Báo cáo khoa học: "Differences in drought resistance among 3 deciduous oak species grown in large boxes"

Chia sẻ: Nguyễn Minh Thắng | Ngày: | Loại File: PDF | Số trang:13

Thêm vào BST

Báo xấu

22
lượt xem 1
download

Download Vui lòng tải xuống để xem tài liệu đầy đủ

Tuyển tập các báo cáo nghiên cứu về lâm nghiệp được đăng trên tạp chí lâm nghiệp quốc tế đề tài: "Differences in drought resistance among 3 deciduous oak species grown in large boxes...

Chủ đề:

Bình luận(0) Đăng nhập để gửi bình luận!

Lưu

Nội dung Text: Báo cáo khoa học: "Differences in drought resistance among 3 deciduous oak species grown in large boxes"

Original article Differences in drought resistance among 3 deciduous oak species grown in large boxes G Aussenac P Vivin G Levy 1 Nancy, Laboratoire de Bioclimatologie et Écophysiologie Forestière, Champenoux; INRA 2 Nancy, Laboratoire Sol et Nutrition, Unité d’Écophysiologie Forestière, INRA 54280 Champenoux, France (Received 7 October 1992; accepted 18 February 1992) Summary — The purpose of this study was to explain large differences in growth and decline of the pedunculate oak (Ouercus robur L) and the sessile oak (Q petraea Liebl) observed in the forest as a result of drought. In addition, northern red oak (Q rubra L) was compared with the 2 indigenous oaks. The effects of controlled soil water deficits on growth and water relations of young plants of these 3 species grown in large boxes have been studied. The plants were old enough to have devel- oped normal root systems. Two species were planted in each box, and submitted to very similar pat- terns of water stress. Predawn leaf water potential, stomatal conductance, net assimilation rates, shoot elongation and mortality were monitored. The effect of an overall improvement in mineral nutri- tion on these parameters was also tested. During water deficit (decrease in predawn leaf water po- tential), the pattern of decrease of gas exchange was similar for the 3 species. Thus, their ability to limit water deficit by reduction of transpiration was similar. On the other hand, shoot growth of Q ru- bra was more reduced than that of Q robur for similar predawn leaf water potential; growth of Q pe- traea was the least sensitive. However, increase of mineral nutrition improved the growth of both Q robur and Q rubra, but not that of Q petraea. For the 3 species, no mortality was noted as long as predawn leaf water potentials remained > -3.6 MPa. Below this limit, the mortality rate was highest in Q robur, Q petraea and lowest in Q rubra. These differences in mortality between species are due to differences in tolerance to water stress, not in avoidance. drought / growth I gas exchange I dieback I fertilization I Quercus Résumé — Différences dans la résistance à la sécheresse de 3 espèces de chêne à feuilles caduques, cultivées en conteneurs. Le but de ce travail était d’expliquer les grandes différences de croissance et de dépérissement observées en forêt suite à des sécheresses, entre le chêne pé- donculé (Quercus robur L) et le chêne sessile (Q petraea Liebl). De plus, le chêne rouge d’Amérique (Q rubra L ) a été comparé aux 2 chênes indigènes. Les effets d’un déficit hydrique édaphique contrôlé sur la croissance et les relations hydriques de jeunes plants de ces 3 espèces, cultivés dans de grandes cuves, ont été étudiés. Les plants étaient assez âgés pour avoir pu développer des systèmes racinaires normaux. Deux espèces ont été plantées dans chaque cuve, subissant ainsi
exactement à chaque moment le même stress hydrique. Les variables suivantes ont été prises en compte : potentiel hydrique foliaire de base, conductance stomatique, assimilation nette, croissance aérienne et mortalité. L’effet d’une amélioration globale de la nutrition minérale sur ces paramètres a été également étudié. En situation de déficit hydrique (diminution du potentiel hydrique de base), le modèle de diminution des échanges gazeux a été similaire pour les 3 espèces; ainsi, la manière dont elles évitent le stress hydrique est quasiment identique. En revanche, la croissance aérienne de Q rubra a été plus réduite que celle de Q robur pour un même potentiel hydrique foliaire de base, la croissance de Q petraea était la moins sensible. Cependant, une amélioration de la nutrition minérale a augmenté la croissance de Q rubra et Q robur, mais non celle de Q petraea. Pour les 3 espèces, -3.6 MPa. En dessous de aucune mortalité n’a été notée pour des potentiels hydriques de base > cette