Agricultural and Forest Meteorology 239 (2017) 141–150

Contents lists available at ScienceDirect

Agricultural and Forest Meteorology journal homepage: www.elsevier.com/locate/agrformet

Poor acclimation to current drier climate of the long-lived tree species Fitzroya cupressoides in the temperate rainforest of southern Chile J. Julio Camarero 1,∗ , Alex Fajardo 2 1

Instituto Pirenaico de Ecología (IPE-CSIC), Avda. Monta˜ nana 1005, 50192 Zaragoza, Spain Centro de Investigación en Ecosistemas de la Patagonia (CIEP) Conicyt–Regional R10C1003, Universidad Austral de Chile, Camino Baguales s/n, Coyhaique 5951601, Chile 2

a r t i c l e

i n f o

Article history: Received 5 October 2016 Accepted 3 March 2017 Keywords: Intrinsic water-use efficiency Isotopes Nothofagus Patagonia Tree growth

a b s t r a c t Climate change and rising atmospheric CO2 concentrations (ca ) are expected to affect forests worldwide. The effects of climate change, however, have not been deeply assessed in humid forest biomes from the southern Hemisphere where climate is not warming but drying. This is the case of the temperate rainforest in southern Chile, where the endemic and threatened long-living gymnosperm Fitzroya cupressoides occurs. We assessed how radial growth, intrinsic water-use efficiency (iWUE) and tree-ring ␦18 O responded to increasing ca and decreasing precipitation in F. cupressoides and companion species. We hypothesized that F. cupressoides, a long-lived and probably less plastic species, will show less acclimation to global-change effects than co-occurring Nothofagus species which show broader climatic niche. Thus, F. cupressoides should display iWUE increases different from the ci /ca constant scenario, which represents an active mechanism to increase intercellular CO2 concentrations (ci ) as ca rises. Although cool and wet conditions during the growing season enhanced growth of all species, particularly in F. cupressoides, growth of F. cupressoides declined noticeably since the 1980s in response to a decrease in precipitation. Current drier conditions led to increased iWUE in Nothofagus species. According to ␦18 O values, this increased in iWUE should be due to a decrease in stomatal conductance. Fitzroya cupressoides, however, displayed a decrease in iWUE in response to drier conditions, shifting from an active ci /ca scenario to a more passive ci /ca scenario, and maintaining a relatively constant stomatal conductance. Using multiple bodies of evidence, our findings indicate a poor adaptability of the long-lived F. cupessoides to drier conditions despite rising ca . Thus, not all species are having similar and expected responses to increasing ca , which should be a call of attention in the case of long-lived, endangered and narrow-distributed species, like F. cupressoides. © 2017 Elsevier B.V. All rights reserved.

1. Introduction In 2015, the Mauna Loa observatory recorded atmospheric CO2 concentrations (ca ) over 400 ppm and global air temperatures were +1.0 ◦ C warmer than during preindustrial times (Blunden and Arndt, 2016). These two global-change milestones may be biased towards the much intensive carbon-emitter northern Hemisphere as compared with southern latitudes. Despite ca over the Antarctica reached also a record value of 399 ppm in 2016 (see http://www. esrl.noaa.gov/gmd/obop/spo/), some austral regions as southern South America are among the few land areas where temperature records did not show a clear warming trend (Bidegain et al., 2016).

∗ Corresponding author. Tel.: +34 976 716031; fax: +34 976 716019. E-mail addresses: [email protected] (J.J. Camarero), [email protected] (A. Fajardo). http://dx.doi.org/10.1016/j.agrformet.2017.03.003 0168-1923/© 2017 Elsevier B.V. All rights reserved.

Such understudied regions may provide valuable information to disentangle long-term forest responses to changes in ca and temperatures. In principle, forests were expected to show enhanced growth rates in response to rising ca and warmer temperatures. This expectation was based on the assumption that trees would exhibit improved photosynthesis and hence growth rates due to an increase in their intrinsic water-use efficiency (iWUE, i.e. the carbon fixed per unit of water transpired through stomata) through a CO2 -fertilization effect (Norby et al., 2005). These relationships, however, were mostly constructed in short-term experimental studies. A much integral approach to understanding the response of plant to changes in ca and temperatures is by comparing decadal to centennial tree-ring records of growth and iWUE through the use of C isotopic ratios (e.g. Linares and Camarero, 2012; Saurer et al., 2004). Such long-term approaches have indicated that despite for-

142

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

est water-use efficiency has generally increased during the 20th century—inferred from tree-ring C isotopes (Saurer et al., 2004) or from eddy-covariance flux data (Keenan et al., 2013), this increase has not led to enhanced forest growth (Lévesque et al., 2014; ˜ Penuelas et al., 2011; Silva and Anand, 2013). Additionally, warmer temperatures may also lead to higher water deficit in droughtprone areas translating this into growth declines or a lack of response to rising ca in some tree species (Camarero et al., 2015; Gómez-Guerrero et al., 2013). Despite this knowledge, we know little about how forest communities less used to dry conditions (e.g. temperate rainforests) will respond to changes in ca and temperature. In fact, increasing concern about the potential negative effects of global change on trees has shifted from semi-arid areas to wet forests that were initially not considered at drought risk. Inferring how changes in ca and temperature will affect forests may become challenging given that some species are more susceptible or resistant-tolerant to drought than others (Granda et al., 2014). In particular, it is not clear at all how long-lived tree species will react to global change. In principle, long-lived tree species could become more susceptible to sudden changes in climate as dry spells because they are expected to have a rather narrow climatic niche. In contrast, we could also expect that long-lived tree species may be more resilient to rather recent changes in ca and temperature. In this respect, long-lived tree species could also function as long-term monitors (sentinels) of global-change components (rising ca and warmer air temperatures) given that long-living tree species respond differently to these two environmental drivers (Voelker et al., 2006). The iWUE of trees can be retrospectively inferred using the 13 C/12 C (␦13 C) isotope ratio from tree-ring wood or cellulose which simultaneously tracks changes in assimilation rates and stomatal conductance under increasing ca (Francey and Farquhar, 1982; McCarroll and Loader, 2004). In normal conditions, C3 carbon fixation plants photosynthesis discriminates against 13 C (the heaviest isotope), but if gas exchange is constrained by stomatal closure (e.g. during warm and dry conditions), 13 C is used along with 12 C and the ␦13 C becomes less negative (Farquhar et al., 1989). In most treering based studies assessing long-term ca and iWUE associations, the temporal variation in ␦13 C supports the prevalence of an active plant mechanism that maintains a constant ratio between intercellular (ci ) and atmospheric (ca ) CO2 concentrations (Leonardi et al., 2012; Saurer et al., 2004). Recently, the analysis of the 18 O/16 O (␦18 O) isotope ratio has also been applied to tree-ring based studies since ␦18 O reflects the changes in stomatal conductance (g) and it is considered a proxy for evaporative flux, relative humidity and ␦18 O source-water signal (Anderson et al., 1998; Barbour, 2007; Voltas et al., 2013). The use of ␦18 O has been helpful to clarify whether the rising iWUE trends are more related to increasing photosynthesis assimilation rates (A) or to decreasing g rates (Saurer et al., 1997). In this study, we were interested in determining how increasing ca , temperature and dryness are impacting the growth (in the form of basal area increment, BAI) and the intrinsic water use efficiency (iWUE) of tree species in a temperate rainforest. We focused on the temperate rainforest (TRF) of western southern South America (Chile), which is known to harbour a relatively high number of tree species, including the endangered, long-lived and high-biomass conifer Fitzroya cupressoides (Cupressaceae) (Armesto and Figueroa, 1987; Urrutia-Jalabert et al., 2015b; Veblen and Schlegel, 1982). This temperate rain forest also includes many endemic species, all adapted to wet and cool conditions (annual precipitations can surpass 4000 mm), which are suitable sentinels to assess how tree species are responding to global-change components throughout the late 20th century, when ca and temperatures have rapidly increased (e.g. Salzer et al., 2009). In contrast to most other regions, the Southern South American temperate rain forest is not experiencing a clear warming trend but a decrease in precipitation during

