Influence of rocks on soil temperature, soil water potential, and rooting ...

Oecologia (1992) 92:90-96

Oecologia 9 Springer-Verlag 1992

Influence of rocks on soil temperature, soil water potential, and rooting patterns for desert succulents Park S. Nobel, Patsy M. Miller, and Eric A. Graham Department of Biology and Laboratory of Biomedical and Environmental Sciences, University of California, Los Angeles, CA 90024, USA Received April 22, 1992 / Accepted in revised form July 7, 1992

Summary. At a site in the Sonoran Desert, subterranean rocks and exposed boulders affected soil water potential as well as root morphology and distribution. For Agave deserti, the number of lateral roots per unit length of main root was 11 times higher under rocks and six times higher alongside rocks than in rock-free regions. Total root length per unit soil volume for Echinocereus engelmannii averaged 3-fold higher within 1 cm of boulders than 5 cm away, where the soil was drier. The total length of lateral roots per unit length of main root for Ferocactus acanthodes was 4.2 m m-1 under rocks but only 0.8 m m - 1in rock-free regions. The number of lateral roots per unit length of main root for Opuntia aeanthocarpa was 7-fold higher alongside rocks than in rock-free regions and even higher under rocks. For transplanted and watered A. deserti, the number of new main roots produced per 1-2 month interval averaged 13 for five plants on the north side of boulders, 8 on the south side, 11 for five plants with half of their roots under rocks, 2 for those with half of their roots over rocks, and 3 for the control plants without rocks. Laboratory experiments showed that the soil water potential under rocks for 10 and 30 mm waterings stayed above - 0 . 5 MPa for 13 and 19 d longer, respectively, than for regions away from rocks. The shortwave absorptance of granitic rocks from the field site was 0.82, the thermal conductivity coefficient was 1.50 W m -1 ~ -1, and the volumetric heat capacity was 1.75 MJ m -3 ~ -1. Field measurements indicated that 5-cm-thick buried rocks decreased the diel variation in soil temperatures on their undersurface by only 0.4 ~ C compared with soil. Thus, the primary influence of rocks at the field site on root proliferation and branching for the four species was apparently caused by influences on soil water content. Key words: Agave - Cactus - Root morphology - Sonoran Desert - Water potential

Correspondence to: P.S. Nobel

Rocks affect the penetration of rainfall into soil and local soil temperatures. Infiltration of rainfall and retention of soil moisture can both be increased by rocks at the soil surface (Lamb and Chapman 1943; Epstein et al. 1966; Agassi and Levy 1991). Rocks can have different thermal conductivities and different heat capacities than soil, thereby influencing the soil temperature in the immediate vicinity of the rocks. Specifically, surface rocks can act as insulators during the hottest part of the day and as heat sources at night (Jury and Bellantuoni 1976a). The thermal influences of rocks on animal activity have been extensively studied. For instance, ectotherms such as garter snakes (Thamnophis elegans) have specific daily periods when they are in close proximity to rocks (Huey et al. 1989). The behavior of various desert animals ranging from insects to mammals is also influenced by rocks, whose subterranean surfaces tend to moderate daily extremes of temperature (Cloudsey-Thompson 1956; Larmuth 1978; Dean and Turner 1991). Roots of various desert succulents are also influenced by rocks. Roots of Opuntia phaeacantha var. discata tend to occur below shallow rocks but alongside deeper rocks 25 cm below the soil surface (Cannon 1911). Its roots can be closely appressed to the rocks, where the roots often branch profusely, presumably because of greater moisture availability. Root nets of Opuntia acanthocarpa adjacent to the lower surfaces of flat rocks may enable this desert succulent to obtain water seeping around the rocks during wet periods as well as water moving upward in the soil by capillarity and collecting under rocks during drought (Nobel et al. 1991). In this regard, a rainfall of 13 mm leads to wet conditions 25 cm below a large rock but dry conditions at the same depth in the absence of rocks (Cannon 1911). Also, soil beneath rocks in the Negev Desert contained 9 to 12% water by weight when adjacent soil contained less than half as much (Evenari et al. 1971), and the moisture content under surface rocks about 0.5 kg in mass averaged 2.2% when the moisture content of adjacent soil averaged 0.6 % (Dean and Turner 1991). Roots tend to proliferate in pockets that retain water, such as subterranean cracks in large rocks (Hell-

