Contents lists available at ScienceDirect Flora journal homepage: www.elsevier.com/locate/flora Do conspecific populations exhibit divergent phenological patterns? A study case of widespread savanna species Dinnie Michelle Assunção Lacerdaa,⁎, Judá Ben-Hur de Araújo Barrosb, Eduardo Bezerra de Almeida Jr.c, Davi Rodrigo Rossattod a Universidade Federal do Maranhão, Programa de Pós Graduação em Biodiversidade e Biotecnologia − Rede Bionorte, Campos Universitário do Bacanga, 1966, 65080- 805, São Luís, Maranhão, Brazil b Instituto Federal de Educação, Ciência e Tecnologia do Maranhão, Campus Barreirinhas, 65590-000, Barreirinhas, Maranhão, Brazil c Universidade Federal do Maranhão, Departamento de Biologia, Laboratório de Estudos Botânicos− LEB, Campos Universitário do Bacanga, 1966, 65080-805, São Luís, Maranhão, Brazil d Faculdade de Cências Agrárias e Veterinárias, Univ. Estadual Paulista (UNESP), Campus de Jaboticabal, 14884-900, Jaboticabal, SP, Brazil A R T I C L E I N F O Edited by Fei-Hai Yu Keywords: Cerrado Day length Phenology Seasonality Temperature A B S T R A C T Widespread savanna tree species can grow and survive at sites that diverge in water availability and seasonality, thus these species may be able to adjust their phenology in response to site variations. Here we evaluated vegetative and reproductive phenology in five woody species whose populations grow at two savannas sites under divergent climatic regimes, inserted in a large transitional zone between the Amazon forest and the semi- arid region. Patterns of leaf fall, leaf flush and flowering were recorded monthly for five woody species growing under longer (LDS) and shorter (SDS) dry seasons. We evaluated the seasonality, the start and peak dates for phenological events and the associations between phenophases and climatic data. We found a close relationship between phenological events and site temperatures, with phenological peaks in the LDS occurring, in general, about one to three months later than at the SDS site. Leaf fall coincides with warmer and drier periods when the day length is shorter. Leaf production and flowering were associated with increased day length in some popu- lations. Our results support the hypothesis that the conspecific populations have a high degree of association with climatic variables, especially temperature and day length, showing distinct phenological responses asso- ciated to the local climatic differences. 1. Introduction Phenological events are associated with a wide variety of ecological processes (Forrest and Miller-Rushing, 2010), and are therefore im- portant to our understanding of species-level relationships (Goulart et al., 2005; Zalamea et al., 2011; Dalmolin et al., 2015) and adaptive mechanisms (Grogan and Schulze, 2012; Worbes et al., 2013; Guan et al., 2014). Phenology can also explain the temporal organization of plant communities (Stevenson et al., 2008; Kushwaha et al., 2011; Diez et al., 2012; Hawes and Peres, 2016; Ryan et al., 2017), because phe- nological events are linked to climatic and biotic factors (Pau et al., 2011; Tang et al., 2016), affecting the capacity of plants to acquire resources for growth and reproduction (Nord and Lynch, 2009; Nord et al., 2011). Phenological responses of plant communities may diverge between sites, even if those sites are similar in species composition (Forrest et al., 2010). This suggests that climatic parameters are a stronger influence than species identity in determining phenological behavior (Pau et al., 2011). This may be especially evident for plant communities of the Brazilian savanna (Cerrado), where it is possible to find a great variety of phenological patterns following climatic variations along an ex- tensive latitudinal and altitudinal gradient (Bulhão and Figueiredo, 2002; Batalha and Martins, 2004; Pirani et al., 2009; Silva et al., 2011; Rossatto 2013). The wide climatic variations in the Cerrado are due to its occurrence in disjunctive areas in the Amazon; in a continuous strip that runs from Brazil’s Northeast to Center-South regions; and in the Southeast region of the country (Bridgewater et al., 2004). This con- tinental occurrence subjects Cerrado plant communities to distinct re- gimes of dry season duration and annual rainfall amount (Alvares et al., 2013). In tropical and sub-tropical regions, the amplitude in phenological events is often correlated with different climatic variables, including water availability (Dalmolin et al., 2015), photoperiod and light in- tensity (Yeang, 2007; Zimmerman et al., 2007) and temperature http://dx.doi.org/10.1016/j.flora.2017.10.001 Received 19 June 2017; Received in revised form 21 August 2017; Accepted 12 October 2017 ⁎ Corresponding author. E-mail address: michellelacerda@yahoo.com.br (D.M.A. Lacerda). Flora 236–237 (2017) 100–106 Available online 18 October 2017 0367-2530/ © 2017 Elsevier GmbH. All rights reserved. T http://www.sciencedirect.com/science/journal/03672530 https://www.elsevier.com/locate/flora http://dx.doi.org/10.1016/j.flora.2017.10.001 http://dx.doi.org/10.1016/j.flora.2017.10.001 mailto:michellelacerda@yahoo.com.br https://doi.org/10.1016/j.flora.2017.10.001 http://crossmark.crossref.org/dialog/?doi=10.1016/j.flora.2017.10.001&domain=pdf variations (Pau et al., 2013). For savannas in particular, water avail- ability − especially its decrease during the end of the wet season − is cited as one of the main factors driving leaf fall patterns (Bulhão and Figueiredo, 2002; Lenza and Klink, 2006; Rossatto et al., 2009; Guan et al., 2014). Flowering and leaf renewal are concentrated during the dry period; but especially during the transition from the late dry to the rainy season, a time of the year when day length and temperatures increase and first rains begin to fall (Figueiredo, 2008; Lenza and Klink, 2006; Pirani et al., 2009; Borges and Prado, 2014; Dalmolin et al., 2015; Ryan et al., 2017). The production of leaves and flowers during this transitional period has some advantages for savanna plants: the canopy reaches full development and is ready for maximal carbon gain just as the first rains start (Franco et al., 2005); an early flowering allows sufficient time for the complete fruit production and seed dispersal, which increases the success of seed germination and establishment (Batalha and Martins 2004). However, if the rainy period was delayed, and plants were not plastic enough to adapt their strategies, they would suffer from prolonged drought. Although a typical regime of alternation between dry and rainy periods prevails in the savannas, the duration of the dry season, dis- tribution of precipitation, average air temperatures