Flora and Annual Distribution of Flowers and Fruits in the Ubajara National Park, Ceará, Brazil

Although the conservation of tropical biodiversity depends on protected areas, there is still a very large ‘gap’ of knowledge on the flora of Brazilian reserves, especially in the Northeast region of Brazil. Field and herbarium surveys of the phanerogamic flora of the Ubajara National Park, located on the Brazilian Northeast, were made and analyses on phenology and dispersal syndromes were performed. 418 taxa (213 trees and shrubs, 100 terrestrial herbs, 68 climbing plants, 33 sub-shrubs, two epiphytes, one hemiparasite and one aquatic herb) were recorded. The most representative families were: Fabaceae, Malvaceae, Asteraceae, Rubiaceae and Euphorbiaceae. The annual flowering / fruiting peak hypothesis was not fully confirmed, therefore, the forest may be an important food resource for the fauna all year long (especially in the moister region). Zoochory was the predominant dispersal syndrome in the moister area, whereas, autochory and anemochory together, predominated in the drier area.


Introduction and objective
The Brazilian semi-arid domain covers an area with a huge physiographic and climatic heterogeneity (associated with continentality, altitude and slope) -which influences moisture and rainfall (Mantovani et al., 2017). In this context, many kinds of vegetation may occur, such as stepic savanna (caatinga and carrasco), savanna (cerrado), coastal semideciduous forest (mata de tabuleiro), semideciduous/deciduous forest (mata seca) and semideciduous/evergreen forest (mata úmida) -the last two can be found mainly in montane regions (Souza & Oliveira, 2006;IBGE, 2012;Moro et al., 2015;Mantovani et al., 2017). According to Zappi et al. (2015), at least 4,659 species of angiosperms were reported for all these plant formations.
The climate in montane regions of the Brazilian semi-arid domain is usually moister than the surroundings, due to orographic precipitation. These habitats are arranged like moist islands in a dry landscape, where forest flora and vegetation develop (Pereira et al., 2010;Nascimento et al., 2012;Pinto et al., 2012;Silva & Figueiredo, 2013). Known as brejos de altitude, they often shelter deciduous, semideciduous and/or evergreen forests, with high species richness in comparison with the surrounding semi-arid landscapes (Pereira et al., 2010;Nascimento et al., 2012;Pinto et al., 2012;Silva & Figueiredo, 2013). Thus, the brejos de altitude may form a patchy and fragmented mosaic of relic vegetation from a moist paleoclimate (Bétard et al., 2007;Mantovani et al., 2017).
The vegetation in the brejos de altitude is also an important resource to local and migratory fauna. The few researches made on these areas show that the seasonality and the water availability rule the phenological patterns and dispersion syndromes. It seems to be a tendency for this kind of biological community to have flower and fruit availabilities all year long (Locatelli & Machado, 2004). Also, it seems to be a predominance of zoochory in moist areas, while autochory and anemochory are predominant in dry areas (Diogo et al., 2016).
The Brazilian semi-arid domain, in which these montane forests are located, is also subjected to overpopulation, about 15% of the national population with 27 million inhabitants (Araújo, 2011), and to an intense rhythm of anthropic exploitation since the Portuguese colonization, especially on the highlands, which are moister than surrounding semi-arid depression (Lopes et al., 2017;Mantovani et al., 2017). Furthermore, protected areas in the Brazilian semi-arid domain account for only 7.8% of its area, of which only 1.3% is under the full protection usage regime, i.e. with restrictions to human intervention (Menezes et al., 2010). Thus, according to Rylands & Brandon (2005), it is one of the most poorly protected regions of Brazil.
Although the conservation of tropical biodiversity depends on protected areas and their surrounding vegetation (Mantovani et al., 2017), there is still a very large 'gap' on the knowledge of the flora of Brazilian protected areas. Most of them are still poorly known, making a consistent analysis of their effectiveness for conservation very difficult (Oliveira et al., 2017;Mattar et al., 2018). According to Moro et al. (2015), floristic studies in the Ubajara mountain forest are highly desirable, since extensive floristic surveys for the Ceará highlands were only performed for Baturité and Araripe areas.
