Plant Ecol (2017) 218:1009–1020 DOI 10.1007/s11258-017-0749-3

Tracking the impact of drought on functionally different woody plants in a Mediterranean scrubland ecosystem Antonio Gazol

. Gabriel Sangu¨esa-Barreda . Elena Granda . J. Julio Camarero

Received: 31 October 2016 / Accepted: 3 July 2017 / Published online: 21 July 2017 Ó Springer Science+Business Media B.V. 2017

Abstract Climate warming is predicted to amplify drought stress. Thus, it is important to understand how coexisting plant species respond to severe droughts. Here we study how seven Mediterranean woody plant species with different evolutionary history and functional characteristics (Pinus halepensis Mill., Juniperus phoenicea L., Pistacia lentiscus L., Rhamnus lycioides L., Rosmarinus officinalis L., Genista scorpius (L.) DC., and Globularia alypum L.) responded to a severe winter drought during 2011–2012 in Spain. The study site is located in the Valcuerna valley, Monegros desert, northeastern Spain. We evaluated how the drought affected the annual growth-ring formation of the species by using dendrochronology and quantified the intensity of drought-induced defoliation and mortality and compared it between species and groups of species with different evolutionary history. Radial growth of all species was strongly reduced by the 2012 drought. The pre-Mediterranean

species (P. halepensis, J. phoenicea, P. lentiscus and R. lycioides) reduced growth more than the Mediterranean species (R. officinalis, G. scorpius and G. alypum). Defoliation was significantly higher in preMediterranean than in Mediterranean species. When species were analyzed separately we found that P. halepensis was the species with the highest growth reduction but J. phoenicea was defoliated more severely and showed higher mortality rates as a consequence of drought. In the case of the Mediterranean shrubs, drought-induced mortality was only noticeable in R. officinalis. Drought impacted growth of all species but this did not induce mortality in all of them. Growth reduction was dependent on evolutionary history. However, functional characteristics of the species such as leaf stomatal regulation and root architecture may be more important than evolutionary history on explaining drought-induced mortality. Indeed, species with shallow root systems such as J. phoenicea and R. officinalis were the most adversely affected by the drought.

Communicated by Jan Wunder.

Electronic supplementary material The online version of this article (doi:10.1007/s11258-017-0749-3) contains supplementary material, which is available to authorized users. A. Gazol (&)  G. Sangu¨esa-Barreda  E. Granda  J. J. Camarero Instituto Pirenaico de Ecologı´a, Consejo Superior de Investigaciones Cientı´ficas (IPE-CSIC), Avda. Montan˜ana 1005, 50192 Saragossa, Spain e-mail: [email protected]

Keywords Annual growth rings  Defoliation  Dendrochronology  Growth decline  Mortality  Shrub

Introduction The Mediterranean region represents a climatic transition between arid and temperate regions, which is

123

1010

regarded as a climate-change hot-spot given its high year-to-year variability in precipitation and the ubiquity of seasonal drought (Giorgi 2006). Interestingly, climate warming could amplify the intensity and frequency of severe droughts in the Mediterranean Basin threatening some plant species (Giorgi and Lionello 2008). In this sense, die-off and mortality events as a consequence of severe water shortage have emerged as common phenomena in many droughtprone ecosystems of the world including Mediterranean plant ecosystems (Allen et al. 2010; Hartmann 2011; Anderegg et al. 2013a, b, 2015). These mortality events may have consequences at the ecosystem level, modifying species composition by favoring some species against others (Granda et al. 2014a; Valladares et al. 2015). Nevertheless, the response of scrubland ecosystems to drought remains a poorly explored field in the Mediterranean region (Doblas-Miranda et al. 2015; but see Moreno-Gutie´rrez et al. 2012). In this study, we apply a dendroecological approach to better understand how woody plant species belonging to different families that coexist in a Mediterranean scrubland respond to a severe drought that occurred in the winter 2011–2012. We argue that such approach allows a better understanding on how these ecosystems respond to drought. In Mediterranean ecosystems, woody species present different traits to cope with drought mainly based on leaf, root and wood-anatomical characteristics (Pen˜uelas et al. 2001; Filella and Pen˜uelas 2003; Quero et al. 2011; Moreno-Gutierrez et al. 2012). For example, Filella and Pen˜uelas (2003) found different root patterns in pre-Mediterranean species, those that evolved before the onset of the dry-summer Mediterranean climate like Pistacia lentiscus L., as compared to other Mediterranean species such as Globularia alypum L. which evolved when Mediterranean climate conditions developed (Verdu´ et al. 2003). Miranda et al. (2010) found that leaf shedding is a common strategy of some Mediterranean species to xylem embolism and low levels of water potential. These evolutionary and functional differences translate into different water-use strategies. Some pre-Mediterranean species have deep roots that allow them to obtain water from deep soil sources (excluding some exceptions such as J. phoenicea), whereas Mediterranean species display higher net photosynthetic rates and stomatal conductance that allow them to respond to water availability more rapidly (Filella and