limite, pour les plants ayant subi des conditions similaires, les taux de mortalité furent plus éle- vés chez Q robur que chez Q petraea, et très faibles chez Q rubra. Ces différences de mortalité entre les espèces semblent dues à des différences de tolérance et non à l’évitement au stress hydrique. échanges gazeux / mortalité fertilisation / Quercus / sécheresse / croissance / differences in radial growth in a great INTRODUCTION large number of stands. These differences might to be due to a greater demand by Q robur After a severe drought in 1976, oak de- than Q petraea for water and to some de- cline occurred in several regions of France gree for nutrients. Another troublesome (Centre, Bourgogne, Pyrénées Atlan- that, although morphological point was tiques). This phenomenon was of concern distinguish the 2 species characteristics on account of its intensity and economic (Dupouey, 1983; Sigaud, 1986), forest consequences. Similar decline in oaks managers have made little distinction be- was also observed both in Europe (Dela- species when reestablish- tween the 2 oak tour, 1983; Osterbaan and Nabuurs, 1991) ing stands. It is therefore not surprising and in the USA (Tainter et al, 1983; Ab- that in many sites Q robur does not seem 1990). rams, to be suitably located from an ecological Initial phytoecological studies carried point of view. out in different regions (Becker and Levy, From an ecophysiological point of view, 1982; Durand et al, 1983; Macaire, 1984) and with regard to the water relations of revealed that only the pedunculate oak these 2 indigenous species, preliminary (Quercus robur L) was subjected to de- experiments on seedlings showed that cline whereas the sessile oak (Quercus Q petraea was better able to avoid both in- petraea Liebl) remained unaffected. Soil ternal water stress and severe soil drought water deficit appeared to be the determin- than Q robur. Hence, according to Colleu ing factor (Becker and Levy, 1983); other (1983) on the one hand an initiation of factors such as mineral nutrition, patho- more numerous secondary roots furthers genic agents and forestry must be consid- root uptake and decreases internal water ered as only secondary and exacerbating. stress, but produces an increase in soil Furthermore, other studies (Becker and drought; and on the other hand, a stomatal Levy, 1990) revealed that both the ecologi- control which occurs at higher water poten- cal differences between the 2 species, and tial and more effectively reduces water the artificial spread (ie planting) of Quer- losses and soil drought. However, the dif- cus robur to unsuitable sites had led to
(2 species and 40 plants per box). In order to ferences in behaviour that have been ob- avoid any possible microclimatic effects due to served in young plants were not as well ex- site conditions, the allocation of species in the pressed in forest stands. Moreover, these different boxes was randomized. However, spe- studies were carried out with seedlings cies having the strongest juvenile growth were grown in pots with a confined root system planted to the north of each box, so as to reduce development, which considerably limited the competition for light. All trees were grown in open conditions and, during the first few years, the practical relevance of the results ob- developed vigorously, creating closed canopy tained. stands. Some Q rubra whose development was Thus, the purpose of this study was to too great and detrimental to the other plants had characterize in a comparative manner the to be pruned. effects of a prolonged soil drought on the In May 1990, 2 greenhouses covered with a ecophysiological functioning of oaks grown transparent plastic sheath and largely opened at their extremities were installed to intercept rain- outdoors in large boxes, allowing us to fall while maintaining sufficient ventilation, thus work on older plants with a normal root avoiding an increase in temperature during hot system development. It must be empha- summer days. sized that this experimental design (ie large boxes, binary mixed species) allowed accurate interspecific comparisons for Plant conditioning characterization of soil drought intensity. Moreover, it was also interesting to investi- The experimental design adopted for each spe- gate the ecophysiological relations of cies association was a 3 x 2 factoral design con- northern red oak (Quercus rubra L) in com- sisting of 3 watering regimes and 2 nutrient parison with the 2 indigenous