Study area

This study was conducted in an area of Caatinga [9, 38, 42, 44] with deciduous and sub-deciduous tree species near a rural community called “Carão” in Altinho city, Pernambuco state, Brazil. The city is located in the Borborema highland, with altitudes ranging from 650 to 1000 meters and generally medium to high soil fertility. The climate is classified as semi-arid hot; the rainy season takes place from February to August. In this context, the influence of different landscape management regimes on the phenology of S. tuberosa was evaluated in five landscape units (LUs) in the region, namely: Mountain (area with vegetation in regeneration for nearly 50 years – 628m of altitude above sea level), Mountain Base (area of vegetation in regeneration for nearly 15 years – 498 m of altitude above sea level), Pastures (areas of native grassland - 486m of altitude above sea level), Cultivated Areas (cultivated areas of maize and beans- 469 m of altitude above sea level), and Homegardens (backyards - 463 m of altitude above sea level). The choice of these landscape units was based on the study conducted by Lins Neto et al. [45] in which it was determined the history of land use, as well as the landscape units, recognized by the main informants of the place, as well as the study of Almeida et al. [9], which evaluated the fruit and flower production, pollinator guild and frequency of floral visitors of Spondias tuberosa individuals submitted to different soil management regimes. It is noteworthy that both works mentioned above were conducted in the same area of the present study. For more information on historical land-use, the reader is referred to the study of Lins Neto et al. [45]. Yet in the latter, a complete analysis of soil fertility has indicated that areas of native pasture and gardens presented high fertility soils [45].

Social and cultural aspects of Carão are available in Araújo et al. [44], Almeida et al. [9] and Lins Neto et al. [38]. However, some information relevant to the present study should be mentioned.

There are currently 189 people living in 61 houses in Carão. The main economic activity is agro-pastoralism with mainly subsistence agriculture (especially corn and bean monocultures and cattle and goat farming). Near Carão there is a hill, locally known as the Serra (mountain) covered almost entirely with native vegetation that has been in a process of regeneration for nearly 50 years [28]. The top of the hill is flat and there are fields of corn, beans and cassava, which are important for the region’s food supply. At the base of the hill is a transitional area between the hill and flatter areas, locally known as the Baixio (lowlands), where some houses are established, most of which are organized into a village. There are also shops, Catholic and Protestant churches and an elementary school. In the flat areas, there are areas of native grasslands and monocultures close to the houses. Other areas of constant activity are the backyards of houses, where residents usually maintain animal and plant (wild and cultivated) cultures.

Phenology of Spondias tuberosaArruda

For evaluated the pattern phenological of S. tuberosa evaluated the activity, intensity and synchronization of reproductive and vegetative phenophases, as well as its relationship with rainfall, in each LU, ten individuals of this specie were randomly selected from each population (totaling 50 individuals) to be included in monthly monitoring between February 2007 and January 2009. The phenophases considered were flowering (flower buds plus flowers at anthesis), fruiting (immature plus mature fruits), leaf flush and leaf fall. The activity was determined by recording the presence or absence of each phenophase. Estimated the percentage of monthly variation of phenological changes in each individual, and was employed the method proposed by Fournier [46] to evaluate the intensity of each phenophase. This method consists of apply a semi-quantitative interval scale with five categories (0–4) and intervals of 25% between each category for each phenophase considered, and calculate the intensity index. The Fournier index per month for each LUs for each phenophase, ranges from 0 to 100%.

To evaluate variation in the synchronization of phenological phases among individuals of the same LU, and among LUs, a flowering synchrony index was applied [47, 48]. The synchronization of each phenophase (Xi) was measured separately for ten individuals of each landscape unit by summing the number of months where there was an overlap of the phenophase between a focal individual and the rest of the sample. This synchronization was calculated using the following formula: [Xi = Σij/(N-1) fi], where Σij is the sum of the number of months in which individuals i and j showed a synchrony in one phenophase (with i being different from j); fi is the number of months in which individual i exhibited a particular phenophase and N is the total number of individuals in the sample. The synchrony index of the species Z is calculated by the arithmetic mean of Xi, as follows: Z = ΣXi/N. This index ranges from 0 (no synchrony) to 1 (perfect synchrony) [47, 48].