123

Plant Ecol (2017) 218:1009–1020

Pen˜uelas 2003; Moreno-Gutie´rrez et al. 2012). Baquedano and Castillo (2007) investigated drought tolerance of two evergreen oak species (Quercus coccifera L. and Quercus ilex L.), Pinus halepensis Mill. and Juniperus phoenicea L. They concluded that the water potential of P. halepensis was mainly dependent on precipitation. According to Chirino et al. (2011) P. halepensis is a species that shows a water-saver mechanism by exhibiting a strong stomatal regulation to reduce water loss. All in all, Mediterranean species present different strategies to avoid drought (Vilagrosa et al. 2012; Altieri et al. 2015) but regarding shrub species little is known on how these strategies impact radial growth. Dendrochronology serves as a tool quantifying age and to reconstruct radial growth in woody plants through the precise annual dating and measurement of ring variables as width (Fritts 1976). In droughtlimited environments, the effect of drought on plant growth can be retrospectively quantified due to the formation of very narrow or missing rings and the presence of declining growth trends following droughts (Sarris et al. 2007; Novak et al. 2011; Camarero et al. 2015). Similarly, dendrochronology has been applied to reconstruct growth patterns and the response to climate of shrubs and dwarf shrubs from different regions of the world (Bu¨ntgen et al. 2014; Lu et al. 2015). Although radial growth of Mediterranean shrubs is strongly determined by microclimatic conditions (Gazol and Camarero 2012a), annual growth rings can be identified and cross-dated precisely (Copenheaver et al. 2010; Gazol and Camarero 2012b; Battipaglia et al. 2014; Zimowski et al. 2014). In this sense, Copenheaver et al. (2010) showed that the Mediterranean Arbutus unedo L. presented very low growth and missing rings as a consequence of severe drought events. However, it is also true that shrubs and dwarf shrubs present difficulties to be cross-dated due to their short life-span, the eccentric radial growth of stems, and the presence of numerous missing rings (Buras and Wilmking 2014). There is a strong potential that Mediterranean shrubs can be used to test the influence of drought events in growth as trees are (cf. Cherubini et al. 2003), but this research line is largely unexplored. A community-based perspective considering not only trees but other dominant woody vegetation as well may help to understand how different plant species respond to drought (Urli et al. 2013; Hartmann et al. 2015).

Plant Ecol (2017) 218:1009–1020

In the winter 2011–2012 an exceptionally extreme drought event occurred in the Iberian Peninsula (Trigo et al. 2013). The Monegros desert is a steppe ecosystem subjected to semiarid conditions and located in Northern Spain that was strongly affected by the 2012 drought. For example, drought caused the mortality of around 50% of the individuals of several J. phoenicea populations (Lloret and Garcia 2016). Nevertheless, no mortality events were observed in nearby P. halepensis stands which constitute the dominant tree species in this area (Braun-Blanquet and Bolo`s 1957). In order to get a better understanding of why some woody plant species were affected by drought, while others presented no conspicuous negative effects (defoliation, shoot dieback, mortality) we decided to use a dendrochronological approach. We studied the response to drought of seven coexisting trees and shrubs, namely: P. halepensis, J. phoenicea, P. lentiscus, Rhamnus lycioides L., Rosmarinus officinalis L., Genista scorpius L. (DC.), and G. alypum. Here we aimed to quantify how the 2012 severe drought event influenced the radial growth of these coexisting woody plants. In addition, we investigated why the J. phoenicea presented the highest mortality and defoliation rates. We hypothesize that structural and functional characteristics in the root system (e.g., dependence on accessible soil water in shallow soil layer in the case of J. phoenicea) may be more important than evolutionary history, on explaining why some species responded more negatively to drought than others.

1011

Materials and methods

(Supporting Information, Fig. S1). The warmest months are July and August (26.0 °C), and the coldest months are December and January (7.0 °C). The driest months are July (13 mm) and February (14 mm), whereas the wettest months are May (39 mm) and October (42 mm). According to the Thornthwaite method, the water balance (difference between precipitation and potential evapotranspiration) is negative from February to October being maximum in July (-149 mm) when summer drought peaks (Appendix S1). The years 2005 and 2012 were characterized by extreme droughts (see Appendix S2). In the study area soils are basic and develop over limestone and gypsum substrates. The content in organic matter of these soils is low (see Appendix S3). To obtain long and reliable series of monthly climate data (mean temperature, precipitation) for the 1951–2014 period, we used the E-OBS gridded climate database (Haylock et al. 2008). Data were selected from the 0.258 grid delimited by the following coordinates: 41°000 -41°250 N, 0°000 -0°250 W. In order to quantify drought severity, we calculated the standardized precipitation-evapotranspiration index (SPEI) using the aforementioned gridded climate data and the SPEI package (Beguerı´a and Vicente-Serrano 2013) in the R statistical language (R Development Core Team 2015). The SPEI is a multiscalar drought index which is sensitive to changes in temperature (evapotranspiration) and precipitation with positive and negative values indicating wet and dry conditions, respectively (Vicente-Serrano et al. 2010). According to the SPEI the 2005 and 2012 droughts were among the most severe and longest in the study area during the 1950–2014 period (Appendix S2).

Study site

Study species

The study site is located in the Valcuerna valley, Monegros desert, northeastern Spain (41°250 2800 N; 0°50 2500 E; Fig. 1). The Valcuerna valley runs from the Northeast to the Southwest. The elevation goes from 200 m a.s.l. in the bottom of the valley to 350 m a.s.l. on the Northwest oriented slopes. Monegros is a semiarid steppe characterized by cold winters and a severe summer drought. According to data from the Caspe meteorological station (418140 1100 N, 0 00 0802 15 W, 149 m a.s.l.), mean annual temperature is 15.9 °C (period 1972–2012) and total annual precipitation is 313 mm (period 1952–2012)

The vegetation in the Valcuerna sites is dominated by scattered trees (P. halepensis and Q. ilex) and tall shrubs (J. phoenicea, P. lentiscus, Q. coccifera, R. lycioides), shrubs (G. scorpius, R. officinalis, Thymus vulgaris L., G. alypum, Helianthemum myrtifolium (Lam.) Samp, Lithospermum fruticosum L.) and grasslands dominated by Lygeum spartum (L.) Kunth. There are strong differences between the south and north slopes and across the slope. Since south-facing slopes have poor soils and a very low vegetation cover we avoided sampling those sites. In the north-oriented slope, lowlands are dominated by sclerophyllous