and day length duration are highly variable (Rivera et al., 2002; Silva et al., 2008; Alvares et al., 2013). Temperature and day length are considered im- portant drivers for phenological events in savanna species (Rivera et al., 2002), especially for vegetative growth. Ascertaining how species re- spond to such differences enables the predictions of changes in phe- nological events and of population dynamics, and contributes im- portantly to the assessment of how climate change affects natural populations (Franco et al., 2014). The use of conspecific populations (represented by individuals of the same species, but growing at dif- ferent sites) is an interesting approach to comprehend local adaptations to climatic aspects (Panchen and Gorelick, 2016). In the present study, we evaluated vegetative and reproductive phenology in five species whose populations grow at two savannas sites (Cerrado), under distinct climatic regimes, in a large transitional zone between the Amazon forest and the semi-arid region. These sites exhibit differences in the duration of the dry season: one of them shows higher total rainfall, but a longer dry season, while the other have lower rainfall, but a shorter dry season. Additionally, short dry season site shows the highest temperature and longest day length variations earlier than the long dry season site. This study aimed to answer whether: 1) differences in the seasonal climatic patterns of these sites influence the period, start and peaks, of the phenological events, 2) phenological events are associated with temperature and day length independently of the dry season duration. We hypothesize that if phenological phe- nomena are intrinsically linked to climatic variables at these sites, the shorter dry season will promote phenological events to occur earlier in comparison to the longer dry season. These responses will occur be- cause leaf events and flower production are supposedly related to temperature and day length, which increase from mid to late dry season on each site and are indicative of the following rainy period. 2. Material and methods 2.1. Study sites and climate We conducted our study at two savanna sites (regionally known as Cerrado sensu stricto) in the northeastern region of Brazil: the first site is located at Barreirinhas municipality (03°01′17″S, 43°06′28″ W) and the second at the Mirador State Park (06°37′55″S, 45°52′38″W), both in Maranhão State, Brazil (Fig. 1). The climate is As in Barreirinhas, and Aw in Mirador State Park (Alvares et al., 2013), both of which are classified as tropical with two distinct seasons − dry and rainy− but differing in terms of total rainfall and in the dry season length, as Fig. 1. Sites selected for phenological observations: Barreirinhas (longer dry season − LDS) and Mirador State Park (shorter dry season − SDS), Maranhão State, Brazil. D.M.A. Lacerda et al. Flora 236–237 (2017) 100–106 101 observed during the study period (Fig. 2). For the sake of simplification and to clearly associate each study site with its respective climate, the Barreirinhas region will be referred to as the longer dry season (LDS) region, and Mirador State Park, as shorter dry season (SDS) region (Fig. 2). The annual historical rainfall average for a 30-year period in LDS (from 1986 to 2016) was 1,700 mm; average monthly air temperature is around 27 °C, with a mean maximum of 34 °C, mean monthly minimum of 23 °C, and a dry period between July and late December, when the first rains begin to fall. The thermal amplitude, considering mean monthly data, varies from 8 to 14 °C (INMET-Instituto Nacional de Meteorologia/Chapadinha Station (n° 82.382), 2017). The day length varies from 11 h and 57 min in June to 12 h and 18 min in December (Source: http://www.sci.fi/∼benefon/sol.html). Altitudes around 90 m are predominant and the soils are sandy yellow latosols, with a high content of coarse sand and clay, being generally poorer than SDS soil (Appendix A, Supplementary material). In SDS, the annual historical average rainfall was 1,200 mm; the monthly mean air temperature of 27 °C, monthly mean maximum of 33 °C and monthly mean minimum of 22 °C. Although SDS and LDS sites show similar ranges of mean temperature variation, mean monthly data show that the SDS region has a greater thermal amplitude − from 7 to 16 °C (INMET, Balsas Station (n° 82.768), 2017) −, and a dry period extending from May to September. Despite little variation, the day length also has greater amplitude earlier in SDS area, going from 11 h and 44 min in June, to 12 h and 31 min, in December (Source: http://www.sci.fi/∼benefon/Sol.html). The altitude ranges from 300 to 600 m, and the predominant soils in the region are litholic and yellow latosols, of the sandy loam type, with a high content of fine sand, and about 10% clay (Appendix A, Supplementary material). The monthly climatic data of total precipitation, mean maximum and mean minimum air temperatures (Source: http://www.inmet.gov.br/portal/) and day length (Source: http://www.sci.fi/∼benefon/sol.html), for the period from April 2014 to March 2016, were obtained for association with the phenological data. 2.2. Phenological study The vegetative (leaf fall and flush) and reproductive (flowering) phenology were recorded monthly in populations of five woody species that are common to the study sites and widely distributed in typical savanna areas of the Brazilian territory: Byrsonima crassifolia (L.) Rich. (Malpighiaceae), Caryocar coriaceum Wittm. (Caryocaraceae), Hirtela ciliata Mart. & Zucc. (Chrysobalanaceae), Qualea grandiflora Mart. (Vochysiaceae) and Plathymenia reticulata Benth. (Fabaceae). We marked and monitored all individuals of these species in 20 plots (20 × 20 m), covering 1.6 ha at each site. We sampled individuals with a minimum of 10 cm of trunk circumference at the ground level. Conspecific populations presented different densities in each area, im- plying a different number of individuals sampled per species. The number of individuals in LDS and SDS were, respectively: 69 and 11 for B. crassifolia; 13 and 26 for C. coriaceum; 19 and 27 for H. ciliata; 28 and 48 for Q. grandiflora and 140 and 11 for P. reticulata. Vouchers for these species are deposited in the Universidade Federal do Maranhão her- barium (Herbarium MAR). Over a period of two years (between April 2014 and March 2016), we made monthly visual estimates of the percentage of crown area in each individual that presented the following signs: losing brown leaves (for leaf fall); flushing new leaves (for leaf flush); flower buds and/or open flowers (for flowering). Estimates for leaf fall, leaf flush and flowering