the leaf area index, which, in turn, alters the generation and expansion of new tissues (Duchini et al., 2017).However, it is important to show that these alterations affect, firstly, the structural characteristics, highlighting the changes that involve the tiller population (Lemaire and Chapman, 1996).According to Cauduro et al. (2006), stocking practices influence morphogenic characteristics (continuous stocking results in higher rates of leaf appearance and elongation and leaf lifespan in the first stocking cycle) and structural characteristics, with the highest tiller population density observed in continuous stocking.The information should be viewed with caution, as the authors describe that the beginning of the experimental protocol occurred when the ryegrass pasture was at an advanced phenological stage.
The similarity observed for the morphogenic traits, leaf appearance rate, phyllochron, leaf expansion, lifetime of ryegrass leaves in different stocking practices can be explained because these variables are genetically determined, although they may be influenced by biotic and abiotic (Lemaire et al., 2009).The highest rate of leaf appearance occurred during the vegetative stage, which is characterized as favorable to plant development (Lemaire and Chapman, 1996).Leaf appearance rate was negatively correlated with pseudostem height (p = 000.2;r = -0.69).The increase in pseudostem height is associated with the greater length that the leaf blade needs to travel before being emitted in the canopy (Duru and Ducrocq, 2000).
The phyllochron was lower in the initial stages of ryegrass development (cycles 1 and 2) and according to Lemaire et al. (2008), this condition is associated with the greater efficiency with which the grass intercepts and converts light energy in leaf tissue.The increase in phyllochron with the approach of the reproductive stage was significant and expected, since at this stage the plant allocates most of the nutrients to the formation of the reproductive structure, considerably reducing the production of new leaves (Cauduro et al., 2006;Duchini et al., 2017).
The similarity of leaf expansion occurred, possibly because forage mass and phyllochron value were also similar for both stockings (Table 1).The leaf expansion seems to be the morphogenic variable that, in isolation, most directly correlates with the forage dry mass (Confortin et al., 2010).The lowest rate of leaf expansion from the third grazing cycle is related to the phenological stage of ryegrass at this time, allocating photo assimilates primarily to reproductive structures and reducing expansion rate (Duru and Ducrocq, 2000).
The lifespan of leaves was similar between stocking practices and stocking cycles.The variable is dependent on the number of live leaves and the phyllochron, which also remained similar during the experimental protocol.It confirms the similarities in maximum forage production potential for the tested stocking practices.The lives lifespan was on average 505.5 degree-days, which was lower than that obtained when ryegrass was managed under rotating stocking (557.3 degree-days) (Confortin et al., 2010).If all tiller leaves had been removed, the potential ceiling of annual ryegrass growth (lifespan of leaves) would occur 40 days after the start of the stocking cycle.
Even though the number of live leaves is a stable genotypic characteristic for a given grass species (Martuscello et al., 2015), with an increase of one degree in the thermal sum accumulated during pasture use, there was a reduction of 0.0016 in the number of live leaves.The reduction in the number of live leaves as the stocking cycle progresses is probably due to the translocation of nutrients to form the reproductive structure at the expense of foliar production (Cauduro et al., 2006).The number of live ryegrass leaves agrees with that described in the literature, between three and four live leaves per tiller (Confortin et al., 2010).
Leaf senescence is a natural process that characterizes the last stage of leaf development, which begins after its complete expansion.The senescence of a leaf in a tiller starts shortly after reaching the balance between the rate of appearance and senescence to keep the number of live leaves per tiller constant (Confortin et al., 2010).The highest value for the leaf senescence rate was concentrated during the initial stage of ryegrass development (cycles 1 and 2) and may Arq.Bras.Med.Vet.Zootec., v.74, n.6, p.1134-1142, 2022 be associated with a higher rate of leaf appearance.
The higher pseudostem height in continuous stocking (cycle 4) is probably a result of increased competition for light between tillers.In this condition, the plant prioritizes the allocation of carbon for the extension of the internodes, to position the young leaves in the upper stratum of the forage canopy (Lemaire, 2001).Stem elongation occurs during the flowering season, and in the present study this change in pasture structure seemed not be a barrier to defoliation, as mentioned by Roman et al. (2007) in ryegrass.
