Introduction
The family Asteraceae has an abundance of diverse specialized metabolites, which have storage or chemical defense functions, and the development of these metabolites is a factor of immeasurable importance in the evolutionary success of the family. Chemically, it is known to produce monoterpenoids, diterpenoids, triterpenoids, sesquiterpenoids and sesquiterpene lactones, polyacetylenes, flavonoids, phenolic acids, benzofurans, coumarins and pyrrolizidine alkaloids, which are, with few exceptions, restricted to the Senecioneae and Eupatorieae tribes (Anderberg et al., 2007; Calabria et al„ 2009).
The genus Vernonia (Asteraceae) is composed of more than 1000 species, which can be found in tropical and subtropical areas of the Asian, African and American continents, and of these approximately 350 are located in South America, occurring frequently in southern Brazil, Argentina, Paraguay and Bolivia (Robinson,1999). Many species of this genus have medicinal properties and are used in some African countries for the treatment of malaria, amebiasis, infections and sexually transmitted diseases (Hamill et al., 2000; Tona et al, 2004).
V. polyanthes has diverse pharmacological properties. According to Lorenzi and Matos (2002) its phytochemical composition includes alkaloids, glycosides, flavonoids and essential oils. However, it is important to study the plant's phenology to develop restoration and conservation, which are essential processes for sustainable management and tools for domestication of the species (Ribeiro et al., 1981; Vieira; Alves, 2003). Species of the Cerrado, under seasonal climates, show variations in seasonality with regard to the production of leaves, flowers and fruits, which determines adaptations to biotic and abiotic factors. These adaptations are linked to functional or structural characteristics, which are analyzed by phenology. Despite this, some authors argue that the seasonal limitation of the flowering period and other phenological events are established phylogenetically, that is, per phylogenetic constraints and strongly induced by the way of life, being independent of biotic or abiotic factors such as precipitation (Wright and Calderon 1995).
The annual development of Cerrado plants has well-defined phenophases, where each phenomenon occurs at a given time of year (Mantovani and Martin, 1998). The sprouting of buds in underground organs begins in the spring, after the rapid growth of the aerial part and development of reproductive organs during the summer and until the beginning of autumn. The stem then stops growing and the leaves begin to show the first signs of senescence, which leads to abscission of the aerial parts and dormancy of the plant. Finally, at the end of the winter, there is new budding as dormancy is broken (Carvalho and Dietrich, 1993).
Therefore, the objective of the present study was to observe and describe the phenology of V. polyanthes, to obtain important information for the domestication of this species.
Material and Methods
Study Area
The present study was carried out at Lageado Experimental Farm, belonging to the Faculdade de Ciências Agronômicas UNESP, Botucatu Campus. The area is located in the Experimental Farm Lageado, of the UNESP, Botucatu Campus, which is about 3 km from the seat of the city of Botucatu - SP, Brazil. The geographical coordinates are 22°52'20" S and 48°26'37" W. According to the Koppen classification (1948), the climate of the region is Cfa (warm humid temperate (mesothermal) climate), with average temperature of the hottest month exceeding 22 °C). The area of Botucatu is predominantly Cerrado, bordered by semideciduous seasonal forest. The only significant watercourse that cuts through the study area is the Lava-pés River, tributary of the Tiete Basin. This watercourse, before going through the Lageado farm, crosses the urban part of Botucatu. The soils of the study area are clayey, dark red in color, with predominantly undulating terrain. In some areas, the topography is smooth, but near the areas of shallow soils (Latosols), the relief becomes more rugged, with a dense drainage network (Sousa; Cavalheiro, 1988).
Phenological Observations
The method of sampling tracks described by d'Eça Neves and Morellato (2004) was used here. We randomly selected 40 adult individuals of V. polyanthes that appeared to be in satisfactory phytosanitary condition (absence of diseases and parasites). Observations and collections of the phenological data were performed monthly, from August 2014 to August 2016, totaling 24 months of observation. The entire population was considered adult, since all the individuals flourished and/or fruited during the study period.
The presence or absence and intensity of the following phenophases were recorded: (1) budding: beginning with the appearance of the leaf shoots and until the total expansion of the new leaves; (2) mature leaves: fully expanded leaves, dark green in color; (3) leaf senescence: indicated by the progressive yellowing of the leaves; (4) flowering: appearance of the flower buds until the last flowers;
(5) fruiting: indicated by the end of anthesis and subsequent confirmation of the existence of seeds;
(6) leaf fall: whenever there were empty spaces on the branches or when fallen leaves were observed on the ground just below the individual.
At the individual level, this method of analysis had a qualitative character. At the population level, however, it assumed a quantitative character, indicating the percentage of individuals in the population that were manifesting a certain phenological event. This method was also used to estimate the synchrony between individuals of a population, considering that the greater the number of units manifesting the phenophases at the same time, the greater the synchrony of this population.
We used the method proposed by Fournier (1974) to quantify the phenophases, which we then evaluated individually, using a semi-quantitative interval scale of five categories (0 to 4) and 25% interval between each category, where: 0 = absence of the phenophase; 1 = the presence of phenophase with magnitude between 1 and 25%; 2 = presence of phenophase between 26 and 50%; 3 = presence of phenophase between 51 and 75% and 4 = presence of phenophase between 76 and 100%. Considering the intensity of each phenophase, the percentage of Fournier then calculated (Fournier, 1974).
Fournier intensity percentage - In this method, the values obtained in the field through a semi-quantitative interval scale of five categories (0 to 4) allowed us to calculate the percentage of intensity of the phenophases. In each month, we determined the intensity values for all individuals of the species, which was divided by the maximum possible value (number of individuals multiplied by four). The value obtained, which corresponds to the proportion, was then multiplied by 100, to convert it to a percentage value (Fournier, 1974).
Population synchrony in relation to the phenophases evaluated was determined using the method of Bencke and Morellato (2002), who considered the following proportion of individuals manifesting a certain phenological event: <20% asynchronous; 20-60% little synchronous; >60% synchronous. This calculation was estimated in the period of maximum activity of each phenophase.
The meteorological data was provided by the Department of Soils and Environmental Resources and obtained from the Meteorological Station of the Department of Soils and Environmental Resources, located in the same place as the study area. The normality of data distribution was tested, and since the data did not show a normal distribution, Pearson correlation analysis (r) was used to determine the relationship between the climatic variables (monthly total precipitation, mean temperature and mean relative humidity) and the intensity of the phenophases.
Results
The species studied displayed seasonal phenological behavior with annual periodicity, where synchrony between the individuals evaluated during 24 months was generally elevated (Table 1). The decrease in precipitation, relative humidity and temperature was accompanied by a gradual increase in leaf fall, thereby demonstrating a positive relation with evapotranspiration.
Table 1 Degree of synchrony according to phenophase, in percentage of individuals, estimated in the period of maximum activity of the species, Botucatu- SP, 2014/2015.

The first shoots of V. polyanthes emerge in the first half of August. In relation to mature leaves, maximum intensity was reached in February, exactly one month after the maximum recorded precipitation, in contrast to leaf fall, which showed the highest intensity in the month of June. These phenophases showed opposite behavior, since the mature leaves correlated positively with the climatic variables, while leaf fall and senescence showed a negative correlation (Figure 1).

Figure 1 Leaf fall and flowering of V. polyanthes in relation to temperature (ºC) and precipitation (mm) in Botucatu-SP.
Flowering was annual with a duration of two months and was negatively correlated (Table 2) with all climatic variables analyzed, but mainly with temperature (Figure 2). However, peak flowering occurred in the month of July, which was in the driest period of the year.

Figure 2 Leaf senescence and mature leaves of V. polyanthes in relation to temperature (ºC) and rainfall (mm) in Botucatu-SP.
Table 2 Pearson correlations between the evaluated phenophases of Vernonia polyanthes Less. and climatic factors (temperature, relative humidity and mean rainfall) at Fazenda Lageado - FCA / UNESP, Botucatu. 2014-2015.

The fruiting phenophase began in August and its maximum intensity occurred in the first half of October.
Discussion
Under conditions of low precipitation, tem perature and relative humidity, there is a tendency for increased evapotranspiration. This increase in evapotranspiration is associated with greater nutrient withdrawal from the soil, where at the end of this season, nutrient absorption is hampered by low air humidity and decreased availability of water in the soil, causing the loss of leaves for most plant species in the Cerrado (Morellato, 1992). It is possible that during the dry season, there is a diversion of resources from the vegetative phase to the reproductive phase, since these deciduous species invest in flower production and fruit formation during the period of intense leaf senescence (Foster, 1990).
Deciduous species show budding immediately after the first rains, where a small amount of available water is enough to break bud dormancy which induces sprouting (Borchert et al., 2002). A decrease in rainfall triggers leaf fall in a Cerrado environment (Santos et al., 2009).
Peak flowering occurs in July in the driest period of the year. This behavior is probably associated with greater protection of the reproductive organs and increased efficiency of pollinators, since intense rains can damage floral parts and hamper the action of pollinators, affecting the production of seeds by plants (Pedroni et al., 2002). Generally, flowering during the dry season occurs after the abscission of the leaves and breaking of dormancy of the apical meristems of the aerial part of the plants, coinciding with the resumption of growth. This sequence of phenological events during the dry period is a typical pattern for deciduous tree species in the savanna and Cerrado (Opler et al. 1980, Miranda, 1995), where it is evident that flowering or anthesis is induced by rehydration due to a reduction of transpiration (Borchet, 1994).
The fruiting phenophase begins in August and its maximum intensity occurs in the first half of October. Fruit dehiscence occurs when the climate is conducive to dispersion of the anemochorous propagules (Schaik et al., 1993). In Cerrado vegetation, woody plants disperse their seeds during the dry season, which seem to germinate readily, whereas the seeds of species with dormancy are disseminated predominantly in the rainy season (Oliveira, 2008).
Conclusion
The phenological events of the population of V. polyanthes studied were synchronous; that is, occurred in the same period. Growth was periodic and seasonal, and the climatic variable that correlated strongly with phenophases was temperature. The vegetative phenophases were concentrated in October to May and the reproductive ones in June to the first half of October.