the second half of the 20th century accompanied by the occurrence of more severe droughts (Fuenzalida et al., 2007; Quintana and Aceituno, 2012). Although Urrutia-Jalabert et al. (2015a) found that growth and iWUE responses to rising ca varied in F. cupressoides between sites and populations, we still do not know how this longlived species reacts to global warming effects when it is compared to other tree species members of the temperate rain forest under a scenario of lower precipitation. We assessed here the temporal change in tree secondary growth (BAI), iWUE (␦13 C) and ␦18 C during the last 50–100 years of F. cupressoides as compared to other tree species in a temperate rain forest in southern Chile. Given that this temperate rain forest is subjected to wet and cool conditions, we first expected than ca -induced changes in iWUE will mainly correspond to changes in photosynthesis rates (A) and not in stomatal conductance (g). Second, since selection pressures for long-lived, tall tree species, such as F. cupressoides, are very strong (Lanfear et al., 2013), and must therefore, select for greater specialization or less plasticity—narrower climatic breadth, we also expected that F. cupressoides will display less acclimation to global-change effects than other companion species with greater climatic breadth; i.e. an increase of iWUE more different from the constant ci /ca scenario, which represents an active mechanism to increase ci as ca rises, than other companion species. Marked differences in how different tree species face drier conditions and rising ca , especially in temperate rainforests, will have consequences for the future structuring of these unique communities where species with low acclimation may be outcompeted by other more resilient species.

2. Material and methods 2.1. Study sites and tree species Temperate rainforests in southern Chile are characterized by a cold-temperate and super-humid climate (Alaback, 1991; Luebert and Pliscoff, 2006), with mean annual temperature of ∼10 ◦ C and annual precipitation of 2500 mm or more, regularly distributed throughout the year (Appendix, Fig. S1). We worked in the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile. Here, the topography and geomorphology are characterized by fjords, steep slopes and glacial valleys. The annual precipitation ranges from 2000 to 6500 mm, with a very wet winter and a less wet summer (Dávila et al., 2002). Soils are rather acid (mean pH = 4.7), well developed (mean N concentration = 0.96%). The aspect of the study area is W–NW (270–335◦ ), where slopes ranged from 5 to 40◦ . The vegetation at Huinay is typical of the temperate rain forest; a dense understory of bamboos (Chusquea spp.) and ferns (Blechnum spp., Lophosoria spp.) (Donoso, 1993; Tecklin et al., 2011), broadleaf evergreen angiosperms as dominant tree species, and the presence of the long-lived conifer species, Fitzroya cupressoides I.M. Johnst (Veblen and Schlegel, 1982; Villagrán, 1985; Lara and Villalba, 1993; Donoso et al., 2006). At low-elevation (140 m a.s.l.), the dominant tree species include Nothofagus dombeyi Mirb. Oerst. and N. nitida (Phil.) Krasser (Nothofagaceae), Weinmannia trichosperma Ruiz & Pav., Eucryphia cordifolia Cav. (Cunoniaceae), Laureliopsis philippiana (Looser) R. Schodde (Atherospermataceae) and Luma apiculata (DC.) Burret (Myrtaceae). At mid-elevation (560 m a.s.l.), the dominant tree species are N. dombeyi and Saxegothaea conspicua Lindl. (Podocarpaceae), N. betuloides (Mirb.) Oerst., and Podocarpus nubigena Lindl. (Podocarpaceae). Finally, at high-elevation (780 m a.s.l.), N. betuloides and F. cupressoides (Cupressaceae) dominate, along with W. trichosperma, P. nubigena, N. antarctica (G.Forst.) Oerst., and Desfontainia spinosa Ruiz & Pav. (Columelliaceae). In short, the low-elevation site is compositionally similar to a TRF sensu stricto, without conifers present and

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

where Nothofagus species are dominant (sensu Veblen and Schlegel, 1982); the mid-elevation site is similar to a North Patagonian forest type with Podocarps (e.g., S. conspicua) coexisting with Nothofagus species; and at the upper elevation forest F. cupressoides and N. betuloides become dominant. Nothofagus betuloides, N. dombeyi and N. nitida are evergreen species, and pre- and neo-shoot formation has been observed in N. dombeyi which dominates moist lowlands of central and southern Chile. The conifer species here grow slowly and can become very old (Donoso et al., 2006); in fact, F. cupressoides is one of the oldest tree species in the world reaching ages of at least 3610 years (Lara and Villalba, 1993). At intra-annual time scales the radial growth of F. cupressoides is enhanced by wet and humid conditions and a low vapour pressure deficit (Urrutia-Jalabert et al., 2015c), whereas the growth of pioneer species as N. nitida species respond negatively to warm and sunny conditions probably increasing the evaporative water demand (Pérez et al., 2009). The elevation treeline here is located at ca. 1350 m a.s.l. and it is dominated by Nothofagus pumilio. 2.2. Field sampling We established sampling plots at three different elevations (low-, mid-, and high-elevation) at the Huinay area (Table 1 and Fig. 1). We avoided sampling in disturbed (recent landslides) forests. We sampled two dominant tree species at each location. Specifically, we sampled N. dombeyi and N. nitida at low-elevation, N. dombeyi and S. conspicua at mid-elevation, and F. cupressoides and N. betuloides at high-elevation. At each of the three locations, in a 1ha large area we selected 8–10 individuals of each species, for which we measured diameter at breast height (DBH, 1.3 m of height) and total height using diameter tape and a clinometer, respectively. We then extracted two to three to-the-pith stem cores per tree at DBH and perpendicular to the main slope using 5.15 mm Pressler increment borers (Haglöf, Långsele, Sweden). We targeted dominant trees of relatively similar size (Table 2) to reflect same-age tree growth responses to climate change throughout the 20th century (e.g. ca ). When trees were thicker than 20 cm diameter at DBH, two cores were extracted in order to get best estimates of growth rates (see below Tree growth determination). An additional core was kept for isotope analyses (see below Carbon and oxygen isotope analyses). 2.3. Climate data To characterize climate conditions (air mean temperature and relative humidity of the air) in situ and at intra-annual scale, three Hobo HOBO Pro RH/Temp data Loggers (Onset, Bourne, USA) data loggers were placed at each site location, protected into corresponding radiation shields. Data were taken every 30 min from February to October 2012 and then converted to daily values. To obtain climate data for the 1951–2010 period, with available tree-ring data for all selected tree species, we used monthly mean temperature and precipitation data from the Climate Research Unit (CRU) version 3.22 with gridded data at a 0.5◦ resolution (Harris et al., 2014). Data were extracted for the grid with coordinates 42.0–42.5◦ S and 72.0–72.5◦ W using the Climate Explorer webpage (https://climexp.knmi.nl/), and then compared with data from the Puerto Montt meteorological station which holds the longest and most complete climate record in the study area for the given period. 2.4. Tree-ring width data Cores were prepared following dendrochronological techniques (Stokes and Smiley, 1996). Cores were dried, mounted and glued in grooved wooden sticks, and then sanded with successively finer grades of sandpaper until optimal surface resolution allowed annual rings to be visible under magnification (×10). Following

143

visual cross-dating, tree-ring widths were measured to the nearest 0.001 mm using a microscope mounted on a Lintab-TSAP measuring device (F. Rinn, Heidelberg, Germany). Calendar dates were assigned to rings according to the southern hemisphere tree-ring dating convention that assigns an annual ring to the calendar year in which the annual formation begins (Schulman, 1956). Cross-dating and ring-width measurements for each site-species combination underwent a quality control with the COFECHA program (Holmes, 1983). To quantify growth rates we calculated the basal area increment (BAI), which was computed as:



