WATER STRESS IN LONGLEAF PINE INDUCED BY LITTER REMOVAL
DEBORAH L. GINTER*, KENNETH W. McLEOD and CASEY SHERROD, JR
Savannah River Ecology Laboratory, Aiken, S.C. 29801 (U.S.A.) and Institute o f Ecology, University of Georgia, Athens, Ga. 30602 (U.S.A.)
*Present address: School of Forestry, University of Missouri, Columbia, Mo. 65201 (U.S.A.)
(Accepted 27 November 1978)
ABSTRACT
Ginter, D.L., McLeod, K.W. and Sherrod, Jr, C., 1979. Water stress in longleaf pine induced by litter removal. Forest Ecol. Manage., 2 : 13--20.
A forest litter layer acts as a nutrient pool to support future growth of trees and as a blan- ket which intercepts and absorbs rainfall and insulates soil from evaporative water loss. Lit- ter disturbance can modify either of these functions. Soil moisture content and xylem pres- sure potential were determined in a longleaf pine plantation for 7 weeks following litter re- moval. Two weeks following treatment, xylem pressure potential of trees in control, red straw removal and total litter removal plots diverged. Trees in the total litter removal plot had the lowest xylem pressure potential (i.e. most stressed), control trees had the highest xy- lem pressure potential, and trees in the red straw removal plot were intermediate. Differen- ces in xylem pressure potential among treatments were consistent throughout the season, regardless of precipitation or drought. Soil moisture content also showed different patterns based on the particular litter removal treatment. Differences in soil water content and xy- lem pressure potential could explain observed growth reductions of trees in plantations where litter had been removed. Therefore consideration of the potential modification of both the hydrologic and nutrient cycles should be made prior to litter removal.
INTRODUCTION
Recent studies illustrate how climatic, edaphic, and species differences in- fluence plant water relationships (Wambolt, 1973; Camacho-B et al., 1974; Heth, 1974; Hail and Kaufmann, 1975). Much of this research has been made possible by the development of portable field instruments which allow rapid determinations of plant water status. However, practical application of these instruments to evaluate impacts of particular treatments or applied stresses has been slow to develop. A growth reduction in longleaf pine (Pinus palustris Mill.) correlated with red straw removal, a common practice in the southeast- ern United States has been shown (McLeod et ai., 1979). Red straw removal involves removal of freshly fallen needles from pine plantations for sale to nurs- eries and landscapers. Remaining litter is left relatively undisturbed. Major alter ations of both the nutrient and hydrologic cycles are highly probable following
14
litter removal. Some aspects of nutr ient cycle modification are discussed by McLeod et al. {1979). The modification of the hydrologic cycle could include decreased water availability to trees by increasing evaporation from the soil due to a reduction in the insulating litter layer and/or by increasing competi t ion with released understory vegetation. Decreased water availability could be at least partially responsible for the observed growth reduction. Therefore, the purpose of this study was to determine the effect of litter removal on soil wa- ter content and tree water status in longleaf pine plantations.
MATERIALS AND METHODS
Extensive longleaf pine plantations exist on the United States Department of Energy's Savannah River Plant located near Aiken, South Carolina. The ex- perimental site was a 19-year-old longleaf pine plantation planted on abandon- ed farmland with a Troup loamy sand soil (Aydelott , 1971). To ascertain the effect of litter removal on tree water potential, three plots were treated on 9 June 1976 in the following manner: control (C), red straw removal (R), and total litter removal (T). Based on five 1 m 2 subplots per t reatment plot, the litter removed amounted to 4647.0 + 1006.2 kg ha- i (~ + 1 S.D.) in the red straw removal plot and 38104.0 + 10317.2 kg ha- 1 in the total litter removal plot. Litter depth before removal was 6.2 + 3.0 cm (X + 1 S.E.) for the three plots. Plots ranged from 0.16 to 0.18 ha in size. To avoid any possible edge effects, all sampling was confined to a central area of the treated plots, with a 7.5 m border surrounding it. This insured that roots existed entirely within treated areas. Due to effects of tree density on xylem pressure potential as described by Wambolt (1973), an a t tempt was made to find equal density plots. Densi- ties still varied (C, 148 trees ha-l; R, 159 trees ha-l; T, 135 trees ha-l).
Beginning 5 days before the plots were treated, soil moisture was measured gravimetrically three times weekly to a 0.90 m depth in each plot to determine the effect of treatment. Locations for soil sampling were randomly chosen: one location along each of three transects per plot.
Trees with healthy needle clumps at a height of approximately 6.1 m and a DBH between 12.7 and 25.4 cm were selected for measuring xylem pressure potential. The needle clumps were marked for repeated sampling. Subsequent measurements {overall height, age, distance from bole to sampled branchlet, branchlet diameter at needle clump and number of live branches below branch sampled) were made to characterize individual trees and were found to be similar between treatments. Important characteristics are shown in Table I.