123

1012

Plant Ecol (2017) 218:1009–1020

1

2

3

1 km

Fig. 1 Study site (symbols show the three sampled locations) in the Valcuerna valley, situation in Europe and example of J. phoenicea defoliation. Numbers indicate the three study sites

scrublands, the abundance of J. phoenicea increases with elevation and the top of the slope is dominated by P. halepensis natural stands. Sampling protocol In 2013, after the extreme 2012 drought event, we collected wood samples of P. halepensis and J. phoenicea. We randomly selected 20 mature P. halepensis trees to quantify their growth trends and two radial cores were extracted at 1.3 m using a Pressler increment borer. Similarly, we selected 20 mature J. phoenicea shrubs and cut a basal section of stem wood from each individual. To compare the growth trends and response to drought of P. halepensis and J. phoenicea we selected five shrub species with recognizable annual growth rings: P. lentiscus (Pistaceae); R. lycioides (Rhamnaceae); R. officinalis (Labiatae); G. scorpius (Fabaceae) and G. alypum (Globulariaceae). These species coexist and are characteristic of the vegetation in the region (Braun-Blanquet and Bolo`s 1957; Bolo`s 1973). In 2014, to estimate mortality rates and to measure structural characteristics of each species (height) we located six square plots (10 m 9 10 m) in three different locations along the valley (two plots per location). We found no evident signs of change in J.

123

phoenicea mortality rates along the altitudinal gradient. Thus, we decided to locate the study plots in the middle part of the slope to increase the number of coexisting shrub species despite we lost P. halepensis. In each plot, we measured the height of each individual of the selected species (including J. phoenicea) using tapes, and assessed the degree of defoliation by counting the number of completely defoliated and non-defoliated stems of each individual. Thus, those stems presenting at least few living leaves were considered as non-defoliated stems. Defoliation was assessed during late spring and early summer. For each individual, we calculated the percentage of stems completely defoliated. Dead plants were completely defoliated in 2014. Basal sections of stem wood samples (except for J. phoenicea) from several individuals were obtained in the field and they were stored and processed in the laboratory. For each selected species, we sampled several individuals of different heights and presenting several defoliation degrees. Dendrochronological analyses In the laboratory, wood samples were carefully sanded until rings were clearly visible. The samples were observed under a magnifying binocular to perform a

Plant Ecol (2017) 218:1009–1020

1013

precise visual cross-dating of annual rings (Fritts 1976). Samples without pith or with very eccentric rings were discarded. The annual rings of the sections were then measured with a 0.001 mm resolution up to the pith using a Lintab-TSAP measuring device (Rinn 2005). This was done for two radii per section or tree. Since some shrubs tend to produce very eccentric stems, which difficult their cross-dating, we followed our own experience with other Mediterranean shrubs which shows that two radii, when carefully selected, can be enough to cross-date the shrubs and to adequately quantify their growth changes (Gazol and Camarero 2012b). The visual cross-dating was checked using the program COFECHA (Holmes 1983), which calculates correlations between ringwidth series and a master chronology for a common period (Table 1). To study the differences in growth patterns between the studied species tree-ring width measurements were transformed to basal area increment (hereafter BAI). The BAI represents a more accurate indicator of growth than ring-width, if we assume an approximately concentric growth of stems, since BAI reduces the variation caused by adding volume to a circular stem (Biondi and Qaedan 2008). We used the following formula:
2 BAI ¼ p rt2  rt1 ; ð1Þ where rt and rt-1 are the radii at time t and t - 1, respectively. Finally, we calculated a mean BAI series for each individual tree.

To obtain ring-width series of each individual and species (species chronology) the cross-dated ringwidth series were standardized and detrended using the dplR library (Bunn 2010) in the R statistical language (R Development Core Team 2015). First, we fitted a spline through the ring-width series of each measured radius. We used a spline where the frequency response is 0.5 at a wavelength of 0.67 multiplied by the length of the series in years (Bunn 2010). Then, the ring-width data were divided by the fitted lines to obtain residual ring-width indices (RWI) that were subjected to autoregressive modelling (pre-whitening) and then averaged for each year using a biweight robust estimation of the mean (Fritts 1976). Calculating RWIs using subtraction instead of division rendered similar results as those here presented. Thus, we obtained mean residual chronologies for each species of prewhitened growth indices. To quantify the influence of drought on growth of each species we calculated the difference in growth between the year of drought and the three preceding years. Specifically, we subtracted the growth of the drought year from the growth in the preceding years and divided it by the growth in the preceding years. For each individual of each species, we calculated the average ring-width index (RWI) for 2009–2011 and 2012. We used detrended data (RWI) instead of BAI in order to remove age and size effects as well as potential bias due to the eccentric growth of some shrub species.

Table 1 Number of individuals of the species found in the plots, mean defoliation (%), height, and estimated age obtained by counting the rings of selected individuals are shown. Defoliation, height, and age data are mean ± SE (standard error) Species

Evolutionary historyb

Growth form

No. individuals

Mean defoliation

Mean height (cm)

Age (years)

R. officinalis G. scorpius

MED MED

Shrub Shrub

240 68

0.45 ± 0.03 0.17 ± 0.03

77.37 ± 1.82 54.21 ± 2.25

12 ± 0.94 14 ± 1.94

G. alypum

MED

Shrub

37

0.11 ± 0.04

59.65 ± 3.09

16 ± 1.73

R. lycioides

pre-MED

Tall shrub

62

0.29 ± 0.04

110.71 ± 6.64

19 ± 3.05

P. lentiscus

pre-MED

Tall shrub

4

0.08 ± 0.08

131.00 ± 15.84

12 ± 3.04a

J. phoenicea

pre-MED

Tall shrub

76

0.66 ± 0.05

153.36 ± 11.21

77 ± 3.84a

P. halepensis

pre-MED

Tree

–

–

–

68 ± 2.43a

a

Data obtained from individuals located nearby but outside the sampled plots. In case of P. lentiscus and J. phoenicea some individuals were sampled outside of the plots b