91

mers et al. 1955; Drew 1979). In addition to decreasing the evaporation of water, rocks on the soil surface can result in water vapor migration to their cooler lower surfaces, where condensation can increase water availability for roots of desert plants (Stark 1970; Jury and Bellantuoni 1976b). The field site chosen to examine the influence of rocks on root distribution and branching is in the northwestern Sonoran Desert and is characterized by numerous rock outcrops, exposed boulders, and many subterranean rocks. The rock outcrops and boulders channel rainfall runoff to their periphery, causing the adjacent soil to become wet after even light rainfall, creating a microhabitat on the north side of boulders that is colonized by a fern, Notholaena parryi (Nobel 1978). Its roots occur in a dense mat mostly within 1 cm of the rocks, where a 7 mm rainfall raises the soil water potential to - 0.03 MPa compared with -0.35 MPa at the same depth (10 cm) away from the rocks. Fine roots and root branching of Echinocereus engelmannii and O. acanthocarpa are also more common adjacent to rocks than in rock-free regions at the field site (Nobel et al. 1991). Root occurrence and branching for these two species plus two other desert succulents (Agave deserti and Ferocaetus acanthodes) were quantified adjacent to rocks in the field. Also, A. deserti was transplanted so that its roots were in specific locations with respect to rocks, whose influence on soil temperature and soil water potential were determined. Laboratory experiments were performed to determine the influence of rocks from the field site on soil water potential in their vicinity and to measure the thermal properties of the rocks.

Materials and methods Field site - Plant species Field measurements were made at the University of California Philip L. Boyd Deep Canyon Desert Research Center near Palm Desert, CA, at Agave Hill (33~ 116~ 850 m elevation). The dominant species there is Agave deserti Engelm. (Agavaceae), which comprises about one-third of the rather sparse plant ground cover of 21% in the autumn (Nobel 1976). Other conspicuous succulent plants at Agave Hill include Echinocereus engelmannii (Parry) Lem. (Cactaceae), Ferocactus acanthodes (Lena.) Britton & Rose (Cactaceae), and Opuntia aeanthocarpa Engelm. & Bigel. (Cactaceae), whose roots were also examined. The site is hilly and rocky, with numerous outcroppings of disintegrating granite. The soil is a loamy sand, with the non-gravel portion of soil consisting of 73 % sand (particle sizes of 0.05-1.0 mm) by mass (Nobel 1977). Based on a survey o f over 200 rocks on the soil surface, the median dimension of a rock in the horizontal direction was 17 cm and the median thickness in the vertical direction was 5 cm.

potential (~soil) could be determined gravimetrically based on volumetric water content (Young and Nobel 1986); soil dry mass was obtained after drying in a forced-draft oven at 70 ~ C for 48 h.