2 BAI = ˘ Rt2 − Rt−1



(1)

where Rt and Rt−1 are the stem radii in years t and t − 1. These radii were computed by considering diameter at coring height, and the distance between the geometric theoretical pith and the innermost ring in the core (Fajardo and McIntire, 2012). Tree-ring width series were converted into ring-width indices and mean chronologies for each site and species. First, the individual series were standardised using a cubic-smoothing spline curve of 60 years with a 50%-frequency response cut-off. This procedure eliminates biological trends as trees enlarge, hence preserving high-frequency growth variability that is potentially related to climate. Next, an autoregressive model was applied to remove the first-order temporal autocorrelation in the detrended series and generate residual or pre-whitened indices. Lastly, a bi-weight robust mean was computed to produce residual chronologies for each site and species. These procedures were done using the ARSTAN program (Cook and Krusic, 2005). All chronologies surpassed the expressed population signal (EPS) threshold of 0.85 for the common 1951–2010 period, which is considered as a reliable criterion of chronology replication (Wigley et al., 1984), excepting the series from the low-elevation site which encompassed the 1972–2010 period. 2.5. Carbon and oxygen isotope analyses For C and O isotope analyses, we considered those tree species with more individuals established before the 20th century at the high and mid-elevation sites (F. cupressoides, N. betuloides and N. dombeyi). Thus, we selected the five oldest trees (age 150–200 yrs.) of these species whose cross-dated tree-ring series presented a high correlation with the residual chronology of each species. Although we used a small number of trees per species at each site in isotopic analyses, this number is considered adequate to represent stand-level trends given the typical consistency of envi˜ ronmental signals (Penuelas et al., 2008). For each core, rings were separated in decadal groups using scalpels. Wood samples were purified to ␣-cellulose to remove the extractives and lignin following methodological recommendations by Voltas et al. (2013). The resulting samples were homogenized to a fine powder with a ball mixer mill (Retsch MM301, Haan, Germany). Then, aliquots were weighed on a microbalance (Sartorius, 0.01 mg) into tin foil capsules and combusted to CO2 using a Flash EA-1112 elemental analyser interfaced with a Finnigan MAT Delta C isotope ratio mass spectrometer (Thermo Fisher Scientific Inc., MA, USA). Similarly, for O isotope analysis, samples were weighed into silver foil capsules and combusted using a Carlo Erba 1108 elemental analyser (Carlo Erba Instruments Ltd., Milan, Italy) interfaced with a Finnigan Deltaplus XP isotope ratio mass spectrometer (Thermo Fisher Scientific Inc.). Both isotopic analyses were conducted at the Stable Isotope Facility of the University of California at Davis (USA). Stable isotope ratios were expressed as per mil deviations using the ␦ notation (␦13 C, ␦18 O) relative to Vienna Pee Dee Belemnite (VPDB) and Vienna Standard Mean Ocean Water standards for C and O, respectively. The standard deviation for the repeated analysis of standard cellulose was better than 0.1‰ for C and better than

144

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

Table 1 Characteristics of the three study sites and tree species sampled at the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile. Site

Latitude (S) ◦

Elevation (m a.s.l.)

Slope (◦ )

Species

◦

Fitzroya cupressoides Nothofagus betuloides Nothofagus dombeyi Saxegothea conspicua Nothofagus dombeyi Nothofagus nitida

Longitude (W)

High

42.388

72.397

780

15

Middle

42.387◦

72.400◦

560

20

Low

42.382◦

72.411◦

140

35

Fig. 1. Location of the three study sites (boxes) sampled along the altitudinal gradient in the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile (a). Representative views of the studied forests in the high- (b), mid- (c) and low-elevation (d). In the (a) plot, the inset shows the location of Huinay in southern Chile, while the buildings in the lower portion of panel (a) correspond to the Huinay Scientific Station.

Table 2 Mean values (±1SE) for growth trends and size characteristics of the different tree species sampled at the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile. Site

Species

DBH (cm)

Height (m)

# Trees (# cores)

Tree-ring width (mm)

Timespan†

Correlation with site mean series§

High

F. cupressoides N. betuloides

47.6 ± 4.9 32.4 ± 2.1

30.1 ± 1.8 25.8 ± 1.7

14 (28) 10 (20)

0.56 ± 0.10 1.07 ± 0.14

1544–2010 (1840–2010) 1809–2010 (1902–2010)

0.43 0.30

Middle

S. conspicua N. dombeyi

41.3 ± 6.7 47.0 ± 4.0

22.2 ± 1.2 27.7 ± 1.6

15 (30) 9 (18)

1.75 ± 0.53 1.54 ± 0.23

1861–2010 (1904–2010) 1867–2010 (1883–2010)

0.42 0.36

Low

N. dombeyi N. nitida

55.2 ± 2.5 37.8 ± 3.5

24.6 ± 0.8 21.2 ± 1.5

17 (34) 11 (22)

5.90 ± 0.42 5.54 ± 0.66

1956–2010 (1972–2010) 1964–2010 (1972–2010)

0.53 0.40

† §

Best-replicated timespan. Mean correlation coefficient calculated from the comparison of all individual series with the site chronology of each species.

0.3‰ for O. Finally, since ca is rising after the beginning of industrialization, particularly after the 1950s, leading to a decrease in ␦13 C of atmospheric CO2 due to the emission of fossil CO2 , which is depleted in 13 C (Suess effect), we applied the atmospheric correction to ␦13 C values to obtain ␦13 Catm. corr values following McCarroll et al. (2009).

Tree-ring ␦18 O is mainly influenced by the ␦18 O of precipitation, and secondarily by plant interception, evaporation from surface retention, and soil evaporative enrichment (McCarroll and Loader, 2004). We could not calculate tree-ring 18 O by subtracting the mean decadal ␦18 O precipitation values from corresponding ␦18 O tree-ring values because reliable ␦18 O measurements in growing-

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

season (from November to March) precipitation are not available for Puerto Montt station since 1989 (see Global Network of Isotopes in Precipitation program, http://www.naweb.iaea.org/napc/ih/GNIP/IHS GNIP.html). 2.6. Models of intrinsic water-use efficiency (iWUE) First, to account for changes in ␦13 C of atmospheric CO2 (␦13 ca ), we calculated C isotope discrimination in cellulose (13 C) from ␦13 ca and plant ␦13 C (␦13 cp ), as described by Farquhar and Richards (1984): 13

13 C =

13

␦ ca − ␦ cp

(2)

13

1 + ␦ cp /1000

␦13 ca was obtained from published data (McCarroll and Loader, 2004). Second, following Farquhar et al. (1982), we estimated intrinsic water-use efficiency (iWUE) using the equation: iWUE =



A = ca 1 − g

 c  i

ca

0.625

(3)

where A is the rate of net photosynthesis, g is stomatal conductance to H2 O, ci is the internal CO2 concentration in the sub-stomatal internal cavity of leaves, ca is the ambient air CO2 concentration, and 0.625 is the relation among the conductance of H2 O compared to the conductance of CO2 . The iWUE (␮mol mol−1 ), defined as the A to g ratio, has been widely related to long-term trends in the internal regulation of C uptake and water loss in tree species (see McCarroll and Loader, 2004). Third, to determine ci , we used the equation proposed by Francey and Farquhar (1982):



ci = ca

13

13

(␦ cp − ␦ ca + 1) (b − a)



