Abstract Aims Addressing plant responses to water stress is critical to understand the structure of plant communities in water-limited environments and to forecast their resilience to future changes in climate. In a semiarid agroforestry system in the Sahelian savannah of Leona (Senegal), we selected nine common tree species and explored their stress-resistance mechanisms. These species represent a variety of life forms and are of high regional socio-economic importance. We hypothesized that different species would show different suites of traits to cope with water stress and expected to identify functional groups differing in strategies to withstand water shortage. Methods Along a dry and a wet season, we monitored four traits reflecting above- and below-ground strategies of resource acquisition such as predawn leaf water potential (ψpd), specific leaf area (SLA), leaf thickness and leaf area index (LAI). We also measured two morphological traits: trunk diameter and tree height. LAI and ψpd were measured six times during the dry and rainy seasons, and the other traits were measured once. Important Findings We identified two functional classes subdivided into two functional groups of each class. The first class included deciduous and semi-deciduous species that generally had large SLA, low leaf thickness and small-to-intermediate inter-seasonal variations in ψpd. The second class included evergreen species of two functional groups that differ in SLA, leaf thickness and the magnitude of inter-seasonal variations of ψpd throughout the year. The four functional groups identified in this study represent plant strategies differing in their response to changing environmental conditions. Acacia tortilis, Adansonia digitata, Balanites aegyptiaca, Celtis integrifolia, Combretum glutinosum, Faidherbia albida, functional traits, Neocarya macrophyla, Sclerocarya birrea, Tamarindus indica, water stress INTRODUCTION Tropical seasonally dry forests and savannahs are subjected to rainfall regimes that greatly vary in frequency and intensity across years where rainfall is unevenly distributed among seasons within a year. Therefore, water availability is one of the most limiting factors for plants in such tropical systems, influencing plant productivity and community structure across large-scale regional gradients and small-scale, topographic gradients (Ogle and Reynolds 2004). The relative success of tree species along these gradients and their fate under potential changes in water availability will depend on the mechanisms through which they respond to water shortage (Markesteijn 2010). Research on plant responses to water stress is critically important as most climate change scenarios suggest an increase in aridity in many areas of the globe, including the tropics (Dai 2013; Diop et al. 2016; Petit et al. 1999), which may result in shifts in the composition of current plant communities and their distribution range. For instance, the severe droughts in the Sahel from the 1970s to the 1990s, followed by relatively low annual rainfall afterward (Funk et al. 2012), changed the ecological baselines in the region (Bell and Lamb 2006; Nicholson 2005). In fact, in this area, there are already significant declines in tree density and species richness as a consequence of increased temperature and decreased rainfall (Gonzalez et al. 2012). Functional traits may result from evolutionary processes (Flores et al. 2014) and are used as indicators of ecophysiological function. For example, they may be used to anticipate plant responses to environmental factors (Cornelissen et al. 2003a; Lavorel and Garnier 2002), including water shortage (Knevel et al. 2005), to highlight adaptive strategies (Grime 2001), and, overall, their analysis help us explain species effects at ecosystem level (Vendramini et al. 2002; Wright et al. 2005). Species differ in drought tolerance and associated traits (Lohbeck et al. 2015; Markesteijn 2010; Valladares and Sánchez-Gómez 2006). Plant sensitivity to water shortage may be evaluated through different indicators of plant physiological status, such as leaf water potential, stomatal conductance, or chlorophyll fluorescence (Armas and Pugnaire 2005, 2009; Gómez-Aparicio et al. 2006; Quero et al. 2011), and therefore, measuring key drought-resistance traits may provide powerful tools to examine interspecific responses to drought (Engelbrecht and Kursar 2003). Such trait measurements, and the trade-offs they may show, can explain much about species performance and community composition (Poorter et al. 2010). Although there are many plant traits reflecting sensitivity to water stress, most studies on trait–environment linkages have focused on leaf traits (Reich 2014) as they reflect strategies of resource uptake and use. For example, species with high specific leaf area (SLA), low-density tissues and high leaf nitrogen content tend to have rapid resource capture (Poorter and Garnier 2007; Ruiz-Robleto and Villar 2005; Wright and Westoby 2001), allowing these plants to be dominant in moist and fertile areas (Grime et al. 1997; Reich et al. 1999). By contrast, species with efficient resource conservation (Chapin et al. 1993; Wright et al. 2004) minimize nutrient loss and increase their competitive abilities in dry and nutrient-poor environments (Aerts 1995; Hobbie 1992). Plant species segregate along natural gradients of water availability according to their capacity to withstand water stress. However, species with contrasting ecological requirements coexist (Valladares and Sánchez-Gómez 2006), as in the semiarid Sahelian zone of Senegal, where evergreen and deciduous species co-occur. Such contrasted functional groups certainly reflect different strategies of the different species in a plant community to deal with water shortage (Fournier 1995), a feature that still puzzles the scientific community (Westoby et al. 2002). Here, we focus on drought-resistance mechanisms in nine Sahelian tree species of high socio-economic importance for local populations. We hypothesized that (i) in the semiarid area of Sahel, different woody species will evidence trade-offs between water uptake traits and use conservation and (ii) the presence of