were made using a scale of five categories: 0, 1, 2, 3 and 4, at intervals of 25% (0: 0%; 1: 1–25%; 2: 26–50%; 3: 51–75% and 4: 76–100%), following the methodology described by Fournier (1974). We calculated monthly averages of the percentage for each studied population, using the sum of the intensity values obtained for all in- dividuals and dividing by the maximum possible value (number of in- dividuals multiplied by four). This value was converted into a percen- tage. 2.3. Statistical analysis We applied circular statistics to detect seasonal trends at the start (the mean date in which each population manifested the phenophase for the first time) and peak dates (the mean date in which each popu- lation showed the highest value for intensity) for a given phenological event. We also compared phenological patterns between the conspecific populations at both sites. For this, the months were converted to angles at 30° intervals. The circular parameters were estimated for each con- specific population (mean start and peak dates and length of mean vector (r)). To test the occurrence of seasonality, the Rayleigh Z test was applied. If the probability value is significant, the concentration in- tensity around the mean angle, represented by the vector r, is con- sidered as a measure of the degree of seasonality, ranging from zero to one. It was not possible to determine the mean date of leaf fall start, since the activity in this phenophase was continuous in all individuals of the evaluated species, though it differed in intensity of occurrence. Fig. 2. Climatic data (April 2014 to March 2016) for LDS − longer dry season site (dark gray lines) and SDS − shorter dry season site (light gray lines), Maranhão State, Brazil. Rainy season indicated by continuous lines below the x axis; dry season indicated by dashed lines below the x axis. D.M.A. Lacerda et al. Flora 236–237 (2017) 100–106 102 http://www.sci.fi/~benefon/sol.html http://www.sci.fi/~benefon/Sol.html http://www.inmet.gov.br/portal/ http://www.sci.fi/~benefon/sol.html We performed the Watson-Williams two-sample (F) test when the mean angle was significant, to determine whether conspecific populations exhibit a similar seasonal pattern across areas (Zar, 2010). The analyses related to circular statistics were performed in the software ORIANA 4.0 (Kovach, 2011). The species P. reticulata had only one flowering record in SDS and was therefore excluded from the statistical evaluations of this phenophase. In addition, we used multiple regression analysis to evaluate the association between phenological events (dependent variables) and monthly climatic variables, such as total precipitation, mean maximum and minimum air temperatures, and day length (in- dependent variables). For these analyses, we showed only the sig- nificant relationships and the respective beta values (the effect of the independent variable on the dependent variable) (Zar, 2010). 3. Results All species exhibited seasonal behavior on their peak dates of leaf fall as well on the start and peak dates of leaf flushing and flowering (Fig. 3 Table 1). In LDS, these phenomena were concentrated between July and December, whereas in SDS, they occurred from August to October, predominantly in the driest months. Independently of the species, the phenological phenomena tended to occur about one to three months earlier in SDS than in LDS (Fig. 3 Table 1). For B. crassifolia, the mean start and peak dates occurred from October to December at LDS, and from August to October in SDS. The events showed significantly different start and peak dates (Table 2), with a delay of few days to two months at LDS (Table 1). The intensity of events exhibited the same trend, with uneven distribution between sites and higher proportions occurring earlier in SDS than in LDS. The duration of leaf flushing and flowering was longer in LDS than in SDS (Fig. 3). For C. coriaceum, the mean dates of phenophases occurred from September to November in LDS, and from August to September in SDS (Table 1), with all phenophases showing significant differences be- tween sites (Table 2). In LDS, start and peak events for this species occurred about one to two months later than in SDS (Table 1), and the duration of leaf flushing and flowering was longer. In SDS, higher in- tensity percentages of all phenophases occurred earlier than in LDS (Fig. 3). For H. ciliata, the mean dates occurred from July to October in LDS, and from June to August in SDS. The events followed a pattern similar to those reported for the previous species, with a delay of one to two months in LDS (Table 1), except to leaf flushing start. As for the dis- tribution of the phenophases percentages, leaf fall had the most irre- gular pattern throughout the year. The highest proportions of leaf flushing and flowering intensities in SDS preceded those in LDS, where their duration was prolonged (Fig. 3). Between Q. grandiflora populations, the mean dates of phenophases Fig. 3. Fournier intensity (percentage) for leaf fall, leaf flushing and flowering in the species B. crassifolia, C. coriaceum, H. ciliata, Q. grandiflora and P. reticulata, for LDS − longer dry season site (dark gray lines) and SDS − shorter dry season site (light gray lines), Maranhão State, Brazil. Rainy season indicated by continuous lines below the x axis; dry season indicated by dotted lines below the x axis. D.M.A. Lacerda et al. Flora 236–237 (2017) 100–106 103 were concentrated between November and February in LDS, and be- tween September and December in SDS. The delay for this species was about two to almost three months (Table 1), with different mean start and peak dates (Table 2). The highest proportions of all SDS pheno- phases preceded those of LDS (Fig. 3). The leaf flushing and flowering pattern in LDS, referring to the phenophases duration, was not verified for this species (Fig. 3). Phenological events in P. reticulata differed between sites (Table 2), with start and peak dates distributed from November to January in LDS and from August to October in SDS, with a delay of three months for all phenophases in LDS (except to Flow-S and Flow-P) (Table 1). Con- sidering the intensity of events, we observed the same pattern of earlier occurrence in SDS and a longer leaf flushing duration in LDS (Fig. 3). We found that maximum (positive trend) or minimum air tem- peratures (mostly negative trend) and day length were significantly associated with the intensity of the phenophases (Table 3). For most species, leaf fall, increased as maximum temperature increase at both sites, but it was also associated with a decrease of minimum tempera- tures in SDS. Day length also had a negative association with leaf fall in C. coriaceum and H. ciliata, for both sites (Table 3). For all species in SDS, except to H. ciliata, leaf flushing was related to the temperature (positive trend). In LDS, B. crassifolia, Q. grandiflora and P. reticulata showed association between leaf flushing and day length (positive trend), while for C. coriaceum and H. ciliata, maximum temperature (positive trend) and precipitation (negative trend) were, respectively, more associated with leaf flushing (Table 3). Flowering was not asso- ciated with climatic variables in C. coriaceum and P. reticulata (Table 3). For B. crassifolia, flowering was related to increases in maximum tem- perature at both sites. For H. ciliata, this event was associated with the decrease of the minimum temperature in SDS and an increase of the maximum temperature in LDS. Q. grandiflora exhibited significant re- lationship between flowering and day length in SDS (Table 3). 