Forest Ecology and Management 320 (2014) 70–82

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Drought induced decline could portend widespread pine mortality at the xeric ecotone in managed mediterranean pine-oak woodlands Guillermo Gea-Izquierdo a,b,⇑, Bárbara Viguera a, Miguel Cabrera c, Isabel Cañellas a a b c

INIA-CIFOR, Crta. La Coruña km 7.5, 28040 Madrid, Spain CEREGE UMR 7330, CNRS/Aix-Marseille Université, Europole de l’Arbois, BP 8013545, Aix-en-Provence cedex 4, France Aranzada Gestión Forestal S.L.P., C/Alonso Heredia, 31, 28028 Madrid, Spain

a r t i c l e

i n f o

Article history: Received 24 October 2013 Received in revised form 12 February 2014 Accepted 19 February 2014

Keywords: Quercus pyrenaica Pinus sylvestris Global change Dendroecology Vulnerability to water stress

a b s t r a c t There is a need to better understand how different biotic and abiotic factors interact to determine climate change enhanced tree mortality. Here, we investigated whether rising water stress determined enhanced Pinus sylvestris L. mortality at the species low-elevation limit in Central Spain. We analyzed the factors determining the health status of pines and compared with co-occurring and more drought-tolerant Quercus pyrenaica Willd along one transect following an elevation gradient. We used ordinal logistic regression to model the susceptibility of a tree to decline in relation to variability in stand competition and individual growth-patterns. The mortality pattern differed with local site conditions. Pine growth was faster but life-span shorter at drier and warmer low-elevations than at high-elevations. However, within stands, healthy trees exhibited less abrupt growth reductions and higher growth-rates but not as a consequence of lower competition, which under present stand conditions did not seem to increase adult mortality risk. Low moisture availability reduced tree-growth and, although P. sylvestris is less tolerant to drought, Q. pyrenaica was more sensitive to year-to-year moisture variability. Previous growth of dead trees from both species declined with rising water stress after the 1970s at low-elevations, which suggests that water stress intensity limited particularly tree-growth of dead trees in the long-term. For pines, widespread symptoms of crown decline (expressed by mistletoe infestation and defoliation) were only observed at low-elevation stands where, in opposition to oaks, weakened and healthy pines also exhibited recent negative growth-trends parallel to those of dead trees. The pervasive growth decline with enhanced water stress in pines from all health status at the species sampled xeric ecotone combined with the abundant crown decline symptoms observed, suggest pine vulnerability and could portend widespread mortality at its current low-elevation limit. ! 2014 Elsevier B.V. All rights reserved.

1. Introduction Climate change related increases in drought frequency and severity have a negative impact on forest ecosystem productivity and tree performance (Boisvenue and Running, 2006; Lenoir et al., 2008; Choat et al., 2012). Therefore it is crucial to better describe the underlying processes governing forest acclimation to water stress (Bréda et al., 2006; Niinemets, 2010; McDowell et al., 2011) particularly at those ecosystems where sustainability is threatened by enhanced mortality caused by recent climatic changes (Adams et al., 2009; Van Mantgem et al., 2009; Allen et al., 2010). Accurate prediction of tree vulnerability to increasing ⇑ Corresponding author at: CEREGE UMR 7330, CNRS/Aix-Marseille Université, Europole de l’Arbois, BP 8013545, Aix-en-Provence cedex 4, France. Tel.: +33 (0) 442971532. E-mail addresses: [email protected], [email protected] (G. Gea-Izquierdo). http://dx.doi.org/10.1016/j.foreco.2014.02.025 0378-1127/! 2014 Elsevier B.V. All rights reserved.

water stress is challenging because many biotic and abiotic factors interact at different time scales and trees with symptoms of low vitality can recover from stress provided they do not fall below thresholds where irreversible damage occurs (Suárez et al., 2004; Dobbertin, 2005; Breshears et al., 2009). In addition, it can be particularly complex to isolate the long-term effect of climate at those sites where land-use has shaped the current state of forests and determine forest dynamics (Gimmi et al., 2010; Van Bogaert et al., 2011; Wischnewski et al., 2011). The interdependent factors that determine species-specific susceptibility to drought induced decline and mortality must be investigated at different temporal scales in order to understand forest vulnerability. Growth can be used as a direct proxy to the tree net carbon pool. Using dendrochronological methods it is possible to analyze long-term environmental stressors (like water stress) predisposing trees to decline and short-term agents or events inciting posterior death of individuals (Suárez et al., 2004;

71

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

McDowell et al., 2010; Linares and Camarero, 2012). The sizedependent mortality risk of trees increases under more resource stressful conditions when stand competition is intense (Zahner et al., 1989; Voelker et al., 2008; Luo and Chen, 2011) and within stands the trees with a higher likelihood to die are generally those with slow growth and negative growth-trends (Bigler and Bugmann, 2004; Suárez et al., 2004; Bigler et al., 2006). Plants co-regulate different functional traits to cope with water stress (Bréda et al., 2006; Niinemets, 2010; McDowell et al., 2011) and complementary functionalities between species can help them to optimize environmental resources in mixed stands (Cavard et al., 2011; Luo and Chen, 2011). A widespread example in the Northern Hemisphere is that of oak-pine forests (Brose and Waldrop, 2010; Wischnewski et al., 2011; Sheffer, 2012). There is variability in drought tolerance among oak and pine species. Often pines behave as isohydric species (i.e. plant maintain relatively constant leaf water potential through stomatal control and reduced conductance despite declining soil and root water potentials) whereas oaks as anisohydric, which means they developed different strategies to cope with water stress (Abrams, 1990; Eilmann et al., 2009). Submediterranean oak species are expected to endure water stress further than less drought tolerant boreal conifers such as P. sylvestris and there could exist already an ongoing shift in species composition in mountain oak-pine forests, e.g. in the Alps (Bigler et al., 2006; Gimmi et al., 2010; Rigling et al., 2013) and the Pyrenees (Galiano et al., 2010; Heres et al., 2012; Vilá-Cabrera et al., 2013). In this study we analyzed how climate and stand related factors determined the health status of two functionally different co-occurring species (Q. pyrenaica, a deciduous submediterranean oak, and P. sylvestris, an evergreen boreal pine) along an altitudinal gradient in Central Spain where pine stands at low-elevations seem to express symptoms of decline and enhanced mortality. The response to climate and the long-term growth-trends of dead, weakened and healthy trees were compared in relation to recent enhancement in water stress. The likelihood for a tree of being within a given health status was modeled to describe variability in the mortality pattern along the altitudinal gradient. By comparing the growth trends of trees with different health status we specifically analyzed whether the growth pattern of live trees resembled that of dead trees to analyze stand vulnerability to climate change. We hypothesize that the two species will exhibit different growth-trends in relation to recent water stress increase, particularly: (1) within stands, dead trees of both species will exhibit slower growth with recent negative trends; (2) weakened trees will show similar growth-trends to those of dead trees, which will portend near mortality; (3) the growth-trends of non-healthy trees at low elevations will match the recent regional increase in water stress, suggesting that long-term drought severity enhancement predisposed trees to decline only at xeric sites; (4) recent growth-trends of healthy trees resembling those of dead trees will express species-specific vulnerability to climate change.