All trees were sampled three times weekly using climbing ladders. Three needles were collected between 10.30 h and 11.30 h from each selected clump. Since needle age has been shown to affect xylem pressure potential (Running, 1976), only needles of the previous season's growth were sampled. Needles were stored in tightly rolled polyethylene bags to minimize transpiration which could result in changed water potential. They were returned to the laboratory, recut to remove resin plugs formed during storage, and xylem pressure poten-
15
TABLEI
Characteristics of sample trees. Treatments: R = red straw removal plot, C = control plot, and T = total litter removal plot
tial measu red using the pressure b o m b m e t h o d (Scho lander e t al., 1965) . The pressure b o m b does n o t y ie ld absolute values fo r x y l e m pressure po ten t i a l (Hsiao, 1973; Ri tch ie and Hinck ley , 1975) . However , on a relat ive scale the pressure b o m b is well adap ted for expe r imen t s requir ing on ly compara t ive values (Boyer , 1969) .
Methods o f s torage and recu t t ing could po ten t i a l l y m o d i f y leaf x y l e m pres- sure po ten t ia l . Since r ecu t t ing p roved to be necessary because o f resin plugs which f o r m e d in the needles, x y l e m pressure po ten t i a l was d e t e r m i n e d for once and twice cu t needles. A t-test ind ica ted no signif icant d i f fe rence b e tw een once and twice cu t needles (X _+ 2 S.E., - -24 .70 + 0.46 and - -24 .43 + 0 .66 bars, re- spect ively) . The x y l e m pressure po ten t i a l o f needles s tored as in the dally sam- pling for up to 8 h was also de t e rmined . The re was no signif icant change in xy- lem pressure po ten t i a l dur ing the 8 h per iod , initial value at t ime zero of - -23 .04 + 0 .88 ()( + S.E.) and final value o f - -23 .83 + 0 .88 bars.
A dawn- to -dusk curve o f x y l e m pressure po ten t i a l was d e t e r m i n e d for th ree t rees in each p lo t on 1 day, which was 29 days fo l lowing t r e a t m e n t and 1 day fo l lowing a rainfall o f 10.7 mm. This d e t e r m i n e d the diurnal pa t t e rn o f x y l e m pressure po ten t i a l which occur red . Three needles per t ree were sampled at h o u r l y intervals f r om 06 .00 (pre-dawn) to 21 .00 h (post~lusk) . Sampling and d e t e r m i n a t i o n o f x y l e m pressure po ten t i a l were as descr ibed previously . Solar rad ia t ion dur ing this pe r iod was measured wi th a py rhe l iome te r . To de te rmine w h e t h e r ou r arbi t rar i ly chosen he ight r ep resen ted a response c o m m o n to the whole t ree , needles f r om three heights on th ree t rees per p lo t were also mea- sured. Sampl ing and d e t e r m i n a t i o n o f po ten t ia l s were as descr ibed above.
Da ta were ana lyzed using one- and two-way analysis o f var iance p rocedures in the Stat ist ical Analysis Sys tem (Service, 1972) . T u k e y ' s Tes t was used to examine d i f fe rences b e t w e e n t r e a t m e n t means (Kirk, 1968) .
RESULTS AND DISCUSSION
Prior to t r e a t m e n t o f the plots , the re was no significant d i f f e rence in soil wa-
16
ter content be tween treatments within each soil depth (P > 0.05). After treat- ment, soil water content varied with respect to rainfall, t reatment, and depth (Fig. 1), although these differences were rarely statistically significant due to sample size. Although three samples per t reatment at each time period may not be totally adequate to show statistical difference in soil water content , we do feel that they are adequate to illustrate the different response in soil moisture of the t reatments to precipitation events. The effects may be at t r ibuted to lit- ter acting in both insulative and absorptive capacities. Higher amounts of wa- ter were found in the surface layer of T plot following moderate rainfall due to removal of the absorptive litter layer. Evaporation from the open soil surface caused less soil water to be present in R and T during dry periods. Response of the control plot was generally slower and more buffered due to the litter layer absorbing light rains and retarding evaporation of soil water. An interesting ob- servation is the consistently higher soil moisture levels in the 61--90 cm depth
~ I0
h i
p-- u~ 5
_ J
I0
I01 Treotment ~ C
B
C
t
~ 4o
~ 2o I I . I i _
TIME (days) Fig. 1. Seasona l soil m o i s t u r e and ra infal l in e x p e r i m e n t a l p lo ts . A, 0 - -30 c m d e p t h ; B, 31- - 60 c m d e p t h ; C, 6 1 - - 9 0 cm d e p t h . Each value r ep resen t s t he m e a n o f t h r ee repl icates . Treat - m e n t : C = c o n t r o l p l o t (e ) , R = r ed s t raw remova l p lo t (o) , and T = to t a l l i t t e r r emova l p l o t (O) .