(4)

where ␦13 cp and ␦13 ca are the plant and atmospheric C isotope compositions, respectively, a is the diffusion fractionation across the boundary layer and the stomata (+4.4‰), b is the Rubisco enzymatic biologic fractionation (+27.0‰), and ci and ca are intercellular and ambient CO2 concentrations, respectively. According to previous studies (Linares and Camarero, 2012; Saurer et al., 2004; Seibt et al., 2008), we considered three theoretical scenarios of iWUE, namely (i) constant ci , (ii) constant ci /ca , and (iii) constant ca − ci . These three scenarios differ in the degree to which the increase in ci follows the increase in ca : either (i) not at all, (ii) in a proportional way or (iii) at the same rate. In scenario (i), the iWUE increases strongly, while in scenario (ii) the iWUE is improved, but not as strongly as in scenario (i), because there is a proportional regulation of A and g. Finally, in scenario (iii), ci and ca follow similar increasing rates and iWUE is not improved. To assess the goodness of fit of the three iWUE theoretical models to actual data (decadal iWUE values inferred from tree-ring C isotopes) we used the Root Mean Squared Error (RMSE), which is the square root of the variance of the residuals. This statistic quantifies the absolute fit of the model to the data, and it can be interpreted as the standard deviation of the unexplained variance. Lower values of RMSE indicate better fit. RMSE is a good measure of model accuracy if the main purpose is prediction because it measures the average deviation of the fitted from the observed data (Chatfield, 2000). 2.7. Statistical analyses Trends in monthly climate data from the Puerto Montt station and in growth (BAI) were assessed using the Kendall correlations

145

( statistic). To compare mean values of several variables (local climate data, BAI, isotope data), we used simple t tests. Correlations between monthly climate variables and growth indices were calculated for the temporal window including the previous August to current March, which was chosen based on previous studies (e.g. Suarez and Kitzberger, 2010), and using Pearson coefficients (r). All statistical analyses were performed in R version 3.3.1 (RDevelopment-Core-Team, 2016).

3. Results 3.1. Regional climatic trends and local climate differences among locations According to the local climate data from the Puerto Montt meteorological station, autumn (April,  = −0.30; May,  = −0.34) and late spring (November,  = −0.39) temperatures have decreased significantly (p < 0.005) since 1951, while winter precipitation (July and August,  = −0.25; September,  = −0.27) has also diminished (Appendix, Fig. S2). Both summer precipitation ( = −0.13) and temperature ( = −0.21) have decreased at Puerto Montt since 1951, albeit only the cooling trend was significant (p = 0.016). The local climate data showed significantly colder conditions at the high-elevation site (mean ± SE, 5.35 ± 0.24 ◦ C) than at the low(8.87 ± 0.24 ◦ C) and mid-elevation (6.89 ± 0.22 ◦ C) sites (high- vs. low-elevation site, t = −10.44, high- vs. mid-elevation site t = −4.80; p < 0.001 in both cases). Temperatures below zero were only detected at the high-elevation site reaching a minimum value of −1.57 ◦ C (Fig. 2). Minimum values for the relative humidity in air (50–70%) were only registered at the high-elevation site. Humidity differed more between the high- (95.0 ± 0.7%) and mid-elevation (99.3 ± 0.2%) sites than between the high- and low-elevation (96.8 ± 0.3%) sites (high- vs. low-elevation site, t = −2.50, high- vs. mid-elevation site t = −6.28; p < 0.01 in both cases), although dayto-day variability was more similar between the two highest sites. The relative humidity in the air reached 100% in 66%, 88% and 43% of all measurements at the high-, mid- and low-elevation sites, respectively.

3.2. Growth trends At high elevation, the mean annual growth rate of F. cupressoides (12.9 ± 0.3 cm2 yr−1 ) was always significantly higher (t = 13.07, p < 0.001) than that of N. betuloides (8.6 ± 0.2 cm2 yr−1 ). At midelevation, the growth rate of N. dombeyi (17.3 ± 0.6 cm2 yr−1 ) was significantly lower (t = −3.34, p < 0.001) than that of S. conspicua (22.3 ± 1.4 cm2 yr−1 ), but this difference appeared only after a shift in the growth ranking which occurred in the 1970s. In contrast, at low elevation, the growth rate of N. dombeyi (60.8 ± 3.8 cm2 yr−1 ) was significantly higher (t = 5.54, p < 0.001) than that of N. nitida (31.6 ± 3.6 cm2 yr−1 ). At the high-elevation site, no significant trend in BAI was observed, although since the 1980s BAI has declined significantly in F. cupressoides ( = −0.79, p = 0.001) and N. betuloides ( = −0.29, p = 0.030). We found a general significant positive increase in BAI for S. conspicua ( = 0.77, p < 0.001; period 1951–2010) at mid-elevation and in both low-elevation Nothofagus species (N. dombeyi,  = 0.72; N. nitida,  = 0.91; in both cases p < 0.001; period 1972–2010; see Fig. 3). Considering the post-1980 period, the mid-elevation tree species also showed significant BAI increases (p < 0.05). The chronologies of all species were significantly related between them for the 1972–2010 period, with the exception of the high-elevation N. betuloides chronology which only showed significant and positive associations with the high-elevation F.

146

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

Months F

M

A

M

J

J

A

20

S

(a)

20 High Mid Low

15

10

10 5

5

0

(b)

2

-1

Basal area increment (cm yr )

Temperature (ºC)

15

0

100

Relative humidity (%)

N. betuloides F. cupressoides

90

80

40

N. dombeyi S. conspicua

30

20

10

0 100

70

(c)

N. dombeyi N. nitida

80

60 60

50

40

30

60

90

120

150

180

210

240

270

Days Fig. 2. Local climate conditions (mean temperature, air relative humidity) measured from February to October 2012 in the three study sites (high-, mid- and low-elevations) located along the altitudinal gradient of the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile.

cupressoides (p = 0.04) and the mid-elevation S. conspicua (p = 0.02) mean series. 3.3. Climate–growth relationships In general, wet and cool conditions from November to March were associated to improved growth in all species (Table 3). At high elevation, F. cupressoides negatively responded to warm and dry conditions, but at low elevation warm and wet conditions were associated to growth enhancement (e.g. N. nitida). Climate conditions prior to the growing season were also correlated to ringwidth indices of S. conspicua and the two low-elevation Nothofagus species. 3.4. Trends in iWUE, ı13 C and ı18 O The observed mean (±1SE) tree-ring ␦13 C decadal values were significantly higher (post-hoc Tukey test, Q = 11.35, p = 0.001) for F. cupressoides (␦13 C = −22.68 ± 0.22) than for the two Nothofagus species (N. betuloides, ␦13 C = −24.96 ± 0.23; N. dombeyi, ␦13 C = −25.57 ± 0.14). The ␦13 C decadal values showed significant (p = 0.002) negative trends for all three species, being more pronounced in N. betuloides (−0.023‰ yr−1 ) and F. cupressoides (−0.021‰ yr−1 ) than in N. dombeyi (−0.012‰ yr−1 ). These rates correspond to mean ␦13 C decreases throughout the 20th century of −2.08‰, −1.79‰, and −0.87‰ for N. betuloides, F. cupressoides, and N. dombeyi, respectively.

20

0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

Year Fig. 3. Trends of mean annual growth rate (expressed as basal area increment) of five tree species sampled at the high- (a), mid- (b) and low-elevation sites at the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile. Values are means ± 1SE.

