Diurnal and seasonal changes in stem radius increment and sap flow density indicate different responses of two co-existing oak species to environmental stress

Acta Silv. Lign. Hung., Vol. 7 (2011) 97–108 Diurnal and Seasonal Changes in Stem Radius Increment and Sap Flow Density Indicate Different Responses of Two Co-existing Oak Species to Drought Stress Ilona M ÉSZÁROS a* – Péter K ANALAS a – András F ENYVESI b – József K IS a – Balázs N YITRAI a – Erzsébet S ZŐLLŐSI a – Viktor O LÁH a – Zita D EMETER a – Ágnes L AKATOS a – István A NDER b a Department of Botany, Faculty of Science and Technology, University of Debrecen, Debrecen, Hungary b Section of Cyclotron Applications, Nuclear Research Institute Debrecen, Hungary Abstract – Using continuous monitoring of stem radius combined with sap flow measurements we assessed the effects of environmental conditions on tree radial growth and water status of two coexisting oak species (Quercus petraea and Quercus cerris) at high resolution time in growing seasons of 2008 and 2009. The forest (95–100 yr) is situated in a xeric site in the transition zone between forested and forest-steppe regions in north-eastern Hungary, Bükk mountains (47 o 90’N, 20 o 46’E, elevation 320–340 m a.s.l.). Weather conditions in the growing season of 2008 (total rainfall 354 mm, mean daily temperature 17.0 o C) was less extreme than in 2009 (total rainfall 299 mm, temperature 17.9 o C). Rainfall strongly determined the course of radial growth increment in trees. Radial growth of trees was limited in 2009 due to the drought in spring. The maximum radial increment of both species was achieved three weeks earlier (4 th week of June) than in 2008 (4 th week of July). We used dendrometer monitoring data for estimation of stem (tree) water deficit (∆W) by measuring water-related changes in stem radius (Zweifel et al. 2005). The magnitude of tree water deficit variation (∆W) was always smaller in Q. cerris than in Q. petraea. In contrast, Quercus cerris always exhibited larger daytime averages and maxima of sap flow density. In August of 2009 when drought became severe there were larger increases in tree water deficit (∆W) (50–55 %) in both species compared to July as it could be expected from the extent of decreases in sap flow density (24–28%). Our data suggested that due to the low SWC the transpiration was supported mainly from the inner water storage of trees during prolonged drought which resulted in high stem water deficit (∆W). drought / forest/potential evapotranspiration / Quercus petraea / Quercus cerris / stem (tree) water deficit Kivonat – Két zonális tölgyfaj törzsnövekedésének és nedváramlásának napi és szezonális dinamikája. A tanulmányban összefoglalt eredmények részét képezik a Bükk-hegységi Síkfőkút Project LTER kutatási terület (47 o 90’N, 20 o 46’E, tszf. 320–340 m) cseres-tölgyes erdőállományban folyó növény-ökofiziológiai kutatási programnak. Vizsgálataink során folyamatos dendrometriás és nedváramlás méréseket végeztünk a 2008 és 2009 évek vegetációs időszakaiban annak megállapítására, hogy a kocsánytalan tölgy (Quercus petraea (Matt.) Liebl.) és a csertölgy (Quercus cerris L.) radiális * Corresponding author: imm[email protected]; H-4032 DEBRECEN, Egyetem tér 1. Mészáros, I. et al. Acta Silv. Lign. Hung. 7, 2011 98 törzsnövekedésében és vízforgalmában az időjárási extrémitások milyen mértékű változásokat okoznak. A dendrometriás mérési eredmények azt jelzik, hogy a csapadék mennyisége és eloszlása döntően meghatározza a törzsek radiális növekedését mindkét fajnál. Minkét faj esetében kimutattuk, hogy a 2009 évi tavaszi aszály hatására a törzsek radiális növekedése három héttel korábban befejeződött és alacsonyabb értéket ért el a 2008 évi növekedéshez képest. A törzs radiális változásának a finom időfelbontású mérése a növekedési ütemen kívül lehetőséget nyújtott arra is, hogy az adatok alapján a törzsek (fák) vízdeficit értékeit (∆W) is megbecsüljük, amelyhez Zweifel et al. (2005) módszerét használtuk fel. A két fafaj közül a vízdeficit a Q. petraea törzsében mindig szignifikánsan nagyobb amplitúdójú változásokat mutatott, mint a Q. cerris esetében, ami alapján az utóbbi fajnál nagyobb törzsbeli vízkészletre következtettünk. Ugyanakkor a nedváramlás nappali középértéke és maximuma mindig a Quercus cerris esetében volt magasabb. 