2. Material and methods 2.1. Study site and data Trees were sampled at four locations in Central Spain along one transect following an altitudinal gradient of increasing rainfall and decreasing temperature with increasing altitude (Table 1). Low-elevation stands (i.e. Sites #0 and #1) were an open oak woodland with some clusters of adult pines at an altitude around 1075 m subjected to typical silvopastoral management in Mediterranean oak woodlands, i.e. thinning to convert open woodland, grazing livestock and copiccing for firewood. Clusters of adult trees were monospecific within stands: Site #0 corresponds to stands with adult pine overstory whereas Site #1 to stands with oak overstory. Low-elevation Site #2 (altitude around 1180 m) was dominated by pine with an irregularly mixed oak saplings understory. High-elevation Site #3 (altitude above 1400 m) was monospecific P. sylvestris forest. Sites #2 and #3 presented higher stand density but generally not full canopy cover. Sites #1 and #2 are within the current local low-elevation xeric ecotone for pine (Table 1), but the distribution of the two species has been likely modified by management in history (López-Sáez et al., 2014). Low elevation stands (Sites #0, #1 and #2) presented low pine regeneration (likely partly as a result of higher grazing pressure and differences in management between stands (Donés and Cabrera, 2009)), widespread symptoms of crown dieback, defoliation and abundant mistletoe infestation, which has been shown to be one of the several agents (with drought) involved in P. sylvestris decline (Dobbertin and Rigling, 2006; SangüesaBarreda et al., 2013; Zweifel et al., 2012). To analyze the relationship between climate and tree health and characterize the mortality pattern under different ecological conditions we sampled trees within three health status classes at the four different elevations: (1) ‘dead’ when trees presented 100% crown dieback; (2) ‘weakened or declining’ were considered those oaks with stem basal rot and partial canopy dieback, and those pines with a certain level of crown defoliation (over 20% compared to neighboring healthy trees) and infested by the semiparasite plant Viscum album L. (mistletoe) in at least one third of the canopy (SangüesaBarreda et al., 2012, 2013); (3) ‘healthy’ were those trees without the symptoms described in the previous two categories. It must be noted that trees with canopy decline presented both defoliation and mistletoe infestation. We sampled dominant adult trees (dbh > 25 cm) to analyze the growth trends in the longest possible interval and avoid oak resprouts of 40–70 years old likely established after last logging for firewood in the mid 1900s. We sampled neighboring trees of different health classes subject to similar stand competition conditions. However, this was not possible in all cases, the reason why we report a different number of plots for the different sites and health status (see below for definition of sampled plot). 2.2. Dendrochronological methods At each site we searched for target individuals of the two species with different health status. Trees were bored twice at

Table 1 Mean site characteristics. Ps = P. sylvestris; Qp = Q. pyrenaica. dbh = diameter at 1.30 m; BA = basal area; Dd = stand density. DBH, BA and Dd are plot values calculated from the characteristics of the neighbor trees (either the closest 10 of dbh over 7.5 cm if they are closer than 10 m or those included within a radius of 10 m) of those cored. Different letters correspond to significant differences using the LSD test on a one-way ANOVA comparing the four sites. Standard deviations are between parentheses. Site

0 1 2 3

Name

QUPY0 PISY1-low PISY2-medium PISY3-high

Sp.

Qp Ps Ps Ps

# Plots

28 17 20 17

Altitude (m)

1056 1077 1179 1596

(57) (17) (92) (105)

Coordinates

40.86"N 40.86"N 40.90"N 40.84"N

4.02"W 4.01"W 4.02"W 4.06"W

Slope

Climate

(")

Ppt (mm)

0–15 15–25 0–15 0–45

660.3 660.3 731.4 944.4

Competition Tmean ("C) 10.3 10.3 9.9 7.9

Tmin ("C) 5.35 5.35 5.15 2.95

dbh (cm) 29.1 14.6 36.2 35.9

a

(15.4) (9.1)b (15.6)a (18.7)a

BA (m2/ha)

Dd (n/ha)

12.0 (11.0)b 5.5 (8.5)c 22.4 (13.2)a 22.0 (11.7)a

213.5 155.9 197.8 263.2

(281.4)a (222.1)a (80.9)a (286.4)a

72

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

1.30 m, with an angle of 180" and perpendicular to the dominant slope to avoid reaction wood. Plots were then laid out around each bored target tree to asses stand competition by measuring distances to the target tree and two times the dbh of the 10 nearest neighbors of trees with dbh > 7.5 cm up to a maximum distance of 10 m away. We did not separate the influence of competition by species because it was homogeneous within altitudes: almost pure oak competition in #0 and #1, 50% of trees from each species in #2 (with pine dominating the canopy and oaks with dbh < 20 cm in the subcanopy) and pure pine competition in #3. Cores were sanded and growth increments visually crossdated, measured and verified with COFECHA (Holmes, 1983). The growth trends in relation to climate were analyzed at different time scales. Firstly, the long-term response was analyzed using basal area increments (BAI) calculated from measured dbh and tree-ring widths. Secondly, to study the response of growth to the inter-annual variability of climate (short-term response) we used prewithened residual chronologies (GIres) calculated from biweight means of ratios between raw growth measurements and individual cubic splines with a 50% frequency cutoff at 30 years (Fritts, 1976). In addition to the dendroecological analysis of the tree overstory, we collected 12 crossections of Cistus laurifolius L. from Sites #0 and #1 (this species was only abundant on the understory of those two sites). This was done to analyze the age distribution of the understory in low-elevation open stands to study management events of shrub removal or changes in grazing intensity. 2.3. Climatic covariates Gridded CRU TS 3.10 data (Mitchell and Jones, 2005) including monthly precipitation and mean, minimum and maximum temperature were obtained from the KNMI explorer (http://climexp.knmi.nl/get_index.cgi) for the period 1901–2004. Gridded data were corrected linearly using data from local stations from AEMET (http://www.aemet.es/es/portada). In addition we calculated the Standardized Precipitation-Evapotranspiration Index (SPEI) using the SPEI package in R (http://sac.csic.es/spei/) with a scale of 6 months. This drought index is calculated using water balances from precipitation and potential evapotranspiration (PET), which we estimated following Thornthwaite (Vicente-Serrano et al., 2010). Water stress increases inversely to SPEI. 2.4. Analysis of the climate-growth response of trees with different health status To analyze the growth response to the year-to-year variability in climate we calculated bootstrap correlations (period 1901– 2010) between growth and climatic periods of maximum growth response for the species studied as selected from a preliminary analysis (Appendix A) and the literature (e.g. Tessier et al., 1994; Eilmann et al., 2009). To analyze the species drought tolerance and see whether very dry years triggered abrupt reductions in mean growth we used superposed epoch analysis (SEA, Prager and Hoenig, 1989) on 7 drought events within the period 1901– 2010 defined as those where precipitation of the hydrological year was below 500 mm (1950, 1957, 1965, 1986, 1992, 1995 and 2005), which were also those of minimum SPEI (Appendix B). Finally we compared smoothed climatic data and BAI chronologies to investigate the growth-response to the long-term variability in drought intensity. 2.5. Effect of stand and individual features on the tree health status We used ordinal logistic regression using the ‘parallel slopes’ model to analyze the influence of individual and stand features

on the probability for a tree to be within a health class (Bender and Grouven, 1997; Canham et al., 2010). To ultimately calculate the probability p of a tree falling into health class k + 1, ordinal logistic regression uses cumulative probabilities as defined by:

logitðPiks Þ ¼ logðPiks =ð1 $ Piks ÞÞ ¼ aks þ b1s & x1i þ & & & þ bns & xni

ð1Þ

The cumulative probability of being ‘dead’ is P3s = 1. Piks is the cumulative probability of an individual i of species s falling into the health class k (healthy or weakened, i.e. ordinal logistic regressions for each species includes two model equations for the cumulative probability differing in the intercept ak). The natural logarithm of the odds ratio (logit) of the cumulative probability is modeled as a linear function of different stand and tree individual covariates (xni) where bns are regression parameters associated to the different covariates xni. P(i 6 Yk|X) is the probability that an observation Y will be less than or equal to ordinal level Yk (k = 1,2; i.e. health class healthy or weakened) given a vector of explanatory variables X. The probability p that an individual i falls into health class k + 1 is: pk M k + 1 = P(i 6 Yk + 1|X) $ P(i 6 Yk|X), where X is a vector of explanatory variables. Since the probabilities calculated from the logistic model (P) are cumulative, the probability p for an individual to be dead would be p2 M 3 = 1–P(i 6 Y2|X). A different ordinal logistic model was fitted for each species. We tested the following hypotheses expressed by covariates (i.e. bns&xni in Eq. (1)) related to stand and tree individual features: (1) Dead trees will exhibit a distinctive negative trend in the last decades previous to the tree death (Bigler and Bugmann, 2004). We used the slope of the regression of growth on year since 1970 (s1970) to test whether the growth reduction previous to death was concurrent to the increase in drought since the early 1970s (Appendix B). (2) Healthy trees will have a lower number of abrupt growth suppressions calculated as the number of years where the negative growth change (NGC) was below $25% (Das et al., 2007). Similarly, different health classes could exhibit a different number of years with positive growth changes (PGC). A greater number of PGCs could reflect that trees were more sensitive to increased available resources after stand disturbances or humid periods (see Appendix C for calculations and references). (3) The relative growth rate of individual trees (relBAIi = BAIi/BA, for tree i) will differ for the different health classes (Bigler and Bugmann, 2004; McDowell et al., 2010). (4) Trees of different health status will exhibit different average growth (Gmean) or juvenile growth (here calculated for the first 40 years, G40) (Bigler and Bugmann, 2004; Bigler and Veblen, 2009). (5) Dead trees will exhibit higher mean sensitivity (msx, a measurement of the year-to-year variation in growth; as in Fritts, 1976) to climate (Suárez et al., 2004; McDowell et al., 2010). (6) We analyzed whether higher levels of competition determined individual adult mortality by comparing three different competition indices: two distance independent indices (density in trees&ha$1 and basal area [BA] in m2 ha$1 of competitors around each target tree) and a distance dependent index analyzed within a 5 m " and 10 m radius ! P HI ¼ nj¼1 ðdbhj =dbhi Þ & ½1=ðdistij þ 1Þ( .