similar water-use strategies will allow the grouping of different tree species into the same plant functional group. MATERIALS AND METHODS Study site and species The study was conducted in the Sahelian savannah of Leona, northwest Senegal, a semiarid environment with Sub-Canarian climate (Wade 1997). It is under the influence of oceanic winds and currents that reduce the extreme seasonal contrasts of the Sudano-Sahelian climate. Therefore, this region has a smooth, atypical climate whose influence diminishes away from the coast. Between February and May, the area is dominated by the Harmattan (hot and dry winds) with huge sand storms and high desiccating effect. The wet monsoon season occurs between June and October with an average annual rainfall that varies between 220 and 350 mm (Gaye and Edmunds 1996), followed by a dry season between November and May. Temperatures are high during most of the year. The hottest periods generally correspond to the months of May and October. Minimum temperatures range between 22 and 28°C and maximum temperatures between 31 and 37°C (Wade 1997). However, temperatures have increased steadily since the 1950s while rainfall has decreased (Sagna et al. 2015), in agreement with the Intergovernmental Panel on Climate Change scenarios. Soils are mostly sandy, little-leached ferruginous tropical soils with poor structure and usually occupied by peanut, cowpea, millet crops and grasslands. There are some areas under intensive cropping all year round, leading to impoverished soils that require large inputs of fertilizer before new crops are sowed. Nonetheless, a large proportion of the area is covered with grassland used for livestock. Being sandy soils, they have low water-holding capacity, low organic matter content and are often subject to wind erosion. The natural landscape is a savannah where scattered big trees and shrubs grow in a matrix of a continuous herbaceous/grass species layer that thrives during the rainy season. Most woody species of this Sahelian savannah ecosystem are thorny. Harvesting of trees and shrubs, grazing, cropping and rainfall all contribute to shape the vegetation regionally (Konaté 2010). We selected nine dominant and ecologically and socio-economically important tree species with multiple uses in these areas: Acacia tortilis spp. raddiana, Adansonia digitata (baobab), Balanites aegyptiaca, Celtis integrifolia, Combretum glutinosum, Faidherbia albida, Neocarya macrophyla, Sclerocarya birrea and Tamarindus indica (Table 1). Table 1: selected woody species of the Sahelian zone of Senegal and habits Species Families Leaves habit Acacia tortilis subsp. raddiana (Savi) Brenan Fabaceae Deciduous Adansonia digitata L Malvaceae Deciduous Balanites aegyptiaca (L.) Del. Balanitaceae Evergreen Celtis integrifolia Lam. Ulmaceae Evergreen to deciduous Combretum glutinosum Perr. Ex DC. Combretaceae Evergreen Faidherbia albida (Del.) Chev. Fabaceae Deciduous Neocarya macrophyla (sabine) Prance Chrysobalanaceae Evergreen Sclerocarya birrea (A. Rich) Hochst Anacardiaceae Deciduous Tamarindus indica L. Fabaceae Evergreen to deciduous Species Families Leaves habit Acacia tortilis subsp. raddiana (Savi) Brenan Fabaceae Deciduous Adansonia digitata L Malvaceae Deciduous Balanites aegyptiaca (L.) Del. Balanitaceae Evergreen Celtis integrifolia Lam. Ulmaceae Evergreen to deciduous Combretum glutinosum Perr. Ex DC. Combretaceae Evergreen Faidherbia albida (Del.) Chev. Fabaceae Deciduous Neocarya macrophyla (sabine) Prance Chrysobalanaceae Evergreen Sclerocarya birrea (A. Rich) Hochst Anacardiaceae Deciduous Tamarindus indica L. Fabaceae Evergreen to deciduous View Large Plant traits We selected four plant traits that are indicators of different functions related to resource use by the plant and are also complementary in representing water-use strategies. They were predawn leaf/twig water potential (ψpd), SLA, leaf area index (LAI) and leaf thickness (Niinemets 2001). The ψpd provides information on plant water status as well as on its capacity to take up soil water. Its value range is species specific and depends, among others, on rooting depth, root architecture and root physiological properties (Pérez-Harguindeguy et al. 2013). The ψpd provides data comparable among species and is less variable than midday water potential, which is affected by temperature, radiation, vapour pressure deficit and other environmental variables. Although midday water potential could make a better index of a species tolerance to water shortage, midday water potential is quite variable as hydraulic properties of highly stressed leaves and twigs may vary significantly from point to point (Koide et al. 1989; Turner et al. 1984). We used a pressure chamber (PMS Instruments, Santa Barbara, CA, USA) to measure ψpd. SLA is one of the most widely used leaf traits to analyse plant responses to the environment. SLA is strongly linked to relative growth rate and the plant resource-use strategy (Poorter and Garnier 2007) and can be used to estimate resource availability (Pérez-Harguindeguy et al. 2013). A related trait is leaf thickness, linked to leaf construction costs, leaf lifespan and gas exchange (Loranger and Shipley 2010). Leaf thickness is one of the key components of SLA. Finally, we also measured the LAI or the total leaf area of the plant per unit ground area (Jonckheere et al. 2004), with an LAI 2000 (Li-Cor Biosciences, Lincoln, NE, USA). LAI is a dimensionless index (m2/m2) and reflects the capacity of the plant to intercept radiation. and LAI and ψpd may be inversely related (Bréda et al. 1995), as higher LAI means higher evaporative surface, which may lead to a decrease in ψpd. In summary, these four traits reflect plant strategies in resource capture and use. Two tree-level morphological traits, diameter at breast height (DBH) and plant height were additionally measured to control for variability associated with tree size. Trait data were collected following the protocols in Cornelissen et al. (2003b), Knevel et al. (2005) and Pérez-Harguindeguy et al. (2013). We measured these traits in six randomly selected healthy, mature trees of each of the nine species, all growing in the field. Leaf measurements were