2009 augusztusában, erősen aszályos időszakban mindkét fajnál jóval nagyobb mértékben (50–55%-kal) emelkedett a fatörzsbeli vízdeficit a júliusi mérési eredményekhez képest, mint amit az ugyanezen időszakban mért nedváramlás csökkenés mértéke alapján várnánk (24–28%). Ez azt sugallja, hogy a fák a tartósan aszályos periódusokban a belső, törzsbeli vízraktárakat hasznosítják a vízszállító pályák feltöltésére és a transzpiráció fenntartására. aszály / nedváramlás / Quercus petraea / Quercus cerris / radiális törzsnövekedés / vízdeficit 1 INTRODUCTION Owing to climate change (IPCC, 2007) the future survival and sustainability of forest ecosystems has become of great concern (Jump et al. 2009, Mátyás 2010). For Carpathian basin climate projections predict a reduction in the total area of climate-zonal forests and the gradual shift „forward” of transition between forest and forest-steppe zones (Mátyás – Czimber 2004). Severe and recurring drought has been identified as major contributing factor to the recently accelerated tree decline and mortality in Europe (e.g. Jakucs et al. 1986, Gibbs – Greig 1997, Siwecki – Ufnalski 1998, Thomas et al. 2002). In Hungary serious tree decline has been reported for the mixed stands of sessile oak (Quercus petraea (Matt.) Liebl) and Turkey oak (Quercus cerris L.) from the 80’s (Jakucs et al. 1986). These forests represent one of the most important vegetation type in the Carpathian basin therefore tree decline has large economic and nature conservation consequences. Sessile oak suffered more drastic decline than Turkey oak in Hungarian forests as well as in whole Europe. Mortality of sessile oak varied with site conditions and became very serious in xeric margins of this forest type suggesting that climate change will threaten sessile oak very strongly. Simulation of future distribution of sessile oak by BIOMOD model (Thuiller 2003) projects that there will be a shift of its bioclimatic envelope as a result of climate change. On regional scale, analyses provided also a very pessimistic scenario for Hungary since the species may loose the majority of the distribution area by 2080 (Czúcz et al. 2011). The objective of this work was to analyse the effects of climatic fluctuations on growth and water status of two co-existing tree species (Quercus petraea (Matt.) Liebl. and Quercus cerris L.) in two subsequent growing seasons, 2008 and 2009. More specifically we intended 1) to estimate the influence of environmental conditions on stem radius increment with high time resolution; 2) to assess drought related responses of tree water deficit for the two species; 3) to describe seasonal and diurnal course of sap flow density and its correlation with environmental factors. Diurnal and seasonal changes in stem radius increment and sap flow density Acta Silv. Lign. Hung. 7, 2011 99 2 MATERIAL AND METHODS 2.1 Study site and experimental trees The study was carried out at Síkfőkút Project LTER forest site (47o90’N, 20o46’E, elevation 320–340 m a.s.l.), Bükk Mountains, north-eastern Hungary in summer of 2008 and 2009. The site is covered by a mixed forest stand (95–100 yr) dominated by sessile oak (Quercus petraea) and Turkey oak (Quercus cerris) in the canopy layer. The soil of the site is a deep brown forest soil formed on miocenic pebble (Jakucs 1985). According to the current climatic conditions the site is close to the forest-steppe limit. The average annual rainfall of the past 30 years is 555 mm and the mean annual temperature is 10.3 oC at the site (Table 1). The growing season usually lasts from mid-April to mid-October. In the northern mountain region of Hungary the annual total of global radiation falls between 4300–4400 MJ m–2 (Jakucs 1985). During the past decade extreme drought and heat-waves have appeared at the site in summer of 2003, 2007 and 2009. The tree species composition of the stand was: Quercus petraea 46.9%, Quercus cerris 22.8%, Carpinus betulus 0.4%, Acer campestre 28.2%, Acer tataricum 0.9%, Cerasus avium 0.8%. Trees of oak species belong to the dominant and co-dominant crown classes, while other tree species represent intermediate and co-dominant crown classes of forest canopy. For our study we selected one mature sessile oak and one Turkey oak (95–100 years old). Both trees represented the dominant crown class growing to uppermost position of the forest canopy (height of experimental trees was 20–22 m, DBH of the sessile oak tree was 29 cm, for Turkey oak 46 cm). 