Rainfall is a relevant variable in the case of the Caatinga, therefore this was correlated with the intensity of phenophases. The rainfall data were obtained from the monitoring carried out daily by an experimental basis the Instituto Agronômico de Pernambucano (IPA), which is set at about 13 km from the study area.

It is worth mentioning that the approach serves the purpose of phenological general characterization of phenophases of Spondias tuberosa Arruda under different conditions of soil management in order to establish a solid connection with the local perception of those phenophases.

Local people’s perceptions of phenophases

A collaborative approach, characterized by the sharing of experiences between the researcher and the residents, was used to assess the perceptions of Carão residents regarding the phenophases of S. tuberosa. This approach valued the knowledge of local people and resulted in determination of the most appropriate strategies to improve planning and actions [49]. This study is part of the research project “Traditional knowledge, phenology and morphological and genetic variability in populations of Spondias tuberosa Arruda (Anacardiaceae) in semi-arid Northeast” that was approved by the Ethics Committee on Research Involving Humans of the Centro de Ciências da Saúde, Universidade Federal de Pernambuco (registry number 401/08). Despite the phenological studies had begun in 2007, the only ethnobotanical approach was developed in 2009, after the aproval of the ethics committee. Everybody that chose to participate in the study sign a Free Consent and Understanding Agreement.

The participative method used in this study is consistent with an approach called “Score Exercises” that is commonly used to determine the relative importance of the environment and/or resources; the approach consists of local people quantifying the importance of a given resource [49, 50]. The method was adapted for this study with the development of a table called the “Phenological Calendar” with the phenophases as the column labels and months as the row labels. All those people (112 informants) known from previous ethnobotanical studies in this community [38] who have mentioned knowledge and/or use of S. tuberosa were invited to participate in the activity (70 people). Only 26 people were able to participate and they were randomly assigned to four groups. Despite this low number of respondents, it is emphasized that the vast majority of informants (10 people) of the community when it comes to knowledge and use of umbuzeiro, identified in a previous study [38], participated in this activity. Thus, four groups were formed (two groups of seven and two to six members) (Figure 1). The formation of four groups was made strictly for reasons of management activity, as well as a good distribution of participants. This activity was conducted at the end of March 2009, a period coinciding with the peak of the harvest of umbuzeiro. It is noteworthy that previous studies demonstrate the reliability of the knowledge people have about the plants of the region [42, 44], especially umbuzeiro [38, 45], a prominent species within the community if compared with other food plants [44].

Each group was coordinated by two members of the research team, the discussion mediator and the person who reported the consensus of the group for each phenophase. The intensity of each phenophase per month was measured on a scale of 0 to 10 (100%), determined by the group’s consensus. The score was determined by the number of “seeds of bean” (the seed number varied from 0, absence of phenophase, 10, maximum intensity, 100%, phenophase of the month) placed in each table cell, (each cell corresponded to one month) (Figure 1). For this analysis, two main habitat types were considered: native vegetation, which comprises the LUs Mountain Base and Mountain, and managed areas, represented by Pasture, Cultivated Areas and Homegardens. These two groups were necessary because most people stated that there is no distinction of phenophases within the managed areas and within native vegetation areas. Similarly, phenological data were grouped into native vegetation and managed areas for comparison with the information on the “Phenological Calendar”.

Data analysis

To the phenological characterization of the species in the different areas of landscape management, we tested the variation in intensity among LUs, within in year; between years, within LUs; and differences in mean intensity (across two years) among LUs were used a one-way ANOVA (significance level of 5%) with a posterior test of Tukey. Concerning the data of proportion, it was necessary its change into angular values (arcoseno√x/100), and only then apply the ANOVA test. A Spearman’s rank correlation coefficient was used to test the correlation between monthly rainfall and the intensity of phenophases in each landscape unit. The Bioestat 5.0 [51] computer program was used for all analyses.