performed on six intact, fully developed leaves from a well-lit branch (~1 cm in diameter and at least 1 m long) per individual tree. The ψpd was measured on one leaf or twig (depending on the species) per branch. LAI and ψpd measurements were carried out five times, two during the dry season (November 2010 and April 2011) and three during the rainy season (July 2010, August and September 2011), whereas SLA and leaf thickness were measured once when leaves were at their best during the rainy season for all species except for Faidherbia; for this species, leaves were collected in the dry season, as it is a rainy season deciduous species. The 2010 wet season was well above average, which in July reached 230 mm, while the dry season was quite below average, with no precipitation in November 2010 and April 2011—our sampling dates. The sampling dates in the 2011 wet season (August and September) were above average. Data analysis Differences in plant traits among species, seasons and months were analysed with repeated-measures general linear mixed models. Main fixed factors were species, season/month and the interaction between them. Each tree (identified with a unique ID) was the repeated-measured unit across months, and we set a compound symmetry correlational structure to account for the relation among repeated measures. We also tested several variance structures to avoid heteroscedasticity. In the case of ψpd and LAI, we selected varExp, which represents an exponential structure of a variance–covariate function structure (Galecki and Burzykowski 2013). For the other traits (SLA, thickness, DBH and height), we used varIdent, which represents a variance structure with different variances for different strata (Galecki and Burzykowski 2013). We selected the best model according to the Akaike information criterion (Akaike 1974), choosing the one with lowest value. Post hoc differences were tested with Fisher’s least significant difference test. We also performed multivariate analyses (principal component analysis [PCA] and cluster analysis) of all functional traits, in order to identify groups of individuals with common functional characteristics. Statistical analyses were performed with Infostat (Di-Rienzo et al. 2015). Reported values throughout the text and figures are means ± 1 standard error (SE). RESULTS There were significant changes in the seasonal course of ψpd in most species (Fig. 1a). It was highest (i.e. less negative) for all species in July 2011, after the onset of the rainy season, where values ranged from −0.24 to −0.65 MPa. Adansonia, Sclerocarya and Neocarya showed rather steady ψpd during both the rainy and dry seasons. In most species, however, there was a decrease in ψpd during the dry season, which remained low for the whole period (Fig. 1b). We recorded important intraspecific variability in November 2010 and April 2011 (both months in the dry season), especially in Acacia, Balanites and Tamarindus. Figure 1: View largeDownload slide seasonal changes in Ψpd (a) and LAI (c) and mean Ψpd (b) and LAI (d) along the rainy and dry seasons of nine tree species in the Sahelian region of Senegal: Acacia tortilis (Acto), Adansonia digitata (Addi), Balanites aegyptiaca (Baae), Celtis integrifolia (Cein), Combretum glutinosum (Cogl), Faidherbia albida (Faal), Neocarya macrophylla (Nema), Sclerocarya birrea (Scbi) and Tamarindus indica (Tain). Data are mean values ± 1 SE, n = 6. Post hoc letters in panels (a) and (c) are not included to improve clarity (see supplementary Tables S1and S2 for post hoc tests). Mean seasonal values shown by horizontal lines (solid, wet; broken, dry) ±1 SE (dotted lines). Figure 1: View largeDownload slide seasonal changes in Ψpd (a) and LAI (c) and mean Ψpd (b) and LAI (d) along the rainy and dry seasons of nine tree species in the Sahelian region of Senegal: Acacia tortilis (Acto), Adansonia digitata (Addi), Balanites aegyptiaca (Baae), Celtis integrifolia (Cein), Combretum glutinosum (Cogl), Faidherbia albida (Faal), Neocarya macrophylla (Nema), Sclerocarya birrea (Scbi) and Tamarindus indica (Tain). Data are mean values ± 1 SE, n = 6. Post hoc letters in panels (a) and (c) are not included to improve clarity (see supplementary Tables S1and S2 for post hoc tests). Mean seasonal values shown by horizontal lines (solid, wet; broken, dry) ±1 SE (dotted lines). Similarly, LAI changed across seasons (Fig. 1), but differences between dry and wet seasons were only significant for Faidherbia and Sclerocarya. Faidherbia was the only species in our data set that was deciduous during the rainy season, and it showed higher LAI values during the dry season than during the rainy season (Fig. 2d). Sclerocarya showed high LAI in the dry season, likely due to a long rainy season in 2010. During the dry season, the evergreen species Neocarya and Combretum had the highest LAI. Large intraspecific variation was observed in Adansonia (April 2011), Combretum and Celtis in August 2010 and also in Neocarya (November 2010 and August 2011) (Fig. 1c). Figure 2: View largeDownload slide SLA (a), leaf thickness (b), plant height (c) and DBH (d) of the nine Sahelian tree species (species legend as in Fig. 1). Data are mean values ± SE, n = 6. Bars with different letters are significantly different (Fisher’s LSD post hoc tests). Figure 2: View largeDownload slide SLA (a), leaf thickness (b), plant height (c) and DBH (d) of the nine Sahelian tree species (species legend as in Fig. 1). Data are mean values ± SE, n = 6. Bars with different letters are significantly different (Fisher’s LSD post hoc tests). SLA differed across species, being smallest in Balanites, Combretum and Neocarya and highest in Faidherbia (Fig. 2a). Leaf thickness (Fig. 2b) also differed among species and, not surprisingly, was highest in the evergreens Balanites, Neocarya and Combretum, whereas it was low in the deciduous and semi-deciduous species Faidherbia and Tamarindus, the other species displaying intermediate values. Finally, the tallest (tree height, Fig. 2c) individuals were those of Celtis and Adansonia, and there was large intraspecific variation in Balanites and Celtis. We aimed to identify functional groups according to variations in