2.2 Measurements of environmental parameters Weather conditions were monitored automatically by Hobo ProSeries RH&Temp sensor (Onset Computer Corporation, Pocasset, USA), and Hobo Micro Station (Onset Computer Corporation, Pocasset, USA) with external sensors (Rain gauge, PAR, atmospheric pressure, wind speed) during the study period. Weather data were recorded at every 30 min. at the top of a meteorological tower (25m above ground). Volumetric soil moisture content (SWC) was measured using ECH2O sensors (Decagon Devices, Pullmann WA, USA) within the upper 30 cm with 15 min sampling interval. To assess the differences in the microclimatic conditions between vegetation seasons of 2008 and 2009 we calculated the cumulative daily mean temperature and cumulative rainfall. Cumulative daily mean temperature and rainfall data and course of soil moisture in 2008 and 2009 are presented in Figure 1. Due to the lack of measurements of global radiation, the Hargreaves-Samani temperature based method was used to estimate the the mean daily potential evapotranspiration (PET mm day–1) (Hargreaves – Samani 1982). 2.3 Sap flow measurements Continuous sap flow measurements began at our site in growing season of 2009. Sap flow density (ml cm–2 min–1) was measured with heat dissipation method developed by Granier (1985). An SF-G sensor (Ecomatik, GmbH, Dachau, Germany) was mounted on the northern side of tree stems to avoid direct solar heating and shielded with aluminum foil to minimize temperature fluctuations in the sapwood. The SF-G sensor consists of two identical needles with copper-constantan thermocouples and a special heating wire. The two needles were inserted 2 cm into the sapwood, one above the other, 15 cm apart. The upper needle was installed at a height of 1.5 m. The top needle was heated with constant energy supply (at 12V with 83 mA). The temperature difference between the two needles (∆T) was the output signal of the sensor and used for calculation of sap flow density according to the formula by Granier (1985): Mészáros, I. et al. Acta Silv. Lign. Hung. 7, 2011 100 u = 0.714*[(∆Tmax–∆T)/∆T]1.231, where ∆T is temperature difference between two needles; ∆Tmax is the maximum value of ∆T when sap flow can be considered as 0 during night. SF-G Sensors were installed on March 26 2009 and monitoring of ∆T was planned in 5 min interval till the end of October but there were some short periods during the growing season when unexpected errors (due to heavy rain events, animal damages etc.) disrupted the continuous measurement. 2.4 Measurements of tree stem radius changes and estimation of tree water deficit Tree stem radius changes (∆r) were monitored with automated DR dendrometers (Ecomatik Gmbh, Dachau, Germany) with resolution up to 0.2 microns. The dendrometers were mounted at 1m above ground on north side of stem of one sessile oak and one Turkey oak tree. Continuous recording of stem radius changes began on June 6 2008. ∆r was recorded in 10 min intervals. A sensor fixed in a frame was installed to the measuring section of the stem after removal of the dead bark. Sensors were installed carefully to avoid damages to the living tissues below the dead bark. The course of ∆r depends mainly on stem radial growth and fluctuation of water-storage. Other factors e.g. temperature and xylem-tension-related fluctuations may contribute only slightly (<10%) to ∆r (Zweifel et al. 2005). From the course of ∆r we estimated the changes in stem water-storage by using the algorithm suggested by Zweifel et al. (2005). We hypothesized that a rainfall event above 10 mm can induce stem hydration and increase of r to maximum. The difference between the trunk radius of maximum hydrated (normally after a rainfall event) state and actual hydration status was used for quantifying the degree of stem water deficit (∆W) during a given period. This is also considered as a measure of water deficit in the whole tree and defined as tree water deficit (Hinckley – Lassoie 1981). 