Field experiments Square flat rocks averaging 16.8 • 0.5 (mean + range) cm in the horizontal direction and 4.7+ 0.3 cm in thickness were obtained from Agave Hill. On 21-22 September 1991, 25 similar mature plants of A. deserti 30 4- 2 cm tall with 14 4- 1 leaves unfolded from the central spike of folded leaves were excavated. These plants had 41 :k 4 main (nodal) roots averaging about 30 cm in length emanating from the base of the stem. The plants were replanted at Agave Hill with five in each of the following configurations: (1) on the north side of an exposed boulder at least 60 cm across, with the center of the plant about 20 cm from the boulder; (2) similarly placed on the south side of an exposed boulder; (3) in an open region, with half of the roots immediately under one of the square flat rocks placed horizontally with its center 10 cm below the soil surface; (4) similarly placed with half of the roots above such a rock; and (5) in an open region without rocks, which served as the control. Wescor PCT-55 15-SF soil thermocouple psychrometers (calibrated with NaC1 solutions) were placed with their axes horizontal at the center of the root zone at a depth of 10 cm for all treatments, either halfway between the plant's center and the boulders, immediately above or below the flat rocks, or at a depth of 10 cm in the regions without rocks. The psychrometers were about 10 cm radially from a plant's center and in all cases were under soiI shaded by the plant canopy (which extended approximately 20 cm radially), thereby minimizing temperature gradients (the resulting offset between sensing and reference junctions before Peltier cooling was < 3 gV). ~r was determined with a Wescor H R - 3 3 T microvoltmeter in the dew-poil~t mode at dawn on dates when the plants were re-excavated; samples for gravimetric analysis of ~t~oil were also obtained during the excavations and were used in the range - 3 to - 5 MPa. Rainfall between excavation dates amounted to 9 mm from 21 September to 26 October 1991, 5 mm from 27 October to 30 November, 42 mm from 1 December to 3 January 1992, and 93 mm from 4 January to 29 February. Copper-constantan thermocouples 0.51 mm in diameter were placed at various locations to monitor soil temperature, which was measured with an Omega HH-25TC digital thermometer. Root water potential was determined on 28-29 February 1992 with a SoilMoisture C950 pressure chamber (Nobel and Lee 1991).

Laboratory - Soil water potential To investigate the influence of rocks on Vsoilunder controlled conditions, four median-sized square flat rocks from Agave Hill were placed horizontally with their centers 10 cm below the soil surface in a 1 m x 1 m wooden box filled to 30 cm with 450 kg of soil from Agave Hill. Calibrated soil psychrometers were placed immediately above, below, and between the rocks. ~.-o~1was determined daily at 07:30 following waterings equivalent to depths of 10 and 30 ram. The box was maintained in a glasshouse with average daily maximal/minimal air temperatures of 26 ~ C/15 ~ C, average daily maximal/minimal relative humidities of 74%/39%, and 80% of the ambient solar irradiation.

Field observations Laboratory - Rock thermal properties Root systems were carefully excavated at Agave Hill on 28-29 February 1992, and depth, length, and proximity to rocks were noted for main roots of the four species considered. In addition, the frequency and the length of lateral roots occurring as branches on the main roots were determined. Soil samples (20-30 cm 3) were taken from the root zone of E. engelmannii so that the soil water

The three principal intrinsic parameters affecting temperature (Nobel 1991) were determined in the laboratory for rocks from Agave Hill. To determine the shortwave absorptance (a), an integrating sphere radiometer (Dunkle et al. 1960) was used with a Wescor M J55 psychrometric voltmeter to measure the thermopile

92 voltage. Calibration was done with standards of known reflectivity (Smith and Nobel 1977). The volumetric heat capacity (Cp; M J m -3 ~ -~) was determined for median-sized rocks whose volumes were determined by Archimedes' principle using water displacement. The rocks were heated to 95.0 ~ C for 24 h in a Haake ASl water bath and then placed in 2.4 litre of water at 20.0 ~ C in a Styrofoam chest within an insulated chamber. The water was stirred and its temperature was recorded with five copper-constantan thermocouples 0.51 m m in diameter. Based on the temperature rise when a rock was immersed in the water and on the known heat capacity of water, Cp was determined for the rocks; corrections for heat losses (less than 3%) were based on taking aluminum blocks of approximately the same volume through the identical procedure. The thermal conductivity coefficient (K; W m -~ ~ -1) was determined for rock slabs about 1 cm thick placed horizontally in the center of a 16-cm depth of soil from Agave Hill. Thermonetics H11-$357 and H11-$359 heat flux plates were placed a few millimeters above and below the slabs and four copper-constantan thermocouples 0.51 m m in diameter were placed on the upper and on the lower surfaces of the rocks. A steady-state heat flux of 200-250 W m -2 was created by an electric heating pad placed underneath the soil 24 h before the measurements. After temperatures stabilized, K for the rock slab was determined from the heat flux across it per unit temperature gradient (Nobel 1991).