traits by PCA and performed a hierarchical clustering, both considering the following variables: ψpd, LAI, SLA and height of all species. We excluded leaf thickness as we had no values for one species. The PCA showed that the absolute value of ψpd (i.e. without sign) was positively correlated with LAI and inversely correlated with SLA and tree height (Fig. 3); i.e. the lower the SLA and tree height, the higher the LAI and the absolute value of ψpd (thus, the more negative the water potential). Thus, Balanites and Combretum, which had more negative ψpd values than most of the species, had also low SLA. Species like Acacia and Adansonia were characterized by high SLA, while Neocarya, Celtis and Tamarindus showed comparatively higher LAI. Figure 3: View largeDownload slide PCA of SLA, absolute value (with no sign) of Ψpd, and tree LAI and height of the nine Sahelian tree species. Figure 3: View largeDownload slide PCA of SLA, absolute value (with no sign) of Ψpd, and tree LAI and height of the nine Sahelian tree species. The hierarchical classification (Fig. 4) allowed us to group species into two classes with two functional groups each. The first class included deciduous and semi-deciduous species and generally had large SLA and low leaf thickness, showing small-to-intermediate inter-seasonal variations in ψpd. This first class was split into two functional groups, the first group was formed by four deciduous species, Acacia tortilis, Adansonia digitata, Faidherbia albida and Sclerocarya birrea, while the second group only included one species, the evergreen-to-semi-deciduous Celtis integrifolia. The second class included all evergreen species and one semi-deciduous, all with overall low SLA, high leaf thickness and generally large inter-seasonal variations of ψpd throughout the year. This class was also subdivided into two groups, one formed only by Neocarya macrophylla while the other by Balanites aegyptiaca, Combretum glutinosum and Tamarindus indica (this last species was the only semi-deciduous in this class). Figure 4: View largeDownload slide cluster analysis of three physiological traits (Ψpd, SLA and LAI) of the nine Sahelian tree species (Cophenetic correlation = 0.925). Figure 4: View largeDownload slide cluster analysis of three physiological traits (Ψpd, SLA and LAI) of the nine Sahelian tree species (Cophenetic correlation = 0.925). DISCUSSION A combination of physiological and morphological traits enabled the grouping of our nine Sahelian tree species into different functional types. This classification implies that species in the same functional group, sharing similar features, will likely display similar responses to the environment (Garnier and Navas 2012) i.e. they are ‘functional response groups’ (Lavorel et al. 1997). We used traits easy to monitor and quantify (Garnier et al. 2004), measured using standardized protocols (Cornelissen et al. 2003b; Knevel et al. 2005; Pérez-Harguindeguy et al. 2013), which are indicators of the mechanisms by which plants make use of water and tolerate water stress. Tropical savannahs are important biomes across the world (Williams et al. 1997), with a high diversity of species and life forms in both the herbaceous and woody layers (Wilson et al. 1996). Numerous woody species in savannahs, dominant and subdominant, are drought deciduous but have developed additional strategies to cope with seasonal, chronic and erratic dry spells. Larcher (1995) described two wide groups that appear to display ‘drought-avoiding’ and ‘drought-tolerant’ strategies, and our data lend support to classifications that sort out drought avoidance and drought-tolerance strategies. Tree species in our study avoid water stress by different means. Our analyses grouped species into two classes with two functional groups each. One of these classes could be considered as that of the water stress avoiders and, overall, included the deciduous and semi-deciduous species, which were divided into two functional groups. One group included Acacia, Adansonia, Faidherbia and Sclerocarya, all deciduous species, and another group included Celtis (semi-deciduous). The two functional groups decrease evaporative surfaces by shedding leaves during the unfavourable season (except Faidherbia), contributing to preserve water within the plant; but these species are also deep-rooted (Logan et al. 2010), which suggests access to deep water storages as a complementary measure to avoid water stress (Logan et al. 2010). In addition, these two functional groups share large SLA and generally low leaf thickness, traits associated with low leaf longevity (i.e. deciduousness) and low construction costs (Westoby et al. 2002; Flores et al. 2014) and are indicative of rapid resource capture or exploitative strategy (Flores et al. 2014). The first functional group of water stress avoiders included the deciduous Acacia, Adansonia, and Sclerocarya characterized by small-to-intermediate inter-seasonal variations in water potential, suggesting that they have access to permanent water sources. Overall, this functional group includes species that avoid water stress first by maximizing water uptake through extended root systems, except Faidherbia, and, when water stress accentuates, by minimizing water loss by progressively shedding their leaves, again, except Faidherbia. Shallow roots in Adansonia occupy up to 0.7 ha, while lateral roots in Sclerocarya extend up to 58 m (mean 40 m) and 20–25 m in Acacia (Breman and Kessler 1995; Soumaré et al. 1994). Faidherbia, also included in this group, has the unique character of losing its leaves in the wet season. Roupsard et al. (1998) proposed that this behaviour is intended to avoid competition with herbs. However, Faidherbia’s strategy may be better regarded as a response that provides protection against herbivores and diseases, as some studies have suggested for other species (William and Thomas 1986). This explanation would be supported by the fact that Faidherbia showed high SLA, which is related to soft, less defended leaves against herbivores. Faidherbia thrives in the dry season based on the efficiency of its root system, able to take up water 30 m deep (Breman and Kessler 1995). As the