3 RESULTS 3.1 Weather and soil moisture conditions Compared to the average weather conditions of the last 30 years (1978–2007), annual mean temperature was higher by 0.9–1 oC in 2008 and 2009 (Table 1). The vegetation period of both study years was warmer and drier than the average. Summer of 2009 was extraordinarily hot and had one of the lowest total amounts of rain during the last decade (Figure 1). However, in 2009 the total annual rainfall did not differ from the 30 year average while in 2008 it was 56 mm lower (Table 1). In 2009 there was a four-week period without rainfall during budburst in April which led to rapid decline of soil moisture content (Figure 1). The whole vegetation period of 2009 was drier and warmer than in 2008 with low volumetric water content. In 2009 the end of vegetation period was without rain events that caused large water deficits in the soil in September. During the vegetation period of 2008 the soil was wet in spring. Volumetric soil water content transiently decreased in May, but it increased in June and July due to frequent rains. Diurnal and seasonal changes in stem radius increment and sap flow density Acta Silv. Lign. Hung. 7, 2011 101 Table 1. Annual mean of air temperature (T), mean daily air temperature of vegetation period ( T04–10) , annual total rainfall (P) and total rainfall of vegetation period ( P04–10) in 2008 and 2009. Long-term mean values of air temperature and rainfall were calculated for 1978–2007 T oC T04–10 oC P mm P04–10 mm 2008 11,3 17,0 499 354 2009 11,3 17,9 554 299 30-year mean (1978–2007) 10,4 16,5 555 393 Cumulative rainfall (mm) 0 50 100 150 200 250 300 350 Cumulative daily mean temperature (oC) 0 1000 2000 3000 4000 5000 rainfall 2008 rainfall 2009 T oC 2009 T oC 2008 DOY 100 120 140 160 180 200 220 240 260 280 300 soil water content (m 3 m -3 ) 0,06 0,08 0,10 0,12 0,14 0,16 0,18 0,20 0,22 0,24 2009 2008 Figure 1. Cumulative values of daily mean air temperature, cumulative daily precipitation (upper figure) and volumetric soil water content (lower figure) during the growing seasons of 2008 and 2009 in the Síkfőkút study area 3.2 Stem radius changes and variation in tree water deficit Dendrometers were installed on tree stems on June 6 2008. We selected a three-month period between DOY158 and 248 in 2008 and 2009 for comparison of stem radius changes (∆r). Although the course of ∆r was different in 2008 and 2009, the stem radial increment did not change after DOY 248 in both years (Figure 2). DOY 158 is considered as reference day and ∆r values show deviations from it. The dendrometer data was set as 0 on the first day of this selected period. During the three-month period in 2009 there was only 750–850 µm maximum increment for Quercus petraea 500–600 µm for Quercus cerris. Stem radius stopped to increase by the first week of July 2009 and then only short-term fluctutations occurred due to the daily transpiration and rainfall events. Mészáros, I. et al. Acta Silv. Lign. Hung. 7, 2011 102 In 2008, however, both species showed three or four times larger stem radius increase in the corresponding period than in 2009 (Figure 2). In the corresponding period of 2008 the stem radius change was 2100–2300 µm for Quercus petraea and 1800–2100 µm for Quercus cerris in the three month period. In 2008 the stem radius showed increases till the end of July. 2008 stem radius change (µm) 0 500 1000 1500 2000 2500 3000 rainfall (mm) 0 5 10 15 20 25 30 Q. cerris Q. petraea rainfall 2009 DOY 160 180 200 220 240 stem radius change (µm) 0 500 1000 1500 2000 2500 rainfall (mm) 0 5 10 15 20 25 Figure 2. Temporal course of cumulative stem radius change ( ∆ r) for Quercus petraea (black line) and for Quercus cerris (gray line) and daily sums of rainfall in summer of 2008 and 2009. DOY 158 is considered as reference day and ∆ r values show deviation from it The three-month long dendrometer data series were also used to estimate fluctuations