MED, Mediterranean species; pre-MED, pre-Mediterranean species. Classification according to Filella and Pen˜uelas (2003) and Moreno-Gutierrez et al. (2012)

123

1014

Plant Ecol (2017) 218:1009–1020

Table 2 Growth statistics of the studied species in the Valcuerna valley, Monegros (Spain) Period

No. individuals (no. radii)

Mean ring-width (mm)

Mean correlation between individuals

R. officinalis

1998–2014

43 (88)

0.775

0.38

G. scorpius

2000–2014

12 (16)

0.359

0.48

G. alypum R. lycioides

1986–2014 1987–2014

20 (40) 20 (40)

0.369 0.513

0.52 0.49

P. lentiscus

1965–2014

14 (28)

0.672

0.46

J. phoenicea

1940–2014

18 (36)

0.336

0.44

P. halepensis

1940–2014

20 (35)

0.359

0.55

The number of sampled individuals and related growth statistics are shown

We correlated the mean residual chronology of each species with the temperature and precipitation from September of the previous year to September of the current year using the Pearson correlation coefficient. We used local monthly climatic data from the Caspe climatic station (mean temperature, total precipitation). The length of the correlation series were determined by the length of the species chronology (see Table 2). Statistical analyses We compared the size, defoliation, age and growth response to drought of the different species by means of the Kruskal–Wallis rank sum test (Sokal and Rohlf 1995). We used non-parametric tests instead of ANOVA because the different sample sizes observed between groups resulted in unequal variances (i.e., heteroscedasticity). When significant differences were observed, we run multiple comparisons using the Dunn’s-test (Dunn 1964). The Bonferroni adjustment was applied to correct significance (P) values for multiple comparisons. The Kruskal–Wallis rank sum test (Sokal and Rohlf 1995) was also applied to search for significant differences in size, defoliation, and growth response to drought between groups of species according to their evolutionary lineage (i.e., preMediterranean vs. Mediterranean species). The analyses were performed with the use of PMCMR package in R (Pohlert 2014).

Results The tall shrub and shrub species studied showed significant differences in size, and percentage of

123

defoliation (Tables 1, 3; Fig. 2; Appendix S4). Most species had a height lower than 1 m except for some individuals of R. lycioides and most individuals of P. lentiscus and J. phoenicea. Defoliation and mortality were low for most of the woody species considered except for R. officinalis and J. phoenicea. In the case of the last species more than 50% of the sampled individuals were dead. When species were grouped according to their evolutionary lineage we found that defoliation was significantly higher in pre-Mediterranean than in Mediterranean plant species (Kruskal–Wallis v2 = 23.88, df = 1, P \ 0.001). The species showed markedly different growth trends (Fig. 3). Small shrubs such as R. officinalis and G. scorpius presented a very strong growth during the first years of life and a very short life-span. Conversely, P. halepensis, J. phoenicea and P. lentiscus had a much lower growth rate and a larger life-span. The exception to this pattern was R. lycioides, a tall shrub that was characterized by relatively high growth rates and a short life-span. All species displayed a marked growth decline during the 2012 and 2005 droughts (Fig. 3b). The growth response to the 2012 drought event showed that P. halepensis was the species with the largest growth reduction due to drought followed by J. phoenicea (Fig. 4). The small shrubs showed lower sensitivity than large ones. When species were grouped according to their evolutionary lineage, we found that pre-Mediterranean species reduced growth after the 2012 drought significantly more than Mediterranean species (Kruskal–Wallis v2 = 22.44, df = 1, P \ 0.001). Most of the species responded positively to precipitation during the previous autumn and winter (Fig. 5).

Plant Ecol (2017) 218:1009–1020

1015

Table 3 Pairwise comparisons between the species size, defoliation, and growth response to drought Size

Defoliation

Drought-induced growth reduction

R. officinalis

b

b

ab

G. scorpius

a

a

a

G. alypum R. lycioides

a c

a ab

ab ab

P. lentiscus

–

–

bc

J. phoenicea

c

c

bc

P. halepensis

–

–

c

We applied the Dunn’s test for multiple comparisons of independent samples with Bonferroni adjustment of significance (P) values for multiple comparisons. Letters are used to indicate significant differences among groups of species

Fig. 2 Characteristics of the studied shrubby plant species. Histograms of shrub height (a), age (b) and percentage of defoliation (c) for the species studied are presented. For each species values are estimated with the data from the six plots studied. Dead individuals correspond to 100% of crown defoliation in 2014 (2 years after the drought). Species are colored according to their evolutionary lineage: pre-Mediterranean species (dark green, blue and light blue) and Mediterranean species (dark red, red and orange). (Color figure online)

In addition, G. alypum, R. lycioides, and J. phoenicea showed a significant (P \ 0.05) positive relationship to July precipitation. High temperatures during the late

Fig. 3 Radial growth for the study plant species presented either as a normalized basal area increment (BAI), i.e., BAI divided by its sum of values, or as b ring-width indices (RWI). Red arrows in plot (b) indicate the severe 2005 and 2012 droughts. Species are colored according to their evolutionary lineage: pre-Mediterranean species (black, dark green, blue and light blue) and Mediterranean species (dark red, red and orange). (Color figure online)

spring and summer had a significant negative influence on the growth of P. halepensis. We also detected significant negative correlations between temperature of previous autumn (September–November) and growth of R. lycioides and R. officianlis. Lastly, we found that the growth of J. phoenicea, was positively related to January temperature.

