( en | de | ru )

 

  2. user
  3. @dianella
  4. Freezing-induced xylem...
@dianella

copydeleteadd this publication to your clipboard
Freezing-induced xylem cavitation and the northern limit of @Larrea tridentata

, and . Oecologia, (1997)

Abstract

We investigated the occurrence of freezing-induced cavitation in the evergreen desert shrub Larrea tridentata and compared it to co-occurring, winter-deciduous Prosopis velutina. Field measurements indicated that xylem sap in L. tridentata froze at temperatures below c. ?5C, and that this caused no measurable cavitation for minimum temperatures above ?7C. During the same period P. velutina cavitated almost completely. In the laboratory, we cooled stems of L. tridentata to temperatures ranging from ?5 to ?20C, held them at temperature for 1 or 12 h, thawed the stems at a constant rate and measured cavitation by the decrease in hydraulic conductivity of stem segments. As observed in the field, freezing exotherms occurred at temperatures between ?6.5 and ?9C and as long as temperatures were held above ?11C there was no change in hydraulic conductivity after thawing. However, when stems were cooled to between ?11C and ?20C, stem hydraulic conductivity decreased linearly with minimum temperature. Minimum temperatures between ?16 and ?20C were sufficient to completely eliminate hydraulic conductance. Record (>20 year) minimum isotherms in this same range of temperatures corresponded closely with the northern limit of L. tridentata in the Mojave and Sonoran deserts.

Links and resources

BibTeX key
PockmanSperry_97
entry type
article
year
1997
journal
Oecologia
pages
19--27
volume
109
citeulike-article-id
1523891
priority
5
posted-at
2007-07-31 06:53:37
citeulike-linkout-0
http://www.springerlink.com/index/NJ2567XT89JB47N1.pdf http://www.biology.utah.edu/sperry/pubs/Oecologia97.pdf
comment
(private-note)Look up: Shreve 1914, George et al. 1974, Sakai and Larcher 1987: freezing tolerance affects distribution. (Tyree et al. 1994)... Compressive surface tension forces are given by laplace's law = 2(t/r) where t = surface tension of water, r=radius (phi)(sub)px tries to expand the bubble. The biggest of these two forces wins. Equilibrium freezing (Beck et al. 1984) - What is it? It's prevents freezing of living cells, but it isn't supercooling. Salleo et al 1996 - an auxin-dependent increase in water potential for refilling Refilling may occur more than once per season, especially for herbs: (Waring et al. 1979; Tyree et al. 1986; Salleo and LoGullo 1989; Ewers et al. 1991; Lewis et al. 1994; Salleo et al. 1996). chucked branch segments in freezer, (or frozen in modified garbage can with copper coil, but supposedly still quickly, left there at min temp for 1 hour, removed After 1 h at minimum temperature, the chamber temperature was increased. For stems at the lowest minimum temperatures, warming rates approached 1C min?1 until the exotherm temperature was approached (?10C). Warming rate for all stems was lower from exotherm temperature until wood temperatures were above 10C. The warming rates from minimum temperature to 0C ranged from 0.13 to 0.35C min?1 (mean = 0.21C min?1, SD = 0.08). This compares to warming rates of 0.1?0.2C min?1 measured in the field on the coldest night of the winter. 12 h vs 1 h at min temp were tested different min temps: ?5.6, ?8.4, ?11.8, ?13.9, ?16.2 or ?18.7C. Diameter distributions were then calculated using 5- and 10-mm classes based on both the actual percent of vessels in each class (frequency distribution) and on the estimated percent of total conductance contributed by each class (hydraulic distribution). The latter was calculated as the percent contribution of each diameter class to the sum of the fourth power of the radius of all conduits (Sperry et al. 1994). We also calculated the means of both distributions. The mean of the hydraulic distribution equals the sum of the fifth power of the vessel radii divided by the sum of the fourth power of the radii (Sperry et al. 1994). We refer to this as the ?hydraulic mean??. Vessel volume distributions were calculated using two models of the relationship between conduit length and diameter. First, it was assumed that the longest conduits were also the widest (model 1) as shown by Zimmermann and Potter (1982). Based on this assumption, we calculated the volume associated with a given pair of diameter and length classes. We started with the largest diameter and longest length class, and using the middle of each size class for the calculation, worked step-wise through progressively smaller size classes. The percentage of vessels in each volume class was taken to be the smaller of the percentages in each pair of length and diameter classes considered. For example, given a length class with 10\% of the total conduits and a diameter class with 15\%, 10\% of the total vessels fell in the corresponding volume class. The remaining 5\% of the total vessels in the diameter class were applied to a volume class calculated using the middle of the initial diameter class and the next shortest length class. This procedure was repeated until all classes of both distributions had been included in the calculation. Since there is only limited empirical basis for the assumption behind model 1 (Zimmermann and Potter 1982), we also calculated the vessel volume distribution based on the more conservative assumption that vessel diameter is random with respect to vessel length (model 2). In this case the diameter distribution was applied to each length class to calculate the range of volumes associated with that class. ---=note-separator=--- (private-note)They found it hard to detect exotherms in the field. They didn't really detect much embolism either. (probably because it had repaired by the time they collected their branches?*) The deciduous tree had far more embolism problems. Yes, there was a temperature effect. Above -9, embolism didn't occur. At -11 and below, it increased linearly. cooling and warming rates didn't have a significant effect. Cooling rates were between 0.14 and 0.24C min?1 (mean = 0.17C min?1, SD = 0.04) Warming rates from minimum temperature to 0C ranged from 0.12 to 1.35C min?1 across all experiments and were not correlated with percentage embolism Total time frozen was also not correlated with embolism. No relationship btw. xylem water potential and incidence of embolism neither. Temperature at which 100\% embolism was observed (-16 to -20) was also the northern limit on a temperature minimum isotherm map of the southern states of the USA. Quercus ilex of the southern mediterranean basin developed increasing embolism with decreasing minimum temperature (Lo Gullo and Salleo 1993), as did Rhus laurina and R. ovata of the California chaparral (S.D. Davis and F.W. Ewers, unpublished work). The temperature dependence might arise from the fact that the water potential of ice decreases steeply with declining temperature (Rajashekar and Burke 1982; Beck et al. 1984). As the water potential of apoplastic ice decreases, water is drawn from surrounding xylem parenchyma into the xylem conduits and intercellular spaces where it freezes. As a result, the colder the ice, the more dehydrated the surrounding cells become. Upon thawing, the water potential gradient into the living cells would be determined by the minimum temperature during the freezing episode. A colder freeze would result in a steeper gradient, and a more rapid uptake of apoplastic water during the thaw. A more rapid uptake could result in a more rapid establishment of low Ypx during thawing, and thus more cavitation. Our work leaves unanswered the question of what is responsible for the high native percentage embolism (\~60\%, Fig. 1a) in the population of L. tridentata that we studied.
Document
http://www.springerlink.com/index/NJ2567XT89JB47N1.pdf http://www.biology.utah.edu/sperry/pubs/Oecologia97.pdf

Tags