Thus, floristic surveys on Brazilian protected areas have a huge role on supporting conservation, management and restoration programs. Although many efforts have been made to assess the flora of Ceará (Loiola, 2013;Tabosa et al., 2016;Ribeiro & Loiola, 2017;Carneiro et al., 2018;Sampaio et al., 2019), there are still very few published studies on the flora of the protected areas of the state of Ceará. From an inventory of the phanerogamic flora, it could be presented some analysis on the floristic composition, the phenological patterns and the dispersal syndromes found in the Ubajara National Park, located in a montane region of the state of Ceará, Brazil. This protected area is considered of extreme biological importance due to its high diversity and for being listed as a priority area for conservation and sustainable use (Brasil, 2007).

Study area
The Ubajara National Park is a full protection reserve, created in 1959. After two additions (the first in 1973 and the second in 2002) to the Ubajara National Park, its protected area reached 6,288 ha. It is located in the north of the Ibiapaba plateau, near the coast (3° 45' S, 40° 54' W). Due to orographic rainfall, most part of the park is under a moister climate (Aw, Köppen-Geiger) than the surrounding valleys (Bsh, Köppen-Geiger). The terrain stands on a sedimentary sandstone on the higher altitude areas with outcrops of calcareous rocks on the windward slope. The Tropical Seasonal Deciduous Forest (TSDF) develops between 400-700 m and accounts for 72.1% of the area of the park. Between 700-900 m, there is a Tropical Seasonal Evergreen Forest (TSEF), accounting for 18% of the park; and in the lowlands (bellow 400 m), there is the stepic savanna, which accounts for 9.9% of the park (Figure 1). The park is well preserved, it has about 88.7% of conserved vegetation and only 1.9% of disturbed vegetation. On the surroundings of the park, there is significant anthropogenic disturbance, especially on the moist side, with permeability of only 36.8%, while on the dryer side the permeability is 56% (Mantovani et al., 2017).
On the lands with altitudes around 400-700 meters, the average annual rainfall is 950 mm, concentrated from January to June, and the temperature is 28.2 °C (Figure 2A). The soil is 3 -19  shallow and stony and could be classified as Dystrophic Regolith Neosol. The main headwaters of the Acaraú River basin are lotaced inside the Ubajara National Park. Additionally, the rainfall of the eastern slope of Ibiapaba infiltrates into calcareous rocks and resurge in the lower altitudes. This water is used to irrigate agriculture and also for household use. However, both the headwaters and the resurgent water can deplete during the dry season of deforested area (inside or outside the park) during consecutive years of drought.
The TSEF area (above 700 m) exhibits average annual rainfall of 1,500 mm, concentrated from January to June, and an average temperature of 26.1 °C ( Figure 2B). The soil was classified as deep Dystrophic Tb Haplic Cambisol (CXBd).

Data collection and analysis
During approximately six years, the team went through the main park trails and vegetation patches to collect samples of the flora. The first collection was made between 2011 and 2014, being part of the project "Efetividade de UCs Federais do estado do Ceará na conservação biológica do semiárido brasileiro" (CNPq/ICMBio 13/2011, Nº. 551998/2011-3). Additional samples were obtained between 2017 and 2018 during the project "Conservação da biodiversidade em nível de paisagem: mudanças climáticas e distúrbios antropogênicos" (CNPq/ICMBio/FAPs 18/2017, Nº 421350/2017-2). Collections were also made in phytosociological and functional survey plots (unpublished data). Whenever possible, samples in reproductive state (with flowerbuds, flowers and/or fruits) were obtained. Vouchers were deposited in the Herbarium Prisco Bezerra (EAC) at the Universidade Federal do Ceará (UFC). The survey was complemented by gathering samples deposited by other researchers into the EAC collection.