oaks, and availability treatments (see table I). determine the former’s drought sensitivity. In fact, Q rubra is one of the most remarka- Water supply regimes ble species introduced in Europe for re- establishing stands in unfavorable ecologi- supply regimes were as follows: The 3 water cal sites (Timbal, 1990) in particular due to control boxes (W) maintained permanently its rapid growth. - near field capacity by frequent watering (3 x 50 I per week); boxes (D) submitted to moderate drought, - MATERIALS AND METHODS then brought back to field capacity whenever av- Experimental design The experimental design was set up near the INRA Research Centre of Nancy (in Lorraine, northeast France). It consisted of 26 large boxes (depth: 100 cm, volume: 1.62 m which were ), 3 partially buried. These boxes were filled with 10 cm of gravel at the bottom to improve water drainage, and 90 cm of a sandy loam soil from the horizon A of a brown soil from the Mon- 2 /A 1 don Forest (France) mixed with peat in the upper 10 cm. In March 1987, 2-yr-old saplings from the Forest Research Centre’s nursery were planted
erage predawn leaf water potential reached - elongation and mortality Shoot 2.0 MPa (each time 2 x 150 I within 2 d); measurements were carried Shoot elongation boxes (DD) submitted to severe drought up to - week on 10 plants per species per out - 4.0 MPa and then brought back to field capaci- once a box. In order to determine whether a part of the ty (2 x 150 I within 2 d). difference in Quercus drought behaviour was The water supply in the control treatments due to a difference in tolerance to low water po- represented≈ 92l m weekly. As a result, irri- -2 tentials, all the boxes were subjected to extreme gation was greater by a factor of=3 than the drought conditions by withholding irrigation after observed ETP rate at Nancy and hence could August 30 (jd 242). When ψ reached -5.0 wp have given rise to nutrient leaching. MPa, the soil was watered to field capacity in or- Drought in the dry treatments began on May der to estimate the survival rate of each species. 22 (Julian day 142). From August 30 onwards Mortality rate was assessed the following year (jd 242), the boxes at field capacity were no on June 10 1991 (jd 160) in all treatments. longer watered. Levels of mineral nutrition RESULTS follows: The 2 levels of mineral nutrition were as Predawn leaf water potential unfertilized boxes (u); - boxes with a supply of NPK mineral fertilizer - (f): 220 g/m patentkali, 160 g/m triple super 2 2 As illustrated in figure 1, the time course of phosphate and 37 g/m ammonitrate were add- 2 predawn leaf water potential (ψ was vir- ) wp ed manually. tually identical for each couple of species grown in a given box up to the lowest val- ψ (-4.0 MPa). This results implied ues of wp Ecophysiological measurements that there was no interspecific heterogene- performed ity in the exploitation of the soil water with the possible exception of some fertilized Water relations treatments submitted to drought (see fig 1, upper right). Predawn leaf water potential (ψ was meas- ) wp (Aussenac Under controlled conditions, ψ ranged and wp ured using a pressure chamber from -0.05 to -0.60 MPa for all species; Granier, 1978) and was determined before sun- rise at least once a week on an average plant the fluctuations were mainly dependent on per species and per box. delays in recovery of field capacity. For the moderately dry treatments, ψ wp Gas exchange reached -2.0 to -3.0 MPa depending on the boxes. Two rehydrations to field capac- Net CO assimilation rate (A, μmol m s and -2 -1 ) 2 ity were carried out according to the exper- stomatal conductance for water diffusion (g, w imental design (jd 183, jd 203). The first mmol m s were performed in situ using a -2 -1 ) drought period lasted 41 d while the sec- portable gas exchange measurement system (Li ond was shorter (23 d). 6200, Li-Cor, USA) under natural climate and ir- radiance, and expressed on a leaf area basis Concerning the severely dry treatments, meter (Li 3000A, Li-Cor, using a portable area the decrease of ψ was also rapid: 64 d, wp was determined once a USA). Gas exchange to reach -4.0 MPa on average. Conse- week from 11 am to 01 pm when the sun was at quently, irrigation to field capacity (jd 207) its zenith, on the leaf of 1 average plant per spe- was essential in order to prevent the plants cies per box. Only leaves exposed to full light from early wilting. Then plants were sub- were selected.