4. Discussion Our results support the hypothesis that phenological phenomena are intrinsically linked to climatic variables, especially temperature and day length, independent of the site. We also confirm that a shorter dry season, in comparison to a longer dry season, promotes earlier occur- rence of phenological patterns, because temperature and day length reaches a greater variation earlier at SDS site than at the LDS. The conspecific populations of the five evaluated species showed distinct behavior for all phenological events between the two sites, with dif- ferences in the start and peak dates. Peak dates generally occurred about one to three months later in LDS. This difference may represent levels of local adaptation and a high degree of phenotypic plasticity related to adjustments for the selective abiotic pressures driving leaf and flower production (Pellissier et al., 2014; Panchen and Gorelick, 2016; Silva Moraes et al., 2017). The start and peak dates and the highest proportions of leaf fall intensity for the conspecific populations occurred at different times during drought in each area, accompanying the variations in tem- perature and/or decrease in day length at both sites. The association between temperature and leaf fall is presumably related to the increase in evaporative demand and water stress, with consequent loss of leaves, which was already reported for some savanna areas in Brazil (Bulhão and Figueiredo, 2002; Silvério and Lenza, 2010). As for the relationship with day length, this variable decreases during the early dry period, and may also function as a factor associated with leaf fall (Garcia et al., 2017), signaling climatic changes. Day lengh is the only variable cor- related with leaf fall in H. ciliata at both sites, and was included among the variables related to leaf fall in C.coriaceum. The species B. crassifolia Table 1 Circular statistics for seasonality occurrence and mean dates for leaf fall peak and the start and peak for leaf flushing and flowering in savanna woody species subjected to longer dry season (LDS) and shoter dry season (SDS), Maranhão State, Brazil. LMV: length of the mean vector; LF = leaf fall; LFl = leaf flushing; Flw = flowering; P = peak; S = start; na = not available. Significance levels: * P < 0.001. Sites Statistics LDS SDS Species LF-P LFl-S LFL-P Flw-S Flw-P LF-P LFl-S LFL-P Flw-S Flw-P B. crassifolia Mean date Oct (295.8°) Oct (270.1°) Dec (330.2°) Oct (284°) Nov (308.7°) Aug (233.6°) Sep (263.9°) Oct (279.5°) Oct (269.9°) Oct (282.8°) LMV (r) 0.92 0.72 0.84 0.89 0.89 0.82 0.83 0.82 0.9 0.93 Rayleigh Test (Z) 17.74* 11.37* 15.54* 16.81* 16.62* 72.04* 91.13* 87.68* 66.78* 71.85* C. coriaceum Mean date Sep (268.5°) Sep (263.6°) Nov (304.1°) Oct (279°) Oct (287.5°) Aug (210.9°) Aug (220.4°) Sep (244.5°) Aug (225.2°) Aug (220.8°) LMV (r) 0.86 0.75 0.65 0.87 0.92 0.85 0.71 0.75 0.96 0.99 Rayleigh Test (Z) 17.86* 14.74* 9.59* 8.89* 10.07* 31.6* 24.08* 26.8* 6.49* 6.92* H. ciliata Mean date Jul (199.6°) Jul (197.6°) Sep (252.6°) Sep (249.7°) Out (274.7°) Jun (176.6°) Aug (212.4°) Aug (222.4°) Aug (223.3°) Aug (236.9°) LMV (r) 0.45 0.67 0.8 0.96 0.91 0.94 0.87 0.92 0.94 0.96 Rayleigh Test (Z) 6.02* 17.86* 23.02* 29.51* 25.77* 7.95* 32.72* 35.66* 30* 32.36* Q. grandiflora Mean date Nov (317.9°) Dec (348.9°) Jan (8.1°) Feb (53.7°) Feb (55.12°) Sep (261.5°) Oct (269.9°) Nov (312.3°) Dec (345.8°) Dec (339.7°) LMV (r) 0.89 0.81 0.86 0.91 0.9 0.88 0.72 0.79 0.93 0.52 Rayleigh Test (Z) 34.6* 37.01* 38.93* 19.03* 18.86* 72.8* 50.26* 54.83* 12.05* 3.84* P. reticulata Mean date Nov (304.8°) Dec (350.6°) Jan (0.9°) Dec (356.9°) Dec (358.2°) Ago (26.9°) Set (257.4°) Oct (276.7°) na na LMV (r) 0.95 0.86 0.85 0.93 0.98 0.94 0.84 0.91 Rayleigh Test (Z) 64.09* 52.98* 51.35* 62.51* 69.50* 23.43* 17.87* 21.72* Table 2 Watson Williams test results for comparisons among the mean start and peak dates for leaf fall, leaf flushing and flowering in savanna species subjected to longer (LDS) and shorter dry season (SDS), Maranhão State, Brazil. LF = leaf fall; LFl = leaf flushing; Flw = flowering; S = start; P = peak; na = not available. F = circular statistical value; p = probability value. Species Statistics LF-P LFl-S LFl-P Flw-S Flw-P B. crassifolia F 61.219 0.45 36.856 3.528 20.243 p <0.001 0.503 < 0.001 0.063 < 0.001 C. coriaceum F 47.601 14.405 22.015 16.67 16.096 p <0.001 <0.001 <0.001 <0.001 <0.001 H. ciliata F 1.155 2.584 18.046 33.029 54.258 p 0.289 0.112 < 0.001 <0.001 <0.001 Q. grandiflora F 115.156 116.974 78.434 65.327 20.071 p <0.001 <0.001 <0.001 <0.001 <0.001 P. reticulata F 76.517 63.753 67.109 na na p < 0.001 <0.001 <0.001 D.M.A. Lacerda et al. Flora 236–237 (2017) 100–106 104 shows different degrees of leaf deciduousness, with greater crown cover and smaller variations in the proportions of leaf fall in LDS. The H. ciliata species also showed lower seasonality for this event (shorter length of vector r) in this area. We can postulate that these differences in the degree of intraspecific deciduousness among Cerrado species are indicative of high levels of plasticity against climatic aspects (Kuster et al., 2017), which allows them to cope with very distinct climatic regions of the Brazilian territory. Leaf flushing showed generally consistent differences in the start and peak dates with distinct distribution of the intensity percentage throughout the year for both sites. This confirms the occurrence of the pre-rain green-up phenomenon reported for some savanna species in Central Brazil (Rivera et al., 2002; Franco et al., 2005; Silvério and Lenza, 2010; Dalmolin et al., 2015). This phenomenon is characterized by the appearance of new leaves before the rainy season. This strategy