S.Afr.J. Bot., 1990, 56(3) 409 Short Communications moisture (riparian zone). Riparian zones in are generally left unafforested with commercial species because of the perceived detrimental effect of affores­ tation on water yield (Malherbe 1968). These zones do, however often contain dense indigenous vegetation, and are frequently invaded by alien . Results from Comparison of the water relations several studies have shown that phreatophytes (plants of capensis and Pittosporum with constant root access to freely available water) use unduJatum in a riparian woodland in the considerable quantities of water in comparison with plants growing outside these moist zones (Rycroft 1955; south-western Cape Banks 1961; Wicht 1971; Nanni 1972; Patric 1980). The question arises as to whether indigenous riparian vegeta­ R.E. Smith tion uses more water than exotic weeds or trees planted South African Forestry Research Institute, Jonkershoek in these areas for commercial reasons. Forestry Research Centre, Private Bag X5011, Stellenbosch, Weed eradication is expensive but can be justified if 7600 Republic of South Africa the weed seriously disrupts natural plant communities (e.g. reducing indigenous species diversity (Richardson Accepted 19 March 1990 et al. 1989) or uses more water than the indigenous An attempt is made to eradicate exotic invasive species from vegetation. Knowledge of the water use of different riparian zones, partly on the assumption that these exotics species in riparian zones is therefore necessary for use more water than indigenous vegetation. This assumption decision making on riparian zone management. was tested by estimating the water use of an indigenous Pittosporum undulatum (sweet pittosporum), which is riparian forest species, Cunonia capensis L., and an invasive native to Australia, has recently invaded forest patches woody species, Pittosporum undulatum Vent. within a riparian zone. Stomatal conductances and xylem pressure potentials of the south-western Cape including the 10nkershoek were determined using a null-balance porometer and a valley near Stellenbosch (Richardson & Brink 1985). It Scholander pressure chamber respectively, and transpiration is a tree similar in growth and form to many of the rates were estimated. Differences in stomatal conductance indigenous forest species. The two main vegetation types and transpiration between species were minimal, with C. invaded by this species at 10nkershoek are tall shrub and capensis having slightly higher values for most of the year. Neither species experienced any water stress in a mediterra­ indigenous forest patches, typically occurring in nean context, and stomatal conductances were moderate to mountain ravines and along the Eerste river and its low throughout the year. Average water loss per unit leaf area tributaries (Richardson & Brink 1985). The rapid during the study period was c. 16 mmol m-2 d-1 for both P. dispersal , fast growth rate and early production are undulatum and C. capensis, illustrating that the two species factors contributing to its invasiveness, while its are both similar and conservative in terms of water demand, despite an unlimited water supply. adaptable root system enables success on a wide variety of soil types (Gleadow & Ashton 1981). Cunonia 'n Poging word aangewend om oewersones skoon te hou van capensis was chosen as a prominent representative uitheemse indringerplantsoorte, omdat daar onder meer species of the indigenous riparian forests of the south­ vermoed word dat hierdie uitheemse plantegroei meer water western Cape. Both trees are broadleaf evergreens, and gebruik as inheemse plantegroei. Hierdie vermoede is getoets deurdat 'n bepaling van die waterverbruik binne 'n in closed canopy situations have approximately equiva­ oewersone van 'n inheemse boomspesie gemaak is, Cunonia lent leaf areas. capensis L., en die van 'n uitheemse houtagtige oewerspesie, In testing the hypothesis that exotic invasives use Pittosporum undulatum Vent. Blaarkonduktansie- en xileem­ water more extravagantly than indigenous plants in areas drukpotensiaallesings is geneem met onderskeidelik 'n with unlimited water supply, xylem pressure potentials nulbalans-porometer en 'n Scholander-drukkamer, en die tempo van transpirasie is bepaal. Verskille in blaarkonduk­ (as an indicator of water stress) and leaf conductances tansie en transpirasie tussen die spesies was minimaal, were measured and transpiration rates were calculated alhoewel C. capensis