123

1016

Fig. 4 Drought-induced growth decline in the studied species. The figure shows the impact of 2012 drought on basal area increment for each species. Drought impact is defined as the growth before drought (2011) minus the growth during drought (2012) divided by the growth before drought (2011) and it is expressed as a percentage. The notches represent confidence intervals for the median. Species are colored according to their evolutionary lineage: pre-Mediterranean species (black) and Mediterranean species (light grey). (Color figure online)

Discussion Our results agreed with the previous observations of a strong defoliation in J. phoenicea in response to the 2012 winter drought (Lloret and Garcı´a 2016), and showed that R. officinalis had a quite strong defoliation rate followed by R. lycioides. However, the defoliation and mortality were very low or almost inexistent in shrubs such as G. alypum and G. scorpius, the two species which better tolerated drought in terms of growth (Figs. 2, 4). J. phoenicea, R. officinalis, and R. lycioides have a different lineage age but they might share common traits that made them more vulnerable to the 2011–2012 severe drought in terms of defoliation. For example, Moreno–Gutie´rrez et al. (2012) distinguished two types of strategies for woody plant species according to their water use in a semiarid scrubland located in SE Spain: (i) those showing a conservative water use with during the growing season (high water-use efficiency through low stomatal conductance) as P. halepensis, which were able to use deep ([40 cm) soil water sources; and (ii) those characterized by a ‘‘wasteful’’ water use (low wateruse efficiency, high stomatal conductance) which depended on shallow soil water resources (e.g., R. officinalis). Junipers are anisohydric species and

123

Plant Ecol (2017) 218:1009–1020

Fig. 5 Growth responses of the studied species to climate variables. For each species we show the marginally significant (P \ 0.10) correlations between monthly climate variables (a, mean temperature; b, total precipitation) and radial growth (ring-width indices). Blue colors indicate marginally significant positive correlations and red colors indicate marginally significant negative correlations. Months abbreviated by lowerand upper-case letters correspond to the previous and current years (the current year includes the period of tree-ring formation). (Color figure online)

maintain gas exchange and higher hydraulic conductance at a significantly lower leaf water potential than isohydric coexisting pine species, which keep constant midday leaf water potential by reducing stomatal conductance; this difference makes ‘‘risk-taking’’ junipers prone to hydraulic failure due to xylem cavitation when drought is intense (Tardieu and Simonneau 1998; Maseda and Fernandez 2006; Plaut et al. 2012). In fact, prolonged droughts and very warm conditions leading to extreme evapotranspiration rates cause juniper dieback and death as we showed here, while P. halepensis presented low defoliation (Fig. 2). Altieri et al. (2015) found that J. phoenicea was more vulnerable to hot and dry summers than P. lentiscus because the latter species displayed a tight stomatal regulation. However, P. lentiscus is more vulnerable than J. phoenicea to freezing during the early growing season due to a long extent of crown phenological activity (Palacio et al. 2005). Similar results were obtained by Baquedano and Castillo

Plant Ecol (2017) 218:1009–1020

(2007) when compared the response to drought of P. halepensis and J. phoenicea in northern Spain. It is plausible to think that the main factor threatening J. phoenicea is the increase in summer temperature rather than drought per se. Paula and Pausas (2011) found that seedlings of resprouter species (e.g., G. scorpius, P. lentiscus) tend to form thicker and poorly ramified roots in contrast to non-resprouting species such as R. officinalis. In this sense, elevated summer temperatures again negatively impact the performance of Mediterranean woody plant species with shallow roots by rising water evapotranspiration from the soil surface. The Mediterranean shrubs studied here were relatively short and presented very steep and short growth trends. Conversely, pre-Mediterranean species were taller and presented larger and less steep growth trends, excepting R. lycioides. Similar patterns of lower growth trends in Mediterranean shrubs than in coexisting tree species were found by Gazol and Camarero (2012b). Similarly, Eugenio et al. (2012) found that gypsophyte shrubs as Helianthemum squamatum and Lepidium subulatum had short life-spans. The short life-span of the shrubs studied here may contradict the idea that plants inhabiting stressful environments tend to present characteristic such as long life-spans due to low growth rates related to stressful environmental conditions (Garcı´a and Zamora 2003). Nevertheless, some authors have proposed that a strategy of a persistent seed bank together with a high reproduction cost and short-lived life cycle may pose an alternative adaptation to stressful environments (Arago´n et al. 2009). This confirms that, except for some remarkable exceptions showing low growth rates as mountain shrub species (often of the Brassicacea family) such as Hormathophylla spinosa (Gazol and Camarero 2012a) or ancient J. phoenicea individuals inhabiting cliffs (Mathaux et al. 2016), Mediterranean shrubs are not usually long-lived species. It should be further investigated if low growth rates are associated to long lifespans in Mediterranean shrub species. Despite the differences in defoliation, mortality, longevity, and growth patterns observed between the Mediterranean and pre-Mediterranean species we found a strong convergence towards a sharp growth reduction in response to water shortage. We found evidence indicating that secondary growth (here measured as width of annual growth rings) of the studied woody species was negatively affected by drought. The growth for the year 2012, characterized

1017

as the most extreme winter drought in the last half of century (Trigo et al. 2013), was much lower than the growth of the preceding years. A growth reduction was also observed in 2005, a very dry year in the region. Ring-width during the 2012 drought was on average 20% lower than that in the three preceding years for all species. These results reinforce the idea that annual growth rings of woody Mediterranean plants are good markers of extreme climate events (Gazol and Camarero 2012b). Somehow, the results also indicate that the growth of pre-Mediterranean species is more sensitive in magnitude to drought than that of the Mediterranean shrubs suggesting evolutionary adaptation in their growth response to drought. When species were analyzed separately we found that while the reduction in growth of P. halepensis and J. phoenicea is around 60–70%, that in the rest of shrubs and tall shrubs studied is in general lower than 40%. This difference in growth response can be explained by two reasons mainly. First, the meristems of shrubs are closer to the soil than the meristems of trees and tall shrubs and thus they might be less influenced by air temperature oscillations (Gazol and Camarero 2012b). Second, some pre-Mediterranean species as P. halepensis and P. lentiscus are mainly characterized by a deep and well-developed root system in contraposition with Mediterranean shrubs that tend to have a more superficial and opportunistic root system (Filella and Pen˜uelas 2003; MorenoGutie´rrez et al. 2012). For example, P. lentiscus is a species that is able to develop roots up to a depth of 5 m and that can obtain water from deep soils during spring (Filella and Pen˜uelas 2003), conversely the Mediterranean G. alypum tends to develop a shallow root system to make a rapid use of short-storm rainfall during summer (Filella and Pen˜uelas 2003). In the case of R. lycioides this shallow root system may be combined with the formation of deep and welldeveloped main roots (Terradas 1991). However, there are exceptions to this rule as is the case of J. phoenicea, a pre-Mediterranean species with a shallow root system. Overall, these results suggest that the differences in growth response to drought of the studied species may be partially explained by their distinct evolutionary history (Pen˜uelas et al. 2001; Filella and Pen˜uelas 2003; Moreno–Gutie´rrez et al. 2012). Our findings showed that the growth of the three species that presented higher rates of defoliation and mortality (J. phoenicea, R. officinalis and R. lycioides)