taproots of adult Faidherbia individuals reach the water table, they ensure water supply all year round. In the same group of water stress avoiders is Acacia, displaying a moderate inter-seasonal variation in ψpd. Many authors have characterized water relations in Acacia (Do et al. 1998; Gebrekirstos et al. 2011; Otieno et al. 2005). Its wide spatial distribution is indicative of a remarkable adaptability to different environmental conditions that can be attributed to three basic elements: water uptake from deep soil layers, low water consumption and optimization of the ratio between assimilation and transpiration (i.e. high water use efficiency; Do et al. 1998). In addition, the bulk of gas exchange does occur in the rainy season where potential water losses are lower (Do et al. 1998). Therefore, the drought-avoidance strategy of Acacia is based on two mechanisms: maximization of water uptake and minimization of water loss. Both mechanisms keep turgor high and, as shown by our data, maintain relatively high water potential all year round. Regarding Adansonia and Sclerocarya, in addition to losing their leaves under water stress, they have relatively short taproots (reaching depths of 2.4 m in Sclerocarya; Orwa et al. 2009) and robust lateral roots. Adansonia roots are relatively shallow (down to ~1.8 m) but spread out to a distance greater than the height of the tree (Fenner 1980). Robust lateral roots allow these species to explore the upper soil horizons and extract the maximum of rainwater before infiltration. Such an extensive shallow root system suggests it is adapted to exploiting erratic rainfall (February et al. 2013; Hodge 2010; Pugnaire et al. 2006). Water may be stored in the trunk and, along with leaf shedding during the dry season, enable the tree to have access to water supplies. Storage organs in Adansonia are large woody stems more or less lignified with succulent tissue (Arbonnier 2004). Baobab trees have long been assumed to depend on water stored in their large, swollen stems (Wickens 1983), but recent reports indicate that only a limited amount of stored water is used for physiological processes buffering daily water deficits (Chapotin et al. 2006b). In contrast, stem water reserves are used by the tree to support new leaf growth and cuticular transpiration, but not to support stomatal opening in the dry season (Chapotin et al. 2006a) since leaves are present only during the rainy season. The second group in the class of water stress avoiders is monospecific and formed by Celtis, the only evergreen-to-semi-deciduous species (Arbonnier 2004). It has intermediate-to-high SLA values and the highest height along with Adansonia. However, it has high LAI, and water loss is minimized by a progressive loss of leaves as water stress intensity progresses (Bourou 2012; Maes et al. 2009). Depending on the environmental conditions where this species grows, it can show an evergreen or semi-deciduous syndrome. As Bai et al. (2015) described, when unfavourable conditions (temperature, water and nutrient) are relatively short, the retention of evergreen leaves in evergreen-to-semi-deciduous species is beneficial for carbon fixation. However, when unfavourable conditions become longer, the maintenance of leaves requires a vast investment of nutrients and energy for these species. Thus, a species such as Celtis integrifolia with high LAI under high water availability tends to have more leaves with higher SLA (Ogaya and Peñuelas 2007), which is a more efficient strategy to maximize photosynthetic gain. Under lower water availabilities, however, Celtis adopts a deciduous strategy instead of investing more resources in strong, sclerophyll leaves to stand climatic adversity. The other functional class is split into two functional groups that showed a drought-tolerance strategy by being able to maintain low leaf water potentials (Valladares et al. 2004). Tolerant species have tissues resistant to dehydration and xylem cavitation, show osmotic adjustment and high cell wall elasticity (Marshall and Dumbroff 1999; Pereira et al. 2004; Villar-Salvador et al. 2013). With the exception of Tamarindus, which is semi-deciduous, species in this class (Balanites, Combretum—Group III) and Neocarya (Group IV) are all evergreen, maintain high LAI all the year round and show low ψpd even during the rainy season, but particularly during the dry season, allowing high rates of light interception (Kool and Lenssen 1997). Maintaining a higher LAI, however, increases transpiration, leading to higher inter-seasonal variations in ψpd which is characteristic of this group. During the dry season, leaf water potential decreased in all evergreen species in our study (except Neocarya), thereby reducing their ability to supply water to cells. Such imbalance between water provision and needs may be explained by a very high resistance to the passage of water in the soil–plant interface (Sobrado 1986). Evergreen species lose their turgor pressure at a total water potential much lower than deciduous species (Fournier 1995). Thereby, the leaf tissue of evergreen species is adapted to stand higher turgor pressure than deciduous species when the water potential decreases, although there is variability (e.g. Balanites and Combretum; Fournier 1995). The species in this group have developed several strategies to withstand water stress. Balanites, one of Sahelian trees with higher tolerance to water stress (Depierre and Gillet 1991), in addition to have deep root system (Hall and Walker 1991), has long spines and sclerophyllous leaves. Tamarindus seems the less water stress-resistant species in the group, as it reduces transpiration through a gradual loss of leaves to almost total defoliation under stressful conditions but maintains water potentials (Bourou 2012). The last functional group included only Neocarya, a species that experienced small ψpd variations throughout the year, with high LAI, leaf thickness and low SLA. Low SLA tend to correspond with high structural defences and long leaf lifespan (Cornelissen et al. 2003b; Wright et al. 2004), which is characteristic of a resource-conservative strategy (Flores et al. 2014). Indeed, Neocarya has leathery and hairy