To test the variation between the perceived intensity of phenophases obtained by "phenological calendar" activity with the intensity measured over two years initially, the perception data were multiplied by 10, since the scale used for phenological timing activity was from 0 to 10. Then both perceived and monitored intensity during two years were transformed to angular values (√ arcoseno x/100) and subsequently tested the variation of the data by ANOVA (one way) with a posteriori Tukey test at 5% probability. The Bioestat 5.0 [51] computer program was used for all analyses. Differences in the perception of intensity of phenophases were also verify using one way ANOVA.

Flowering

During the two years of the study, flowering started in September and ended in April (with a peak in November) in all LUs, (Figure 2). There were variations in the intensity of flowering among landscape units (Figure 2) but they were not significant (first year: F = 0.30; P > 0.05, second year: F = 0.27; P > 0.05). There were no significant differences in the intensity of phenophases in the same landscape unit between the monitoring years (Mountain: F = 0.002; P > 0.05; Mountain Base: F = 0.19; P > 0.05; Pasture: F = 0.04; P > 0.05; Cultivated Areas: F = 0.67; P > 0.05 and Homegardens: F = 0.06; P > 0.05) or in the mean intensity of the two years among landscape units (F = 0.29; P > 0.05). There were no differences in mean intensity between Native vegetation and Managed areas. There was a strong negative correlation between the intensity of flowering and rainfall (Figure 2) in all areas, both in the first (Mountain: rs = - 0.71; Mountain Base: rs = - 0.77; Pasture: rs = - 0.75; Cultivated Areas: rs = - 0.74 and Homegardens: rs = - 0.72; P <0.01 for all tests) and second year (Mountain: rs = - 0.59; Mountain Base: rs = - 0.80; Pasture: rs = - 0.58; Cultivated Areas: rs = - 0.73 and Homegardens: rs = - 0.62, P <0.01 for all tests).

Fruiting

The fruiting period was also similar among LUs, occurring from November to June in the Mountain Base, Cultivated Areas and Pasture, and extending to July in the Mountain and Homegardens. The fruiting peak occurred in March in all LUs. As observed for flowering, there were no significant differences in the intensity of fruiting among individuals in the LUs, either in the first (F = 0.20; P > 0.05) or second year (F = 0.36; P > 0.05, Figure 3), or between the two years (Mountain: F = 0.45; Mountain Base: F = 0.51; Pasture: F = 0.34; Cultivated Areas: F = 0.69 and Homegardens: F = 0.22; P > 0.05 for all tests). The mean intensity of fruiting within the two years did not vary significantly (F = 0.28; P > 0.05, Figure 3). There also significant differences (F = 0.25, P > 0.05) in the intensity of phenophases between native vegetation and managed areas.

Unlike the results for the flowering period, there was no correlation between rainfall and fruiting intensity in the LUs during the first year (Mountain: rs = 0.07; Mountain Base: rs = - 0.05; Pasture: rs = - 0.2005; Cultivated Areas: rs = - 0.05 and Homegardens: rs = - 0.03; P > 0.05 for all tests; Figure 3). In the second year, only the fruiting individuals of the Mountain were significantly correlated with rainfall (rs = 0.58; P < 0.05).

Leaf flush

In the two years of monitoring, leaf flush occurred continuously in all landscape units with peak intensity in March (Figure 4). In the first year and in the second year, there were no significant differences in the intensity of this phenophase among LUs (F=0.55; P > 0.05, to fist year; F = 0.36; P > 0.05, to second year) (Figure 4). As observed for other phenophases, there was no significant variation in the intensity of leaf flush in the same LU between years (Mountain: F = 1.68, Mountain Base: F = 3.61, Pasture: F = 1.95; Cultivated Areas: F = 1.78 and Homegardens: F = 1.25; P > 0.05 for all tests) (Figure 4). The mean intensity of leaf flush between the two years did not vary significantly among landscape units (F = 0.43; P > 0.05).