effense hOEk lesings vir die grootste for both species to determine the seasonal and diurnal gedeelte van die jaar gehad het. Nie een van die spesies het water use patterns. enigsins onder waterspanning (in 'n mediterreense konteks) Pittosporum undulatum and C. capensis are found verkeer nie, en blaarkonduktansie was gematig tot laag growing together in Vergetenkloof in the 10nkershoek dwarsdeur die jaar. Gemiddelde waterverlies per eenheid blaaroppervlakte vir die studietydperk was ongeveer 16 Valley (33°57'S, 18°55'E) near Stellenbosch in the mmol m-2 d- 1 vir albei spesies. Dit wys dat hulle, ten spyte van Cape Province. Climate is mediterranean, with dry 'n onbeperkte watervoorraad, eenders sowel as konserwatief summers and wet winters. The area has a mean annual is in terme van hul watervereiste. temperature of 16.1°C and receives a mean annual precipitation of 1 530 mm yr- 1. The vegetation on the Keywords: Alien plants, Cunonia capensis, Pittosporum study site comprises a riparian forest community, similar undulatum, riparian zone, water relations to the Rapanea melanophloeos - Cunonia capensis high forest community in the 10nkershoek Valley (McDonald This study assesses the water use of an indigenous 1988). riparian forest tree, Cunonia capensis L. , and an invasive Comparison between species was limited to one pair alien tree, Pittosporum undulatum Vent. of equivalent of trees as the weed eradication program had so reduced structure, under conditions of continuously available soil the numbers of P. undulatum in the valley that only one 410 S.-Afr.Tydskr. Plantk. , 1990 , 56(3) suitable pair of trees could be found . Both trees had browsing eland, reducing the leaf area by a third. A their roots in a small perennial stream and were approx­ substitute tree was found 300 m from the original site. imately 7 m tall. Stomatal conductance and xylem This tree was not strictly riparian being 15 m from the pressure potential were measured at monthly intervals nearest stream and partly shaded by a pine plantation. for both species from October 1986 to October 1987. Temperature and relative humidity conditions between Conductances were measured with a MSC 301 null­ the trees on subsequent measurement runs were virtually balance diffusion porometer (Beardsell et al. 1972). identical. Measurements on the second P. undulatum Replicates of leaves within the two sample trees were commenced in April 1987 at the start of the wet season. selected from similar positions with regard to canopy Examples of diurnal conductance patterns for summer layer, height and aspect, with two readings on each of and winter are given in Figure 1 and seasonal mean and three leaves per species being taken at 1- to 11/2-h peak stomatal conductance rates are shown in Figure 2. intervals through the day from predawn until dusk. Cunonia capensis had on average, slightly higher mean Transpiration rate per unit leaf area was then approx­ and peak daily stomatal conductances throughout the imated from the following formula (Licor Products year than P. undulatum, except during the dry summer Manual 1981): months (December to February). Mean transpiration Tr = FILA x (p c-pa) (g cm-2 S- I) rates were very similar for the two species throughout where F (cm-3 S-I) is the flow rate within the parometer, the year, but C. capensis generally had higher peak rates LA is the leaf area (cm-2) , pc (g cm-3) is the water vapour throughout the year with the exception of the dry density in the cuvette and pa (g cm-3) is the water vapour summer months. density of the dry air stream entering the cuvette (2% The seasonal variation in pre-dawn and midday xylem relative humidity). Xylem pressure potential, as an pressure potentials for both species is illustrated in indicator of plant water stress, was also measured at 1- to Figure 3, and an example of diurnal patterns for winter 11/2-h intervals from predawn until dusk using a Schol­ and summer is given in Figure 1. Diurnal (hourly) xylem ander pressure Chamber (Ritchie & Hinckley 1975). pressure potentials for P. undulatum were slightly lower Three leaves per species were measured from the same than those of C. capensis throughout the year (Figure 1) . canopy layer, height, and aspect as the leaf conductance Pre-dawn potentials did not change seasonally for either samples. species with values remaining above -0.6 MPa through­ In April 1987, the P. undulatum tree was damaged by out the year (Figure 1) . Midday poterltials were c. 1 MPa

a) b) 0 ..•....•...... •....:. ...• -D,5 ...... ~...... ,.::::.:: rn ...... ~ ...... :::.v· Q) ...... 0 ...... Q. -1,0 .... ~ro !i ::J[L ~~ -1,5 .' (1'- ~ , Q...... E Q) -2,0 ;;, x -2,5 -.. :l

-3,0 150

Q) 125 u ffi;:- o I (/) ::I N 100 "0 , § E u_ 75 2~ E_roE 0 50 U5

25

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Time (h) Figure 1 Diurnal patterns of leaf stomatal conductances (solid line) and xylem pressure potentials (stippled line) for Cunonia capensis (dots) and Pittosporum undulatum (squares) during (a) summer (February 1987) and (b) winter (June 1987). S.AfrJ. Bot., 1990, 56(3) 411

300 a) Cii 250 E 0 OJ 200 0 -D,5 0. 150 e! ro -1,0 OJ :::J(L U ~::;; -1 ,5 c 100 OJ- '" .0- gloo 50 0. -2,0 TIN E c l 0 OJ -2,5 E >, 8 x -3,0 Cii (5 300 b) - E E.s 250 o N D J F M A M J J A SON '86 '87 a 200 150 Date 100 Figure 3 The seasonal variation in the predawn (solid line) 50 and midday (stippled line) xylem pressurc potcntials for Pitto­ o ;--'~'--'r-'---r-~--r--r--r--r--~~~ sporum undulatum (squares) and Cunonia capensis (dots) for ONDJ FMAMJ JASON the period October 1986 to October 1987. The move to a new '86 '87 Pittosporum undulatum is indicatcd by an arrow on the x-axis. Date Figure 2 Seasonal patterns of mean (stippled Iinc) and pcak (solid line) daily leaf conductance rates for (a) Cunonia 1987; von Willert et al. 1989). The lack of any notable capensis and (b) Pittosporum undulatum for thc period water stress can probably be ascribed to the riparian October 1986 to October 1987. The move to a new Pitto­ situation and free availability of water. Miller et al. sporum undulatum is indicated by an arrow on the x-axis. (1983) suggested that low potentials are either the result of adequate water supply to roots or low transpiration rates. Both of these possibilities may be valid explan­ lower than the pre-dawn potentials for both species, ations for the low potentials in this study. except during the dry months of February and March There is no evidence of excessive consumption of where P. undulatum midday potentials dropped to - 2.7 water by either species despite free access to soil MPa (Figure 1). moisture. Pittosporum undulatum appears to be similar The hypothesis that invasive exotics use more water to C. capensis in terms of water use. Tn this case than indigenous species does not appear to be valid in therefore, the invasion of a riparian zone by P. undula­ this instance. Pittosporum undulatum and C. capensis tum, an exotic woody species, will affect water yield no had similar leaf conductances and transpiration rates, more than the development of an indigenous forest with mean rates of conductance per unit leaf area for C. species (c. capensis) resulting from total protection of capensis being if anything, slightly higher than for P. the zone. Although an invasion by P. undulatum may be undulatum. The lower xylem potentials for P. undula­ hydrologically 'safe' , the ecological implications of tum in summer possibly result from slightly higher root invasion by P. undulatum must be considered before a resistances to water uptake. This may cause moderate final management decision is made. The similarity of stress levels in response to their comparatively higher hydrological behaviour observed between an exotic transpiration rates during summer. Miller et al. (1983) invader and an indigenous forest species is perhaps not consider leaf conductance values of 130 to 350 mmol m- 2 surprising, given similarities in growth and structure. S- I as moderate (at least within mediterranean regions). Nevertheless, it is apparent that generalized assumptions In these terms, the conductances of both P. undulatum with regard to water use require verification. Given the and C. capensis are generally low, with the maximum limitations of this study (lack of replication, only two conductance for C. capensis being 200 mmol m- 2 S- I (P. species), further research in comparative physiology is undulatum is lower). The average water loss per unit leaf therefore needed to extend these results to different area (over a year) was estimated at 16.04 mmol m- 2 d- I habitats and species. for C. capensis and 15.88 mmol m- 2 d- I for P. undulatum, which is much less than, for example, the 190 mmol m- 2 Acknowledgements d-I per unit leaf area determined for Eucalyptus grandis, This work forms part of the conservation forestry a commercial forest species (Dye 1987) , research programme of the Forestry Branch of the Neither of the species experienced water stress in a Department of Environment Affairs. I thank W.J. mediterranean-climate context (cf. Blake-Jacobsen Anthony for assistance with field work, and my 1987; Roberts et al. 1981; Roberts 1982 where xylem colleagues at Jonkershoek Forestry Research Centre for potentials for selected chaparral shrub species in summer useful comments on the manuscript. dropped to -7 MPa). Xylem potentials of C. capensis compare well with those measured by Kruger et al. References (1988) on the same species and others in a riparian zone BANKS, C.H. 1961. The hydrological effects of riparian and in Swartboskloof, nearby in the Jonkershoek Valley. adjoining vegetation. For. S. Afr. 1: 31 - 45. The potentials for both species generally fall well within BEARDSELL, M.F. , JARVIS, P.G. & DAVIDSON, B. the range of other tall (unstressed) fynbos shrubs and 1972. A null-balance diffusion porometer suitable for use trees (Miller et al. 1983; Miller et al. 1984; Jeffrey et al. with leaves of many shapes. 1. Appl. Ecol. 9: 677-690. 412 S.-Afr.Tydskr. Plantk. , 1990, 56(3)