leaves that decrease transpiration and allow coping with water stress (Arbonnier 2004). CONCLUSION Our data show that different tree species display contrasting suites of traits reflecting different mechanisms to cope with water stress even within an apparently homogeneous environment, suggesting that they evolved under different environmental conditions. Functional traits allowed us to identify plant strategies and group species into four functional groups based on their responses to water stress. Overall, evergreen and semi-deciduous species root deeper than deciduous species and rely on deep-stored water reserves during the dry period. Changes in rainfall amount and the length of drought spells will likely be critical for these species, being the species with high LAI more sensitive to these changes. On the other hand, deciduous and semi-deciduous species cope with drought by avoiding activity in the periods of water shortage. They make a less efficient use of water and have low capacity to control water losses. Changes in the length of the wet season and of drought spells within the wet season will likely be the most critical for the persistence of these species. SUPPLEMENTARY DATA Supplementary data are available at Journal of Plant Ecology online. FUNDING This work was carried out within the FUNCITREE Project, an EU 7FP (contract KBBE-227265). It was also partly funded by the Spanish MICINN (grant CGL2014-59010-R). J.S.D. received a PhD fellowship from the Spanish Agency of International Cooperation for Development (AECID). C.A. is grateful to the Spanish Government for her ‘Ramón y Cajal’ contract (RYC-2012-12277). ACKNOWLEDGEMENTS We thank Mr Abdou Faye for technical assistance and the local people in the study area for allowing us to work on their fields and for their support during the entire period of data collection. We also thank Fernando Casanoves for statistical advice. Conflict of interest statement. None declared. REFERENCES Aerts R( 1995) The advantages of being evergreen. Trends Ecol Evol 10: 402– 7. Google Scholar CrossRef Search ADS PubMed Akaike H( 1974) A new look at the Statistical Model Identification. IEEE Trans Autom Control 74: 716– 23. Google Scholar CrossRef Search ADS Arbonnier M( 2004) Trees, Shrubs and Lianas of West African Dry Zones . éditions Quæ, 574. Armas C Pugnaire FI( 2005) Plant interactions govern population dynamics in a semiarid plant community. J Ecol 93: 978– 89. Google Scholar CrossRef Search ADS Armas C Pugnaire FI( 2009) Ontogenetic shifts in interactions of two dominant shrub species in a semi-arid coastal sand dune system. J Veg Sci 20: 535– 46. Google Scholar CrossRef Search ADS Bai K He C Wan Xet al. . ( 2015) Leaf economics of evergreen and deciduous tree species along an elevational gradient in a subtropical mountain. AoB PLANTS 7: 1– 15. Google Scholar CrossRef Search ADS Bell MA Lamb PJ( 2006) Integration of weather system variability to multidecadal regional climate change: the West African Sudan–Sahel zone, 1951–98. J Clim 19: 5343– 65. Google Scholar CrossRef Search ADS Bourou S( 2012) Étude Éco-physiologique du Tamarinier (Tamarindus indica L.) en milieu Tropical Aride . Belgium: Université de Gand. Bréda N Granier A Aussenac G( 1995) Effects of thinning on soil and tree water relations, transpiration and growth in an oak forest (Quercus petraea (Matt.) Liebl.). Tree Physiol 15: 295– 306. Google Scholar CrossRef Search ADS PubMed Breman H Kessler JJ( 1995) Woody Plants in Agro-Ecosystems of Semi-arid Regions, with an Emphasis on the Sahelian Countries . Berlin: Springer Verlag. Google Scholar CrossRef Search ADS Chapin FSI Autumn K Pugnaire F( 1993) Evolution of suites of traits in response to environmental stress. Am Nat 142: S78– 92. Google Scholar CrossRef Search ADS Chapotin S Razameharizaka JH Holbrook NM ( 2006a) Baobab trees (Adansonia) in Madagascar use stored water to flush new leaves but not to support stomatal opening prior to the rainy season. New Phytol 169: 549– 59. Google Scholar CrossRef Search ADS Chapotin S Razanameharizaka JH Holbrook NM( 2006b) Water relations of baobab trees (Adansonia L.) during the rainy season: does stem water buffer daily water deficits? Plant Cell Environ 29: 1021– 32. Google Scholar CrossRef Search ADS Cornelissen JHC Cerabolini B Castro-Díez Pet al. . ( 2003a) Functional traits of woody plants: correspondence of species rankings between field adults and laboratory-grown seedlings? J Veg Sci 14: 311– 22. Google Scholar CrossRef Search ADS Cornelissen JHC Lavorel S Garnier Eet al. . ( 2003b) A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Aust J Bot 51: 335– 80. Google Scholar CrossRef Search ADS Dai A( 2013) Increasing drought under global warming in observations and models. Nat Clim Chang 3: 52– 8. Google Scholar CrossRef Search ADS Depierre D Gillet H( 1991) L’arbre désertique source de vie. Bois Forets des Trop 227: 43– 50. Di-Rienzo JA Casanoves F Balzarini MGet al. . ( 2015) InfoStat versión 2013 . Córdoba, Argentina: Grupo InfoStat, FCA, Universidad Nacional de Córdoba. Diop L Bodian A Diallo D ( 2016) Spatiotemporal trend analysis of the mean annual rainfall in Senegal. European Scientific Journal 12: 231– 45. Do F Rocheteau A Diagne ALet al. . ( 1998) Flux de sève et consommation en eau d’Acacia tortilis dans le Nord Ferlo. In Claudine C Grigon CC Gueye Met al. (eds). L’ Acacia au Sénégal . Paris, France: Éditions de I’Orstom, 63– 8. Engelbrecht BMJ Kursar TA( 2003) Comparative drought-resistance of seedlings of 28 species co-occuring tropical woody plants. Oecologia 136: 383– 93. Google Scholar CrossRef Search ADS PubMed February EC Matimati I Hedderson TAet al. . ( 2013) Root niche partitioning between shallow rooted succulents in a South African semi desert: implications for diversity. Plant Ecol 214: 1181– 7. Google Scholar CrossRef Search ADS Fenner M( 1980) Some measurements on the water relations of baobab trees. Biotropica 12: 205– 9. Google Scholar CrossRef Search ADS Flores O Garnier E Wright IJet al. . ( 2014) An evolutionary perspective on leaf economics: phylogenetics of leaf mass per area in vascular plants. Ecol Evol 4: 2799– 811. Google Scholar CrossRef Search ADS PubMed Fournier C( 