17
in C plot than removal plots. The R plot was usually not different from the T plot in soil moisture. Differences in soil water content at the lower depth are probably not due to evanorative loss. Increased evaporative demand within the canopy caused by change in forest floor albedo may have increased transpira- tion and be responsible for the soil water differences at 61--90 cm. Another potential consequence of the observed changes in soil moisture of the 0--30 cm soil depth is a dehydration of the remaining litter in the R plots. This de- hydration would then inhibit nutrient release from the decomposition process. The softs have low cation exchange capacity due to their sandy nature and are not able to contribute much to the plant's nutrition. Therefore continual in- put of nutrients from decomposition is very important.
Up to 14 days following treatment, no significant difference in xylem pres- sure potential existed between trees in treated plots (Fig. 2). After 14 days, potentials for trees in the three plots were significantly different (P < 0.01).
3 0
TIME SINCE TREATMENT {days)
Fig. 2. Seasonal xylem pressure potential of longleaf pine following litter removal treat- ments. Each value represents the mean of three needles from each of ten trees. [C (e); R (a);T(o)].
0
,0t E
.J
I- z
~ 2c
Since the soil--plant water continuum is already well-documented (Gardner, 1960; Camacho-B et al., 1974; Lopushinsky and Klock, 1974; Kaufmann, 1975), these differences may reflect differences occurring in the soil. Trees in T plot always showed the more severe stress, R trees showed moderate stress and C trees showed the least stress. These differences between treatments ex- isted throughout the period sampled with stress increasing during dry periods (e.g. days 30--36) and stress declining immediately following rainfall events (e.g. days 37:-40, days 47--49). Trees in R plot and T plot responded to rain- fall but never recovered to the xylem pressure potential of trees in C plot be- yond day 12. Furthermore, the three plots as a group never returned to early season (prior to day 14) xylem pressure potential levels, indicating seasonal wa- ter stress.
The dawn-to-dusk pattern of water potential in longleaf pines (Fig. 3) agrees
18
1.6 Z O A
< ~ 0.8
u
~ ' 5 0.4 j u
0.0
'f . . . . . t
oor. o.f,
0600 0800 I0001200 1400 16001800 200( TIME OF DAY
75 i,,. O
I0.0 "
12.5 ,~ I..-
15.0 ~ i--
17.5 ~
20.0 ~-
22.5
25.0
Fig. 3. Diurnal pa t te rn of xylem pressure potent ia l of longleaf pine and solar radiat ion in li t ter removal plots, 1 day following a rainfall of 10.7 mm. Each value represents the mean of three needles from each of three trees. [C (e); R ([:3); T ( o ) ]
with that found by Cline and Campbell (1976) and Hellkvist and Parsby (1976). After first light in the morning, xylem pressure potential began a rapid decline. This is controlled by the opening of the stomata and the physical en- vironment acting on the trees. Trees in all three treatments were similar in their response in the morning hours. This is probably due to the cloudiness of the morning as shown by the solar radiation record for the day and the preceed- ing 10.7 mm rainfall. Maximum water stress in all three treatments occurred at noon. By 18.00 h the rate of water recovery began to decline. At 21.00 h the trees had nearly returned to early morning water potential values. Although few differences between treatments existed during the day, differential res- ponse was observed between treatments. Based on the pattern of xylem pres- sure potentials observed in T plot, stomatal closure probably occurred. This would not appear as a statistical difference but nevertheless represents a differ- ent response than the trees in the other treatments. Statistical differences were found at 17.00, 18.00, and 21.00 h. Differences between treatments at dusk suggest tha t length of recovery during the night is longer for T trees than C or R trees. Since the order of sampling was rerandomized for each hourly mea- surement, these differences are not felt to be error due to sampling pattern.
The relationship between photosynthet ic rate and xylem pressure potential is undocumented for longleaf pine. Bacone et al. (1976) and Ormsbee et al. (1976) show --15 bars to be the point beyond which photosynthesis is reduced
19
in Ulmus alata and Juniperus virginiana. If one considers --15 bars to be the point beyond which photosynthesis is impaired, then trees in the C, R, and T plots had impaired photosynthesis for 6.2, 6.4, and 7.9 h per day, respectively. Thus stress induced by litter removal t reatments could contr ibute to the ob- served growth reduction due to this reduction in photosynthet ic activity.