4 a

=

7 cm

3

o n"

I

O

[

I

I

b -0.1

-o2

....

"

-}

-0.3

-0.4

0

1

2

3

4

5

Horizontol distance from boulder (cm)

Results

Root associations with rocks Agave deserti. Agave deserti had a higher frequency of lateral roots branching from its main roots within 1 cm of rocks than in regions without rocks (Table 1). The mean lengths of the excavated roots were approximately the same in soil regions without rocks, and where the roots were either above, alongside, or under subterranean rocks. The mean root depth below the soil surface was 9 to 10 cm, except for roots above rocks for which the depth was about half as great (Table 1). The frequency of lateral roots per unit length of main root was about 6-fold greater above or alongside rocks and about l 1-fold greater under rocks compared with soil regions without rocks. Consequently, the total length of lateral roots per unit length of main roots from which they branched was 1.11 m m -~ under rocks, about 0.25 m m -~ above or alongside rocks, and only 0.05 m m -z in soil regions without rocks (Table 1). Moreover, essentially no secondary lateral roots branched from the (primary) lateral roots occurring above, alongside, or without rocks; under rocks, 17% of

Fig. la, b. Influence of horizontal distance from exposed boulders at Agave Hill on (a) root length per unit soil volume for Echinocereus engelmannii and (b) soil water potential. Plants, averaging 24 cm in height with 9 stems in a clump, were immediately adjacent to the southeast, south, or southwest side of exposed boulders that were at least 60 cm across in the horizontal direction. Soil centered at the indicated mean depths of 7 cm (o) and 14 cm (~) was removed in layers approximately 1 cm thick in the horizontal direction away from the boulder, 4 cm in the vertical direction, and 5 to 7 cm horizontally along the rock surface. Root length refers to the total length of main roots, primary lateral roots, and secondary lateral roots. Water content was determined gravimetrically, and soil water potential (~oi~) was calculated using relationships for soil from Agave Hill (Young and Nobel 1986). Data are means • SE for four plants and were obtained on 28-29 February 1992

the 86 primary lateral roots had 2.14-0.3 (mean Jz SE, n-- 15) secondary lateral roots each.

Echinocereus engelmannii. Roots of Echinocereus engelmannii were more common alongside boulders than at increasing distances horizontally away (Fig. la). At a depth in the soil of 7 cm, the total length of main roots, primary lateral roots, and secondary lateral roots within 1 cm of a boulder was 3.2 cm per cm 3 of soil, decreasing 4-fold by 5 cm horizontally away. At a depth of 14 cm,

Table 1. Root properties of Ayave deserti whose roots were associated with rocks. Data are means _+SE (number of measurements) for main roots from seven plants averaging 38 cm in height, except

above rocks where data are means for the two roots observed. Rocks were below the soil surface and averaged 14.2_+2.3 (mean • SE for n = 14) cm in the horizontal direction along the roots

Quantity (units)

No rock

Above rock

Alongside rock

U n d e r rock

Length of main root (cm) Depth below soil surface (cm)

16.9 8.8

17.2 4.5

15.0 9.2

15.4 10.4

28 0.23

31 _+5 (4) 0.27 -I-0.05 (4)

Lateral roots per m of main root Number Length (m)

_+1.0 (16) _+0.4 (16)

5 __4 (16) 0.05 -t-0.04 (16)

-t-1.7 (4) +0.7 (4)

__ 1.1 (10) __ 0.4 (10)

56 _+11 (10) 1.11 _+ 0.14 (10)

93

the root length per unit soil volume was 1.8 cm cm -3 close to a boulder, decreasing over 2-fold at 5 cm away (Fig. la). Soil water potential was highest near the boulders and decreased horizontally away (Fig. lb). Compared with the soil interval within 1 cm of a boulder, where ~o~j averaged - 0.13 MPa, tV~o~~decreased more than 2-fold by 5 cm away. For all intervals, the soil was wetter at a depth of t4 cm than at 7 cm (Fig. lb).

and the lateral roots tended to be longer under rocks. Consequently, the total length of lateral roots per unit length of main root was greater than 4 m m-1 under rocks and less than 1 m m-1 in regions without rocks (Table 2).