1995) Fonctionnement Hydrique de Six Espèces Ligneuses Coexistant dans une Savane Sahélienne (Région du Ferlo, Nord Sénégal) . Paris, France: ORSTOM. Funk C Michaelsen J Marshall M( 2012) Mapping recent decadal climate variations in precipitation and temperature across eastern Africa and the Sahel. In Wardlow B Anderson M Verdin J (eds). Remote Sensing of Drought—Innovative Monitoring Approaches . London: Taylor and Francis, 25. Galecki A Burzykowski T( 2013) Lineal Mixed-Effects Models Using R: A Step-by-Step Approach . New York: Springer-Verlag. Google Scholar CrossRef Search ADS Garnier E Cortez J Billès Get al. . ( 2004) Plant functional markers capture ecosystem properties during secondary succession. Ecol Monogr 85: 2630– 7. Garnier E Navas ML( 2012) A trait-based approach to comparative functional plant ecology: concepts, methods and applications for agroecology. Agron Sustain Dev 32: 365– 99. Google Scholar CrossRef Search ADS Gaye CB Edmunds WM( 1996) Groundwater recharge estimation using chloride, stable isotopes and tritium profiles in the sands of the north-western Senegal. Environ Geol 27: 246– 51. Google Scholar CrossRef Search ADS Gebrekirstos A van Noordwijk M Neufeldt Het al. . ( 2011) Relationships of stable carbon isotopes, plant water potential and growth: an approach to asses water use efficiency and growth strategies of dry land agroforestry species. Trees 25: 95– 102. Google Scholar CrossRef Search ADS Gómez-Aparicio I Valladares F Zamora R( 2006) Comparative response of tree saplings to light: implications for understanding inter-specific differences in nurse-plant relationship. Tree Physiol 26: 947– 58. Google Scholar CrossRef Search ADS PubMed Gonzalez P Tucker CJ Sy H( 2012) Tree density and species decline in the African Sahel attributable to climate. J Arid Environ 78: 55– 64. Google Scholar CrossRef Search ADS Grime JP( 2001) Plant Strategies, Vegetation Processes and Ecosystem Properties . New York: Wiley. Grime JP Thompson K Hunt Ret al. . ( 1997) Integrated screening validates primary axis of specialisation in plants. Oikos 79: 259– 81. Google Scholar CrossRef Search ADS Hall JB Walker HD( 1991) Balanites Aegyptiaca: a Monograph , 3rd edn. Bangor, Royaume Uni: University of Wales. Hobbie SE( 1992) Effects of plant species on nutrient cycling. Trends Ecol Evol 7: 336– 69. Google Scholar CrossRef Search ADS PubMed Hodge A( 2010) Roots: the acquisition of water and nutrients from the heterogeneous soil environment. Prog Bot 71: 307– 37. Jonckheere I Fleck S Nackaerts Ket al. . ( 2004) Review of methods for in situ leaf area index determination: Part I. Theories, sensors and hemispherical photography. Agric For Meteorol 121: 19– 35. Google Scholar CrossRef Search ADS Knevel IC Bekker RM Kunzmann Det al. . ( 2005) The LEDA Traitbase collecting and measuring standards of life-history traits of the NW European flora. J Ecol 96: 1266– 74. Koide RT Robichaux RH Morse SRet al. . ( 1989) Plant water status, hydraulic resistance and capacitance. In Pearcy RW Ehleringer JR Mooney HAet al. (eds). Plant Physiological Ecology. Field Methods and Instrumentation . London, New York: Chapman and Hall, 161–84. Konaté NM( 2010) Diversité Interspécifique D’efficience D’utilisation de L’eau des Acacias Sahéliens et Australiens . Nancy, France: Université Henri Poincaré. Kool MTN Lenssen EF( 1997) Basal-shoot formation in young rose plants: effects of bending practices and plant density. J Hortic Sci 72: 635– 44. Google Scholar CrossRef Search ADS Larcher W. 1995. Physiological Plant Ecology . Berlin: Springer. Lavorel S Garnier E( 2002) Predicting changes in community composition and ecosystem function from plant traits-revising the holy grail. Funct Ecol 16: 545– 56. Google Scholar CrossRef Search ADS Lavorel S Mcintyre S Forbes TDA( 1997) Plant functional classifications: from general groups to specific groups based on response to disturbance. Trends Ecol Evol 12: 474– 8. Google Scholar CrossRef Search ADS PubMed Logan K Brunsell N Jones ARet al. . ( 2010) Assessing spatiotemporal variability of drought in the U.S. central plains. J Arid Environ 74: 247– 55. Google Scholar CrossRef Search ADS Lohbeck M Lebrija-Trejos E Martínez-Ramos Met al. ( 2015) Functional trait strategies of trees in dry and wet tropical forests are similar but differ in their consequences for succession. PLOS ONE 10: e0123741. Google Scholar CrossRef Search ADS PubMed Loranger J Shipley B( 2010) Interspecific covariation between stomatal density and other functional leaf traits in a local flora. Botany 88: 30– 8. Google Scholar CrossRef Search ADS Maes WH Achten WMJ Reubens Bet al. . ( 2009) Plant-water relationships and growth strategies of Jatropha curcas L. seedlings under different levels of drought stress. J Arid Environ 73: 877– 84. Google Scholar CrossRef Search ADS Markesteijn L( 2010) Drought Tolerance of Tropical Tree Species: Functional Traits, Trade-offs and Species Distribution . Wageningen, The Netherlands: Wageningen University. Marshall JG Dumbroff EB( 1999) Turgor regulation via cell wall adjustment in white spruce. Plant Physiol 119: 313– 20. Google Scholar CrossRef Search ADS PubMed Nicholson S( 2005) On the question of the ‘recovery’ of the rains in the West African Sahel. J Arid Environ 63: 615– 41. Google Scholar CrossRef Search ADS Niinemets U( 2001) Global-scale climatic controls of leaf dry mass per area, density, and thickness in trees and shrubs. Ecol Monogr 82: 453– 69. Ogaya R Peñuelas J( 2007) Leaf mass per area ratio in Quercus ilex leaves under a wide range of climatic conditions. The importance of low temperatures. Acta Oecologica 31: 168– 73. Google Scholar CrossRef Search ADS Ogle K Reynolds JF( 2004) Plant responses to precipitation in desert ecosystems: integrating functional types, pulses, thresholds and delays. Oecologia 141: 282– 94. Google Scholar CrossRef Search ADS PubMed Orwa C Mutua A Kindt Ret al. . ( 2009) Agroforestree Database: A Tree Reference and Selection Guide Version 4.0 . Kenya: World Agroforestry Centre. Otieno DO Schmidt MWT Kinyamario JIet al. . ( 2005) Responses of Acacia tortilis and Acacia xanthophloea to seasonal changes in soil water availability in the savanna region of Kenya. J