Nested models including the previous covariates were compared based on the Akaike Information Criterion corrected for small sample size (AICc) reported as the index Di = AICci–AICcmin, where AICcmin is the minimum AICc of all formulations compared. The model with Di = 0.0 is the best (Burnham and Anderson, 2004). The maximum likelihood parameters and two-unit support intervals for each parameter were estimated in the probabilistic

73

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

model using simulated annealing and R likelihood package (http:// www.sortie-nd.org/lme/lme_R_code_tutorials.html). Two-unit support intervals for specific parameters are analogous to 95% confidence intervals but calculated as those reducing two units the maximum likelihood function maximized for the other parameters (Bolker, 2008). To assess the goodness of fit in the final ordinal logistic model we followed Canham et al. (2010). First using the fitted parameters in the best model we calculated the predicted probability for each tree to fall within each of the three health classes. Then we divided the predicted probability of each class into 10% intervals and compared the proportion of predicted trees within each class with the proportion of observed trees that actually fell within a given health class. 3. Results Average tree and stand characteristics were similar among health classes within site except for growth and dbh for both species, and height for oaks (Table 2; Fig. 1). The trees were older and growth was slower at high elevation than at low elevations (Figs. 1 and 2; Tables 2 and 3). We did not observe symptoms of canopy decline at high-elevation stands, therefore at that site we only sampled ‘healthy’ and ‘dead’ trees. Conversely, more than 30% of pines presented symptoms of canopy decline at the pine forest ecotone (Site #2) whereas at the lowest elevation site (Site #1) all pine trees were ‘weakened’ or ‘dead’. Although we report results from 148 trees (Table 3) we sampled a higher number of dead trees because often wood decay made rings indistinguishable. This was the reason why the ‘dead’ pine chronology at the lowest site was less replicated and why most of the dead trees analyzed had died in the last 3 years. Since at Site #1 the growth trends of ‘dead’ trees were similar to those of the ‘weakened’ trees (Fig. 1) we built a single chronology (PS1 in Table 3) for the climate response analysis. At low-elevation all stands exhibited a major establishment date likely related to some management event (Table 2, Fig. 2), which was in the mid 19th century for pines and oaks at the lowest elevation, and in the early 1900s at the pine forest ecotone. A bimodal age distribution of C. laurifolius (peaks around 1955 and 1985, Appendix D) suggests periodic shrub removal (i.e. silvopastoral management) at Sites #0 and #1. 3.1. Drought increase determines long-term growth of different health classes The response of the three health classes to year-to-year climatic variability (i.e. short-term) was similar and differences in this response were mostly explained by differences in species and

altitude (Fig. 3; Appendix A). Oak growth was more sensitive to year-to-year moisture availability than pines, which lowest sensitivity to moisture was expressed at high-elevation (Fig. 3). PET was calculated using Thornthwaite and correlations between growth and calculated PET were very similar to those with temperature (not shown). The pines were positively affected by warm conditions in February, while oak growth did not seem to be affected by winter temperature (Fig. 3). June and July were the months where the growth response to climate was maximized for both species, which expressed a negative impact of high early summer temperatures (Appendix A). Drought events triggered minimum growth for all trees but this was not significant at high elevation (Fig. 4). Oak growth recovered faster than that of pines the year after drought. It took two years for pines to recover from drought events but then growth was higher than the average, which suggests that pines were more affected than oaks by drought. The response of trees to climate in the long-term presented some differences to that in the short-term. The three health classes expressed a different long-term response to climate. The mortality pattern was not homogeneous among chronologies as shown by the different long-term growth trends between the three health classes and the different mean age of dead trees observed at different altitudes (Fig. 1). Healthy oaks and high-elevation healthy pines were the only chronologies which did not exhibit a recent negative growth trend. Dead trees generally exhibited minimum absolute growth during recent years. Growth of weakened oaks diverged from that of healthy oaks since the 1960s. Dead oaks consistently exhibited lower growth rates during the last century, and their recent growth decline matched the increase in drought intensity since the 1970s (Fig. 5). Growth of weakened oaks and dead pines at high-elevation declined concurrently to increased drought severity since approximately year 2000. All pine trees at low-elevation exhibited a marked recent growth decline. At Site #2 weakened and dead pines showed similar abrupt negative trends starting in the late 1970s, whereas growth of healthy pines was above that of non-healthy pines but also exhibited a parallel growth decline (Figs. 1 and 5). At Site #1 a negative trend along the life span of pines (Fig. 5) was mixed with mid-term disturbances likely related to more intense silvopastoral management at the lowest elevation (e.g. peaks around 1950, Appendix D). 3.2. Tree and stand factors influence health status and vulnerability to drought increase The fit of the logistic model was good but only some of the tested covariates were significant (Fig. 6). The most significant

Table 2 Mean characteristics of trees included in the chronologies. Standard deviations are shown between parentheses. Site is as in Table 1. Age corresponds to estimated age at 1.30 cm. Different letters mean significant differences between health categories within each site according to a LSD test of mean differences in a one-way ANOVA between the 10 chronologies. DBH, tree height (H) and age correspond to the cored trees; BA, density, and HI corresponds to plot values calculated for each cored tree in relation to the neighbor trees within plot. Ps = P. sylvestris; Qp = Q. pyrenaica. HI10 and HI5 are the distance dependence indices (HI) calculated for a radius of 10 and 5 m respectively. Site

Name

Sp.

Sanitary status

Sampled trees DBH (cm)

Competition H (m)

Age (years)

BA (m2/ha)

Density (trees/ha)

HI10

HI5

0

QPH QPW QPD

Qp Qp Qp

Healthy Weakened Dead

56.0 (15.5) 48.6 (16.4)de 43.4 (6.8)e

15.4 (3.0) 14.2 (3.6)d 11.1 (2.8)e

159 (15) 140 (36)bc 137 (23)c

13.6 (10.3)bcde 9.4 (10.2)def 13.4 (12.4)cde

221.0 (306.8)ab 181.0 (222.2)ab 226.7 (306.0)ab

0.41 (0.31)bc 0.50 (0.50)ab 0.51 (0.50)ab

0.24 (0.27)bc 0.36 (0.42)ab 0.35 (0.53)abc

1

PS1W PS1D

Ps Ps

Weakened Dead

66.2 (18.0)abc 71.8 (27.9)a

16.0 (2.6)cd 16.0 (4.4)cd

147 (21)bc 147 (12)bc

6.5 (9.6)ef 2.3 (2.1)f

181.2 (247.6)ab 75.1 (78.0)b

0.19 (0.19)c 0.14 (0.19)c

0.09 (0.12)c 0.09 (0.17)bc

2

PS2H PS2W PS2D

Ps Ps Ps

Healthy Weakened Dead

69.6 (21.0)ab 57.1 (12.9)cd 58.8 (16.7)bcd

20.1 (2.5)ab 20.9 (2.8)ab 18.5 (4.4)bc

106 (17)d 102 (7)d 95 (23)d

22.5 (16.3)ab 23.1 (9.6)a 18.4 (12.1)abcd

147.8 (64.5)ab 232.1 (75.9)ab 175.3 (75.5)ab

0.51 (0.43)ab 0.77 (0.37)a 0.61 (0.47)ab

0.31 (0.42)abc 0.54 (0.54)a 0.34 (0.48)abc

3

PS3H PS3D

Ps Ps

Healthy Dead

54.2 (10.3)cde 50.7 (12.8)de

22.0 (4.1)a 21.4 (3.1)a

207 (39)a 202 (40)a

21.1 (12.5)abc 21.1 (12.1)abc

247.8 (279.6)a 245.1 (271.0)ab

0.58 (0.39)ab 0.63 (0.42)ab

0.21 (0.24)bc 0.22 (0.26)bc

cd

d

b

74

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

Fig. 1. BAI trends of the different mean chronologies according to health status at the four sites. Low, medium and high refer to Sites #1, #2 and #3, respectively. QUPY = Q. pyrenaica; PISY = Pinus sylvestris. Dashed vertical lines highlight years 1950 and 1970. Mean site altitude as from Table 1 is specified in each graph.