is reported to avoid plant nutrient losses due to herbivory or leaching (Van Schaik et al., 1993); to maximize carbon gain by increasing CO2 assimilation (Rossatto et al, 2009; Dalmolin et al., 2015) and to make the best use of available nutrients in the soil as rainfall increases (Scholes and Walker, 2004; Nord and Lynch, 2009). During the pre-rain green-up, plants take advantage of the higher temperatures to quickly produce and expand their leaves. The influence of temperature on leaf flushing has been reported in tropical seasonally dry regions, such as savannas in Brazil (Pirani et al., 2009; Silvério and Lenza, 2010) and Africa (Seghieri et al., 2012) and may be a key variable for this process (Chambers et al., 2013). The delayed leaf production at LDS site (compared to plants in SDS) may be an adjustment to avoid over- exposure of leaves produced during the dry season, which can decrease productivity in savanna plants. The influence of day length has been reported as a trigger factor for leaf flush and flowering events in tropical regions, including savanna areas (Rivera et al., 2002; Yeang, 2007; Renner, 2007; Zimmerman et al., 2007; Pau et al., 2013; Borchert et al., 2015; Ryan et al., 2017). Although the results of the multiple regression did not show a clear relationship between the phenophases and day length, the leaf flush for all the species in LDS, except H. ciliata, increased considerably during the end of the dry period, when day length increases. In SDS, the in- crease in leaf flushing during the middle and end of the dry period occurs when there is a decrease in clouds and a higher incidence of solar radiation. Thus, as photosynthetic rates are higher in young leaves, the production of new leaves during the period of greater day length and/or irradiance favors photosynthesis and carbon gain (Wright and Van Schaik, 1994), since savanna plants has specific mechanisms to use and store water during the dry season (Goldstein et al., 2008). Therefore, the occurrence of production and growth phenophases during the period of low or no precipitation, indicates the existence of mechanisms that allow plants to rehydrate or maintain internal water levels (Worbes et al., 2013) and are favored by the transfer of assimi- lated products, which link directly to the growth organs and promote the production of leaves (Ryan et al., 2017) flowers and fruits (Pau et al., 2013; Zimmerman et al., 2007; Camargo et al., 2011; Borchert et al., 2015). Flowering, which followed the trend of earlier occurrence in SDS, was less related to the climatic variables of each site. However, in B. crassifolia and H. ciliata, flowering was associated with variations in temperature during the dry period, and in Q. grandiflora, it was asso- ciated with the increased day length in SDS during the early rainy season. Although the populations of other species did not show sig- nificant results in relation to climatic variables, they showed a degree of seasonality compatible with different flowering strategies. Flowering occurred in C. coriaceum at both sites during the dry period, under higher irradiance and temperatures, and in Q. grandiflora in LDS, during early rainy season, with increased day length. P. reticulata in LDS had its flowering peak in the final days of the dry period. Thus, for both the evergreen species and the deciduous species, the water restriction in the environment does not limit flowering in general (Lenza and Klink, 2006; Tannus et al., 2006; Figueiredo, 2008; Pirani et al., 2009; Camargo et al., 2011; Silva et al., 2011; Barbosa et al., 2012). Re- garding temperature, it is widely recognized as a variable that affects plant distribution and metabolism, and is an important requirement for flowering (Grace, 1987). Its influence on leaf production and flowering in savannas appears to be more important than the influence of water availability in dry and wet season (Seghieri et al., 2012). In summary, we characterized the phenological responses in con- specific populations under distinct climates, showing that temperature and day length are the major drivers for phenological events, in- dependent of the site. The duration of the dry season does not limit growth phenomena, but shorter dry seasons allows plants to produce leaves and flowers earlier than under longer dry seasons, because temperature and day length start to increase earlier in SDS site. This strategy indicates higher phenotypic plasticity in savanna woody plants, aligned with adjustments to maximize carbon gain and re- productive success. Acknowledgments We thank Prof. Patricia Maia Albuquerque, Maria Antonia de Melo and Éville Ribeiro for support in the execution of the project; Zoe Panchen for initial reading and suggestions; Davi B. Muniz for the map illustration; and the members of the Laboratorio de Estudos Botânicos, Table 3 Multiple regression analyses results assessing the effect of climatic variables on phenology in savanna species subjected to longer dry season (LDS) and shorter dry season (SDS), Maranhão State, Brazil. NA − not available, NS − not significant. R2 = coefficient of determination; p = probability value and β= standardized regression coefficients. Variables: MinT = mean minimum air temperature; MaxT = mean maximum air temperature; DL = day length; Ppt = precipitation. Species LDS SDS Leaf fall Leaf flushing Flowering Leaf fall Leaf flushing Flowering B. crassifolia R2 = 0.70 p ≤ 0.01 R2 = 0.71 p ≤ 0.01 R2 = 0.74 p ≤ 0.01 R2 = 0.77 p ≤ 0.01 R2 = 0.63 p ≤ 0.01 R2 = 0.62 p ≤ 0.01 MaxT β= 4.3 DL β = 97.27 MaxT β = 7.3 MaxT β = 2.82 MaxT β = 2.65 MaxT β = 2.69 MinT β = 7.3 MinT β= −4.35 C. coriaceum R2 = 0.57 p ≤ 0.01 R2 = 0.75 p ≤ 0.01 NS R2 = 0.77 p ≤ 0.01 R2 = 0.72 p ≤ 0.01 NS MaxT β= 6.0 MaxT β= 4.4 MinT β= −3.23 MaxT β = 4⋅7 DL β =−13.2 DL β= −26.61 MinT β = −4⋅6 H. ciliata R2 = 0.44 p ≤ 0.05 R2 = 0.55 p ≤ 0.01 R2 = 0.61 p ≤ 0.01 R2 = 0.62 p ≤ 0.01 NS R2 = 0.47 p ≤ 0.05 DL β =−81.56 Ppt β = −0.01 MaxT β = 4.12 DL β= −31.64 MinT β =−5.02 Q. grandiflora R2 = 0.67 p ≤ 0.01 R2 = 0.71 p ≤ 0.01 NS R2 = 0.67 p ≤ 0.01 R2 = 0.73 p ≤ 0.01 R2 = 0.65 p ≤ 0.01 MaxT β= 5.57 DL β = 178.3 – MaxT β = 11.52 MaxT β = 4.44 DL β = 0.45 P. reticulata R2 = 0.92 p ≤ 0.01 R2 = 0.91 p ≤ 0.01 NS R2 = 0.92 p ≤ 0.01 R2 = 0.59 p ≤ 0.01 NA MaxT β= 11.32 DL β = 0.55 MaxT β = 7.66 MaxT β = 4.64 MinT β= −10.9 NS = not significant. D.M.A. Lacerda et al. Flora 236–237 (2017) 100–106 105 especially Ariade Nazareth da Silva, Bruna Correia, and Gustavo Lima for their participation in the collection and identification of species. This work was supported by Fundação de Amparo à Pesquisa e ao Desenvolvimento Científico e Tecnológico do Maranhão – FAPEMA (Projeto APP-UNIVERSAL – 00744/13). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.flora.2017.10.001. References Alvares, C.A., Stape, J.L., Sentelhas, P.C., de Moraes, G., Leonardo, J., Sparovek, G., 2013. Köppen's climate classification map for Brazil. Meteorol. Zeitschrift 22, 711–728. Barbosa, R.I., Mourão Jr., M., Casadio, G.M.L., da Silva, S.J.R., 2012. Reproductive phenology of the main tree species in the Roraima savanna, Brazilian Amazon. Ecotropica 18, 81–91. Batalha, M.A., Martins, F.R., 2004. Reproductive phenology of the cerrado plant com- munity in Emas National Park (central Brazil). Aust. J. Bot. 52, 149–161. Borchert, R., Calle, Z., Strahler, A.H., Baertschi, A., Magill, R.E., Broadhead, J.S., Kamau, J., Njoroge, J., Muthuri, C., 2015. Insolation and photoperiodic control of tree de- velopment near the equator. New Phytol. 205, 7–13. Borges, M.P., Prado, C.H.B., 2014. Relationships between leaf deciduousness and flow- ering traits of woody species in the Brazilian neotropical savanna. Flora 209, 73–80. Bridgewater, S., Ratter, J.A., Ribeiro, J.F., 2004. Biogeographic patterns, β-diversity and dominance in the cerrado biome of Brazil. Biodivers. Conserv. 13, 2295–2317. Bulhão, C.F., Figueiredo, P.S., 2002. Phenology of leguminous trees in an area of cerrado in the northeast of Maranhão. Braz. J. Bot. 25, 361–369. Camargo, M.G.G., Souza, R.M., Reys, P., Morellato, L.P., 2011. Effects of environmental conditions associated to the cardinal orientation on the reproductive phenology of the cerrado savanna tree Xylopia aromatica (Annonaceae). An. Acad. Bras. Cienc. 83, 1007–1020. Chambers, L.E., Altwegg, R., Barbraud, C., Barnard, P., Beaumont, L.J., Crawford, R.J., Durant, J.M., Hughes, L., Keatley, M.R., Low, M., Morellato, P.C., Poloczanska, E.S., Ruoppolo, V., Vanstreels, R.E.T., Woehler, E.J., Wolfaardf, A.C., 2013. Phenological changes in the southern hemisphere. PLoS One 8. Dalmolin, Â.C., de Almeida Lobo, F., Vourlitis, G., Silva, P.R., Dalmagro, H.J., Antunes, M.Z., Ortíz, C.E.R., 2015. Is the dry season an important driver of phenology and growth for two Brazilian savanna tree species with contrasting leaf habits? Plant Ecol. 216, 407–417. Diez, J.M., Ibánzes, I., Miller-Rushing, A.J., Mazer, S.J., Crimmins, T.M., Crimmins, M.A., Bertelsen, C.D., Inouye, D.W., 2012. Forecasting phenology: from species variability to community patterns. Ecol. Lett. 15, 545–553. Figueiredo, P.S., 2008. Fenologia e estratégias reprodutivas das espécies arbóreas em uma área marginal de cerrado, na transição para o semi-árido no nordeste do Maranhão, Brasil. Rev. Trop. Cienc. Agrar. Biol. 2, 8–21. Forrest, J., Miller-Rushing, A.J., 2010. Toward a synthetic understanding of the role of phenology in ecology and evolution. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 365, 3101–3112. Forrest, J., Inouye, D.W., Thomson, J.D., 2010. Flowering phenology in subalpine mea- dows: does climate variation influence community co-flowering patterns? Ecology 91, 431–440. Fournier, L.A., 1974. A quantitative method for the measurement of phenological char- acteristics in trees. Turrialba 24, 422–423. Franco, A.C., Bustamante, M.M.C., Caldas, L.S., Goldstein, G., Meinzer, F.C., Kozovits, A.R., Rundel, P., Coradin, V.T.R., 2005. Leaf functional traits of Neotropical savanna trees in relation to seasonal water deficit. Trees 19, 326–335. Franco, A.C., Rossatto, D.R., Silva, L.D.C.R., da Silva Ferreira, C., 2014. Cerrado vege- tation and global change: the role of functional types: resource availability and dis- turbance in regulating plant community responses to rising CO2 levels and climate warming. Theor. Exp. Plant Physiol. 261, 19–38. Garcia, L.C., Barros, F.V., Lemos-Filho, J.P., 2017. Environmental drivers on leaf phe- nology of ironstone outcrops species under seasonal climate. An. Acad. Bras. Cienc. 89, 131–143. Goldstein, G., Meinzer, F.C., Bucci, S.J., Scholz, F.G., Franco, A.C., Hoffmann, W.A., 2008. Water economy of Neotropical savanna trees: six paradigms revisited. Tree Physiol. 28, 395–404. Goulart, M.F., Lemos Filho, J.P., Lovato, M.B., 2005. Phenological variation within and among populations of Plathymenia reticulata in Brazilian Cerrado, the Atlantic Forest and transitional sites. Ann. Bot. 96, 445–455. Grace, J., 1987. Climatic tolerance and the distribution of plants. New Phytol. 106, 113–130. Grogan, J., Schulze, M., 2012. The impact of annual and seasonal rainfall patterns on growth and phenology of emergent tree species in southeastern Amazonia, Brazil. Biotropica 44, 331–340. Guan, K., Wood, E.F., Medvigy, D., Kimball, J., Pan, M., Caylor, K.K., Shefield, J., Xu, X., Jones, M.O., 2014. Terrestrial hydrological controls on land surface phenology of African savannas and woodlands. J. Geophys. Res. Biogeosci. 119, 1652–1669. Hawes, J.E., Peres, C.A., 2016. Patterns of plant phenology in Amazonian seasonally flooded and unflooded forests. Biotropica 48, 465–475. Kovach, W.L., 2011. Oriana–circular Statistics for Windows, Ver. 4. Kovach Computing Services, Pentraeth. Kushwaha, C.P., Tripathi, S.K., Tripathi, B.D., Singh, K.P., 2011. Patterns of tree pheno- logical diversity in dry tropics. Acta Ecol. Sinica 31, 179–185. Kuster, V.C., de Castro, S.A.B., Vale, F.H.A., 2017. Environmental conditions modulate plasticity in the physiological responses of three plant species of the Neotropical savannah. Acta Physiol. Plant 39, 103. Lenza, E., Klink, C.A., 2006. Comportamento fenológico de espécies lenhosas em um cerrado sentido restrito de Brasília, DF. Rev. Bras. Bot. 29, 627–638. Nord, E.A., Lynch, J.P., 2009. Plant phenology: a critical controller of soil resource ac- quisition. J. Exp. Bot. 60, 1927–1937. Nord, E.A., Shea, K., Lynch, J.P., 2011. Optimizing reproductive phenology in a two- resource world: a dynamic allocation model of plant growth predicts later re- production in phosphorus-limited plants. Ann. Bot. 108, 391–404. Panchen, Z.A., Gorelick, R., 2016. Canadian arctic archipelago conspecifics flower earlier in the high arctic than the mid-Arctic. Int. J. Plant Sci. 177, 661–670. Pau, S., Wolkovich, E.M., Cook, B.I., Davies, T.J., Kraft, N.J., Bolmgren, K., Betancourt, J.L., Cleland, E.E., 2011. Predicting phenology by integrating ecology, evolution and climate science. Global Chang. Biol. 17, 3633–3643. Pau, S., Wolkovich, E.M., Cook, B.I., Nytch, C.J., Regetz, J., Zimmerman, J.K., Wright, S.J., 2013. Clouds and temperature drive dynamic changes in tropical flower pro- duction. Nat. Clim. Chang. 