mitted to very severe water deficits (-5.0 These kinetics revealed differences in MPa). In this case, predawn leaf water po- the evolution of the soil drought according tential decreased rapidly (in 2 od it again to treatments. As far as fertilization was reached -4.0 MPa). concerned, one can only assume that be-
(ψ is displayed in figure 2. Each point of a greater biomass in the fertilized ) wp cause represents a measurement performed on a boxes (especially in controlled conditions) sunny day from 11 am to 01 pm on an the total transpiration was higher and in- average plant per species per box. Values duced faster soil water depletion. corresponding to PPFD < 800 μmol m -2 -1 s or to recent rehydration were not plot- ted. Net photosynthesis and stomatal conductance When ψ was not limiting (ψ > -0.75 wp wp MPa), the stomatal conductance and pho- tosynthesis values showed wide variability. Evolution of CO assimilation rates (A) and 2 This heterogeneity could be explained by a stomatal conductance for water (g with ) w wide intraspecific variability (choice of respect to predawn leaf water potential
plant, of leaf, genetic factors) and also MPa to -2.50 MPa); and 5) very severely by different daily microclimate conditions. stressed (< -2.50 MPa). Selection of these classes was based on an assessment of As ψ decreased, so did A and g . w wp scattered plots of gas exchange versus Nearly complete stomatal closure was predawn water potentials. Analysis of vari- reached at -1.8 MPa for both indigenous = ance (with Fisher PLSD) was used to de- oak species and at -1.6 MPa for Q rubra. termine the significance of relationships During the drought period, A fell to nearly between gas exchange values and water for ψ values -2.8 MPa for the 3 zero < wp status. oak species. As shown in figure 3, the 2 indigenous Because of the wide variation in gas ex- oaks displayed a similar mean net assimi- change of plants in response to ψ se- , wp lation rate (respectively 8.09 ± 0.29 μmol lected analyses were performed on plants -2 -1 m s for Q petraea and 7.68 ± 0.41 grouped according to predawn water po- μmol m s for Q robur) in well watered -2 -1 tential classes: 1) well watered (0 to -0.75 conditions (class 1).In contrast, Q rubra MPa); 2) moderately stressed (-0.75 MPa to -1.25 MPa); 3) stressed (-1.25 MPa to presented a significant lower value (5.73 ± 0.30 μmol m s Analog findings were -2 -1 ). - 1.75 MPa); 4) severely stressed (-1.75
Shoot elongation obtained with the mean value of g (237 ± w 14 mmol m s for Q petraea, 226 ± -2 -1 22 mol m s for Q robur, and 142 ± -2 -1 Relative daily elongation (RDE) rate was 11 mmol m s for Q rubra). It must be -2 -1 calculated by first dividing weekly elonga- emphasized that Q petraea again showed tion rates of each plant by days separating significantly higher values of A and g in 2 measurements. This absolute daily shoot w classes 2 and 3. increment was then divided by the maxi- mum value for each plant, in order to yield The effects of fertilization on gas ex- a relative daily elongation rate expressed change are summarized in table II. When as a percentage of the maximum value. ψ was > -0.75 MPa (class 1),nutrition wp Predawn leaf water potential was meas- supply increased mean values of A in the ured as median value between 2 weekly 3 oak species. However, Q rubra still dis- measurements of shoot elongation. When played a significantly lower value than the it was not available, ψ was estimated by 2 indigenous oak species. As ψ de- wp wp linear interpolation. creased < -0.75 MPa (other classes), fer- tilization apparently had no more effect on RDE rate/ψ relationships are plotted wp A/ψ relationship. It did not affect mean in figure 4. Zero RDE values were not re- wp values of g in any species except in Q ported in the figure. For each ψ hand- w wp petraea at high leaf water potential. drawn contour indicated maximum curves
As illustrated in figure 4, fertilization had values of RDE rate. Below plotted points positive effect on growth of Q rubra and readily explained by unfavorable a were Q robur whatever the water treatment, but growth conditions. For simplicity, this had no significant effect on Q petraea. graphic representation (ie contour curve) Growth decreased only beyond -1.0 MPa provided a possible guide for understand- for the 3 species, and became very low ing drought effects on growth while avoid- (< 10%) when ψ exceeded -2.0 MPa for ing, at least in part, phenological effects. wp Q rubra, and -2.5 MPa for both indigenous For unfertilized plants, maximum RDE oak species. rate decreased rapidly from -0.3 MPa for These findings suggested that nutrition Q rubra, -0.6 MPa for Q robur and -0.9 supply had little influence on growth of MPa for Q petraea. Growth became non Q petraea in water deficit conditions; this significant (< 10% of maximum) beyond was to be expected, because Q petraea - 1.4 MPa for Q rubra -2.0 MPa for Q robur was already resistant enough to water defi- and -2.5 MPa for Q petraea.
cit. Conversely, fertilization improved Fertilization had an unexpected nega- growth of both other oak species. Hence, tive effect on survival of the 3 species (see Q robur displayed growth similar to that of fig 5). Death rates were increased while Q petraea. keeping initial ranking. This effect might be due at least in part to difference in biomass productivity. Death rate DISCUSSION From August 30 (jd 242), of the trees none were watered. As soon as ψ exceeded wp This study, carried out under semi-natural the minimum value measurable with the conditions on Q petraea, Q robur and Q ru- pressure chamber (-5.0 MPa), plants were bra saplings grown in boxes and submitted irrigated to field capacity with the aim of to soil drought cycles, had 2 aims: i) to an- observing their survival rates the following alyse differences in drought responses of year. In Spring 1991, an inventory was both pedunculate and sessile oaks, so as made to calculate death rate linked to the to understand differences observed in the 1990 imposed drought. Hence, figure 5 forest; and ii) to compare northern red oak shows that there was no mortality in con- with the 2 indigenous oaks. trol treatments, which were only submitted to a late short water stress (ψ > -3.6 Generally, most oaks have deep- wp MPa) after jd 242. In contrast, in other un- penetrating root systems, enabling them to fertilized treatments, Q robur showed the maintain relatively high predawn potentials highest death rate (18.0%); Q petraea during drought (Abrams, 1990). Thus, a (5.6%) and above all Q rubra (0.8%) had a deep root system may be considered as a lower death rate. primary adaptation which allows oaks to
avoid dessication during drought. In the phyll effects took place, causing an altera- present study, trees had an available soil tion in photosynthetic capacity. However, depth of 1 metre, so the root system of our Epron and Dreyer (1990) revealed that the photosynthetic system strongly resisted 5-yr-old plants was less confined than if leaf water deficits, and considered that they had been in small-sized pots. Conse- photoinhibition could be an important fac- quently, it was possible to extrapolate from tor in explaining photosynthetic system these results to natural conditions. sensitivity to drought. But according to the The experimental design allowed new results of Weber and Gates (1990) on information to be obtained, especially Q rubra and those of Epron et al (1992) on since species were studied in pairs. In Q petraea, it seemed that no photoinhibito- each box, ψ temporal evolution was the wp ry damage could be detected in water- same for the 2 species, thus allowing inter- stressed oak before total reduction of A. specific comparison of avoidance and re- So in the present study it appeared that sistance. early drought effects were mainly mediated by stomatal closure. As ψ decreased, so did growth in all Drought effects wp oak species as noted above. In the case of the unfertilized plants, the RDE fell quickly, Leaf gas exchange in Quercus was sensi- usually from -0.3 MPa for Q rubra, -0.6 tive to water stress, as drought clearly in- MPa for Q robur and -0.9 MPa for Q pe- duced a decrease in net CO assimilation 2 traea and became non significant (< 10%) rate and stomatal conductance. Net photo- below -1.4 MPa for Q rubra, -2.0 MPa for synthesis became non significant as ψ wp Q robur and -2.5 MPa for Q petraea. In reached -2.8 MPa for any oak species. previous experiments on Q robur, Ausse- Very similar results have been reported nac and Levy (1983) found a total growth with Q petraea seedlings, showing an iden- inhibition when ψ reached -1.1 MPa. wp tical decrease of gas exchange during Could this difference be meaningful? In drought, with a total inhibition of photosyn- fact in the present