For the tiller population to be considered stable, there must be a balance between death and tiller emergence (Lemaire and Chapman, 1996;Sbrissia et al., 2010).This condition can be monitored from the tiller population density (Matthew and Sackville-Hamilton, 2011).In continuous and rotational stocking, the highest (2176.1 tillers m -2 ) and the lowest tiller population density (1919.0tillers m -2 ) were obtained, respectively.The results were lower than those obtained by Cauduro et al. (2006) in continuous stocking (3684.80tillers m -2 ) and rotating stocking (2666.18tillers m -2 ).The lower tiller population density in the rotating stocking was expected, due to the presence of a rest period, which determines a free growth of the plants in the absence of grazing.In this way, the plant allocates the reserves and production of photoassimilates to form leaves and reproductive structures of the main stem, reducing tiller production (Cauduro et al., 2007).
The inverse relationship between tiller population density and tiller weight was observed (Calsina et al., 2012).The highest tiller population density observed in continuous stocking is associated with lighter tillers (0.029 g), the opposite being evidenced for rotating stocking (0.033g).It is evident that the variable tiller population density allowed ryegrass greater flexibility and adaptation to the stocking practices employed.Changes at the tillering pattern allow the plant to adapt more quickly to environmental and management conditions when compared to changes in morphogenic and other structural characteristics (Lemaire and Chapman, 1996;Duchini et al., 2017).
On the other hand, the tiller population can be analyzed in more depth, considering information related to tiller generations (Matthew and Sackville-Hamilton, 2011), as well as the tiller stability index.Thus, a low tiller population alone should not be considered an indicator of loss of productive potential and reduced plant persistence, since the pasture can be stable even with low tiller population (Sbrissia et al., 2010).
From the monitoring of the stability index of the tiller population, it is possible to infer about the stability of plants in the pasture.In the sward the lower population density of tillers for rotating stocking is according with the literature consulted (Lemaire and Chapman, 1996;Caminha et al., 2010).It is important to emphasize, however that defoliation management based on rest periods per degreedays may have contributed favorably to the maintenance of the tiller population stability index, since the adopted practice allows a direct association with the tiller´s morphogenic (Ongaratto et al., 2020).
The evaluation of the pasture stability index allows the joint assessment of tiller appearance and survival rates, that is, the replacement capacity of dead tillers (Bahmani et al., 2003;Duchini et al., 2017).According to Bahmani et al. (2003), stability values below 1.0 indicate unstable pastures.The tiller population stability index was less than 1.0 in the G1 and G6 generations.In the G1 generation, the lowest value for the tiller population stability index may be related to the initial stage of ryegrass development.In the G6 generation, the lowest value may be associated with the end of the ryegrass cycle, which, being an annual forage species, was in the flowering stage (Sbrissia et al., 2010).The results show that the highest rates of tiller stability occurred in generations 2 and 3, and in G2 the value in was numerically higher than in G3.This information it is related to the higher rate of leaf appearance during cycle 2, as each emerged leaf has the potential to originate new tillers (Duchini et al., 2017(Duchini et al., , 2018)).The results suggest that regardless of the management practice adopted, in most of the evaluated period, it was possible to observe the renewal capacity of ryegrass tillers (higher rate of tiller emergence in relation to tiller death), which is confirmed by the index of stability.

CONCLUSION
In the continuous stocking method, the highest population density of tillers was observed.The highest tiller weight occurred in the rotational stocking method.The morphogenic variables and the other structural variables were not altered by the stocking methods and showed differences during the grazing cycles.The tiller population stability index was similar in the two pasture stocking strategies, and both can be used for ryegrass stocking considering this parameter.Bras. Zootec., v.36, p.282-290, 2007 Rev. Bras. Zootec., v.35, p.1298-1307, 2006 Field Crops Res., v.105, p.253-265, 2008.

Figure 1 .
Figure 1.Average monthly rainfall (mm), insolation (h) and temperature (ºC) from July to October 2019 and normal historical data.Santa Maria/RS.The experimental site had an area of 4.8 ha, divided into six paddocks and of those six, three were subdivided into five plots of 0.16 ha.The ryegrass pasture (Lolium multiflorum Lam.) was