On average, F. cupressoides showed a significantly (p < 0.01 in all comparisons between couple of species) higher iWUE value (98.6 ± 1.4 ␮mol mol−1 ) than N. betuloides (81.8 ± 1.5 ␮mol mol−1 ) and N. dombeyi (70.0 ± 1.8 ␮mol mol−1 ) (Fig. 4). The average ␦18 O values also differed significantly among species with F. cupressoides showing the lowest value (30.12 ± 0.06‰), followed by N. betuloides (30.87 ± 0.09‰) and N. dombeyi (31.64 ± 0.07 ␮mol mol−1 ) (Fig. 5). Overall, these results correspond to higher iWUE and lower ␦18 O values at the high-elevation for F. cupressoides but lower iWUE and more enriched tree-ring ␦18 O values in Nothofagus species at highand mid-elevation sites (N. dombeyi, r = 0.80, p = 0.003; N. betuloides, r = 0.74, p = 0.009; Fig. 5). Note that ␦18 O values of the mid-elevation N. dombeyi trees are becoming progressively more enriched after 1950 than that of high-elevation N. betuloides trees (t = 5.07) and F. cupressoides (t = 12.10) trees (Figs. 5 and 6). 3.5. iWUE models and relationships with ı18 O Both Nothofagus species showed convergent iWUE values towards the present decade (Fig. 4). In fact, during the first half of the 20th century, ␦13 C mean values (±1SE) for N. dombeyi (␦13 C = −25.28 ± 0.11) were significantly lower (t = −4.82, p = 0.008) than those of N. betuloides (␦13 C = −24.26 ± 0.18). However, since the 1950s onwards this difference was not significant (N. dombeyi,

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

147

Table 3 Climate-growth associations (Pearson coefficients) calculated by relating tree-ring width indices and monthly climate variables (T, mean temperature; P, precipitation) considering the 1951–2010 period common to all tree species excepting the low-elevation species (period 1972–2010). The months with shaded backgrounds correspond to the Austral growing season. Significant correlations (p < 0.05) are indicated in bold font. Site

High

Species

A

F. cupressoides N. betuloides

T 0.19 0.08

P 0.20 0.10

T −0.12 0.15

P −0.10 0.01

T −0.13 −0.15

−0.20 0.23

−0.02 −0.09

−0.19 0.16

−0.14 −0.01

0.23 0.01

0.22 0.27

−0.05 0.02

−0.04 0.17

Middle S. conspicua N. dombeyi Low

N. dombeyi N. nitida

S

O

N P

D P

0.10 −0.02

T −0.07 −0.11

−0.23 −0.06

0.06 −0.03

−0.18 −0.06

0.07 0.10

J

F

M

0.17 0.21

T P −0.21 0.22 −0.19 0.14

T −0.28 0.06

P T 0.20−0.24 0.09−0.02

−0.38 0.05

0.24 −0.10

−0.41 0.17 −0.22 0.23

−0.19 −0.11

0.32 −0.19

−0.14 0.12

−0.17 0.16 −0.35 0.30

−0.16 −0.13

33

P 0.03 0.01

T −0.21 −0.10

P 0.34 0.09

−0.14−0.33 0.11−0.01

0.10 0.10

−0.22 −0.21

−0.05 0.19

0.01 0.22 −0.27 0.23

0.27 −0.08

−0.26 0.19

0.32 0.10

F. cupressoides N. betuloides N. dombeyi

31

18

δ O (‰)

32

30

29 1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

2010

Decade Fig. 5. Decadal mean values (± 1SE, n = 5 trees per species) of tree-ring O isotopes (␦18 O) for the three study species at the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile. Continuous and dashed lines indicate significant (p < 0.05) and non-significant trends for the two Nothofagus species and F. cupressoides, respectively.

Fig. 4. Decadal values of intrinsic water-use efficiency (iWUE, symbols) during the 1900–2010 period for three tree species (a) and theoretical models (continuous and dashed lines) fitted to iWUE values (b). In the upper panel, the right y axis shows the atmospheric CO2 concentration (grey-filled area). Abbreviations: ca , atmospheric CO2 concentration; ci , internal CO2 concentration in the sub-stomatal internal cavity of leaves. Values are means ± 1SE (n = 5 trees per species).

␦13 C = −25.81 ± 0.19; N. betuloides, ␦13 C = −25.53 ± 0.17; t = −1.07, p = 0.314). For the period 1900–2010, the constant ci /ca model provided the best fit (lowest RMSE values) for all species, particularly in the two Nothofagus species (Table 4). These differences were more pronounced for the period 1960–2010, with the exception of F. cupressoides whose iWUE values were better explained by the constant ca − ci model. Considering the three species, ␦18 O and iWUE were negatively associated in agreement with the observed ranking of highest iWUE in F. cupressoides and highest ␦18 O in N. dombeyi, while N. betuloides occupied an intermediate position (Fig. 6). 4. Discussion 4.1. Drier conditions override the potential positive effects of rising CO2 on the long-lived Fitzroya cupressoides We uncovered contrasting responses to global-change drivers among the temperate rainforest species under study. Contrasting growth responses to increasing ca and drier conditions also suggest contrasting physiological mechanisms among species to deal with these changes. In particular, the long-lived F. cupressoides appeared to have a poor acclimation to changes in climate; neither growth (Fig. 3) nor iWUE (Fig. 4) have steadily increased in F. cupressoides as it did occur in coexisting Nothofagus species. In fact, we observed a recent growth decline in F. cupressoides, which may be associated

to drier conditions in the study region and to the high sensitivity of this species to dry and warm conditions during the growing season (negative growth-climate associations, Table 3). The decline in growth of F. cupressoides occurred when iWUE changes implied that maximum photosynthesis rates should increase (as suggested by Urrutia-Jalabert et al., 2015a), but not as much as expected in response to rising ca . According to ␦13 O and ␦18 O tree-ring results, F. cupressoides exhibits a more passive response to increasing ca and dry conditions (shifts from constant ci /ca to constant ca − ci scenarios) and a low stomatal regulation of transpiration as compared with Nothofagus species (Fig. 5). Given these evidences, we, therefore, found only partial support to our expectation that F. cupressoides would show an increasing iWUE more different from the constant ci /ca scenario; i.e. the ci /ca scenario was the best model for iWUE changes in all tree species, but F. cupressoides (Fig. 4). In fact, during the 1960–2010 period, the ca − ci scenario was selected for F. cupressoides (Table 4). Since F. cupressoides is more abundant at the high-elevation site and atmospheric pressure decreases with elevation, the effect of any increase in the partial pressure of CO2 on the rate of photosynthesis would amplify at high than at low elevations (Hultine and Marshall, 2000). In contrast, Nothofagus betuloides and N. dombeyi showed a more active response and a more rapid increase in iWUE as climate dries through a decrease in stomatal conductance and a rise in photosynthesis rates (Figs. 4 and 6). In principle, increasing ca should lead to a reduction in g and consequently lead to an increase in tree iWUE if A remains constant. However, reduced g and transpiration are expected to occur under water deficit and, as a result, both treering ␦18 O and ␦13 C should increase reflecting a greater evaporative enrichment in 18 O and improved iWUE. Here, we only observed long-term changes in tree-ring ␦18 O of Nothofagus species, not in F. cupressoides (Fig. 5). This differentiation among species may suggest that changes in g are probably not related to the progressive

148

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

Table 4 Goodness-of-fit root mean squared error (RMSE, in ␮mol mol−1 ) for the three different theoretical models of intrinsic water-used efficiency (iWUE) fitted to three tree species (see Fig. 4) and considering the periods 1900–2010 and 1960–2010. Abbreviations: ca , atmospheric CO2 concentration; ci , internal CO2 concentration in the sub-stomatal internal cavity of leaves. Species

iWUE models ci

F. cupressoides N. betuloides N. dombeyi

1900–2010 4.80 3.76 4.66

1960–2010 10.52 8.14 9.70

1900–2010 1.70 0.83 1.05

120 F. cupressoides N. betuloides N. dombeyi

-1

iWUE (μmol mol )

110 100 90 80 70 60 29

ca − ci

ci /ca

30

31

32

33

18

δ O (‰) Fig. 6. Decadal values of intrinsic water-use efficiency (iWUE) and tree-ring O isotopes (␦18 O) for the three study species at the Huinay Private Park (42◦ 23 S, 72◦ 25 W, sea level), near the Comau fjord in southern Chile.