There was no correlation between the intensity of leaf flush and rainfall in the first year (Mountain: rs = - 0.02; Mountain Base: rs = - 0.16; Pasture: rs = - 0.11; Cultivated Areas: rs = - 0.03 and Homegardens: rs = - 0.14; P > 0.05 for all tests; Figure. 4). However, in the second year there was a positive correlation in all LUs (Mountain: rs = 0.69; Mountain Base: rs = 0.60; Pasture: rs = 0.64 and Cultivated Areas: rs = 0.64; P < 0.05 for all tests), except in Homegardens (rs = 0.53; P > 0.05).

Leaf fall

Leaf fall occurred throughout the year in Homegardens and from May to February in Cultivated Areas, Pasture and the Mountain Base. In the Mountain, leaf fall occurred from July to February with peaks in November 2007 and September 2008. As for other phenophases, the intensity of leaf fall did not vary significantly among LUs in either the first (F = 0.67; P > 0.05) or second years (F = 0.06; P > 0.05) or within the same LU (Mountain: F = 0.24; Mountain Base: F = 1.04; Pasture: F = 0.06; Cultivated Areas: F = 0.14 and Homegardens: F = 0.36; P > 0.05 for all tests) (Figure 5). The mean intensity of leaf fall also did not vary significantly among LUs (F = 0.24; P > 0.05).

There was a strong negative correlation between leaf fall and precipitation in the first year for individuals from areas of Pasture (rs = - 0.60; P < 0.05), Mountain Base (rs = - 0.60; P < 0.05) and Mountain (rs = - 0.61; P < 0.05; Figure 4B) and in the second year for all individuals in the LUs Mountain (rs = 0.87), Mountain Base (rs = - 0.74), Pasture (rs = - 0.77), Cultivated Areas (rs = - 0.68) and Homegardens (rs = - 0.76; P < 0.05 for all tests; Figure 4B).

Synchrony

Perception of phenophases of Spondias tuberosaArruda

To assess significant differences between the intensity of S. tuberosa phenophases perceived by people in phenological activity calendar with the average intensity of phenophases of the species over two years of phenological monitoring, there has been generated a single calendar from the four built, with mean intensities perceived by the informants. Thus, we have found that as for areas of native vegetation no significant differences between the perceived intensity and average intensity of the umbuzeiro phenophases exist (flowering, fruiting F = 0.05, F = 0.002; fall leaf: F = 0.01; and leaf flush : F = 2.53, P > 0.05) (Figures 6 and 7). Similarly, there was no significant differences between the intensity perceived and the intensity monitored over two years for flowering phenophases (F = 1.9, P > 0.05) and fruiting (F = 1.81, P > 0.05) of individuals in the areas managed, however there were significant differences when it comes to foliage fall (F = 5:39, P < 0.05) and budding (F = 4.64, P < 0.05).

Leaf flush was the phenophase that showed the greatest difference between the local people’s perception and the monitoring data, especially concerning the duration of the phase. However, the peak leaf flush was very similar between the Phenological Calendar and the monitoring data (Table 2). The perception of leaf fall throughout the year was similar among groups, mainly because they do not recognize differences in leaf fall between native vegetation and managed areas. Compared with field data, there were differences in the period of occurrence of this phenophase, but there was some consensus about its duration (Table 2). In areas of native and managed vegetation, the same average length of the phenophase was recorded (eleven months), whereas local people’s perceptions indicated a mean of eight months for areas of native vegetation and nine months for managed areas (Table 2). The peak of this phenophase was the variable of greatest similarity when comparing the two methodological approaches (Table 2).

Finally, people who participated in the activity, when asked about aspects that would influence the phenology behaviour of the species, were categorical in pointing out the rains during the dry season as the only variable that can negatively affect the flowering and therefore the fruiting of the species.