3, 838–842. Pellissier, L., Roger, A., Bilat, J., Rasmann, S., 2014. High elevation Plantago lanceolata plants are less resistant to herbivory than their low elevation conspecifics: is it just temperature? Ecography 37, 950–959. Pirani, F.R., Sanchez, M., Pedroni, F., 2009. Fenologia de uma comunidade arbórea em cerrado sentido restrito, Barra do Garças, MT, Brasil. Acta Bot. Bras. 23, 1096–1109. Renner, S.S., 2007. Synchronous flowering linked to changes in solar radiation intensity. New Phytol. 175, 195–197. Rivera, G., Elliott, S., Caldas, L.S., Nicolossi, G., Coradin, V.T., Borchert, R., 2002. Increasing day-length induces spring flushing of tropical dry forest trees in the ab- sence of rain. Trees Struct. Funct. 16, 445–456. Rossatto, D.R., Hoffmann, W.A., Franco, A.C., 2009. Differences in growth Patterns be- tween co-occuring forest and savanna tress affect the forest-savanna boundary. Funct. Ecol. 23, 689–698. Rossatto, D.R., 2013. Seasonal patterns of leaf production in co-occuring trees with contrasting leaf phenology: time and quantitative divergences. Plant Species Biol. 28, 138–145. Ryan, C.M., Williams, M., Frace, J., Woollen, W., Lehmann, C.E., 2017. Pre-rain green up is ubiquitos acrros tropical Africa: implications for temporal niche separation and model representation. New Phytol. 213, 625–633. Scholes, R.J., Walker, B.H., 2004. An African Savanna: Synthesis of the Nylsvley Study. Cambridge University Press, Cambridge, UK. Seghieri, J., Carreau, J., Boulain, N., de Rosnay, P., Arjounin, M., Timouk, F., 2012. Is water availability really the main environmental factor controlling the phenology of woody vegetation in the central Sahel? Plant Ecol. 213, 861–870. Silvério, D.V., Lenza, E., 2010. Phenology of woody species in a typical cerrado in the Bacaba Municipal Park, Nova Xavantina, Mato Grosso, Brazil. Biota Neotropica 10, 205–216. Silva Moraes, A.C., da Vitória, A.P., Rossatto, D.R., de Miranda, L.D.A.P., Funch, L.S., 2017. Leaf phenology and morphofunctional variation in Myrcia amazonica DC. (Myrtaceae) in gallery forest and campo rupestre vegetation in the Chapada Diamantina, Brazil. Braz. J. Bot. 1–12. Silva, F.A.M., Assad, E.D., Evangelista, B.A., 2008. Caracterização Climática do Bioma Cerrado. In: Sano, S.M., Almeida, S.P., Oliveira, P.E. (Eds.), Cerrado: ecologia e flora. EMBRAPA-CPAC, Planaltina, Brasil, pp. 71–88. Silva, I.A., da Silva, D.M., de Carvalho, G.H., Batalha, M.A., 2011. Reproductive phe- nology of Brazilian savannas and riparian forests: environmental and phylogenetic issues. Ann. For. Sci. 68, 1207–1215. Stevenson, P.R., Castellanos, M.C., Cortés, A.I., Link, A., 2008. Flowering patterns in a seasonal tropical lowland forest in western Amazonia. Biotropica 40, 559–567. Tang, J., Körner, C., Muraoka, H., Piao, S., Shen, M., Thackeray, S.J., Yang, X., 2016. Emerging opportunities and challenges in phenology: a review. Ecosphere 7, 1–17. Tannus, J.L.S., Assis, M.A., Morellato, P.C., 2006. Reproductive Phenology in Dry and Wet Grassland in an Area of Cerrado at Southeastern Brazil, Itirapina. –SP. Biota Neotropica. http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1676- 06032006000300008. Van Schaik, C.P., Terborgh, J.W., Wright, S.J., 1993. The phenology of tropical forests: adaptive significance and consequences for primary consumers. Annu. Rev. Ecol. Syst. 24, 353–377. Worbes, M., Blanchart, S., Fichtler, E., 2013. Relations between water balance, wood traits and phenological behavior of tree species from a tropical dry forest in Costa Rica-a multifactorial study. Tree Physiol. 33, 527–536. Wright, S.J., Van Schaik, C.P., 1994. Light and the phenology of tropical trees. Am. Nat. 143, 192–199. Yeang, H.Y., 2007. Synchronous flowering of the rubber tree (Hevea brasiliensis) induced by high solar radiation intensity. New Phytol. 175, 283–289. Zalamea, P.C., Munoz, F., Stevenson, P.R., Paine, C.T., Sarmiento, C., Sabatier, D., Heuret, P., 2011. Continental-scale patterns of Cecropia reproductive phenology: evidence from herbarium specimens. Proc. R. Soc. Lond. B. Biol. Sci. 278, 2437–2445. Zar, J.H., 2010. Biostatistical Analysis. Prentice-Hall, New Jersey. Zimmerman, J.K., Wright, S.J., Calderón, O., Pagan, M.A., Paton, S., 2007. Flowering and fruiting phenologies of seasonal and aseasonal neotropical forests: the role of annual changes in irradiance. J. Trop. Ecol. 23, 231–251. D.M.A. Lacerda et al. Flora 236–237 (2017) 100–106 106 http://dx.doi.org/10.1016/j.flora.2017.10.001 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0005 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0005 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0010 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0010 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0010 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0015 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0015 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0020 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0020 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0020 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0025 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0025 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0030 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0030 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0035 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0035 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0040 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0040 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0040 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0040 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0045 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0045 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0045 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0045 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0050 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0050 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0050 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0050 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0055 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0055 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0055 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0060 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0060 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0060 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0065 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0065 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0065 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0070 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0070 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0070 