Arid Environ 62: 377– 400. Google Scholar CrossRef Search ADS Pereira JS David JS David TSet al. . ( 2004) Carbon and water fluxes in Mediterranean-type ecosystems-constraints and adaptations. In Esser K Luttge U Beyschlag Wet al. . (eds). Progress in Botany , 65th edn. Berlin, Germany: Springer-Verlag, 467– 98. Google Scholar CrossRef Search ADS Pérez-Harguindeguy N Díaz S Garnier Eet al. . ( 2013) New handbook for standardised measurement of plant functional traits worldwide. Aust J Bot 61: 167– 234. Google Scholar CrossRef Search ADS Petit JR Jouzel J Raynaud Det al. . ( 1999) Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399:429–36. Piot J Diaite I( 1983) Systemes de Production D’élevage au Sénégal—Etude du Couvert Ligneux . Paris, France: GERDAT. Poorter H Garnier E( 2007) Ecological significance of inherent variation in relative growth rate and its components. In Pugnaire FI Valladares F (eds). Handbook of Functional Plant Ecology . New York: Marcel Dek, 82– 114. Poorter H Niinemets U Walter Aet al. . ( 2010) A method to construct dose–response curves for a wide range of environmental factors and plant traits by means of a meta-analysis of phenotypic data. J Exp Bot 8: 2043– 55. Google Scholar CrossRef Search ADS Pugnaire F Chapin FSIII Hardig M( 2006) Evolutionary changes in correlations among functional traits in Ceanothus in response to Mediterranean conditions. Web Ecol 6: 17− 26. Google Scholar CrossRef Search ADS Quero JL Sterck FJ Villar Ret al. . ( 2011) Water use strategies of six co-existing Mediterranean woody species during a summer drought. Oecologia 166: 45– 57. Google Scholar CrossRef Search ADS PubMed Reich PB( 2014) The world-wide ‘fast-slow’ plant economics spectrum: a traits manifesto. J Ecol 102: 275– 301. Google Scholar CrossRef Search ADS Reich PB Ellsworth DS Walters MBet al. . ( 1999) Generality of leaf traits relationships: a test across six biomes. Ecology 80: 1955– 69. Google Scholar CrossRef Search ADS Roupsard O Ferhi A Granier Aet al. . ( 1998) Fonctionnement hydrique et profondeur de prélèvement de l’eau de Faidherbia albida dans un parc agroforestier soudanien. In Campa C Grignon C Gueye Met al. . (eds). L’Acacia au Sénégal . Paris, France: ORSTOM, 81– 104. Ruiz-Robleto J Villar R( 2005) Relative growth rate and biomass allocation in ten woody species with different leaf longevity using phylogenetic independent contrasts (PICs). Plant Biol 7: 484– 94. Google Scholar CrossRef Search ADS PubMed Sagna P Ndiaye O Diop Cet al. . ( 2015) Are recent climate variations observed in Senegal in conformity with the descriptions given by the IPCC scenarios? Pollution Atmosphérique 227. Sobrado MA( 1986) Aspects of tissue water relations and seasonnal changes of leaf water potential components of evergreen and deciduous species coexisting in tropical dry forests. Oecologia 68: 413– 6. Google Scholar CrossRef Search ADS PubMed Soumaré A Groot JJR Koné Det al. . ( 1994) Structure spatiale du système racinaire de deux arbres du Sahel: Acacia sayel and Sclerocarya birrea. Rapport PSS No. 5. Wageningen, The Netherlands. Turner NC Spurway RA Schulze ED( 1984) Comparison of water potentials measured by in situ psychrometry and pressure chamber in morphologically different species. Plant Physiol 74: 316– 9. Google Scholar CrossRef Search ADS PubMed Valladares F Sánchez-Gómez D( 2006) Ecophysiological traits associated with drought in mediterranean tree seedlings: individual responses versus interspecific trends in eleven species. Plant Biol 8: 688– 97. Google Scholar CrossRef Search ADS PubMed Valladares F Vilagrosa A Peñuelas Jet al. . ( 2004) Estrés hídrico: ecofisiología y escalas de la sequía. In Valladares F (ed). Ecología del Bosque Mediterráneo en un Mundo Cambiante . Madrid, España: Ministerio de Medio Ambiente, EGRAF, S.A, 165–92. Vendramini F Díaz S Gurvich DEet al. . ( 2002) Leaf traits as indicators of resource-use strategy in floras with succulent species. New Phytol 154: 147– 57. Google Scholar CrossRef Search ADS Villar-Salvador P Penuelas JL Jacobs DF( 2013) Nitrogen nutrition and drought hardening exert opposite effects on the stress tolerance of Pinus pinea L. seedlings. Tree Physiol 33: 221– 32. Google Scholar CrossRef Search ADS PubMed Wade CT( 1997) Inventaire Floristique dans la Zone Écologique Littoral Nord (Tare Tound Maleye): Analyse de la Dégradation, Probleme de Conservation . Senegal: Université Gaston Berger. Westoby M Falster DS Moles ATet al. . ( 2002) Plant ecological strategies: some leading dimensions of variation between species. Annu Rev Ecol Syst 33: 125– 59. Google Scholar CrossRef Search ADS Wickens GE( 1983) The baobab: Africa’s upside-down tree. Kew Bull 37: 173– 209. Google Scholar CrossRef Search ADS William AG Thomas GW( 1986) Premature Leaf abscission: an induced plant defense against gall aphids. Ecology 67: 1619– 27. Google Scholar CrossRef Search ADS Williams RJ Myers BA Muller WJet al. . ( 1997) Leaf phenology of woody species in a North Australian tropical savanna. Ecol Monogr 78: 2542– 58. Wilson BA Russell-Smith J Williams RJ( 1996) Terrestrial vegetation. In Finlayson CM Oertzen IV (eds). Landscape and Vegetation of the Kakadu Region . Dordrecht, The Netherlands: Kluwer Academic, 57– 79. Google Scholar CrossRef Search ADS Wright IJ Reich PB Cornelissen JHCet al. . ( 2005) Assessing the generality of global leaf trait relationships. New Phytol 166: 485– 96. Google Scholar CrossRef Search ADS PubMed Wright IJ Reich PB Westoby Met al. . ( 2004) The worldwide leaf economics spectrum. Nature 428: 821– 7. Google Scholar CrossRef Search ADS PubMed Wright IJ Westoby M( 2001) Understanding seedling growth relationships through specific leaf area and leaf nitrogen concentration: generalisations across growth forms and growth irradiance. Oecologia 127: 21– 9. Google Scholar CrossRef Search ADS PubMed © The Author(s) 2017. Published by Oxford University Press on behalf of the Institute of Botany, Chinese Academy of Sciences and the Botanical Society of China. All rights reserved. For permissions, please email: firstname.lastname@example.org
Journal of Plant Ecology – Oxford University Press
Published: Jun 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera