Tree Physiology 20, 893–900 © 2000 Heron Publishing—Victoria, Canada Influence of weather on cork-ring width ANTONIA CARITAT,1 EMILIA GUTIÉRREZ2 and MARISA MOLINAS1,3 1 Laboratori del Suro, University of Girona, Campus Montilivi, E-17071 Girona, Spain 2 Department of Ecology, University of Barcelona, Avinguda Diagonal 645, E-08028 Barcelona, Spain 3 Author to whom correspondence should be addressed Received August 4, 1999 Summary Ring-width series of cork from Quercus suber L. trees growing at two sites in Extremadura (southwestern Spain) were analyzed in relation to monthly precipitation and temperature, and to climatic indices combining both variables. Ring width of cork showed strong positive correlations with precipitation, especially during the fall and winter. Moderately low temperatures were favorable for cork growth, except in winter and during the onset of phellogen activity. We conclude that drought or temperature, or both, can limit cork growth during the annual drought period. Keywords: climatic variability, cork growth, cork-ring chronology, dendroclimatology, phellogen activity, Quercus suber. Introduction Quercus suber L., which is well known for its cork production, is an evergreen oak that grows on siliceous substrates and is widely distributed in the western Mediterranean region. The xerophytic character of the cork tree is seen in features such as sclerophylly, the presence of a lignotuber (Molinas and Verdaguer 1993a, 1993b), litterfall pattern (Caritat et al. 1996a), and stomatal response to water deficit (Tenhunen et al. 1984, Oliveira et al. 1992). In Q. suber, the cork cambium (phellogen) is a permanent layer that adds a new layer of cork to the outer bark (phellem) of the tree each year. The phellogen usually becomes active in April and may remain active until the end of October (Natividade 1950). Cork ring boundaries are clearly demarcated because spring cells have thinner walls and larger diameters than those formed later in the season. Cork from wild trees is deeply furrowed and irregular, making it difficult to identify the rings. However, the natural or wild cork must first be removed to obtain commercial cork, and the new growth is much more uniform and forms more regular rings. Wild cork is removed when the tree is about 30 cm in diameter (about 30–40 years old). Henceforth, commercial cork is removed each time the bark reaches at least 25 mm in thickness, resulting in harvest intervals (referred to as cork peel-off turns) of 8–14 years, depending on the area. The cork is harvested only during the season when the tree is in sap and the layers separate more easily, usually beginning in June and lasting until August. For a more regular distribution of the cork harvest, an alternate system is followed in which only some trees are stripped each season. Because most phellems do not show annual growth rings, little is known about interannual variability in phellem growth or about how climatic factors affect phellogen activity (Waisel 1995). Nevertheless, cork rings of Q. suber can be dated and subjected to climatic analysis. A previous study by Caritat et al. (1996b) established a 14-year cork-ring chronology and reported evidence of the influence of precipitation and temperature on ring width variability of cork. The present study extends this previous chronology and constructs a new one from a different site in the same area. Both chronologies were analyzed in relation to precipitation, temperature and combined climatic indices. The objectives were: (1) to determine the influence of climate on the interannual pattern of ringwidth variability of cork, and (2) to analyze the relationship between cork growth and climatic variables. The overall objective was to assess to what extent summer drought affects cork growth in an area characterized by Mediterranean climatic conditions. Materials and methods Study area The two study sites, La Herguijuela (Campo Arañuelo, Cáceres) and Torre Sirgada (Jerez de los Caballeros, Badajoz), were chosen for their different topographic conditions in Extremadura (southwestern Spain) (Table 1). At both sites, the cork woodland is a multipurpose forest or dehesa used for livestock and cork production. At La Herguijuela, the cork trees are partially stripped each year, whereas at Torre Sirgada cork-stripping campaigns take place every 2 years. At both sites, cork harvests occur at 10-year intervals. Samples and data collection Cork samples of 10 × 10 cm 2 were taken at breast height and in the same southern orientation during the stripping of the trees. To observe the growth rings, the samples were boiled in water for about 45 min to prepare them for thin sectioning (Figure 1). Total bark thickness was measured, rings were counted and dated, and a visual synchronization was carried out. The wid- 894 CARITAT, GUTIÉRREZ AND MOLINAS Table 1. Topographical and forestry parameters of the La Herguijuela and Torre Sirgada sites in Extremadura (southwestern Spain). Parameter La Herguijuela Torre Sirgada Geographic coordinates Altitude (m a.s.l.) Annual precipitation (mm)1 Soil parent material Slope (%) Tree age (range, years) Tree density (ha –1)2 Tree height (mean ± SD, m)2 DBH (mean ± SD, cm)2 Bifurcation height (mean ± SD, m)2 Cork ring width (range, mm) Cork ring width (mean ± SD, mm) Shrub stratum 05°55′ W, 39°6′ N 300–400 777.5 Alluvial debris 5–7 150–170 30 6.89 ± 0.95; n = 15 62 ± 4; n = 15 2.47 ± 0.20; n = 15 1.85–5.25 3.30 ± 0.96 No shrub stratum present Herbaceous stratum Annual grasses with Tuberarietalia guttatae (L.) Fourr. being the most abundant species. 06°50′ W, 38°23′ N 500–600 606.2 Shale 20 25–160 57 8.91 ± 2.20; n = 18 69 ± 15; n = 18 3.24 ± 1.25; n = 18 1.95–4.70 3.56 ± 0.95 Low shrub stratum with Cistus salvifolius L. and C. crispus L. Annual grasses with Aegylops geniculata, Plantago sp. and Briza maxima L. equally abundant. 1 2 Annual precipitation for the period 1979–1995. Values for mature trees under cork extraction. est and narrowest rings were used as markers. The initial and final rings were rejected as being associated with incomplete growth periods. For each cork sample, cork-ring series were identified along three radii. Ring widths of the cork were measured with a 10× binocular microscope equipped with a micrometer scale. Climatic data were obtained from nearby weather stations belonging to the Centro Metereológico Territorial de Extremadura (Spain). Precipitation and temperature series from the Malpartida weather station (23 km from the La Herguijuela site) and from Jerez de los Caballeros station (10 km from the Torre Sirgada site) were used (Figure 2). To assure the quality Figure 1. Cork specimen from La Herguijuela, Cáceres, Spain. Note the presence of cork rings of different width. of the records, climatic series from Malpartida were checked against Salto de Torrejón (220 m a.s.l., 0.5 km from La Herguijuela); the series from Jerez de los Caballeros were checked against Valencia de Mombuey (297 m a.s.l.) and Fregenal de la Sierra (580 m a.s.l.), located 25 and 16 km from Tore Figure 2. Climate diagrams for the study period (1979–1995) for Malpartida (Campo Arañuelo, Cáceres) and Jerez de los Caballeros (Badajoz). Symbols: 䊊, temperature; 䊉, precipitation. Note the occurrence of the drought season in summer and to a lesser extent in spring. Note the marked drop in precipitation at the beginning of the summer drought season in La Herguijuela. Temperature and precipitation values at the top of the graphs are mean values. TREE PHYSIOLOGY VOLUME 20, 2000 SENSITIVITY OF CORK GROWTH TO CLIMATE Sirgada, respectively. Pearson correlation coefficients (r) between the monthly climatic variables of Malpartida and Jerez were significant for precipitation (r = 0.84, P < 0.001, n = 17) and temperature (r = 0.63, P < 0.01, n = 17). Cork-ring series and data analysis To establish the mean ring width chronology of the cork for each site, we obtained mean harvest-year series and then overlapped and averaged the successive harvest-year series. To obtain a mean harvest-year series, a cork growth curve was built for each sampled tree and the individual series were compared with each other using multiplot representations for visual corroboration (Cook and Kairiukstis 1990). The Pearson correlation coefficient (r) between the series was calculated to verify synchronization and the series showing statistically significant coefficients were averaged. For La Herguijuela, six harvest-year series (1989, 91, 93, 94, 95 and 96) were overlapped. The 93 series was formed from 46 out of 57 individual-tree series (Caritat el al. 1996b). The 89, 91, 94, 95 and 96 series were formed from 13 out of 15, 11 out of 13, 5 out of 10, 9 out of 12, and 10 out of 13 individual series, respectively. For Torre Sirgada, three harvest-year series (1989, 92 and 95) were overlapped. At this site, the 89 and 92 series were formed from four out of 10 individual series and the 95 series was formed from eight out of 12 individual series. A series was selected if its coefficient of correlation (Pearson’s r) was significant at P < 0.05. Cork-ring chronologies showed a decreasing trend in ring width with time that was attributed to aging (Natividade 1950). To minimize this effect and maximize the interannual fluctuations due to climate, each series of raw data was standardized by means of the empirical growth function: Yt = atbe –ctEt, where Yt = ÿt E t, where Yt is the actual measurement of ring width (RW), ÿt is the estimated value, and E t is the residual or remainder, called growth index (GI) in dendrochronological studies (a, b and c = constants; t = time) (Warren 1980). The standardized series of growth indices were assumed to be constant with respect to the mean and variance. Effect of climate on cork rings The influence of climatic factors on the interannual pattern of ring width was measured by means of the Pearson correlation coefficient (r). The relationship between cork growth and climatic factors was analyzed by means of paired analyses and bivariate scatterplots in which ring width (RW) or growth index (GI) was used as the dependent variable. The mathematical functions showing the best fit were then selected. The Bonferroni method (Sokal and Rohlf 1995) was used to limit the overall experimental error rate. The climatic variables considered were monthly precipitation (mm), precipitation accumulated during selected periods, mean monthly temperature (Tm), mean maximum temperature (Tx-M), the mean minimum temperature (Ti-M), absolute maximum temperature (Tmax), and absolute minimum temperature (Tmin). We also considered the drought index formed by the quotient between annual precipitation and mean annual temperature (P/T + 10) and several 895 other indices designed to unmask the effects of water deficits on cork growth. These indices were formed by using precipitation in periods with high positive effect (see Table 2) divided by temperatures with high negative effect (spring or summer months) or divided by combinations made by subtracting temperatures with positive effect (usually April or May) from temperatures with high negative effect. Multiple linear regression analysis was also used to assess the relative predictive capabilities of climatic independent variables. Results Chronologies At both sites, the individual tree series in each harvest year were closely correlated with one another (La Herguijuela mean r ± SD = 0.81 ± 0.08, P < 0.05; Torre Sirgada = 0.74 ± 0.09, P < 0.05). The correlation coefficients between sites for the overlapping years were r = 0.703 (n = 13 years in common, P < 0.01) for growth indices and r = 0.678 (n = 13, P < 0.02) for ring width. The range of ring widths (Table 1) was typical of mature cork trees (100–150 years old) in the dehesa of southwestern Spain (Caritat et al. 1996b), and comparable with wood-ring chronologies established in other regions of Spain (PérezAntelo 1994, Gutiérrez et al. 1998). Generally, the years of maximum and minimum cork growth were common to both sites (r = 0.678, P < 0.02) (Figure 3). Interannual variability in cork growth indices, as measured by the mean sensitivity coefficient MSx (Fritts 1976), was high. The MSx was 0.185 (RW) and 0.167 (GI) for trees at La Herguijuela, and 0.222 (RW) and 0.214 (GI) for trees at Torre Sirgada. Response to precipitation The effects of monthly precipitation on cork growth indices are given in Figure 4. Most correlation coefficients were not significant; however, precipitation in January had a significant positive effect at both sites, and precipitation in June and July had a positive effect that was significant at La Herguijuela. Precipitation in October showed a negative correlation that was significant at Torre Sirgada. However, because October rainfall is thought to contribute to cork growth, we tested its correlation with growth indices of the following year (t + 1) and found a positive correlation at both sites that was significant at La Herguijuela. The responses to accumulated precipitation over certain months were more striking than the effect of monthly precipitation, especially for trees at La Herguijuela. The highest correlations were obtained when rainfall from autumn and winter months of the preceding year was added to that of the spring and summer of the current year (Table 2). As shown in Figure 5, cork growth was synchronous with accumulated precipitation. A shift in the mean amount of precipitation occurred in 1989, and this shift was also observed in the cork series, most strikingly in the ring-width chronologies. The scatterplots pairing ring-width data with accumulated precipitation showed a strong and positive relationship that TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com 896 CARITAT, GUTIÉRREZ AND MOLINAS Table 2. Pearson correlation coefficients (r) between annual cork growth indices (GI) and ring width (RW) and the accumulated mean monthly precipitation for certain periods. Asterisks indicate statistically significant coefficients (* = P < 0.05, ** = P < 0.01, and *** = P < 0.001). (The Bonferroni adjusted α was 0.0045). Period of precipitation January to June January to December February to April March to June May to September June to September November (t – 1) to June November (t – 1) to July November (t – 1) to September October (t – 1) to June October (t – 1) to September October (t – 1) to December (t – 1) La Herguijuela (n = 17) Torre Sirgada (n = 13) RW GI RW 0.666* 0.437 0.163 0.464 0.572* 0.677** 0.705** 0.746*** 0.781*** 0.572* 0.668** 0.285 0.796*** 0.351 0.150 0.481 0.668** 0.454 0.691** 0.725*** 0.721** 0.712* 0.770*** 0.384 0.088 0.011 0.149 –0.406 –0.244 –0.098 0.528 0.545 0.598* 0.314 0.395 0.320 was generally linear (Figures 6a, 6b and 6d). However, there was a nonlinear relationship between ring width and accumulated precipitation from November (t – 1) to June (t) at Torre Sirgada (Figure 6c), indicating that, after a threshold amount of precipitation, the increase in ring width decreased with further increases in precipitation. GI 0.345 –0.137 0.147 –0.237 –0.141 –0.123 0.538 0.544** 0.691** 0.684** 0.757** 0.566* Figure 8 depicts selected examples of bivariate analysis of temperature and cork growth indices. The relationship with November Tmin showed a gradual decrease in cork growth as temperature increased (Figure 8b). In the scatterplot that Response to temperature Temperature responses were built using the growth indices to minimize the effect of precipitation. The correlation coefficients with monthly temperatures were low, being mostly negative and not statistically significant (Figure 4). There was a high negative correlation for almost all March temperatures, which was significant for Ti-M in trees at Torre Sirgada. The negative effect of March temperatures on the cork growth indices might be attributed to the early spring dry period that usually occurs in March in this area (Figure 2). The general inverse relationship between cork growth and temperature was more pronounced in trees at Torre Sirgada than at La Herguijuela (Figure 7). At Torre Sirgada, cork growth was lowest in 1993 when March Ti-M was highest. At La Herguijuela, cork growth was highest when July Tm reached its minimum value for the study period (1987 and 1988). Figure 3. Cork growth indices from La Herguijuela and Torre Sirgada. Figure 4. Response of cork growth indices to precipitation (P) and temperature (T ) based on Pearson correlation coefficients (r) at La Herguijuela (n = 17 years, 1979–95) and at Torre Sirgada (n = 13 years, 1982–94). Symbols: 䊊, Tm, monthly mean; ⵧ, Tmin, absolute minimum; 䉭, Tmax, absolute maximum; 䉮, Ti-M, mean minimum; 䉫, Tx-M, mean maximum temperatures. Asterisks indicate statistically significant correlation coefficients (P < 0.05). TREE PHYSIOLOGY VOLUME 20, 2000 SENSITIVITY OF CORK GROWTH TO CLIMATE 897 even if precipitation was high. The nonlinear relationship with March Ti-M at Torre Sirgada indicates that cork growth rate was almost constant between 5 and 7 °C, but decreased at higher temperatures (Figure 8c). Joint effect of precipitation and temperature Figure 5. Series of ring widths of cork (䊉) and growth indices (䊏) in relation to accumulated precipitation (䊊) at La Herguijuela. (a) Precipitation from November (t – 1) to June (t), and (b) from January to June. paired cork growth indices and July Tm at La Herguijuela, an exponential pattern emerged when the boundary points (circled dots) were selected (Figure 8a). There was a marked decrease in cork growth when the temperature was above 28 °C, The correlation coefficients between cork growth and the drought index, P/T + 10, were low and not significant. However, for the indices designed to emphasize soil water deficits, the correlation coefficients for the drought index were usually greater than those for precipitation or temperature considered separately (Figure 9). In general, the relationship followed a common pattern that fitted the function Y = ax b (Figures 9b–f). However, in one case it was a linear relationship (Figure 9a), indicating that, at La Herguijuela, cork growth was proportional to an increase in accumulated precipitation for the January to June period, or to a decrease in July Tmax. The most common nonlinear behavior was caused by the limiting effect of precipitation or temperature. If temperature was high compared with precipitation, a narrow ring was formed even if precipitation was abundant. Years with such behavior differed between the two sites because of small local climatic differences. The relationship between annual cork growth and the independent climatic variables, precipitation and mean temperature, exhibited adjusted R 2 values of 0.66 (La Herguijuela) and 0.64 (Torre Sirgada). The relationship was explained by the equations: GR = 0.519 + 0.00152P1–9 for La Huerguijuela and GR = –0.024 + 0.00109P11–9 + 0.035T3 for Torre Sirgada, where GR is annual cork growth, P1–9 is precipitation from Figure 6. Relationship between cork ring width (RW) and precipitation. Scatterplots for the periods (a) January to June and (b) November (t – 1) to September (t) at La Herguijuela; and (c) November (t – 1) to June and (d) November to September at Torre Sirgada. The goodness of fit, R 2, and the significance level, P, are given in each figure. TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com 898 CARITAT, GUTIÉRREZ AND MOLINAS Figure 7. Series of ring widths of cork (䊉) and growth indices (䊏) in relation to temperature (䊊). (a) Monthly mean July temperature (Tm) for La Herguijuela, and (b) March mean minimum temperature (Ti-M) for Torre Sirgada. January to September, P11–9 is precipitation from November of the previous year to September and T3 is the mean temperature in March. Discussion Cork cambium dynamics are sensitive to climatic factors. Cork-ring growth was high in rainy years (Figure 5) and during years in which temperature was moderately low (Figure 7). The decrease in precipitation recorded after 1989 at La Herguijuela and the increase in March Ti-M recorded after 1984 at Torre Sirgada could explain the trend of decreasing cork growth with time at both sites. The ring-width series were more affected by long-term changes caused by precipitation, whereas the series of growth indices were more affected by inter-annual variability caused by temperature. Precipitation exerted a large positive influence on cork growth and drought was a limiting factor, indicating the importance of the water holding capacity of the soil. Precipitation during November and December always had a positive influence on cork growth in the following year. October appeared to be a transitional month, probably because phellogen becomes dormant at this time. Differences in precipitation between sites could account for the higher sensitivity of cork growth to climate in trees at La Herguijuela than at Torre Sirgada. There was more late autumn and winter precipitation and more severe drought in the late spring at La Herguijuela than at Torre Sirgada (Figure 2). At Torre Sirgada, cork growth was less restricted by lack of water (Figure 4). A saturation effect of soil water content could explain the nonlinear relationship between cork growth and precipitation during the November to June period at this site (Figure 6c). Figure 8. Paired relationships between growth indices and (a) monthly mean July temperature (Tm) and (b) absolute minimum November temperature (Tmin) at La Herguijuela; and (c) mean minimum March temperature (Ti-M) at Torre Sirgada. The goodness of fit, R 2, and the confidence level, P, are given in each figure. In (a), the circled points indicate the values used to fit the exponential function. Uncircled points correspond to narrow rings formed in years in which July Tm was lower than 28 °C but precipitation was low (the narrowest one corresponds to 1983). Narrow rings were also formed in years in which precipitation was higher but the temperature was above the 28 °C threshold, such as in 1989 and 1992. Temperature generally exerted a negative effect on cork growth, which is typical for trees growing in a Mediterranean climate. Moderately low temperatures enhanced cork growth, except in January and February and at the onset of phellogen activity (April to May). During this latter period, cork growth was stimulated by an increase in temperature, but later in the season high temperatures slowed cork growth until the beginning of autumn. The significantly negative coefficients of the late autumn and winter temperatures (Figure 4) could be related to carbon balance, because respiration rate can exceed photosynthetic rate at this time of year (Edwards and Hanson 1996). At La Herguijuela, summer temperatures showed the high- TREE PHYSIOLOGY VOLUME 20, 2000 SENSITIVITY OF CORK GROWTH TO CLIMATE 899 Figure 9. Examples of the relationships between cork growth indices (GI) and different climatic indices: (a) quotient of precipitation from January to June and July Tmax; (b) quotient of precipitation from November (t – 1) to September (t) and July Tmax; (c) quotient of precipitation from November (t – 1) to September (t) and July Tmax minus May mean minimum temperature (Ti-M); (d) quotient of precipitation from January to June and March Ti-M; (e) quotient of precipitation from November (t – 1) to September (t) and August Tmax minus May Ti-M; and (f) quotient of precipitation from November to September and March Ti-M. Ti-M, mean minimum; Tmax, absolute maximum temperature. est correlation coefficients. This, together with the positive effect of precipitation in June and July, indicates that cork growth was restricted by temperatures above 28 °C or by the lack of water, or both (Figures 2 and 7). The effect of temperature was less pronounced at Torre Sirgada than at La Herguijuela. At Torre Sirgada, cork-ring growth was controlled more by high temperatures during winter and at the beginning of the growing period (Figure 4). The positive temperature effect observed in April at La Herguijuela was probably equivalent to the positive effect observed in May at Torre Sirgada. This could indicate a certain delay in the activation of the phellogen that may be partially attributable to the higher altitude of the La Herguijuela site. A similar delay was observed for the July and August temperatures (Figure 4). To understand cork growth, the effects of precipitation and temperature should be considered together (Figure 9). The variance explained by drought indices is higher than that explained by precipitation or temperature alone, except for July Tm, which appears to be the most restrictive factor during the growing period at La Herguijuela (Figure 8). Multiple linear regression analysis corroborated the general trends observed, especially the relationship with precipitation. In years of severe summer drought, cork trees can lose all their leaves (Natividade 1950). Also, acorns can abort if temperature or precipi- tation, or both, are unfavorable during their development period. For the Mediterranean basin region, dendroclimatology has been performed for beech (Gutiérrez 1988), some pine species (Gutiérrez 1989) and a few deciduous oaks (Serre-Bachet 1982, Tessier et al. 1994). These dendroclimatological studies have demonstrated that summer drought is the main factor limiting wood ring growth. The wood rings of evergreen oaks (Quercus ilex L. and Q. suber), the predominant tree species in the Mediterranean forest communities (Specht 1988), are hard to identify. Few dendroclimatological data are available (Liphschitz and Lev-Yadun 1986, Zhang and Romane 1991, Oliveira et al. 1994, E. Gutiérrez, unpublished data). These data suggest that cambial activity is highly sensitive to summer precipitation and, as well as during the winter dormant period, can be interrupted several times during the year, resulting in the formation of false rings. However, in semi-arid regions dominated by extensive cork woodlands and dehesas, as for example, certain areas in the south of the Iberian peninsula and in northern Africa, drendrochronology of cork (suberochronology) could provide additional information. A practical application would be for the analysis and prediction of forest decline due to anthropogenically induced climatic change. A detailed knowledge of how climate influences cork production TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com 900 CARITAT, GUTIÉRREZ AND MOLINAS and quality would aid in the management of this important Mediterranean forest resource. Furthermore, a model could be built for a given area to predict cork growth. We could, therefore, adjust the timing of the cork-harvest years in accordance with the growth predicted from the interannual variations of precipitation and temperature. Acknowledgments Members of the Forestry Section of the Instituto para el Corcho, la Madera y el Carbón, Junta de Extremadura, Ap. 437, E-06800 Mérida, Spain, have collaborated in the selection of the sites, in sample and data collection and discussion of the results, and should be considered co-authors of this paper. Dr. Miguel Elena (Institute director) and Dr. Enrique Cardillo deserve particular acknowledgement. The authors thank the owners of La Herguijuela for their kind collaboration and Dr. Emili Garcia-Berthou for his help in the preparation of the manuscript. References Caritat, A., M. Molinas and E. Gutiérrez. 1996a. Annual cork-ring width variability of Quercus suber L. in relation to temperature and precipitation (Extremadura, southwestern Spain). For. Ecol. Manage. 86:113–120. Caritat, A., G. Bertoni, M. Molinas, M. Oliva and A. DominguezPlanella. 1996b. Litterfall and mineral return in two cork-oak forests in northeast Spain. Ann. Sci. For. 53:1049–1058. Cook, E.R. and L.A. Kairiukstis. 1990. Methods of dendrochronology. Applications in the environmental sciences. Kluwer Academic Publishers, Dordrecht, pp 3–19. Edwards, N.T. and P.J. Hanson. 1996. Stem respiration in a closed canopy upland oak-forest. Tree Physiol. 16:433–439. Fritts, H.C. 1976. Tree rings and climate. Academic Press, New York, 576 p. Gutiérrez, E. 1988. Dendroecological study of Fagus silvatica L. in the Montseny mountains (Spain). Acta Oecol. Oecol. Plant. 9: 301–309. Gutiérrez, E. 1989. Dendroclimatological study of Pinus sylvestris L. in southern Catalonia (Spain). Tree-Ring Bull. 49:1–9. Gutiérrez, E., J.J. Camarero, J. Tardiff, O. Bosch and M. Ribas. 1998. Tendencias recientes del crecimiento y la regeneración en bosques subalpinos del Parque Nacional de Aigüestortes. Ecologia 12: 75–84. Lisphschitz, N. and S. Lev-Yadun. 1986. Cambial activity of evergreen and seasonal dimorphics around the Mediterranean. Int. Assoc. Wood Anatomist Bull. 7:145–153. Natividade, V. 1950. Subericultura. Direcçao General dos Serviços Florestais e Aquicolas, Lisboa, pp 83–170. Molinas, M. and D. Verdaguer. 1993a. Lignotuber ontogeny in the cork-oak (Quercus suber; Fagaceae). I. Late embryo. Am. J. Bot. 80:172–181. Molinas, M. and D. Verdaguer. 1993b. Lignotuber ontogeny in the cork-oak (Quercus suber; Fagaceae). II. Germination and young seedling. Am. J. Bot. 80:182–191. Oliveira, G., O.A. Correira, M.A. Martins-Louçao and F.M. Catarino. 1992. Water relations of cork-oak (Quercus suber L.) under natural conditions. Vegetatio 99/100:199–208. Oliveira, G., O. Correira, M.A. Martins-Louçao and F.M. Catarino. 1994. Phenological and growth patterns of the Mediterranean oak Quercus suber L. Trees 9:41–46. Pérez Antelo, A. 1994. Nota de revisión de la investigación dendrocronológica en España. Investigación Agraria, Sistemas y Recursos Forestales 3:221–235. Serre-Bachet, F. 1982. Analyse dendroclimatologique comparée de quatre espècies de pins et du chene pubescent dans la région de la Gardiole prés Rians (Var, France). Ecol. Medit. 8:167–183. Specht, R.L. 1988. Natural vegetation—ecomorphological characters. In Mediterranean-Type Ecosystems: A Data Source Book. Ed. R.L. Specht. Kluwer Academic Publishers, Dordrecht, pp 14–61. Sokal, R.R. and F.J. Rohlf. 1995. Biometry. The principles and practice of statistics in biological research. Freeman, New York, 887 p. Tessier, L., P. Nola and F. Serre-Bachet. 1994. Deciduous Quercus in the Mediterranean region: tree-ring/climate relationships. New Phytol. 126:355–367. Tenhunen, J.D., O.L. Lange, J. Gebel, W. Beyschlag and J. A.Weber. 1984. Changes in photosynthetic capacity, carboxylation efficiency and CO2 compensation point associated with midday stomatal closure and midday depression of net CO2 exchange of leaves of Quercus suber. Planta 162:193–203. Waisel, Y. 1995. Developmental and functional aspects of the periderm. In The Cambial Derivatives. Ed. M. Iqbal. Encyclopaedia of Plant Anatomy, Gebrüder Borntraeger, Berlin, pp 293–311. Warren, W.G. 1980. On removing the growth trend from dendrochronological data. Tree-Ring Bull. 40:35–44. Zhang, S.H. and F. Romane. 1991. Variations de la croissance radiale de Quercus ilex L. en function du climat. Ann. Sci. For. 48: 225–234. TREE PHYSIOLOGY VOLUME 20, 2000
© Copyright 2026 Paperzz