The botanical determination was based on specialized literature, expert opinions and comparisons with other specimens from the EAC collection. The Angiosperm Phylogeny Group IV (APG IV, 2016) classification system was adopted. Names of the families, genera, species and authors names were confirmed on the International Plant Names Index (IPNI, 2019) and in the Flora of Brazil 2020 under construction website (JBRJ, 2019). The classification of exotic species was adopted, as suggested by Moro et al. (2012). It could be gathered data on family, species, vernacular name, main collector, plant formation, growth-form, flowering / fruiting period, fruit size, fruit type and dispersal syndrome only for native flora.
Classification of growth forms followed the Instituto Brasileiro de Geografia e Estatística (IBGE, 2012): tree -woody plants higher than 3 m; shrub -woody plants with the main branch up to 50 cm above the ground level and generally shorter than 3 m; sub-shrub -plants with woody main stem and secondary herbaceous branches, generally having a height of less than 2 m; terrestrial herb -terrestrial plants with herbaceous aerial stem; epiphytic herb -plants with herbaceous stem which develop with another plant's support; climbing plants -lianas or vine plants with elongated stems, that usually are supported by a substrate or epiphytes that rely on their phorophyte for support, and also hemiparasite -plants that use the resources of their host, but have chlorophyll and perform photosynthesis.
The reproductive phenophases data were compiled of exsiccates of the EAC. Fruit size information and diaspore dispersal syndromes were obtained from herbarium sheets whenever possible, and from the literature when not. Diaspores were classified in small, medium and large following the length/width ratio criterion, as proposed by Tabarelli & Peres (2002). The dispersal syndromes were classified as anemochorous, autochorous or zoochorous, followed by Van der Pijl (1982).
The number of species per growth form, dispersal syndromes and diaspore size were analyzed through frequency diagrams. Flowering and fruiting patterns throughout the year were analyzed through circular histograms made with ORIANA 4 software (Kovach, 2011), based on the percentage of species which were flowering and fruiting (there was data on flowering for 80% of the species and on fruiting for only 30%). The following parameters were calculated for the phenology evaluation (with months corresponding to 30° angles): mean angle (μ), representing the period of the year in which a particular phenophase occurred in most taxa; circular standard deviation (sd); Rayleigh (Z) test, which determines the significance of the mean angle; and vector (r), which is a measure concentration around the mean angle. The null hypothesis on the (Z) test is that flowering and fruiting are distributed evenly throughout the year, and therefore there is no seasonality of each phenophase. The alternative hypothesis is that there is seasonality (if the mean angle is significant). The vector (r) varies from 0 (when the phenological activity is evenly distributed throughout the year) to 1 (when the phenological activity is seasonal/concentrated in a period of the year).

Results and discussion
Were found 418 specific/sub-specific taxa (84 families; 274 genera) could be found, from which 335 were found in the TSEF, 53 in TSDF, and 30 in both vegetations. Due to the absence of flowers or fruits, from this total number of registered data, only 22 species were identified up to genus and four were only identified up to family (Appendix A). This floristic richness was higher than any other ever recorded in montane forests of Brazilian semi-arid domain. According to the study of Nascimento et al. (2012), the researchers could find 293 species in Planalto da Borborema (Paraíba), nevertheless, Pereira et al. (2010) found 136 species in Serra Negra (Pernambuco), and Silva & Figueiredo (2013) only 100 species in Serra da Meruoca (Ceará). However, this disparity in the species richness of the Ubajara National Park may be due to the high number of terrestrial herbs recorded, in other words, 100 species that comprise 23.9% of the species; on the other hand, other studies did not indicate no more than 10% of herbs in the species richness. In addition, the other forests cited may have different areas and the respective surveys in each study may have different sampling effort, so the comparison may not be conclusive.
In addition to the species listed in the Appendix A, 28 exotic species (26 genera, 17 families) could be cataloged -Appendix B. The families with greatest numbers of exotic species were Araceaee (6 spp.), Apocynaceae and Amaryllidaceae (3 spp. each). These species are mostly ornamental and were planted on the park entrance and along the pathway to the cable car. Only five exotic species were found inside the park: Cryptostegia madagascariensis Bojer ex Decne., Coffea arabica L., Artocarpus heterophyllus Lam., Mangifera indica L. and Bambusa sp. In this case, the exotic species probably were present in the park before of its creation because it was a human occupied area. In the case of zoochorous plant species, exotic fruitful species may compete for the dispersers with native flora. Once the perpetuation of the native flora depends on both preservation and dispersal of diaspores (see Oliveira et al., 2017;Mattar et al., 2018), indigenous biodiversity may be somehow threatened.