in water status of tree stems. Seasonal variations in stem water deficit calculated by means of method suggested by Zweifel (2005) are presented in Figure 3. The general seasonal course of ∆W was similar for the two species in both years. Seasonal amplitude of stem water deficit differed in the two species, it was smaller for Quercus cerris than for Quercus petraea. Variation of stem water deficit was, however, usually the most significant during dry periods and approached zero after heavy rain events. Even one-two week dry periods could induce rapid increases of tree water deficit depending on the soil water availability and VPD. In 2008 large stem water deficit developed within the period from DOY 175 to DOY 190 (4th week of June and 1st week of July) up to 250 µm for Quercus cerris, and 500 µm for Quercus petraea and between DOY 223 and DOY 236 up to 230 µm and 420 µm for Quercus cerris and Quercus petraea, respectively. In 2009 a long-lasting stem water deficit period appeared from DOY 200 to DOY 234 with 300 and 450 µm maximum values of ∆W for for Quercus cerris and Quercus petraea, respectively. There was only a short (3 days) interruption of this period when ∆W of stem approached 0 in both species. The diurnal fluctuation of ∆W were also different in the two species (Figure 3). Diurnal and seasonal changes in stem radius increment and sap flow density Acta Silv. Lign. Hung. 7, 2011 103 2008 stem water deficit (µm) 0 250 500 Q. cerris Q. petraea 2009 DOY 160 170 180 190 200 210 220 230 240 stem water deficit (µm) 0 250 500 Figure 3. Variation in stem (tree) water deficit (∆W, µ m) with time for Quercus petraea (black line) and Quercus cerris (gray line) On most sampling days in 2008 both species showed lower diurnal amplitudes of ∆W than in 2009. In 2008 diurnal amplitude of ∆W reached 80–120 µm in Quercus petraea and 40–80 µm in Quercus cerris. In 2009 the diurnal variation of stem radius extended to 130–200 µm in Quercus petraea and to 40–100 µm in Quercus cerris. 3.3 Variation of main daytime sap flow density and correlation with PET in summer of 2009 In the second half of the vegetation period of 2009 we experienced severe drought at the site. During the experimental period from DOY 209 to DOY 283 mean daytime sap flow density ranged between 0.11 and 0.04 ml cm–2 min–1 in Quercus cerris, and 0.07 and 0.02 ml cm–2 min–1 in Quercus petraea. During the same period the daily PET changed between 12 and 4 mm day–1. DOY 209 211 215 221 222 223 224 225 226 227 228 229 230 231 232 233 234 238 239 240 244 245 246 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 mean daytime sap flow density (ml cm -2 min -1 ) 0,00 0,02 0,04 0,06 0,08 0,10 0,12 PET (mm day -1 ) 0 4 8 12 Q. cerris Q. petraea PET Figure 4. Daytime mean sap flow density of Quercus petraea (black line) and Quercus cerris (gray line) and potential evapotranspiration estimated by Hargreaves-Samani temperature based method (Hargreaves-Samani 1982) during the experimental period from DOY 209 to DOY 283. Rainy days are excluded from the data set Mészáros, I. et al. Acta Silv. Lign. Hung. 7, 2011 104 A general decreasing trend of mean daytime sap flow density was observed for both species as the drought proceeded in August and September (Figure 4). Analysis of a dataset from this period (50 days after exclusion of rainy days) showed that the daytime sap flow density of both species changed in close positive correlation (P<0.01) with PET which also showed declining trend in the second part of growing season. The correlation was closer for Q. petraea (R2=0.6248) than for Q. cerris (R2=0.7410). 