F. cupressoides’ iWUE decoupling with respect to the ci /ca scenario and the recent coupling to the ca − ci scenario that corresponds to a less homeostatic behaviour in response to rising ca (Fig. 4). In mesic conditions, in general, the constant ci /ca scenario explains better iWUE increases than other scenarios (Maseyk et al., 2011). Similar values of iWUE increases over time (F. cupressoides, 14%; Nothofagus species, 23%) have been observed in other forest types and biomes and are commonly attributed to a physiological effect of increasing ca (Andreu-Hayles et al., 2011; Gómez-Guerrero et al., 2013; Leavitt, 1994; Linares et al., 2009; Maseyk et al., 2011; Saurer et al., 2004; van der Sleen et al., 2014). However, the apparent reduced response of g to changes in ca in F. cupressoides has also been observed in temperate Abies alba forests subjected to drought-induced dieback and growth decline (Linares and Camarero, 2012). Relict species with reduced positive response of g to ca , like F. cupressoides and A. alba, may become replaced by other more adjustable tree species, like Nothofagus. Thus, in the Mediterranean mixed forest, conifer species (Pinus spp.) show higher iWUE values to increasing ca than coexisting broadleaf hardwood species (Quercus spp.), which makes them dominate in more xeric sites whereas the latter are restricted to mesic sites (Ferrio et al., 2003; Granda et al., 2014). Lastly, despite the fact that the dual-isotope approach aids to elucidate whether an iWUE improvement depends more on A than on g changes, we did not apply it given that too strong assumptions must be met when comparing different sites or time periods, namely: similar ca values and environmental influences (ambient humidity, leaf temperature, vapour pressure deficit, source water ␦18 O) affecting evaporative enrichment (Roden and Siegwolf, 2012). According to ␦18 O tree-ring values, F. cupressoides and N. dombeyi experience the weakest (highest g) and the strongest (lowest g) stomatal regulation of transpiration, respectively (Figs. 5 and 6). The relationship between relative humidity of air and tree-ring ␦18 O reflects the effect of atmospheric humidity on g and transpiration because lower humidity reduces stomatal conduc-

1960–2010 3.86 0.71 1.25

1900–2010 1.89 2.71 1.91

1960–2010 3.47 5.50 3.97

tance, leading to increased evaporative enrichment of 18 O, although ␦18 O also records other effects as the relative isotopic composition of rainfall and soil water sources (Cullen and Grierson, 2007; Yakir, 1998). Similar difference values in g between F. cupressoides and Nothofagus have been documented in temperate forests from the northern hemisphere, which has been explained as a lower g of conifers (e.g. Pinus) than broadleaf species (e.g. Quercus) (Hemming et al., 1998). In the northern hemisphere, among-species differences in tree-ring ␦18 O have also been attributed to different g rates and root depths in relation to the ␦18 O soil-water depth gradient (Battipaglia et al., 2008). For instance, in drought-prone Mediterranean ecosystems plant species with higher tree-ring ␦18 O values are assumed to use shallower more-isotopically enriched soil water sources (Moreno-Gutiérrez et al., 2012), or to experience a higher evaporative enrichment of leaf water due to transpiration (Barbour, 2007). In these Mediterranean ecosystems, leaf ␦13 C and ␦18 O values converged across coexisting plant species suggesting that g was the main driver of iWUE (Moreno-Gutiérrez et al., 2012; Wang et al., 1998). In the TRF, water stress can also occur during the growing season triggering stomatal closure in Nothofagus species (Piper et al., 2007), which maintain a high photosynthetic activity through both osmotic adjustment and by a well-developed root systems (Read and Farquhar, 1991). Although we need information on the soil water sources of adult trees in the TRF, our results point to state that iWUE improvement are more determined by changes in g rather than in A in Nothofagus species when compared to F. cupressoides (Fig. 6). 4.2. The growth–climate association Even though temperatures in southern Chile have not increased during the last part of the 20th century, precipitations have declined (Fig. S2). This drying trend is marked by the growingseason rainfall (Urrutia-Jalabert et al., 2015a); after the 1970s, there has a been a shift in the precipitation regime leading to drier conditions since then (Jacques-Coper and Garreaud, 2015). All this is consistent with a southward displacement of the midlatitudes westerly winds over the Southern Hemisphere (Quintana and Aceituno, 2012). Although we did not detect any long-warming trend, abnormal warm conditions have characterized some periods ˜ events (Barichivich, 2005), during the last decades due to El Nino which could also contribute to drought stress. Our results suggest that climate stressors as drought can also act in less expected places like the TRF, being responsible of the post-1980 drying trend that produced the growth reductions in F. cupressoides (Figs. 3 and S2). It is known that the growth of F cupressoides is sensitive to December–March temperatures (austral summer) prior to tree-ring formation, responding negatively to warm conditions during those months, and enhanced by wet and cloudy conditions (Devall et al., 1998; Lara and Villalba, 1993; Urrutia-Jalabert et al., 2015a). The post-1980 drying trend could also explain why the growth of old F. cupressoides trees (1238 years old) has decreased in drier sites at the Chilean Coastal Range (Urrutia-Jalabert et al., 2015a), but cannot explain why growth

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

increased in rainier Andean sites, similar to the Huinay area, despite iWUE has been improved during the 20th century in both sites. The fact that we worked with younger age classes of F. cupressoides (300 years old), i.e. trees established under the post-industrial ca period, may explain why we did not detect those previously documented growth enhancement. Furthermore, Urrutia-Jalabert et al. (2015a) note that significant decreasing trends in growth have also been observed in other Fitzroya stands which may even portend drought-induced dieback (Barichivich, 2005). ˜ In most studies tree growth is decoupled from iWUE (Penuelas et al., 2011), and a growth decline has been even observed despite increasing ca and improved iWUE (Silva et al., 2010). Long-term growth-iWUE relationships are often neutral or negative and they seem to be site- and species-dependent (Silva and Anand, 2013), which seems to be the case also for the Fitzroya-Nothofagus association in the TRF. For instance, in temperate European Fagus sylvatica forests a growth decline was observed in low-elevation stressed trees, whereas in high-elevation, trees under more mesic condi˜ et al., 2008). Here tions showed a growth enhancement (Penuelas we described a reverse pattern linked to lower relative air humidity at higher elevations (Fig. 2, see also Appendix, Fig. S3), and decreasing F. cuppresoides growth (Fig. 3). Overall, the decoupling between growth and iWUE observed also in other forest types may indicate that climate warming drier conditions override any potential CO2 -induced fertilization of growth. Our results likewise suggest that drying conditions can lead to growth decline in temperate rainforests regardless of temperature trends. 5. Conclusions Using multiple bodies of evidence, we found that the long-lived conifer F. cupressoides, an emblematic tree species of temperate rainforests in southern America, shows a poor acclimation to recent drier conditions, especially when compared to Nothofagus species. Nothofagus betuloides and N. dombeyi appear to improve more efficiently their use of water through changes in stomatal conductance, whereas F. cupressoides shows a more passive response to rising atmospheric CO2 concentrations. Hence, not all species are having similar and expected responses to increasing ca , which should be a call of attention in the case of long-lived, endangered and narrowdistributed species, like F. cupressoides. Finally, we urge to search for a better understanding of the mechanisms determining C and O isotopes fractioning and composition of tree rings in old-living tree species, such as F. cupressoides, through intra-annual assessments of environmental variables (soil and air temperature and humidity, radiation) and isotopes (as has been done with growth; cf. Urrutia-Jalabert et al., 2015c). Acknowledgements This study was financially supported by the CSIC–Huinay project “Crecimiento y funcionamiento de los bosques valdivianos de Nothofagus en respuesta al cambio global”. The authors thank G. Sangüesa-Barreda for doing the dendrochronological analysis. The authors also thank the logistic support of Reinhard Fitzek and the personnel working at the Huinay Scientific Field Station (Fundación San Ignacio de Huinay-ENDESA). J.J. Camarero would like to dedicate this work to his close family and descendants living at Rancagua (Chile). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.agrformet.2017. 03.003.