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0075 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0075 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0080 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0080 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0080 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0085 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0085 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0085 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0085 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0090 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0090 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0090 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0095 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0095 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0095 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0100 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0100 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0100 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0105 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0105 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0110 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0110 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0110 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0115 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0115 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0115 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0120 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0120 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0125 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0125 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0130 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0130 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0135 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0135 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0135 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0140 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0140 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0145 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0145 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0150 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0150 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0150 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0155 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0155 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0160 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0160 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0160 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0165 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0165 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0165 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0170 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0170 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0170 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0175 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0175 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0180 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0180 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0185 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0185 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0185 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0190 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0190 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0190 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0195 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0195 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0195 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0200 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0200 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0200 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0205 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0205 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0210 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0210 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0210 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0215 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0215 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0215 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0220 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0220 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0220 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0220 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0225 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0225 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0225 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0230 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0230 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0230 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0235 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0235 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0240 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0240 http://www.scielo.br/scielo.php?script=sci_arttext%26pid=S1676-06032006000300008 http://www.scielo.br/scielo.php?script=sci_arttext%26pid=S1676-06032006000300008 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0250 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0250 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0250 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0255 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0255 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0255 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0260 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0260 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0265 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0265 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0270 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0270 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0270 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0275 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0280 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0280 http://refhub.elsevier.com/S0367-2530(17)33345-5/sbref0280 Do conspecific populations exhibit divergent phenological patterns? A study case of widespread savanna species Introduction Material and methods Study sites and climate Phenological study Statistical analysis Results Discussion Acknowledgments Supplementary data References