Regarding the growth form, considering only indigenous species, it could be recorded 213 trees and shrubs, 100 terrestrial herbs, 68 climbing, 33 sub-shrubs, two epiphytic herbs, one hemiparasite and one aquatic herb ( Figure 3A). Forests usually have a low proportion of herb richness in comparison to trees and shrubs. One possible cause of the high proportion of herbs in the flora of the Ubajara National Park may be a consequence of the compilation of data from the herbarium collection; while rainfall forest trees may exhibit supra-annual flowering/fruiting patterns (see Pereira et al., 2008), herbs usually have annual patterns (increasing the chances of being incorporated to the herbarium -once EAC have not accepted vegetative samples in the last years). A complete floristic checklist for the Ubajara National Park will only be possible after a long-term phytosociological survey or reproductive phenodinamic study on the future.
On spectrum of dispersal syndromes ( Figure 3B), as expected, zoochory was predominant on the total flora (55%) and in the TSEF alone (201 species, 56%); while in the TSDF alone, the proportion of zoochory was lower (37 species, 46%). This predominance of zoochory in forests was found by Locatelli & Machado (2004) and by Diogo et al. (2016), showing that the Brazilian montane regions are important to preserve the richness of native species and available resource supply for local and migratory fauna. In both vegetations (TSEF and TSDF), the large and very large diaspores predominated, followed by medium and small ones ( Figure 3C). The predominant type of fruit are capsule, drupe, bacca, legume, achene, samara, follicle and schizocarp ( Figure 3D).
The Rayleigh test was only significant for fruit production ( Figure 3F), seeing that this production is concentrated on June (in the end of the rainy season) with standard deviation from April to July (μ = 150.334°, sd = 105.854°, Z = 4.215, p = 0.02, r = 0.181). This fruiting peak in the beginning of the dry season seems to be an adaptation of plants from seasonal and dry climates, whose seeds have the following wet season to germinate and establish (Locatelli & Machado;2004;Pau et al., 2011). On the other hand, the vectors r exhibited low results, indicating that the frequencies of the fruiting species are distributed evenly all year long. In addition, data on fruiting was available for only 30% of the species, so conclusions are not supported by these evidences. However, our results allowed generalizing that Ubajara National Park (as other montane forest from Northeast region of Brazil) offers resources for the fauna all year long.
Our results also showed that the Ubajara National Park has an important contribution to conservation of biodiversity, because the number of flowering plant species protected by this area is higher than those of other montane forests in Northeast region of Brazil (Carnaval et al., 2009;Homeier et al., 2010;Leite et al., 2016;Kamimura et al., 2017). The high diversity found in the Ubajara National Park may be attributed to its environmental heterogeneity, seeing that its topographic variation provides a range of different habitats, sheltering different vegetations and therefore a high floristic diversity.

Conclusions
The Ubajara National Park exhibits a huge species richness and an outstanding floristic composition -which makes this protected area very important for preservation of the local biodiversity. Beside the preservation of the plant species, the diversity of phenological patterns, types of fruits and dispersal syndromes create a substantial resource supply for local fauna almost all year long.
These results derived from a long-term sampling effort -an unparalleled study in the whole Brazilian semi-arid region. They represent not only an important contribution for the knowledge on the biodiversity of the semi-arid domain in northeast Brazil, but also a fundamental subside to the management of the park (including the formation of seed banks and seedlings production for restoration) and to further research on biodiversity, ecology, ethnobiology and effectiveness of protected areas.
Our results highlight the importance of studying the flora of the Brazilian protected areas. This is a fundamental step in assessing the effectiveness of protected areas in maintaining biodiversity.