Fig. 2. Distribution of estimated dates of establishment (at 1.30 m) of sampled trees at the four sites. Low, medium and high refer to Sites #1, #2 and #3 respectively. Mean site altitude as from Table 1 is specified in each graph.

covariate selected for both species was NGC, i.e. the number of years when trees exhibited abrupt growth reductions. Selection of NGC in the model expressed that trees suffering a higher number of growth suppressions (i.e. more decreasing growth) along

their life had a lower likelihood of being healthy. Conversely, a higher number of positive growth episodes and higher initial growth increased the likelihood of being healthy for pines. The decline in growth since 1970 separated the different health classes

75

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

Table 3 Mean characteristics of chronologies. # Cores = number of cores (number of trees between parentheses); interval = years replicated with more than five series; RW = mean anual ring-width; Rbs = mean correlation between series; AR = mean autocorrelation of raw series; MS = mean sensitivity; EPS = mean expressed population signal. Names are as in Table 2. PS1 = PS1W + PS1D. Name

# Cores (trees)

Interval (n > 5)

Length (years)

RW (mm)

MS

AR

Rbs

EPS

QPH QPW QPD

37 (18) 29 (15) 26 (12)

1864–2011 1869–2011 1879–2011

148 143 139

1.53 1.47 1.35

0.242 0.258 0.251

0.746 0.717 0.820

0.254 0.234 0.217

0.906 0.862 0.855

PS1W PS1D PS1

32 (16) 11 (5) 43 (21)

1878–2011 1886–2011 1870–2011

134 126 142

1.85 1.85 1.85

0.311 0.291 0.305

0.786 0.747 0.776

0.320 0.282 0.320

0.915 0.783 0.915

PS2H PS2W PS2D

34 (17) 34 (18) 28 (14)

1915–2011 1915–2011 1915–2011

97 97 97

2.76 2.49 2.55

0.279 0.279 0.304

0.765 0.810 0.764

0.261 0.306 0.321

0.908 0.923 0.916

PS3H PS3D

35 (16) 35 (17)

1785–2011 1792–2011

227 219

1.08 1.07

0.297 0.311

0.785 0.810

0.295 0.208

0.916 0.875

Fig. 3. Bootstrap correlations between growth indices (GIres) and selected climatic variables of maximum tree response to climate for the period 1901–2010: (1) hydrological year goes from September of the previous year to September of the growth year; (2) spring (April, May and June); (3) June-July; (4) August; (5) winter (December previous year to February current year); and (6) February. Nomenclature of chronologies as in Table 3. Ppt = precipitation; Tmin = minimum temperature; Tmax = maximum temperature.

only for oaks (Tables 4 and 5) because for pines the negative growth decline since that date was only evident at low-elevation and equally expressed by all health classes. Oaks with higher mean sensitivity had a higher likelihood to die. Remarkably, higher stand density for oaks and higher stand basal area for pines increased the likelihood for trees of being healthy (Tables 4 and 5). 4. Discussion 4.1. Rising water stress threatens pines in managed oak-pine woodlands at the xeric ecotone A long-term increase in water stress can predispose trees to decline (Suárez et al., 2004; Bigler et al., 2006; Breshears et al., 2009; McDowell et al., 2011). The negative trend observed in dead trees concurrent to the increase in drought intensity since the 1970s suggests that enhanced drought severity could have determined mortality of Q. pyrenaica and P. sylvestris at low-elevation. In addition the sensitivity of trees to moisture availability, the negative growth-trends of healthy pines and the abundance of trees with symptoms of canopy decline only at low-elevation stands suggest

that P. sylvestris could be threatened by climate change at its xeric limit in the studied area. This would not be a local phenomenon because enhanced P. sylvestris mortality with or without colonization by more drought tolerant oaks has also been reported in dry valleys in the Alps (Weber et al., 2008; Gimmi et al., 2010; Rigling et al., 2013) and the Pyrenees (Galiano et al., 2010; Heres et al., 2012; Vilá-Cabrera et al., 2013). Tree mortality driven by enhanced drought severity has been observed in other ecosystems worldwide (e.g. Breshears et al., 2009; Allen et al., 2010; Choat et al., 2012). Currently, the climate in the Mediterranean seems to be within one of the driest periods in the last centuries (e.g. Luterbacher et al., 2006; García-Ruiz et al., 2011). Similar to other oak forests in Spain (Gea-Izquierdo et al., 2011, 2012; Gea-Izquierdo and Cañellas, 2014) the low-elevation stands showed a major establishment peak after the first half of the 19th century. This means that most of the sampled trees were established when temperatures were likely 2–3 "C below present values (and hence annual ETP above). The mortality pattern differed with mean climatic conditions at different altitudes (Shifley et al., 2006) probably because complex physiological drivers determine the relationship between tree

76

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

Fig. 4. Superposed epoch analysis with a lag of five years of the seven drought event years with Ppt < 500 mm within the period 1901–2010: 1950, 1957, 1965, 1986, 1992, 1995 and 2005. SEI = superposed epoch scaled index (growth departures). Solid bars correspond to significant values calculated using 10,000 random simulations. Mean site altitude as from Table 1 is specified in each graph.

Fig. 5. Low-frequency analyses between growth (basal area increment, BAI, cm2/year) and mean SPEI (Standardized Precipitation Evaporation Index) of August. Annual data were smoothed using cubic splines with a 50% frequency cutoff at 30 years at the four sites (nomenclature as in Table 3). We show the linear correlation r between smoothed BAI and SPEI for the period 1950–2009 as a ‘naive’ estimator of the linear relationship between the two time series.

productivity and mortality (Breshears et al., 2009; McDowell et al., 2011; Anderegg et al., 2013). As in the literature, the growth decline previous to death showed certain variability ranging from

10 to 40 years which could be related to differences in the actual factors triggering death (e.g. Jenkins and Pallardy, 1995; Bigler and Bugmann, 2004; Suárez et al., 2004). Drought driven mortality

77

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

Fig. 6. Goodness of fit for the ordinal logistic regression model. Trees were grouped in intervals of probability = 0.1 according to the probability calculated for the three health classes and then that probability was plotted against the observed proportion of trees that showed that health status for a given probability class. We show in grey the 1:1 line for reference.

will be likely patchy and concentrated in more xeric and shallower soils (Lloret et al., 2004). Although with exceptions (Jenkins and Pallardy, 1995; Levanic et al., 2011), generally the risk of mortality within stands is maximized for trees with low growth and abrupt growth declines as surrogates of a less positive tree carbon balance (Bigler and Bugmann, 2004; Shifley et al., 2006; Das et al., 2007; Heres et al., 2012). In agreement, healthy trees at low-elevation exhibited faster growth which could reflect that they maintained more positive carbon pools and profited from more favorable microtopographic conditions. However, fast growth rates reduce plant species longevity (Bugmann and Bigler, 2011; Di Filippo et al., 2012). The lowest growth and oldest ages were present at higher elevations where trees were likely more limited by low winter temperatures, a shorter phenological period and/or nutrient availability than by long-term water stress (Bigler and Veblen, 2009). P. sylvestris can attain ages over 500 years in the area but average lifeexpectancy should generally be lower. The lower response to water stress and the older age of dead trees at high-elevation compared to those at low-elevation could express senescence independent of climate change. In this sense it would be interesting to

Table 4 Results of the ordinal logistic regression. Model: logitðPiks Þ ¼ aks þ b1s & x1i þ & & & þ bns & xni , where Piks is the cumulative probability of health status for the i observation, species s and sanitary status k (k ) [1, 2], Pi3s = 1, with 3 = dead; and the probability to fall between one of the three (k + 1) health classes pk$1 M k = P(i 6 Yk|X)–P(i 6 Yk$1|X))). xn are covariates; aks and bns are regression parameters. ML = maximum likelihood estimation; # pars = number of parameters; Di as in the text; s1970 = slope of regression of BAI on age for the period 1970–2010; NGC = number of years with negative growth changes below $25%; PGC = number of years with positive growth changes over 25%; ms = mean tree sensitivity; G40 = mean BAI growth of the series first 40 years of growth (cm2/year); Gmean = mean tree growth (cm2/year); density = stand density (trees/ha); BA = basal area (m2/ha). The best model for each species is in bold. Model #

Covariates

Species Q. pyrenaica

0 1 2 3 4 5 6 6 7 8 9 10 11 12 13 14 15

aks aks + b1s&NGC aks + b1s&NGC + b2s&RelBAI aks + b1s&NGC + b2s&density aks + b1s&NGC + b2s&s1970 aks + b1s&NGC + b2s&ms aks + b1s&NGC + b2s&s1970 + b3s&ms aks + b1s&NGC + b2s&s1970 + b3s&PGC aks + b1s&NGC + b2s&s1970 + b3s&G40 aks + b1s&NGC + b2s&s1970 + b3s&Gmean aks + b1s&NGC + b2s&s1970 + b3s&HI aks + b1s&NGC + b2s&s1970+ + b3s&ms + b4s&Gmean aks + b1s&NGC + b2s&s1970 + b3s&ms + b4s&density aks + b1s&NGC + b2s&PGC ak + b1&NGC + b2&PGC + b3&BA ak + b1&NGC + b2&PGC + b3&BA + b4high-med&G40 Model # 14 with a2 = 0

P. sylvestris

# Pars

ML

Di

# Pars

ML

Di

2 3 3 4 4 4 5 5 5 5 5 6 6 4 – – –

$44.8 $42.9 $42.5 $42.1 $39.3 $41.0 $37.6 $39.0 $38.6 $38.4 $39.2 $37.1 $34.4 $40.9 – – –

19.0 9.4 11.0 10.3 4.5 8.0 3.7 9.5 5.8 5.3 6.9 5.3 0.0 7.7 – – –

2 3 3 4 4 4 5 5 5 5 5 6 6 4 5 6 5

$113.1 $99.0 $98.8 $99.0 $98.1 $98.8 $98.1 $93.6 $98.1 $98.0 $98.1 $98.0 $98.1 $93.5 $88.8 $86.1 $86.1

47.5 21.4 23.2 23.5 21.7 23.2 23.9 15.0 24.0 23.9 24.1 26.6 26.3 12.7 5.4 1.3 0.0

Table 5 Best model parameters and two unit support intervals (2 S.I.) as estimated in Table 4. All acronyms are the same as those in Table 4. Pk is the cumulative probability of health status k (healthy, weakened) where the probability to fall within one of the three health status (i.e. k + 1, 3 = dead) is: pk$1 M k = P(i 6 Yk|X)–P(i 6 Yk $ 1|X), with P3s = 1; q = Quercus pyrenaica; p = Pinus sylvestris. Species

Model

Parameter

Estimate

2 S.I.

Q. pyrenaica

logitðP kq Þ ¼ akq þ b1q & NGC þ b2q & s1970 þ b3q & ms þ b4q & Density

ahealthy-q aweakened-q b1q b2q b3q b4q

8.197 10.544 $0.192 4.040 $26.521 0.004

(7.309, 8.792) (9.832, 11.335) ($0.239, $0.155) (1.469, 7.905) ($29.493, $24.000) (0.002, 0.006)

P. sylvestris

logitðP kp Þ ¼ akp þ b1p & NGC þ b2p & PGC þ b3p & BA þ b4p-high=med & G40

ahealthy-p aweakened-q b1p b2p b3p b4phigh-med

$2.532 0.0 $0.057 0.091 0.046 0.028

(3.116, $2.114) – ($0.078, $0.042) (0.068, 0.106) (0.027, 0.062) (0.012, 0.043)

78

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

determine average life expectancy at different environmental conditions in the absence of a long-term climatic stressor. Size dependent competition reduces the annual variability of growth in dense forests but increases the sensitivity to long-term water stress and the risk of mortality (Voelker et al., 2008; Cavard et al., 2011; Dietze and Moorcroft, 2011). The studied oak-pine woodland was simple in structure and within stands the dominant canopy was generally monospecific. In this sense, positive interactions among species are likely underrepresented (Cavard et al., 2011; Luo and Chen, 2011) but competition did not seem to be responsible for the mortality observed. This situation is very different to that in abundant Mediterranean coetaneous dense stands established after land-use abandonment (Di Filippo et al., 2010; Gea-Izquierdo and Cañellas, 2014; Vilá-Cabrera et al., 2013). The observed negative effect of low competition levels for the tree health may reflect indirectly a negative impact of more intensive management at low-elevation stands predisposing trees to other inciting agents of mortality (Oliva and Colinas, 2007; Camarero et al., 2011; Vilá-Cabrera et al., 2013). During the main establishment period of the sampled stands in the nineteenth century the studied forest experienced great socio-economical changes, the establishment of forest management and the foundation of one local sawmill (Breñosa and Castellarnau, 1884; Franco Mugica et al., 1998). As for other oak-pine ecosystems, the dynamics of the studied stands are not in long-term equilibrium and depend on the interaction between management and climate (Wischnewski et al., 2011; Weber et al., 2008; Gimmi et al., 2010; Sheffer, 2012). Although we did not study oak at its dry limit (it was used as an offset to compare with co-occurring declining pine), other studies suggest that it will be substituted by more drought tolerant species such as Q. ilex L. with climate change (Ruiz-Labourdette et al., 2013). 4.2. Species-specific growth responses to climate and vulnerability to drought increase The relationship between tree performance and climate can depend on the time scale analyzed. In the long-term, trees with different health status exhibited a different response to drought intensity, and this response was less intense for healthy oaks and for pines at high elevation. In the short-term, differences in the tree response to climate were mostly explained by phenological differences imposed by elevation along climatic gradients (Tessier et al., 1994; Gea-Izquierdo et al., 2011, 2012; Gea-Izquierdo and Cañellas, 2014; Lloyd et al., 2011). The maximum growth response to climate for deciduous Mediterranean Quercus sp. is generally between May and June (Tessier et al., 1994; Gea-Izquierdo et al., 2012; Gea-Izquierdo and Cañellas, 2014). In the cold and continental (in relative terms) Mediterranean forest studied, the key period for xylogenesis seemed to be June–July for both species (Eilmann et al., 2009; Michelot et al., 2012). Growth was enhanced under favorable humid conditions. Evergreen pines were benefited by warmer temperatures in winter but not the deciduous oak analyzed (Rozas et al., 2009; Gea-Izquierdo and Cañellas, 2014). Oaks were more sensitive than pines to the year-to-year variability in moisture availability contrary to results in Eilmann et al. (2009). Healthy trees were more sensitive to drought events, in opposition to results from Sangüesa-Barreda et al. (2012) who reported that trees infested by mistletoe suffered more after drought than healthy trees. Similar to results for P. nigra J.F. Arnold (MartínBenito et al., 2008), P. sylvestris required two years to restore positive growth anomalies after drought events whereas oaks needed only one year. Ring-porous oaks need to rebuild their conductive apparatus every year (Abrams, 1990) but not pines. Different species-specific hydraulic regulation implies different strategies to cope with susceptibility to drought induced dysfunctions in the