123

1018

was positively correlated with January temperature, despite this relationship was only significant in the case of J. phoenicea. The positive impact of winter temperatures on J. phoenicea and R. officinalis growth may indicate that these species start forming their xylem earlier and are more sensitive to air temperature during the early growing season. The published research on cambial activity in some Mediterranean shrubs shows that most secondary growth occurs from early to late spring (Liphschitz and Lev-Yadun 1986, Camarero et al. 2013). Furthermore, Camarero et al. (2013) hypothesized that shrubs from dry Mediterranean regions such as Linum suffruticosum and Lepidium subulatum finish most of their xylem development before summer drought. Growth responses to winter climate may also indicate increased winter photosynthesis and/or reduced respiration affecting spring growth or lagged growth responses to latewinter climate conditions. Winter stress can also play a role in these species’ responses since low temperature during this season constrain photosynthesis in some continental Mediterranean areas and cause xylem embolism (Granda et al. 2014b), albeit this is not expected under the relatively mild climate conditions of the site. Studies on cambial activity of most of the study Mediterranean shrub species are lacking, and those available indicate that growth should respond to climate conditions from the previous late winter until the early summer (Liphschitz and Lev-Yadun 1986; Camarero et al. 2013). In this sense, Cherubini et al. (2003) hypothesized that malacophyllous plant species from dry sites should stop their cambial activity in summer and thus they may be able to grow in winter. The results provided in this study should be interpreted with caution due to some limitations related with the visual cross-dating of shrubs. First, shrubs are younger than tree species and this complicates building robust and long ring-width chronologies. This is particularly true for species such as G. scorpius and R. officinalis which present very short life-spans. Second, we used BAI to analyze growth trends which assumes a concentric growth of the stem and may be a problematic variable for shrub species showing eccentric rings (Buras and Wilmking 2014). The short life-span and the potential growth eccentricity complicate building robust ring-width chronologies, and consequently drawing strong conclusions on the long-term influences of climate on shrub secondary growth (but see Lu et al. 2015).

123

Plant Ecol (2017) 218:1009–1020

Our study suggests that the responses to drought of seven coexisting woody plant species vary depending on the species’ evolutionary lineage, but that there are other factors that determine drought sensitivity. In this sense, the three species that presented higher rates of defoliation and mortality (J. phoenicea, R. officinalis, and R. lycioides) differ in their evolutionary origin but they share a common trait: they have a system of shallow roots to make an opportunistic use of water from the surface. Elevated temperatures may negatively impact the performance of Mediterranean woody plant species with shallow roots by rising water evapotranspiration from the soil surface. Our findings indicate that if dry spells become more severe and frequent, as expected, some species as J. phoenicea and R. officinalis, could likely become less abundant in favor of species with a deeper root system such as R. lycioides and P. lentiscus. However, the population dynamics of shrub species with short life-span and high drought-triggered mortality and defoliation rates, as for instance R. officinalis, may be less influenced by drought if the drought return period is shorter than the time needed for the species to produce seeds. Further studies considering a whole set of leaf, stem, and root functional traits for these coexisting species are required to draw strong conclusions. It must be also acknowledged that studying an individual response to drought based on individual characteristics may help to understand the species-specific responses to drought. For example, a consideration of intraspecific variation in size, age and growth response to drought as well as the relationships between these variables could elucidate which individuals are more vulnerable to drought. Acknowledgements We sincerely thank Prof. F. Lloret for guiding us through the interesting Valcuerna site and for alerting us on the juniper die-off. A. Gazol is supported by a Postdoctoral grant from Spanish Ministry of Economy and Competitiveness (Contrato Formacio´n Postdoctoral MINECO, FPDI 2013-16600). We acknowledge the funding provided by the FUNDIVER project (CGL2015-69186-C2-1-R, Spanish Ministry of Economy and Competitiveness and FEDER funds).

References Allen CD, Macalady A, Chenchouni H et al (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. For Ecol Manag 259:660–684