study, by taking a con- thesis at -3.0 MPa (Colleu, 1983; Epron tour curve, the RDE rate of Q robur and Dreyer, 1990). reached 20% when ψ dropped to -1.2 wp Nearly complete midday stomatal clo- MPa. Furthermore, in a tree with short attained when ψ reached -1.8 shoot elongation, each error in its meas- sure was wp MPa. However, Q rubra stomata closed urement (± 1 mm) resulted in high variation earlier (-1.6 MPa); and these species of RDE rate. In other words, points should seemed more sensitive to water deficit be regarded cautiously due to possible than both indigenous oak species. Yet variations on the X and Y axes. Neverthe- such claims could be dubious. The diffe- less, this representation seemed suitable rentiation between species via gas ex- for characterization of growth response to change responses to drought was rather drought. difficult. Concerning resistance to very high wa- Concerning the mechanisms involved ter deficit (many drought cycles), large dif- the stomatal effect was critical to initial re- ferences occurred between species. Death rate was higher in Q robur than in Q pe- duction of A through decreasing intercellu- lar concentrations of CO as ψ fell. traea; Q rubra remained unaffected. How- 2 wp Moreover, according to a number of scien- ever, for all ob- species, mortality no was control served tists, it was likely that simultaneous meso- treatments were as
species. Thus, nutrition supply seem to fa- submitted to a short drought period at the growth at lower water potential, indi- end of summer. Thus for the first time in vour cating possible influence of osmoregula- such experiments, results closely resem- tion phenomena, in particular for Q robur bled forest observations. In fact, data not and Q rubra. With reference to the above- quoted above revealed that at identical mentioned results observed by Aussenac ψ yellowing and withering status oc- , wp and Levy (1983), mineral nutrition level curred earlier in Q roburthan in Q petraea, was presumably higher in the present suggesting that Q robur avoided drought. study even in the unfertilized treatments. But given the mortality rate, the higher sensitivity of Q robur seems mainly due to In conclusion, it was now possible, at lower tolerance to water stress. least for young plants, to put forward a hy- pothesis about the differential behaviour of In connection with the tolerance hypoth- indigenous oak species with respect to wa- esis and survival rate for drought, it must ter stress. In particular, there was no differ- be emphasized that Cochard et al (1992) ence in gas exchange regulation between showed a difference in the sensitivity of Q robur and Q petraea. The 2 species dif- vessels to embolism, providing a possible fered in their survival rate to very severe explanation of forest observations. water stress, and this agreed with observed differences by Cochard et al (1992) on the sensitivity of their vessels to embolism. Nutrition supply effects Finally, results of this study confirmed commonly held opinion that Q rubra is the Fertilization only increased A in all oak a drought-resistant species. Nevertheless, species when plants were well watered. It its growth could be strongly affected by a did not affect A or g in any species when w water deficit. In addition, contrary to earlier ψ was < -0.75 MPa, except for Q pe- wp claims (Kolb et al, 1990), Q rubra had a traea (see Results). Some researchers re- good response to nutrients: fertilization ported similar results on different plant had very positive effect on its growth, es- species grown with high or low nitrogen when this species is confronted pecially supply: a large difference in A at high leaf with soil drought. water potential and practically no differ- ence at low . wp ψ the contour curve, RDE rate By taking ACKNOWLEDGMENTS still 100% when ψ reached -1.0 was wp MPa in any species. Under well watered The authors thank E Dreyer for helpful discus- conditions, fertilization had only a positive sions during the preparation of this article, and effect on RDE rate for Q robur and Q ru- TB Lefevre, JF Muller, J Clerc and F Willm for bra. This result could be due to the fact technical assistance on the site. that, as shown in the forest, growth of Q petraea is less affected by mineral defi- cit than that of Q robur. REFERENCES As drought increased, RDE rate de- creased less rapidly than in the case of un- Abrams MD (1990) Adaptations and responses fertilized plants, except in Q petraea. to drought in Quercus species of North Amer- Hence, under water stress conditions, fer- ica. Tree Physiol7, 227-238 tilization had an essentially positive effect Aussenac G, Granier A (1978) Quelques résul- on both northern red and pedunculate oak tats de cinétiques journalières du potentiel de
sève chez les arbres forestiers. Ann Sci For Epron D, Dreyer E (1990) Stomatal and non 35, 19-32 stomatal limitation of photosynthesis by leaf water deficits in three oak species: a compar- Aussenac G, Levy G (1983) Influence du ison of gas exchange and chlorophyll fluores- dessèchement du sol sur le comportement cence data. Ann Sci For 47, 435-450 hydrique et la croissance du chêne pédoncu- le (Quercus pedunculata Ehrl) et du frêne Epron D, Dreyer E, Bréda N (1992) Photosyn- (Fraxinus excelsior L) cultivés en cases de thesis of oak trees (Q petraea (Matt) Liebl) végétation. Ann Sci For 40, 251-264 during drought under field conditions: diurnal of net CO assimilation and photo- 2 course Becker M, Levy G (1982) Le dépérissement du chemical efficiency of photosystem II. Plant chêne en forêt de Tronçais : les causes éco- Cell & Environ 15, 809-820 logiques. Ann Sci For 39, 439-444 Kolb TE, Steiner KC, McCormick LH, Bowersox Becker M, Levy G (1983) Le dépérissement du TW (1990) Growth response of northern red chêne : les causes écologiques. Rev For Fr oak and yellow poplar seedlings to light, soil 35, 341-356 moisture and nutrients in relation to ecologi- Becker M, Levy G (1990) Le point sur l’écologie cal strategy. Forest Ecol Manag 38, 65-78 comparée du chêne sessile et du chêne pé- (1984) Le dépérissement du chêne Macaire A donculé. Rev For Fr 52, 148-154 pédonculé en forêt communale d’Amance Cochard H, Bréda N, Granier A, Aussenac G (Aube). Rev For Fr 36, 201-205 (1992) Vulnerability to air embolism of three Osterban A, Nabuurs GJ (1991) Relationships European oak species (Quercus petraea, Q between oak decline groundwater class in pubescens, Q robur). Ann Sci For 49, 225-233 The Netherlands. Plant Soil 136, 87-93 Colleu S (1983) Contribution à l’étude de la résis- Sigaud P (1986) Ne parlons plus du chêne mais tance à la sécheresse de jeunes plants de des chênes. Rev For Fr 38, 376-384 chêne sessile, de chêne pédonculé et de chêne rouge. DEA, Université de Nancy I, 61 p Tainter FH, Williams TM, Cody JB (1983) Drought as a cause of oak decline and death Delatour C (1983) Le dépérissement des on South Carolina coast. Plant Dis 67, 195- chênes en Europe. Rev For Fr 35, 265-282 197 Dupouey JL (1983) Analyse multivariable de Timbal J (1990) Le chêne rouge d’Amérique: quelques caractères morphologiques de pop- écologie et facteurs limitants. Rev For Fr 52, ulation de chêne (Quercus robur L and Q pe- 174-181 traea (Matt) Liebl de Hurepoix). Ann Sci For 40, 265-282 Weber JA, Gates DM (1990) Gas exchange in Quercus rubra (northern red oak) during a Durand P, Gelpe J, Lemoine B, Riom J, Timbal J (1983) Le d6p6rissement du chêne pédon- drought: analysis among photosynthesis, transpiration, and leaf conductance. Tree culé dans les Pyrénées-Atlantiques. Rev For Fra 35, 357-368 Physiol 7, 215-225