Phenological pattern and comparison between managed and native vegetation areas

The intensity of all phenophases, as well as the timing of these events were similar between the landscape units, suggesting that human management seems not to have changed significantly the phenological pattern of species. We are going to discuss some ecological aspects that may explain this uniformity, as well as the discussion of the phenological variation of individuals in managed and unmanaged areas.

The phenological results of flowering (occurring in the dry season) and fruiting (late dry season and early rainy season) obtained here are similar to those reported in studies with S. tuberosa in other Caatinga areas [32–34]. Because they only occur once a year and last a few weeks, the phenological patterns of flowering and fruiting of S. tuberosa can be classified as cornucopia, according to Gentry [52], or annual, according to Newstrom et al. [53].

The flowering and leaf flush of S. tuberosa observed during the dry season, as observed in other Caatinga plants [14, 54, 55] can be explained by the presence of organs that can accumulate water (such as roots) [56] and by the stem, which has a low wood density [18]. The phenology of plants that accumulate water does not depend on rainfall [57, 58], even in ecosystems such as arid environments where rainfall strongly influences the phenology [14, 18, 54, 59]. It is believed that some species of tropical dry forest plants flourish in the dry season because the wet season is used for vegetative growth [60, 61]. Moreover, the typical loss of leaves during the dry season favors flower visibility for pollinators [60].

Although flowering and fruiting may vary within and among populations [62], individuals of the same species that occupy adjacent regions tend to have similar phenological events during similar periods due to phylogenetic constraints [39]. This explains the similarity in synchronization and intensity of phenological events of S. tuberosa among the LUs. The similarity in phenological patterns observed among areas under different management regimes has been reported by Oaxaca-Villa et al. [43] in wild and managed populations of Escontria chiotilla (Cactaceae) in a semi-arid region of Mexico. However, other studies emphasize that in managed areas the intensity of fruit is higher when compared to unmanaged areas [23, 24, 27]. Individuals of cultivated populations of Stenocereus stellatus, species in an advanced state of domestication, have presented an increased number of flowers in anthesis if compared to individuals of wilderness areas [24]. Barreto et al. [32], while studying populations of Spondias tuberosa recorded more flowers and less fruits in natural areas in relation to managed areas. They attributed these differences to the fact that there is major pollinator activity (and therefore less competition for pollinators among individual plants) in managed areas, resulting in higher chances of fruit formation. Another study conducted with populations of S. tuberosa has also found that individuals located in corn fields, areas under intensive soil management, produced significantly more inflorescences if compared to other landscape units studied [9]. However, these authors found that there is no significant difference in the fruit production of the species if considered the landscape units studied, in addition to that, despite the intense flowering, reproductive success was limited to only one fruit per inflorescence in most individuals. The authors conclude that the possible explanation is the sharing of pollinators with individuals of other species that flower during the same period, such as Myracrodruon urundeuva Fr All (Anacardiaceae), Prosopis juliflora (Sw.) DC. and Mimosa tenuiflora Bent. (Fabaceae - Mimosoideae), making the reproductive success so low. Accordingly, the present results, at first sight, suggest that populations of S. tuberosa are not reflecting in their phenological behavior, mainly, in flowering and fruiting, the changes related to environmental changes, however, when considering the findings of Almeida et al. [9], seems to be clear that ecological interactions maintained by the species indicate a new structure for communities, suggesting an indirect effect of the domestication of landscapes.

As for phenophases overlap, the strong synchrony among individuals of S. tuberosa within and among the LUs supports the idea that in dry tropical forests there is a high seasonal synchrony in phenological patterns [57]. Phenological patterns possibly result from selective pressure exerted either by abiotic factors (such as intensity and duration of the dry season) [63, 64] or by biotic factors (including interactions with animals such as herbivores, pollinators and seed dispersers) [65, 66]. Thus, the synchrony observed here is especially important for flowering, since it favors gene flow among individuals and thus contributes to the maintenance of genetic variability in populations [19]. A study conducted in isolated populations of Spondias mombin in Panama also recorded a high synchrony in flowering and fruiting [11]; this synchrony may therefore be a characteristic of the genus.