hydraulic system (Eilmann et al., 2009; McDowell et al., 2011; Anderegg et al., 2013). 4.3. Can we forecast future mortality in pine woodlands using simple crown observations? Describing simple observations which can be used to forecast future tree death is important for forest management under global change (Dobbertin, 2005). Drought can induce forest decline through interdependent physiological processes related to carbon and hydraulic constraints (Breshears et al., 2009; McDowell et al., 2010, 2011; Anderegg et al., 2013), predisposing trees to other biotic and abiotic agents affecting at different temporal scales (Bigler et al., 2006; Galiano et al., 2010; Gaylord et al., 2013). In this study we analyzed the influence of climate (an abiotic factor) on the health status of trees. Although we do not have data to analyze in detail the interaction of climate with other biotic factors, to our knowledge there has not been any recent important insect outbreak in the area. Other biotic agents such as semi-parasitic plants were abundant and used to classify weakened pine trees. Mistletoes preferentially infest weak trees already affected by long-term predisposing factors to decline like water stress (Dobbertin and Rigling, 2006; Cailleret et al., 2013). They drive host trees to carbon starvation in the long-term as a consequence of excessive stomatal control to prevent xylem cavitation produced by high mistletoe transpiration (Zweifel et al., 2012). Therefore mistletoe infestation is negative for the host carbon balance (i.e. reduces growth). Mistletoes could preferentially infest those trees with previous poorer carbon balances as expressed by a lower past-growth like in Sangüesa-Barreda et al. (2012) and our results. However other authors reported higher levels of infestation in trees with high-past growth levels, showing that the relationship between tree decline, crown defoliation and mistletoe infestation is modulated by yet not totally understood additional environmental factors (Cailleret et al., 2013; Sangüesa-Barreda et al., 2013). We did not investigate in depth this relationship but used a classification of tree health status based on simple crown observations similar to other studies (e.g. Dobbertin, 2005). We observed identical recent growth-trends of weakened and dead pines at low elevation, which suggests that abundant mistletoe infestation and defoliation reflect pine vulnerability. Mistletoe presence is considered to be a major indicator of decline in P. sylvestris (Galiano et al., 2010; Zweifel et al., 2012; Sangüesa-Barreda et al., 2012, 2013) and other conifers such as Abies alba Mill. (Oliva and Colinas, 2007; Cailleret et al., 2013). Mistletoe also infests other Mediterranean pines such as P. pinaster Ait. and P. nigra and the relationship between tree decline and mistletoe ecology needs to be investigated in more detail (Sangüesa-Barreda et al., 2013; Cailleret et al., 2013). Nevertheless, independent of the agent eventually causing mortality, widespread mistletoe infestation seems to be a symptom anticipating enhanced pine mortality in our study area. Thus it could be used as a simple indicator of the tree vitality to assess sustainability of pine stands (Dobbertin, 2005). 5. Conclusions The mortality pattern varied along the elevation gradient analyzed with average site conditions but enhanced drought severity since the 1970s seemed to predispose certain trees of both species to mortality at low elevation. Pine trees with symptoms of canopy decline were abundant at more xeric low elevation stands where life span of adult trees was much shorter. At low-elevation, regardless of their health status, the trees exhibited a recent decline in growth concurrently with an enhancement of longterm drought intensity, which suggest vulnerability to water

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

stress increase of P. sylvestris growing at its local low-elevation xeric limit. Within stands, the trees of both species analyzed growing under better conditions (as expressed by higher growth-rates and a lower number of negative growth depressions, i.e. likely with more positive tree carbon pools) reduced their susceptibility to decline and mortality. In agreement with other studies, our results suggest that the canopy symptoms used to classify pine health status can be used to characterize vulnerability of stands in forest inventories. We do not know whether the present distribution of the studied species expresses the local potential niche but the present realized low-elevation limit of P. sylvestris in the Central Mountains of Spain seems to be threatened by climate change. In agreement to results from other similar mountain oak-pine woodlands, these results could derive important ecological implications and negative economical consequences for local populations. Acknowledgments We thank Javier Donés for permissions and support to do research in the studied area. This research was supported by project AGL2010-21153-01 funded by MICINN. It is also a contribution to the Labex OT-Med (no ANR-11-LABX-0061) funded by the «Investissements d’Avenir» program of the French National Research Agency through the A*MIDEX project (no ANR-11-IDEX-0001-02). Appendix A Bootstrap correlations between calculated growth indices and monthly climatic covariates. P = precipitation; SPEI = standardized precipitation evaporation index; Tmin = minimum temperature; Tmax = maximum temperature. Dashed horizontal lines at ±0.22 correspond to the minimum value where all correlations are significant at a = 0.05.

79

Appendix B Climatic trends for the period 1901–2010 in low-elevation plots. Hydrological year precipitation and maximum monthly temperature of June-July are shown above whereas calculated mean annual potential evapotranspiration (PET) and SPEI for the hydrological year are shown below. Horizontal dashed lines correspond to mean climatic values for the same period. Vertical black dashed lines above correspond to the key events of minimum precipitation (below 500 mm) namely 1950, 1957, 1965, 1986, 1992, 1995 and 2005.

80

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

Appendix C

References

Since we lacked historical management data we assessed standhistory indirectly from growth. We built sitewise disturbance chronologies using RW (Nowacki and Abrams, 1997) to identify abrupt positive (releases) or negative (suppressions) changes in growth. Growth change (GC) was calculated annually using 10 year windows either as positive growth changes (PGC) or negative growth changes (NGC):

Abrams, M.D., 1990. Adaptations and responses to drought in Quercus species of North-America. Tree Physiol. 7, 227–238. Adams, H.D., Guardiola-Claramonte, M., Barron-Gafford, G.A., Villegas, J.C., Breshears, D.D., Zou, C.B., Troch, P.A., Huxman, T.E., 2009. Temperature sensitivity of drought-induced tree mortality portends increased regional dieoff under global-change-type drought. PNAS 106, 7063–7066. Allen, C.D., Macalady, A.K., Chenchouni, H., Bachelet, D., McDowell, N., Vennetier, M., Kitzberger, T., Rigling, A., Breshears, D.D., Hogg, E.H., Gonzalez, P., Fensham, R., Zhang, Z., Castro, J., Demidova, N., Lim, J.-H., Allard, G., Running, S.W., Semerci, A., Cobb, N., 2010. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecol. Manag. 259, 660–684. Anderegg, W.R.L., Plavcova, L., Anderegg, L.D.L., Hacke, U.G., Berry, J.A., Field, C.B., 2013. Drought’s legacy, multiyear hydraulic deterioration underlies widespread aspen forest die-off and portends increased future risk. Global Change Biol. 19, 1188–1196. Bender, R., Grouven, U., 1997. Ordinal logistic regression in medical research. J. Roy. Coll. Phys. Lond. 31, 546–551. Bigler, C., Bugmann, H., 2004. Predicting the time of tree death using dendrochronological data. Ecol. Appl. 14, 902–914. Bigler, C., Veblen, T.T., 2009. Increased early growth rates decrease longevities of conifers in subalpine forests. Oikos 118, 1130–1138. Bigler, C., Braker, O.U., Bugmann, H., Dobbertin, M., Rigling, A., 2006. Drought as an inciting mortality factor in Scots pine stands of the Valais, Switzerland. Ecosystems 9, 330–343. Boisvenue, C., Running, S.W., 2006. Impacts of climate change on natural forest productivity – evidence since the middle of the 20th century. Global Change Biol. 12, 862–882. Bolker, B.M., 2008. Ecological Models and Data in R. Princeton University Press, NJ. Breda, N., Huc, R., Granier, A., Dreyer, E., 2006. Temperate forest trees and stands under severe drought, a review of ecophysiological responses, adaptation processes and long-term consequences. Ann. Forest Sci. 63, 625–644. Breñosa, R., Castellarnau, J.M., 1884. Guía y descripción del Real Sitio de San Ildefonso. Sucesores de Rivadeneyra, Madrid, p. 335.

8 h i > < PGC i ¼ ðG2$G1Þ + 100; if G2 > G1 G1 h i GC i ¼ ; ðG2$G1Þ > : NGC i ¼ + 100; if G1 > G2 G2

Pði$1Þ Pðiþ9Þ with G1 ¼ ði$10Þ RW i ; G2 ¼ i RW i and RWi being ring-width from year i of an individual tree-ring series. Site disturbance chronologies were constructed by averaging GCi annually (Camarero et al., 2011). Appendix D Mean disturbance chronologies (%) of trees with different health status at the four sites. See text and Appendix C for details on Mean Growth Change (MGC) calculations. In the left graph above (that of Q. pyrenaica, QUPY0) we show the frequency distribution of the establishment dates estimated for competing Cistus laurifolius (CILA) sampled in plots at Sites #0 and #1 (i.e. PISY1-low).