3.4 Diurnal course of sap flow density From the continuous sap flow measurements of 2009 we have selected days from two characteristic periods to assess the relationship between sap flow density and environmental conditions at daily scale. i) DOY 209, DOY 211 and DOY 215 represent sunny days in a moderately rainy period with small rains on DOY 210 (1 mm), DOY 212 (7mm) and DOY 214 (3.4mm), decreasing SWC (from 0.110 to 0.099 cm3 cm–3), high daytime VPD values (VPDmax 3–4 KPa) and light intensity (PPFDmax 1800–1900 µmol m–2 s–1). There were only two small rains (<4 mm) rain during the previous 10 days (Figure 5). ii) DOY 231, DOY 232 and DOY 233 in the driest period of summer with SWC 0,080–0,083 cm3 cm–3, high daily VPD (VPDmax 2.9–3.6 kPa) and PAR (PPFD 1700–1800 µmol m–2 s–1). There was no rainfall event in the previous 10 days (Figure 6). This period represented typical stage of the drought stress. The magnitude of diurnal sap flow amplitude depended significantly on the environmental conditions on selected days and decreased as the drought stress proceeded on sunny and hot days (DOY 209, 211 and 215) of a moderately rainy period in July, the mean SWC value on the selected three days reached 0.10 cm3 cm–3. The low SWC suggests that in July a significant depletion of soil water reserves occurred and small rains were not enough for soil refilling. Comparing with other periods of growing season both tree species exhibited relatively high stem water deficit on these days (Figure 3). 209 PPFD (µmol m -2 s -1 ) 0 500 1000 1500 2000 211 VPD PPFD 215 VPD (kPa) 0 1 2 3 4 209 sap flow density (ml cm -2 min -1 ) 0,02 0,04 0,06 0,08 0,10 0,12 0,14 215 0,02 0,04 0,06 0,08 0,10 0,12 0,14 211 209 daily stem radial changes (µm) -150 -100 -50 0 50 100 150 200 215 -150 -100 -50 0 50 100 150 200 DOY 211 Q. petraea Q. cerris Figure 5. Diurnal course of VPD, light intensity (PPFD) and sap flow density and daily stem radial changes of Quercus petraea and Quercus cerris on sunny days of a moderately rainy period in July 2009 Diurnal and seasonal changes in stem radius increment and sap flow density Acta Silv. Lign. Hung. 7, 2011 105 On DOY 209–Doy 215 the mean daily stem water deficit (∆W) was 169 µm for Q. cerris and 251 µm for Q. petraea (Figure 3). Between the two selected periods sap flow density of both tree species was the highest on these days. The daytime mean sap flow was 0.10 ml cm2 min–1 in Q. cerris and 0.066 ml cm2 min–1 in Q. petraea. The daily maximum of sap flow was relatively high and approached 0.14 ml cm2 min–1 for Q.cerris and 0,09 ml cm2 min–1 for Q. petraea (Figure 5). VPD was very high on these days and showed daily maximum between 3.1 and 3.9 kPa. Sap flow density was in close correlation with VPD but showed a maximum earlier during the day. The stem radius reached maximum in the morning (between 6 and 8 a.m.) and minimum values in the afternoon (between 4 and 6 p.m.). Temporal appearance of minimum was closely related to the maximum of VPD. In 2009 a progressive drought appeared from the first week of August which lasted almost for the whole month and was interrupted only by short rain events (Figure 1). SWC reached its minimal value (0.0819 cm3 cm–3) in this month considering the whole summer (Figure 1). On the three selected representative days of drought period (DOY 231–233) the daily maximum of sap flow was low in both tree species (0.101 ml cm2 min–1 and 0.064 ml cm2 min–1 (Figure 6). These sap flow density values were 28 % (Q. cerris) and 24 % (Q. petraea) lower than in July (Figure 5). Q. cerris exhibited very similar diurnal course of sap flow density as in July in correlation with PPFD and partly with VPD too. While Q. cerris showed maximum sap flow density at midday, Q. petraea had a short maximum of sap flow density in the morning (at VPD 2–2.5 kPa) followed with a gradual reduction later on the day. The amplitude of diurnal variation in the stem radius was higher in Q. petraea than in Q. cerris. Stem water deficit (∆W) was 55 % (Q. cerris) and 49 % (Q. petraea) larger were deduced than in July. Mean daily value of ∆W was 263 µm for Q. cerris and 374 µm for Q. petraea (Figure 3) suggesting the reduction of stem water storage due to the drought. 231 PPFD (µmol m-2 s-1) 0 500 1000 1500 2000 232 VPD PPFD 233 VPD (kPa) 0 1 2 3 4 231 sap flow density (ml cm-2 min-1) 0,02 0,04 0,06 0,08 0,10 0,12 0,14 Q. cerris Q. petraea 233 0,02 0,04 0,06 0,08 0,10 0,12 0,14 232 231 daily stem radial changes (µm) -150 -100 -50 0 50 100 150 200 233 -150 -100 -50 0 50 100 150 200 DOY 232 Q. cerris Q. petraea Figure 6. Diurnal course of VPD, light intensity (PPFD) and sap flow density and daily stem radial changes of Quercus petraea and Quercus cerris on sunny days during the drought in August 2009