These observations allow us to conclude that beyond the truism that species respond differently to the effects of handling, due to genetic and environmental heterogeneity [41], the strong synchrony observed between individuals located in different management areas indicate that timing mechanisms as for reproductive isolation have not occurred [24]. Probably this explains what we have observed in this study. This finding reinforces that an interspecific sharing of pollinators is the most likely explanation for the low reproductive success observed in populations of S. tuberosa under different soil management ways [9]. On the other hand, populations of Polaskia chichipe has its flowering peak ranging among wild populations, cultivated and managed ones, and being later in the latter [23]. This observed pattern provides reproductive isolation of populations, since individuals within populations are visited by pollinators at different times of individuals from other populations [23].

Phenological calendar versus phenological monitoring

The similarity between phenological patterns recorded in the field and those obtained from the Phenological Calendar (especially in relation to flowering and fruiting) reveal how traditional knowledge may be useful for understanding biological phenomena. The greatest similarity in relation to reproductive phenology may be explained by the fact that S. tuberosa occupies a prominent place in the local community, with well-known characteristics and food uses (of fruits) [38]. Moreover, the species provides shade for animals during the dry season in the Caatinga by maintaining its leaves, which reinforces the importance of S. tuberosa for the local population [38]. Nevertheless, the leaf flush phase had the lowest similarity with the phenological monitoring data.

In the Yucatán peninsula, a study conducted with Spondias pupurea also found that people are familiar with the phenophases of this kind, highlighting the flowering and fruiting, showing the existence of a clear relationship between the local classification of fruit types with the phenology of Spondias pupurea[28]. For example, people recognize and classify S. purpurea in three main classes, which are consistent with the period of the year in which the fetching of the fruit is done. However, this study advances if related to the latter establishing a direct comparison of the perception of activity and intensity of S. tuberosa phenophases with concrete data as for the species phenology monitoring. Thus, the phenological calendar proposed here reveals itself as an innovative and efficient concerning the access to knowledge on biological phenomena.

Another interesting aspect reported by Ruenes-Morales et al. [28] is the influence of rain on the productivity of S. purpurea individuals. According to the authors, informants reported that when the rains come during the dry season they affect the annual harvest, reducing fruit production. This perception was also evidenced by Carão people, who have also pointed the rain as the environmental variable that would affect the production of umbuzeiro individuals.

The aspects presented above, together with the studies conducted by Lins Neto et al. [38] and Lins Neto et al. [45] reinforce the prominence of S. tuberosa in Carão community, highlighting the close relationship people have with this plant. This way, our study provides strong evidences that the popular knowledge accumulated over generations is an excellent tool for rapid diagnosis of the phenology of a plant species. Phenological studies require long periods of observation to generate robust data and to increase predictive power [29, 53, 67, 68]. However, when decisions must be made quickly and long-term studies are infeasible, the use of traditional knowledge may facilitate appropriate and immediate conservation strategies.

Studies carried out in different part of the world, be concluded phenological studies satisfactory indicators of environmental changes, especially changes in ecosystem processes [69–72]. Thus, the perception of variations and reflections of these phenological changes within a global perspective reveals itself as importatnt tool for rapid diagnosis environment, assisting in the construction and development of strategies for the conservation of natural resources. However, it should be noted that the Phenological Calendar does not replace conventional phenological analysis, but complements it, especially for constructing future projections. This is because local people have knowledge of climate cycles that are not always detected in short term phenological studies, as they may occur at time intervals that do not coincide with the research period. Most importantly, more studies need to be developed in order to confirm the usefulness of traditional knowledge in the inference of plant phenology, especially reproductive phenophases, for other species and other ecosystems. Therefore, other systems and other approaches must be considered for strengthen the proposal that traditional knowledge is strong enough for diagnoses of phenology as confident indicator for several purposes such as climate change and conservation.