Opuntia acanthocarpa. Main roots of Opuntia acanthocalTa tended to occur alongside rocks, with 40% as many occurring under rocks (Table 3). Alongside rocks and in regions without rocks, the mean root depth was 9 to 10 cm compared with 15 cm under rocks. The length and especially the frequency of lateral roots were greater near rocks than in rock-free regions. Consequently, the total length of lateral roots per unit length of main root was 1.25 m m-1 under rocks, 0.34 alongside rocks, and only 0.04 in soil regions without rocks (Table 3).

Ferocactus acanthodes. For Ferocactus acanthodes, 37% of the length of relatively long main roots occurred under rocks (Table 2). Although the mean root depth was similar with and without rocks (9-10 cm), about 3-fold as many lateral roots occurred per unit length of main root under rocks compared with regions without rocks, TaMe 2. Root properties ofFerocactus acanthodes whose roots were associated with rocks. Data are means i SE for six main roots from different plants averaging 41 cm in height. Rocks were all below the soil surface and averaged 18.6 :t: 1.1 (mean 4- SE for n = 22) cm in the horizontal direction along the roots. Less than 1% of the root length was above or alongside rocks Quantity (units)

No rock

U n d e r rock

Length of main root (cm) Depth below soil surface (cm)

123 9.3

72 10.2


4- 13 :t: 0.4

Responses of transplanted Agave deserti For the five locations of the transplanted Agave deserti in the field, the wettest location at a depth of 10 cm in the soil was on the north side of exposed boulders (Table 4). Soil at the south side of exposed boulders had a ~oii similar to that under subterranean rocks, whereas the control without rocks was drier and that above subterranean rocks was the driest. For the four observation intervals of 1-2 months duration each, the number of new main roots emanating from the base of the stem was 13+-2 (mean+ SE) for the five plants on the north side of rocks, 8 4-2 on the south side, 11 +- 2 for plants with

4- 10 4- 0.7

23 4- 3 63 4- 9 0.76 4- 0.15 4.16 4- 0.56

Table 3. Root properties of Opuntia acanthocatTa whose roots were associated with rocks. Data are means-+ SE for 14 main roots from four plants averaging 36 cm in height. Excluding one large unex-

cavated boulder ( > 6 0 cm across), rocks were all below the soil surface and averaged 14.3-+2.9 (mean-+SE for n = l l ) cm in the horizontal direction along the roots

Quantity (units)

No rock

Alongside rock

Under rock

Length of main root (cm) Depth beIow soil surface (cm)

21.5 9.2

19.2 10.4

7.8 14.6


+_3.5 (14) ___1.0 (14)

3 + 1 (14) 0.04 _+0.02 (14)

Table 4. Soil water potential and new main roots of transplanted A. deserti in the field. Five plants were excavated and then replanted on 21-22 September 1991 with their roots in each of the five indicated locations with respect to boulders or subterranean rocks. The plants were watered over a l-m 2 area centered on the plants

+_2.9 (10) + 1.2 (10)

21 -+2 (10) 0.34 _+0.09 (10)

+1.7 (4) -+ 1.7 (4)

54 _+6 (4) 1.25 _+0.21 (4)

with a 30 m m depth of water initially and after three subsequent excavations (all except the last). Data for soil water potential (XCsoi~ are means • SE (n = 5); on 3 January and 29 February 1992, the soil was wet at all locations (g~oi~> - 0 . 2 MPa). Data for the new main roots are the total for the five plants under each treatment

Excavation date

Quantity (units)

Location of plants North South side of side of boulder boulder

Control

Half of roots over rock

Half of roots under rock

26 Oct 91

~soil (MPa) New roots

-1.5+0.4 17

--2.7_+0.7 5