Plant Ecol (2017) 218:1009–1020 Altieri S, Mereu S, Cherubini P, Castaldi S, Sirignano C, Lubritto C, Battipaglia G (2015) Tree-ring carbon and oxygen isotopes indicate different water use strategies in three Mediterranean shrubs at Capo Caccia (Sardinia, Italy). Trees 29:1593–1603 Anderegg WRL, Kane JM, Anderegg LDL (2013a) Consequences of widespread tree mortality triggered by drought and temperature stress. Nat Clim Change 3:30–36 Anderegg WRL, Plavcova L, Anderegg LDL, Hacke UG, Berry JA, Field CB (2013b) Drought’s legacy: multiyear hydraulic deterioration underlies widespread aspen forest die-off and portends increased future risk. Glob Change Biol 19:1188–1196 Anderegg WRL, Schwalm C, Biondi F, Camarero JJ, Koch G, Litvak M, Ogle K, Shaw JD, Shevliakova E, Williams AP, Wolf A, Ziaco E, Pacala S (2015) Pervasive drought legacies in forest ecosystems and their implications for carbon cycle models. Science 349:528–532 Arago´n CF, Me´ndez M, Escudero A (2009) Survival costs of reproduction in a short-lived perennial plant: live hard, die young. Am J Bot 96:904–911 Baquedano FJ, Castillo FJ (2007) Drought tolerance in the Mediterranean species Quercus coccifera, Quercus ilex, Pinus halepensis and Juniperus phoenicea. Photosynthetica 45:229–238 Battipaglia G, De Micco V, Brand WA, Saurer M, Aronne G, Linke P, Cherubini P (2014) Drought impact on water use efficiency and intra-annual density fluctuations in Erica arborea on Elba (Italy). Plant Cell Environ 37:382–391 Beguerı´a S, Vicente-Serrano SM. (2013) SPEI: calculation of the standardised precipitation-evapotranspiration index. R package version 1.6. http://CRAN.R-project.org/package= SPEI Biondi F, Qaedan F (2008) A theory-driven approach to treering standardization: defining the biological trend from expected basal area increment. Tree-Ring Res 64:81–96 Bolo`s O (1973) La vegetacio´n de la Serreta Negra de Fraga. Mem Real Acad Cienc Artes Barcelona 6:269–313 Braun-Blanquet J, Bolo`s O (1957) Les groupements ve´ge´taux du bassin de l’Ebre et leur dynamisme. An Estacio´n Exp Aula Dei 5:1–266 Bunn AG (2010) Statistical and visual crossdating in R using the dplR library. Dendrochronologia 28:251–258 Bu¨ntgen U, Psomas A, Schweingruber FH (2014) Introducing wood anatomical and dendrochronological aspects of herbaceous plants: applications of the xylem database to vegetation science. J Veg Sci 25:967–977 Buras A, Wilmking M (2014) Straight lines or eccentric eggs? A comparison of radial and spatial ring width measurements and its implications for climate transfer functions. Dendrochronologia 32:313–326 Camarero JJ, Palacio S, Montserrat-Martı´ G (2013) Contrasting seasonal overlaps between primary and secondary growth are linked to wood anatomy in Mediterranean sub-shrubs. Plant Biol 15:798–807 Camarero JJ, Gazol A, Sangu¨esa-Barreda G, Oliva J, VicenteSerrano SM (2015) To die or not to die: early-warning signals of dieback in response to a severe drought. J Ecol 103:44–57 Cherubini P, Ga¨rtner BL, Tognetti R, Bra¨ker OU, Schoch W, Innes JL (2003) Identification, measurement and

1019 interpretation of tree rings in woody species from Mediterranean climates. Biol Rev 78:119–148 Chirino E, Bellot J, Sa´nchez JR (2011) Daily sap flow rate as an indicator of drought avoidance mechanisms in five Mediterranean perennial species in semi-arid southeastern Spain. Trees 25:593–606 Copenheaver CA, Ga¨rtner H, Scha¨fer I, Vaccari FP, Cherubini P (2010) Drought-triggered false ring formation in a Mediterranean shrub. Botany 88:545–555 ´ lvarez A, Doblas-Miranda E, Martı´nez-Vilalta J, Lloret F, A ´ vila A, Bonet FJ, Brotons L, Castro J, Curiel Yuste J, Dı´az A M, Ferrandis P, Garcı´a-Hurtado E, Iriondo JM, Keenan TF, Latron J, Llusia` J, Loepfe L, Mayol M, More´ G, Moya D, Pen˜uelas J, Pons X, Poyatos R, Sardans J, Sus O, Vallejo VR, Vayreda J, Retana J (2015) Reassessing global change research priorities in mediterranean terrestrial ecosystems: how far have we come and where do we go from here? Glob Ecol Biogeogr 24:25–43 Dunn OJ (1964) Multiple comparisons using rank sums. Technometrics 6:241–252 Eugenio M, Olano JM, Ferradis P, Martıı´nez-Duro E, Escudero A (2012) Population structure of two dominant gypsophyte shrubs through a secondary plant succession. J Arid Environ 76:30–35 Filella I, Pen˜uelas J (2003) Partitioning of water and nitrogen in co-occurring Mediterranean woody shrub species of different evolutionary history. Oecologia 137:51–61 Fritts HC (1976) Tree rings and climate. The Blackburn Press, Caldwell, p 584 Garcı´a D, Zamora R (2003) Persistence, multiple demographic strategies and conservation in long-lived Mediterranean plants. J Veg Sci 14:921–926 Gazol A, Camarero JJ (2012a) The performance of Mediterranean subshrubs depends more on microsite than on regional climate conditions. J Veg Sci 23:1062–1070 Gazol A, Camarero JJ (2012b) Mediterranean dwarf shrubs and coexisting trees present different radial-growth synchronies and responses to climate. Plant Ecol 213:1687–1698 Giorgi F (2006) Climate change hot-spots. Geophys Res Lett 33:L08707 Giorgi F, Lionello P (2008) Climate change projections for the Mediterranean region. Glob Planet Change 63:90–104 Granda E, Escudero A, Valladares F (2014a) More than just drought: complexity of recruitment patterns in Mediterranean forests. Oecologia 176:997–1007 Granda E, Scoffoni C, Rubio-Casal AE, Sack L, Valladares F (2014b) Leaf and stem physiological responses to summer and winter extremes of woody species across temperate ecosystems. Oikos 123:1281–1290 Hartmann H (2011) Will a 385 million year-struggle for light become a struggle for water and for carbon?—How trees may cope with more frequent climate change-type drought events. Glob Change Biol 17:642–655 Hartmann H, Adams HD, Anderegg WRL, Jansen S, Zeppel MJB (2015) Research frontiers in drought-induced tree mortality: crossing scales and disciplines. New Phytol 205:965–969 Haylock MR, Hofstra N, Klein Tank AMG, Klok EJ, Jones PD, New M (2008) A European daily high-resolution gridded dataset of surface temperature and precipitation. J Geophys Res 113:D20119