149

References Alaback, P.B., 1991. Comparisons of temperate rainforests of the Americas along analogous climate gradients. Rev. Chil. Historia Nat. 64, 399–412. Anderson, W.T., Bernasconi, S.M., McKenzie, J.A., Saurer, M., 1998. Oxygen and carbon isotopic record of climatic variability in tree ring cellulose (Picea abies): an example from central Switzerland (1913–1995). J. Geophys. Res. 103 (D24), 31625–31636. Andreu-Hayles, L., et al., 2011. Long tree-ring chronologies reveal 20th century increases in water-use efficiency but no enhancement of tree growth at five Iberian pine forests. Global Change Biol. 17 (6), 2095–2112. Armesto, J.J., Figueroa, J.A., 1987. Stand structure and dynamics in the temperate rain forest of Chiloé Archipelago, Chile. J. Biogeogr. 14 (4), 367–376. Barbour, M.M., 2007. Stable oxygen isotope composition of plant tissue: a review. Funct. Plant Biol. 34 (2), 83–94. Barichivich, J., 2005. Muerte apical episódica en bosques de alerce (Fitzroya cupressoides [Mol.] Johnston) de la Cordillera de la Costa de Valdivia. Facultad de Ciencias Forestales, Universidad Austral de Chile, Valdivia, Chile. Battipaglia, G., Jäggi, M., Saurer, M., Siegwolf, R.T.W., Cotrufo, M.F., 2008. Climatic sensitivity of ␦18 O in the wood and cellulose of tree rings: results from a mixed stand of Acer pseudoplatanus L. and Fagus sylvatica L. Palaeogeogr. Palaeoclimatol. Palaeoecol. 261 (1–2), 193–202. Bidegain, M., Stella, J.L., Bettolli, M.L., Quintana, J., 2016. Southern South America [in “State of the Climate in 2015”]. Bull. Am. Meteorol. Soc. 97 (8), S185–S186. Blunden, J., Arndt, D.S., 2016. State of the climate in 2015. Bull. Am. Meteorol. Soc. 97 (8), S1–S275. Camarero, J.J., Gazol, A., Tardif, J.C., Conciatori, F., 2015. Attributing forest responses to global-change drivers: limited evidence of a CO2 -fertilization effect in Iberian pine growth. J. Biogeogr. 42 (11), 2220–2233. Chatfield, C., 2000. Time-series Forecasting. Chapman and Hall/CRC, Boca Raton. Cook, E.R., Krusic, P.J., 2005. Program ARSTAN, A Tree Ring Standardization Program based on Detrending and Autoregressive Time Series Modeling, with Interactive Graphics. Columbia University, Palisades, NY, USA. Cullen, L.E., Grierson, P.F., 2007. A stable oxygen, but not carbon, isotope chronology of Callitris columellaris reflects recent climate change in north-western Australia. Clim. Change 85 (1), 213–229. Dávila, P.M., Figueroa, D., Müller, E., 2002. Freshwater input into the coastal ocean and its relation with the salinity distribution off austral Chile (35◦ –55◦ S). Cont. Shelf Res. 22 (3), 521–534. Devall, M.S., Parresol, B.R., Armesto, J.J., 1998. Dendroecological analysis of a Fitzroya cupressoides and a Nothofagus nitida stand in the Cordillera Pelada, Chile. For. Ecol. Manage. 108 (1), 135–145. Donoso, C., 1993. Bosques templados de Chile y Argentina: variación, estructura y dinámica. Ecología Forestal. Editorial Universitaria, Santiago, Chile, pp. 484. Donoso, C., Lara, A., Escobar, B., Premoli, A., Souto, C., 2006. Fitzroya cupressoides (Molina). In: Johnst, I.M., Donoso, C. (Eds.), Las especies arbóreas de los bosques templados de Chile y Argentina. María Cuneo Ediciones, Valdivia, Chile, pp. 68–81. Fajardo, A., McIntire, E.J.B., 2012. Reversal of multicentury tree growth improvements and loss of synchrony at mountain tree lines point to changes in key drivers. J. Ecol. 100 (3), 782–794. Farquhar, G.D., Ehleringer, J.R., Hubick, K.T., 1989. Carbon isotope discrimination and photosynthesis. Ann. Rev. Plant Physiol. Plant Mol. Biol. 40, 503–537. Farquhar, G.D., O’Leary, M.H., Berry, J.A., 1982. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Aust. J. Plant Physiol. 9 (1), 121–137. Farquhar, G.D., Richards, R.A., 1984. Isotopic composition of plant carbon correlates with water-use efficiency of wheat genotypes. Aust. J. Plant Physiol. 11, 539–552. Ferrio, J.P., Florit, A., Vega, A., Serrano, L., Voltas, J., 2003. 13 C and tree-ring width reflect different drought responses in Quercus ilex and Pinus halepensis. Oecologia 137 (4), 512–518. Francey, R.J., Farquhar, G.D., 1982. An explanation of 13 C/12 C variations in tree rings. Nature 297 (1), 28–31. Fuenzalida, H., et al., 2007. Estudio de la variabilidad climática en Chile para el siglo XXI. Informe Final CONAMA. Gobierno de Chile, Santiago, Chile. Gómez-Guerrero, A., et al., 2013. Growth decline and divergent tree ring isotopic composition (␦13 C and ␦18 O) contradict predictions of CO2 stimulation in high altitudinal forests. Global Change Biol. 19 (6), 1748–1758. Granda, E., Rodrigo Rossatto, D., Camarero, J.J., Voltas, J., Valladares, F., 2014. Growth and carbon isotopes of Mediterranean trees reveal contrasting responses to increased carbon dioxide and drought. Oecologia 174 (1), 307–317. Harris, I., Jones, P.D., Osborn, T.J., Lister, D.H., 2014. Updated high-resolution grids of monthly climatic observations the CRU TS3.10 Dataset. Int. J. Climatol. 34 (3), 623–642. Hemming, D.L., Switsur, V.R., Waterhouse, J.S., Heaton, T.H.E., Carter, A.H.C., 1998. Climate variation and the stable carbon isotope composition of tree ring cellulose: an intercomparison of Quercus robur, Fagus sylvatica and Pinus silvestris. Tellus 50 (1), 25–33. Holmes, R.L., 1983. Computer-assisted quality control in tree-ring dating and measurement. Tree-Ring Bull. 44, 69–75. Hultine, K.R., Marshall, J.D., 2000. Altitude trends in conifer leaf morphology and stable carbon isotope composition. Oecologia 123 (1), 32–40. Jacques-Coper, M., Garreaud, R.D., 2015. Characterization of the 1970s climate shift in South America. Int. J. Climatol. 35 (8), 2164–2179.

150

J.J. Camarero, A. Fajardo / Agricultural and Forest Meteorology 239 (2017) 141–150

Keenan, T.F., et al., 2013. Increase in forest water-use efficiency as atmospheric carbon dioxide concentrations rise. Nature 499, 324–327. Lanfear, R., et al., 2013. Taller plants have lower rates of molecular evolution. Nat. Commun 4, 1879. Lara, A., Villalba, R., 1993. A 3620-year temperature record from Fitzroya cupressoides tree rings in southern South America. Science 260 (5111), 1104–1106. Leavitt, S.W., 1994. South American tree rings show declining ␦13 C trend. Tellus 46B, 152–157. Leonardi, S., et al., 2012. Assessing the effects of nitrogen deposition and climate on carbon isotope discrimination and intrinsic water use efficiency of angiosperm and conifer trees under rising CO2 conditions. Global Change Biol. 18 (9), 2925–2944. Lévesque, M., Siegwolf, R., Saurer, M., Eilmann, B., Rigling, A., 2014. Increased water-use efficiency does not lead to enhanced tree growth under xeric and mesic conditions. New Phytol. 203 (1), 94–109. Linares, J.C., Camarero, J.J., 2012. From pattern to process: linking intrinsic water-use efficiency to drought-induced forest decline. Global Change Biol. 18 (3), 1000–1015. Linares, J.C., Delgado-Huertas, A., Camarero, J.J., Merino, J., Carreira, J.A., 2009. Competition and drought limit the response of water-use efficiency to rising atmospheric carbon dioxide in the Mediterranean fir Abies pinsapo. Oecologia 161, 611–624. Luebert, F., Pliscoff, P., 2006. Sinopsis bioclimática y vegetacional de Chile. Editorial Universitaria, Santiago, Chile, pp. 316. Maseyk, K., Hemming, D., Angert, A., Leavitt, S.W., Yakir, D., 2011. Increase in water-use efficiency and underlying processes in pine forests across a precipitation gradient in the dry Mediterranean region over the past 30 years. Oecologia 167, 573–585. McCarroll, D., et al., 2009. Correction of tree ring stable carbon isotope chronologies for changes in the carbon dioxide content of the atmosphere. Geochim. Cosmochim. Acta 73 (6), 1539–1547. McCarroll, D., Loader, N.J., 2004. Stable isotopes in tree rings. Quat. Sci. Rev. 23 (7–8), 771–801. Moreno-Gutiérrez, C., Dawson, T.E., Nicolás, E., Querejeta, J.I., 2012. Isotopes reveal contrasting water use strategies among coexisting plant species in a Mediterranean ecosystem. New Phytol. 196 (2), 486–496. Norby, R.J., et al., 2005. Forest response to elevated CO2 is conserved across a broad range of productivity. Proc. Natl. Acad. Sci. U.S.A. 102, 18052–18056. ˜ Penuelas, J., Canadell, J., Ogaya, R., 2011. Increased water-use efficiency during the 20th century did not translate into enhanced tree growth. Global Ecol. Biogeogr. 20 (4), 597–608. ˜ Penuelas, J., Hunt, J.M., Ogaya, R., Jump, A.S., 2008. Twentieth century changes of tree-ring ␦13 C at the southern range-edge of Fagus sylvatica: increasing water-use efficiency does not avoid the growth decline induced by warming at low altitudes. Global Change Biol. 14 (5), 1076–1088. ˜ J.M., Armesto, J.J., 2009. Pérez, C.A., Carmona, M.R., Aravena, J.C., Farina, Environmental controls and patterns of cumulative radial increment of evergreen tree species in montane, temperate rainforests of Chiloé Island, southern Chile. Austral Ecol. 34 (3), 259–271. Piper, F., Corcuera, L.J., Alberdi, M., Lusk, C., 2007. Differential phosynthetic and survival responses to soil drought in two evergreen Nothofagus species. Ann. For. Sci. 64 (4), 447–452. Quintana, J.M., Aceituno, P., 2012. Changes in the rainfall regime along the extratropical west coast of South America (Chile): 30–43◦ S. Atmósfera 25 (1), 1–22. R-Development-Core-Team, 2016. R: A Language and Environment for Statistical Computing. Version, 3.3.0. R Foundation for Statistical Computing, Vienna, ISBN 3-900051-07-0 http://www.R-project.org. Read, J., Farquhar, G.D., 1991. Comparative studies in Nothofagus (Fagaceae). I. Leaf carbon isotope discrimination. Funct. Ecol. 5, 684–695.

Roden, J., Siegwolf, R., 2012. Is the dual-isotope conceptual model fully operational? Tree Physiol. 32 (10), 1179–1182. Salzer, M.W., Hughes, M.K., Bunn, A.G., Kipfmueller, K.F., 2009. Recent unprecedented tree-ring growth in bristlecone pine at the highest elevations and possible causes. Proc. Natl. Acad. Sci. U.S.A. 106 (48), 20348–20353. Saurer, M., Borella, S., Leuenberger, M., 1997. ␦18O of tree rings of beech (Fagus sylvatica) as a record of ␦18O of the growing season precipitation. Tellus 49B, 80–92. Saurer, M., Siegwolf, R.T.W., Schweingruber, F.H., 2004. Carbon isotope discrimination indicates improving water-use efficiency of trees in northern Eurasia over the last 100 years. Global Change Biol. 10 (2), 2109–2120. Schulman, E., 1956. Dendroclimatic Change in Semiarid America. University of Arizona Press, Tucson, AZ. Seibt, U., Rajabi, A., Griffiths, H., Berry, J.A., 2008. Carbon isotopes and water use efficiency: sense and sensitivity. Oecologia 155 (3), 441–454. Silva, L.C.R., Anand, M., 2013. Probing for the influence of atmospheric CO2 and climate change in forest ecosystems across biomes. Global Ecol. Biogeogr. 22 (1), 83–92. Silva, L.C.R., Anand, M., Leithead, M.D., 2010. Recent widespread tree growth decline despite increasing atmospheric CO2 . PLoS ONE 5 (7), e115543. Stokes, M.A., Smiley, T.L., 1996. An Introduction to Tree-Ring Dating. The University of Arizona Press, Tucson, AZ. Suarez, M.L., Kitzberger, T., 2010. Differential effects of climate variability on forest dynamics along a precipitation gradient in northern Patagonia. J. Ecol. 98 (5), 1023–1034. Tecklin, D., DellaSala, D.A., Luebert, F., Pliscoff, P., 2011. Valdivian temperate rainforests of Chile and Argentina. In: DellaSala, D.A. (Ed.), Temperate and Boreal Rainforests of the World: Ecology and Conservation. , pp. 132–153. Urrutia-Jalabert, R., et al., 2015a. Increased water use efficiency but contrasting tree growth patterns in Fitzroya cupressoides forests of southern Chile during recent decades. J. Geophys. Res.: Biogeosci. 120, 2505–2524. Urrutia-Jalabert, R., Mahli, Y., Lara, A., 2015b. The oldest, slowest rainforest in the world? Massive biomass and slow carbon dynamics of Fitzroya cupressoides temperate forests in southern Chile. PLoS ONE 10 (9), e0137569. Urrutia-Jalabert, R., Rossi, S., Deslauriers, A., Mahli, Y., Lara, A., 2015c. Environmental correlates of stem radius change in the endangered Fitzroya cupressoides forests of southern Chile. Agric. For. Meteorol. 200 (1), 209–221. van der Sleen, P., et al., 2014. No growth stimulation of tropical trees by 150 years of CO2 fertilization but water-use efficiency increased. Nat. Geosci. 8 (1), 24–28. ˜ ecológica de los bosques del sur de Chile. Veblen, T.T., Schlegel, F.M., 1982. Resena Bosque 4 (2), 73–115. Villagrán, C., 1985. Análisis palinológico de los cambios vegetacionales durante el Tardiglacial y Postglacial en Chiloé, Chile. Rev. Chil. Historia Nat. 58, 57–69. Voelker, S.L., Muzika, R.-M., Guyette, R.P., Stambaugh, M.C., 2006. Historical CO2 growth enhancement declines with age in Quercus and Pinus. Ecol. Monogr. 76 (4), 549–564. Voltas, J., et al., 2013. A retrospective, dual-isotope approach reveals individual predispositions to winter-drought induced tree dieback in the southernmost distribution limit of Scots pine. Plant Cell Environ. 36 (8), 1435–1448. Wang, X.-F., Yakir, D., Avishai, M., 1998. Non-climatic variations in the oxygen isotopic compositions of plants. Global Change Biol. 4 (8), 835–849. Wigley, T.M.L., Briffa, K.R., Jones, P.D., 1984. On the average value of correlated time series, with applications in dendroclimatology and hydrometeorology. J. Clim. Appl. Meteorol. 23, 201–213. Yakir, D., 1998. Oxygen-18 of leaf water: a crossroad for plant-associated isotopic signals. In: Griffiths, H. (Ed.), Stable Isotopes: Integration of Biological, Ecological and Geochemical Processes. BIOS Scientific, Oxford, pp. 147–168.