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82 Breshears, D.D., Myers, O.B., Meyer, C.W., Barnes, F.J., Zou, C.B., Allen, C.D., McDowell, N.G., Pockman, W.T., 2009. Tree die-off in response to global change-type drought, mortality insights from a decade of plant water potential measurements. Front Ecol. Environ. 7, 185–189. Brose, P.H., Waldrop, T.A., 2010. A dendrochronological analysis of a disturbancesuccession model for oak–pine forests of the Appalachian Mountains, USA. Can. J. Forest Res. 40, 1373–1385. Bugmann, H., Bigler, C., 2011. Will the CO2 fertilization effect in forests be offset by reduced tree longevity? Oecologia 165, 533–544. Burnham, K.P., Anderson, D.R., 2004. Multimodel inference – understanding AIC and BIC in model selection. Sociol. Methods Res. 33, 261–304. Cailleret, M., Nourtier, M., Amm, A., Durand-Gillmann, M., Davi, H., 2013. Droughtinduced decline and mortality of silver fir differ among three sites in Southern France. Ann. Forest Sci., doi:10.1007/s13595-013-0265-0. Camarero, J.J., Bigler, C., Linares, J.C., Gil-Pelegrin, E., 2011. Synergistic effects of past historical logging and drought on the decline of Pyrenean silver fir forests. Forest Ecol. Manag. 262, 759–769. Canham, C.D., Thompson, J., Zimmerman, J.K., Uriarte, M., 2010. Variation in susceptibility to hurricane damage as a function of storm intensity in puerto rican tree species. Biotropica 42, 87–94. Cavard, X., Bergeron, Y., Chen, H.Y.H., Pare, D., Laganiere, J., Brassard, B., 2011. Competition and facilitation between tree species change with stand development. Oikos 120, 1683–1695. Choat, B., Jansen, S., Brodribb, T.J., Cochard, H., Delzon, S., Bhaskar, R., Bucci, S.J., Feild, T.S., Gleason, S.M., Hacke, U.G., Jacobsen, A.L., Lens, F., Maherali, H., Martinez-Vilalta, J., Mayr, S., Mencuccini, M., Mitchell, P.J., Nardini, A., Pittermann, J., Pratt, R.B., Sperry, J.S., Westoby, M., Wright, I.J., Zanne, A.E., 2012. Global convergence in the vulnerability of forests to drought. Nature 491, 752–755. Das, A.J., Battles, J.J., Stephenson, N.L., van Mantgem, P.J., 2007. The relationship between tree growth patterns and likelihood of mortality, a study of two tree species in the Sierra Nevada. Can. J. Forest Res. 37, 580–597. Di Filippo, A., Alessandrini, A., Biondi, F., Blasi, S., Portoghesi, L., Piovesan, G., 2010. Climate change and oak growth decline, Dendroecology and stand productivity of a Turkey oak (Quercus cerris L.) old stored coppice in Central Italy. Ann. Forest Sci. 67, 706. Di Filippo, A., Biondi, F., Maugeri, M., Schirone, B., Piovesan, G., 2012. Bioclimate and growth history affect beech lifespan in the Italian Alps and Apennines. Global Change Biol. 18, 960–972. Dietze, M.C., Moorcroft, P.R., 2011. Tree mortality in the eastern and central United States, patterns and drivers. Global Change Biol. 17, 3312–3326. Dobbertin, M., 2005. Tree growth as indicator of tree vitality and of tree reaction to environmental stress, a review. Eur. J. Forest Res. 124, 319–333. Dobbertin, M., Rigling, A., 2006. Pine mistletoe (Viscum album ssp austriacum) contributes to Scots pine (Pinus sylvestris) mortality in the Rhone valley of Switzerland. Forest Pathol. 36, 309–322. Donés, J., Cabrera, M., 2009. Proyecto de 7ª Revisión de la Ordenación del monte número 2 del Catálogo de los de Utilidad Pública de la Provincia de Segovia ‘‘Pinar’’ de Valsaín. Unpublished management plan, Segovia. Eilmann, B., Zweifel, R., Buchmann, N., Fonti, P., Rigling, A., 2009. Drought-induced adaptation of the xylem in Scots pine and pubescent oak. Tree Physiol. 29, 1011–1020. Franco Mugica, F., Anton, M.G., Ollero, H.S., 1998. Vegetation dynamics and human impact in the Sierra de Guadarrama, Central System, Spain. Holocene 8, 69–82. Fritts, H.C., 1976. Tree Rings and Climate. Blackburn Press, p. 567. Galiano, L., Martinez-Vilalta, J., Lloret, F., 2010. Drought-induced multifactor decline of scots pine in the pyrenees and potential vegetation change by the expansion of co-occurring oak species. Ecosystems 13, 978–991. Garcia-Ruiz, J.M., Lopez-Moreno, J.I., Vicente-Serrano, S.M., Lasanta-Martinez, T., Begueria, S., 2011. Mediterranean water resources in a global change scenario. Earth – Sci. Rev. 105, 121–139. Gaylord, M.L., Kolb, T.E., Pockman, W.T., Plaut, J.A., Yepez, E.A., Macalady, A.K., Pangle, R.E., McDowell, N.G., 2013. Drought predisposes pinon-juniper woodlands to insect attacks and mortality. New Phytol. 198, 567–578. Gea-Izquierdo, G., Cañellas, I., 2014. Local climate forces instability in long-term productivity of a Mediterranean oak along climatic gradients. Ecosystems 17, 228–241. Gea-Izquierdo, G., Cherubini, P., Cañellas, I., 2011. Tree-rings reflect the impact of climate change along a temperature gradient in Spain over the last 100 years. Forest Ecol. Manag. 262, 1807–1816. Gea-Izquierdo, G., Fonti, P., Cherubini, P., Martín-Benito, D., Chaar, H., Cañellas, I., 2012. Xylem hydraulic adjustment and growth response of Quercus canariensis Willd. To climatic variability. Tree Physiol. 32, 401–413. Gimmi, U., Wohlgemuth, T., Rigling, A., Hoffmann, C.W., Buergi, M., 2010. Land-use and climate change effects in forest compositional trajectories in a dry CentralAlpine valley. Ann. Forest Sci. 67 (7), 701. Heres, A.-M., Martinez-Vilalta, J., Claramunt Lopez, B., 2012. Growth patterns in relation to drought-induced mortality at two Scots pine (Pinus sylvestris L.) sites in NE Iberian Peninsula. Trees Struct. Func. 26, 621–630. Holmes, R.L., 1983. Computer-assisted quality control in tree-ring dating and measurement. Tree – Ring Bull. 43, 69–78. Jenkins, M.A., Pallardy, S.G., 1995. The influence of drought on red oak group species growth and mortality in the Missouri Ozarks. Can. J. Forest Res. 25, 1119–1127. Lenoir, J., Gegout, J.C., Marquet, P.A., de Ruffray, P., Brisse, H., 2008. A significant upward shift in plant species optimum elevation during the 20th century. Science 320, 1768–1771.