In addition to t reatment effect, a height effect was observed (Fig. 4). Based on three heights examined, water stress increased as height increased. This is probably due to xylem resistance which increases as distance increases (Ritchie and Hinckley, 1975). Thus the observed t reatment effects at a 6.1 m height are reflecting effects observed throughout the tree.
25 I I ¥ 6.1 9.1 12.2
HEIGHT (m)
Fig. 4. Xy lem pressure potent ia l o f longleaf pine as affected by height of needle and l i t te r removal treatment. Each value represents the mean of three needles from each of three trees. [C (e); R ( D ) ; T ( o ) ]
15 A o
. a
i I
Z W
e¢ W
In summary, soils responded to the treatments in a manner such that litter removal plots showed relativelv high water content after rainfall bu t exhibited lower water content during dry periods which dominated the season's climate. Trees in the plots appeared to respond to their own particular t reatment. Trees of the control plot remained at higher xylem pressure potentials through- out most of the season, trees of the total litter removal plot showed the great- est stress by having the lowest water potentials, and trees in the red straw re- moval plot were intermediate. It is expected that due to low water holding ca- paci ty of sandy soils, the response to litter manipulation would be greater than if clay soils were involved. Although many other modifications are occurring because of litter removal, such as disruption of nutrient cycling, greater water stress is induced by litter removal and could be responsible for some of the growth reduction observed in longleaf pine. Therefore one must also consider this impact before litter manipulations are performed.
20
ACKNOWLEDGEMENTS
This research was supported by contract EY-76-C-09-0819 between the De- par tment of Energy and the University of Georgia. D.L. Ginter was supported as an Undergraduate Research Participant. The authors wish to acknowledge Ms Jean Turner-Mobley and the entire SREL technical and senior staff for the assistance and advice.
REFERENCES
Aydelo t t , DIG., 1971. Soils of the Savannah River Project. U.S.D.A.--U.S.F.S., 35 pp. Bacone, J., Bazzaz, F.A. and Boggess, W.R., 1976. Correlated photosynthet ic responses and
habitat factors of two successional tree species. Oecologia (Berlin), 23: 63--74. Boyer, J.S., 1969. Measurement of the water status of plants. Ann. Rev. Plant Physiol., 20:
351--364. Camacho-B, S.E., Hall, A.E. and Kaufmann, M.R., 1974. Efficiency and regulation of water
t ransport in some woody and herbaceous species. Plant Physiol., 54: 169--172. Cline, R.G. and Campbell, G.S., 1976. Seasonal and diurnal water relations of selected forest
species. Ecology, 57: 367--373. Gardner, W.R., 1960. Dynamic aspects of water availability to plants. Soil Sci., 89: 63--73. Hall, A.E. and Kaufmann, M.R., 1975. Regulation of water t ransport in the soil-plant-atmo-
sphere continuum. In: D.M. Gates and R.B. Schmere (Editors), Perspectives of Biophys- ical Ecology. Springer-Verlag, Berlin, pp. 187--202.
Hellkvist, J. and Parsby, J., 1976. Water relations of Pinus sylvestris. Physiol. Plant., 38: 61--68.
Heth, D., 1974. Water potentials of stressed pine seedlings under controlled climatic condi- tions. Isr. J. of Bot., 28: 127--131.
Hsiao, T.C., 1973. Plant responses to water stress. Ann. Rev. Plant Physiol., 24: 519--570. Kaufmann, M.R., 1975. Leaf water stress in Engelmann spruce. Plant Physiol., 56: 841--
844. Kirk, R.E., 1968. Experimental Design: Procedures for the Behavioral Sciences. Brooks/
Cole Publ. Co., Belmont, Calif., 577 pp. Lopushinsky, W. and Klock, G.O., 1974. Transpiration of conifer seedlings in relation to
soil water potential . For. Sci., 20: 181--186. McLeod, K.W., Sherrod, Jr, C. and Porch, T.E., 1979. Response of longleaf pine plantations
to l i t ter removal. Forest Ecol. Manage., 2: 1--12. Ormsbee, P., Bazzaz, F.A. and Boggess, W.R., 1976. Physiological ecology of Juniperus
virginiana in old fields. Oecologia (Berlin), 23: 75--82. Ritchie, G.A. and Hinckley, T.M., 1975. The pressure chamber as an instrument for ecolog-
ical research. Adv. Ecol. Res., 9: 165--254. Running, S.W., 1976. Environmental control of leaf water conductance in conifers. Can. J.
For. Res., 6: 104--112. Scholander, P.F., Bradstreet, E.D. and Hemmingsen, E.A., 1965. Sap pressure in vascular
plants. Science, 148: 339--346. Service, J., 1972. A User's Guide to the Statistical Analysis System. North Carolina State
University, Raleigh, N.C., 260 pp. Wambolt , C.L., 1973. Conifer water potential as influenced by stand density and environ-