123

1020 Holmes RL (1983) Computer-assisted quality control in treering dating and measurement. Tree Ring Bull 43:69–78 Liphschitz N, Lev-Yadun S (1986) Cambial activity in evergreen and seasonal dimorphics around the Mediterranean. IAWA Bull 7:145–153 Lloret F, Garcı´a C (2016) Inbreeding and neighbouring vegetation drive drought-induced die-off within juniper populations. Funct Ecol 30:1696–1704 Lu X, Camarero JJ, Wang Y, Liang E, Eckstein D (2015) Up to 400-year-old Rhododendron shrubs on the southeastern Tibetan Plateau: prospects for shrub-based dendrochronology. Boreas 44:760–768 Maseda PH, Fernandez RJ (2006) Stay wet or else: three ways in which plants can adjust hydraulically to their environment. J Exp Bot 57:3963–3977 Mathaux C, Mandin J-P, Oberlin C, Edouard J-L, Gauquelin T, Guibal F (2016) Ancient juniper trees growing on cliffs: toward a long Mediterranean tree-ring chronology. Dendrochronologia 37:79–88 Miranda JD, Padilla FM, Martı´nez-Vilalta J, Pungaire FI (2010) Woody species of a semi-arid community are only moderately resistant to cavitation. Funct Plant Biol 37:828–839 Moreno-Gutie´rrez C, Dawson TE, Nicola´s E, Querejeta JI (2012) Isotopes reveal contrasting water use strategies among coexisting plant species in a Mediterranean ecosystem. New Phytol 196:489–496 Novak K, de Luis M, Cˇufar K, Ravento´s J (2011) Frequency and variability of missing tree rings along the stems of Pinus halepensis and Pinus pinea from a semiarid site in SE Spain. J Arid Environ 75:494–498 Palacio S, Milla R, Montserrat-Martı´ F (2005) A phenological hypothesis on the thermophilous distribution of Pistacia lentiscus L. Flora 6:527–534 Paula S, Pausas JG (2011) Root traits explain different foraging strategies between resprouting life histories. Oecologia 165:321–331 Pen˜uelas J, Lloret F, Montoya R (2001) Drought effects on Mediterranean vegetation and taxa evolutionary history. For Sci 47:214–218 Plaut JA, Yepez EA, Hill J, Pangle R, Sperry JS, Pockman WT, McDowell NG (2012) Hydraulic limits preceding mortality in a pin˜on–juniper woodland under experimental drought. Plant Cell Environ 35:1601–1617 Pohlert T (2014) The pairwise multiple comparison of mean ranks package (PMCMR). R package, http://CRAN.Rproject.org/package=PMCMR Quero JL, Sterck FJ, Martı´nez-Vilalta J, Villar R (2011) Wateruse strategies of six co-existing Mediterranean woody species during a summer drought. Oecologia 166:45–57

123

Plant Ecol (2017) 218:1009–1020 R Development Core Team (2015) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna Rinn F (2005) TSAP-WIN Time-series analysis and presentation for dendrochronology and related applications. Rinn Tech, Heidelberg Sarris D, Christodoulakis D, Ko¨rner C (2007) Recent decline in precipitation and tree growth in the eastern Mediterranean. Glob Change Biol 13:1187–1200 Sokal RR, Rohlf FJ (1995) Biometry: the principles and practice of statistics in biological research. Freeman, New York, p 887 Tardieu F, Simonneau T (1998) Variability among species of stomatal control under fluctuating soil water status and evaporative demand: modelling isohydric and anisohydric behaviours. J Exp Bot 49:419–432 Terradas J (1991) Mediterranean woody plant growth-forms, biomass and production in the eastern part of the Iberian Peninsula. Oecol Aquat 10:337–349 Trigo RM, An˜el J, Barriopedro D, Garcı´a-Herrera R, Gimeno L, Nieto R, Castillo R, Allen MR, Massey N (2013) The record winter drought of 2011–2012 in the Iberian Peninsula. In: Peterson TC, Hoerling MP, Stott PA, Herring S (eds) Explaining extreme events of 2012 from a climate perspective. American Meteorological Society, Boston, pp S41–S45 Urli M, Porte AJ, Cochard H, Guengant Y, Burlett R, Delzon S (2013) Xylem embolism threshold for catastrophic hydraulic failure in angiosperm trees. Tree Physiol 33:672–683 Valladares F, Bastias CC, Godoy O, Granda E, Escudero A (2015) Species coexistence in a changing world. Front Plant Sci 6:866 Verdu´ M, Da´vila P, Garcı´a-Fayos P, Flores-Herna´ndez N, Valiente-Banuet A (2003) ‘Convergent’ traits of mediterranean woody plants belong to pre-mediterranean lineages. Biol J Lin Soc 78:415–427 Vicente-Serrano SM, Beguerı´a S, Lo´pez-Moreno JI (2010) A multiscalar drought index sensitive to global warming: the standardized precipitation evapotranspiration index. J Clim 23:1696–1718 Vilagrosa A, Chirino E, Peguero-Pina JJ, Barigah TS, Cochard H, Gil-Pelegrı´n E (2012) Xylem cavitation and embolism in plants living in water-limited ecosystems. In: Aroca R (ed) Plant responses to drought stress. Springer, Berlin, pp 63–109 Zimowski M, Leuschner HH, Ga¨rtner H, Bergmeier E (2014) Age and diversity of Mediterranean dwarf shrublands: a dendrochronological approach along an altitudinal gradient on Crete. J Veg Sci 25:122–134