81

Levanic, T., Cater, M., McDowell, N.G., 2011. Associations between growth, wood anatomy, carbon isotope discrimination and mortality in a Quercus robur forest. Tree Physiol. 31, 298–308. Linares, J.C., Camarero, J.J., 2012. From pattern to process, linking intrinsic wateruse efficiency to drought-induced forest decline. Global Change Biol. 18, 1000– 1015. Lloret, F., Siscart, D., Dalmases, C., 2004. Canopy recovery after drought dieback in holm-oak Mediterranean forests of Catalonia (NE Spain). Global Change Biol. 10, 2092–2099. Lloyd, A.H., Bunn, A.G., Berner, L., 2011. A latitudinal gradient in tree growth response to climate warming in the Siberian taiga. Global Change Biol. 17, 1935–1945. López-Sáez, J.A., Abel-Schaad, D., Pérez-Díaz, S., Blanco-González, A., Alba-Sánchez, F., Dorado, M., Ruiz-Zapata, B., Gil-García, M.J., Gómez-González, C., FrancoMúgica, F., 2014. Vegetation history, climate and human impact in the Spanish Central System over the last 9000 years. Quatern. Int.. http://dx.doi.org/ 10.1016/j.quaint.2013.06.034. Luo, Y., Chen, H.Y.H., 2011. Competition, species interaction and ageing control tree mortality in boreal forests. J. Ecol. 99, 1470–1480. Luterbacher, J., Xoplaki, E., Casty, C., et al., 2006. Mediterranean climate variability over the last centuries, a review. In: Lionello, P., Malanotte-Rizzoli, P., Boscolo, R. (Eds.), The Mediterranean Climate, An Overview of the Main Characteristics and Issues. Elsevier, Amsterdam, pp. 27–148. Martín-Benito, D., Cherubini, P., del Rio, M., Canellas, I., 2008. Growth response to climate and drought in Pinus nigra Arn. trees of different crown classes. Trees Struct. Funct. 22, 363–373. McDowell, N.G., Allen, C.D., Marshall, L., 2010. Growth, carbon-isotope discrimination, and drought-associated mortality across a Pinus ponderosa elevational transect. Global Change Biol. 16, 399–415. McDowell, N.G., Beerling, D.J., Breshears, D.D., Fisher, R.A., Raffa, K.F., Stitt, M., 2011. The interdependence of mechanisms underlying climate-driven vegetation mortality. Trends Ecol. Evol. 26, 523–532. Michelot, A., Simard, S., Rathgeber, C., Dufrene, E., Damesin, C., 2012. Comparing the intra-annual wood formation of three European species (Fagus sylvatica, Quercus petraea and Pinus sylvestris) as related to leaf phenology and nonstructural carbohydrate dynamics. Tree Physiol. 32, 1033–1045. Mitchell, T.D., Jones, P.D., 2005. An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int. J. Clim. 25, 693–712. Niinemets, U., 2010. Responses of forest trees to single and multiple environmental stresses from seedlings to mature plants, Past stress history, stress interactions, tolerance and acclimation. Forest Ecol. Manag. 260, 1623–1639. Nowacki, G.J., Abrams, M.D., 1997. Radial-growth averaging criteria for reconstructing disturbance histories from presettlement-origin oaks. Ecol. Monogr. 67, 225–324. Oliva, J., Colinas, C., 2007. Decline of silver fir (Abies alba Mill.) stands in the Spanish Pyrenees, Role of management, historic dynamics and pathogens. Forest Ecol. Manag. 252, 84–97. Prager, M.H., Hoenig, J.M., 1989. Superposed epoch analysis – a randomization test of environmental effects on recruitment with application to chub mackerel. T. Am. Fish. Soc. 118, 608–618. Rigling, A., Bigler, C., Eilmann, B., Feldmeyer-Christe, E., Gimmi, U., Ginzler, C., Graf, U., Mayer, P., Vacchiano, G., Weber, P., Wohlgemuth, T., Zweifel, R., Dobbertin, M., 2013. Driving factors of a vegetation shift from Scots pine to pubescent oak in dry Alpine forests. Global Change Biol. 19, 229–240. Rozas, V., Lamas, S., Garcia-Gonzalez, I., 2009. Differential tree-growth responses to local and large-scale climatic variation in two Pinus and two Quercus species in northwest Spain. Ecoscience 16, 299–310. Ruiz-Labourdette, D., Schmitz, M.F., Pineda, F.D., 2013. Changes in tree species composition in Mediterranean mountains under climate change, indicators for conservation planning. Ecol. Indic. 24, 310–323. Sangüesa-Barreda, G., Carlos Linares, J., Camarero, J.J., 2012. Mistletoe effects on Scots pine decline following drought events, insights from within-tree spatial patterns, growth and carbohydrates. Tree Physiol. 32, 585–598. Sangüesa-Barreda, G., Linares, J.C., Camarero, J.J., 2013. Drought and mistletoe reduce growth and water-use efficiency of Scots pine. Forest Ecol. Manag., doi:10.1016/j.foreco.2013.01.028. Sheffer, E., 2012. A review of the development of Mediterranean pine–oak ecosystems after land abandonment and afforestation, are they novel ecosystems? Ann. Forest Sci. 69, 429–443. Shifley, S.R., Fan, Z.F., Kabrick, J.M., Jensen, R.G., 2006. Oak mortality risk factors and mortality estimation. Forest Ecol. Manag. 229, 16–26. Suarez, M.L., Ghermandi, L., Kitzberger, T., 2004. Factors predisposing episodic drought-induced tree mortality in Nothofagus – site, climatic sensitivity and growth trends. J. Ecol. 92, 954–966. Tessier, L., Nola, P., Serrebachet, F., 1994. Deciduous Quercus in the mediterranean region – tree-ring/climate relationships. New Phytol. 126, 355–367. Van Bogaert, R., Haneca, K., Hoogesteger, J., Jonasson, C., De Dapper, M., Callaghan, T.V., 2011. A century of tree line changes in sub-Arctic Sweden shows local and regional variability and only a minor influence of 20th century climate warming. J. Biogeog. 38, 907–921. Van Mantgem, P.J., Stephenson, N.L., Byrne, J.C., Daniels, L.D., Franklin, J.F., Fule, P.Z., Harmon, M.E., Larson, A.J., Smith, J.M., Taylor, A.H., Veblen, T.T., 2009. Widespread increase of tree mortality rates in the Western United States. Science 323, 521–524.

82

G. Gea-Izquierdo et al. / Forest Ecology and Management 320 (2014) 70–82

Vicente-Serrano, S.M., Begueria, S., Lopez-Moreno, J.I., 2010. A multiscalar drought index sensitive to global warming, the standardized precipitation evapotranspiration index. J. Clim. 23, 1696–1718. Vila-Cabrera, A., Martinez-Vilalta, J., Galiano, L., Retana, J., 2013. Patterns of forest decline and regeneration across scots pine populations. Ecosystems 16, 323– 335. Voelker, S.L., Muzika, R.M., Guyette, R.P., 2008. Individual tree and stand level influences on the growth, vigor, and decline of red oaks in the Ozarks. Forest Sci. 54, 8–20. Weber, P., Rigling, A., Bugmann, H., 2008. Sensitivity of stand dynamics to grazing in mixed Pinus sylvestris and Quercus pubescens forests, a modelling study. Ecol. Model. 210, 301–311.

Wischnewski, J., Kramer, A., Kong, Z., Mackay, A.W., Simpson, G.L., Mischke, S., Herzschuh, U., 2011. Terrestrial and aquatic responses to climate change and human impact on the southeastern Tibetan Plateau during the past two centuries. Global Change Biol. 17, 3376–3391. Zahner, R., Saucier, J.R., Myers, R.K., 1989. Tree-ring model interprets growth decline in natural stands of loblolly-pine in the Southern United States. Can. J. Forest Res. 19, 612–621. Zweifel, R., Bangerter, S., Rigling, A., Sterck, F.J., 2012. Pine and mistletoes, how to live with a leak in the water flow and storage system? J. Exp. Bot. 63, 2565– 2578.

Forest Ecology and Management

ted parameters in the best model we calculated the predicted probability for each tree ..... logitðPkqÞ ¼ akq þ b1q Á NGC þ b2q Á s1970 þ b3q Á ms þ b4q Á Density ahealthy-q .... tissements d'Avenir» program of the French National Research.

2MB Sizes 2 Downloads 116 Views

Recommend Documents

pdf-1449\forest-ecology-by-stephen-hopkins-barnes-burton-verne ...
pdf-1449\forest-ecology-by-stephen-hopkins-barnes-burton-verne-spurr.pdf. pdf-1449\forest-ecology-by-stephen-hopkins-barnes-burton-verne-spurr.pdf. Open.

plant: A package for modelling forest trait ecology ... - GitHub
Department of Biological Sciences, Macquarie University, Sydney, Australia .... of leaf area, including construction of the leaf itself and various support structures.

The Impacts of Joint Forest Management on Forest Condition ...
Sep 1, 2010 - Figure 5.4 Asset nets, types of assets compared across wealth percentile groups ..... Has JFM created sustainable forest governance institutions at the village level ... while Chapter 7 focuses on the impact on household livelihoods and

The Impacts of Joint Forest Management on Forest Condition ...
Sep 1, 2010 - completed this thesis without the support of a number of people. .... being established in 3.6 million ha of forest land and in 1,800 villages ...... In Maseyu village, 57% of the respondents had business income (Table A5.19,.