BLAKE-JACOBSEN, M.E. 1987. Stomatal conductance and VON WILLERT, D.J. , HERPPICH, M. & MILLER, 1.M. water relations of shrubs growing at the chaparral-desert 1989. Photosynthetic characteristics and leaf water relations ecotone in California and Arizona. In: Plant responses to of mountain fynbos vegetation in the Cedarberg area stress, eds Tenhunen, 1.D., Catorino, F.M. , Lange, O.L. & (South Africa). S. Afr. 1. Bot. 55 : 288- 289. Oechel, W.e. , pp. 575- 596 , Springer, Berlin. WICHT, e.L. 1971. The management of mountain catchments DYE, P .J. 1987. Estimating water use by Eucalyptus grandis by Forestry. S. Afr. For. J. 77 : 6-12. with the Penman- Monteith equation. Forest Hydrology and Watershed Management. Proceedings of the Vancouver Symposium, August 1987, IAHS Pub!. no. 167, 1987. GLEADOW, R.M. & ASHTON, D .H. 1981. Invasion by Pittosporum undulatum of the forests of Central Victoria. 1. Invasion patterns and plant morphology. Aust. 1. Bot. 29: 705- 720. JEFFREY, D., MOLL, E.l. & VAN DER HEYDEN, F. 1987. Comparative water potentials of 4-month old and 4- year old Cape fynbos plants. S. Afr. J. Bot. 53: 32- 34. KRUGER, F.J. , RICHARDSON, D .M. & SMITH, R.E. New records of four synanthropic plants 1988. Plant water relations in sclerophyllous trees and found in shrubs of riparian and hillslope habitats at Swartboskloof, south-western Cape Province , South Africa. In: Time scales and water stress. Proc. 5th international conference on A. Danin mediterranean ecosystems, eds Di Castri, F., Floret, Ch. , Department of Botany, The Hebrew University of Jerusalem, Rambal , S. & Roy, J. , pp. 575- 581 , International Union of Jerusalem, Israel 91904 Biological Sciences, Paris. LI-COR Inc. 1981. Li-cor instrumentation for biological and Accepted 8 March 1990 environmental sciences. Li-cor Inc. Ltd. , Nebraska (USA). MALHERBE, H.L. 1968. Afforestation and water supplies in Four adventive synanthropic plant species of Holarctic origin South Africa. Department of Forestry Report of the are recorded for the first time from southern Africa. These are Interdepartmental Committee of Investigation into Conyza albida Sprengel, Stellaria pal/ida (Dumort.) Pire, Sagina apetala Ard. and Sagina maritima G. Don fil. They Afforestation and Water Supplies in South Africa, p. 115. occupy human disturbed habitats such as road sides and MC DONALD, D.J. 1988. A synopsis of the plant cracks in pavements in urban areas. communities of Swartboschkloof, Jonkershoek, Cape Province. Bothalia 18: 233- 260. Vier adventiewe sinantropiese plantspesies van holarktiese MILLER, 1.M., MILLER, P.e. & MILLER, P.M. 1984. Leaf oorsprong is vir die eerste keer in Suidelike Afrika aange­ conductances and xylem pressure potentials in fynbos plant teken. Die plante [Conyza albida Sprengel, Stel/aria pal/ida species. S. Afr. J. Sci. 80: 381 - 385. (Dumort.) Pire, Sagina ape tala Ard. en Sagina maritima G. MILLER, P.e., MILLER, 1.M. & MILLER, P.M. 1983. Don fil.] kom in mens-versteurde habitatte langs paaie en in krake in plaveisel in stedelike gebiede voor. Seasonal progression of plant water relations in fynbos in the western Cape province, South Africa. Oecologia 56: Keywords: Adventive, Asteraceae, Caryophyllaceae, 392- 396. southern Africa, synanthropic NANNI, U.W. 1972. Water use by riparian vegetation at Cathedral Peak. S. Afr. For. J. 80: 1- 10. The rich indigenous flora of sothern Africa draws the PATRIC, J.H. 1980. Effects of wood products harvest on attention of most botanists. However, foreign plants that forest soil and water relations. 1. Env. Qual. 9: 73- 80. invade synanthropic habitats are sometimes overlooked. RICHARDSON, D.M. & BRINK, M.P. 1985. Notes on The following is a list of a few such adventive species, Pittosporum undulatum in the south-western Cape. Veld not mentioned by Gibbs-Russell et at. (1987) , that I and Flora 71: 75- 77. found during a visit to South Africa in July-August RICHARDSON, D.M., MACDONALD, LA. & FORSYTH, G.G. 1989. Reductions in plant species 1989. richness under stands of alien trees and shrubs in the fynbos Conyza aibida Sprengel. Three adventive species of biome. S. Afr. For. J. 149: 1- 8. Conyza, found throughout many European and Mediter­ RITCHIE, G.A. & HINCKLEY, T.M. 1975. The pressure ranean countries are C. bonariensis (L.) Cronquist, C. chamber as an instrument for ecological research. Adv. canadensis (L.) Cronquist, and C. aibida Sprengel Ecol. Res. 9: 165-254. (Guedes & lovet 1975 ; Danin 1976a, b, 1981a, b; Danin ROBERTS, S.W. 1982. Some recent aspects and problems of et ai. 1982). Of these C. aibida is not recorded by Gibbs Chaparral plant water relations. United States Department Russell et ai. (1987). This species has a set of morpholog­ of Agriculture. General Technical Report PSW-58: ical characters somewhat intermediate between those of 351- 357. C. bonariensis and C. canadensis (Table 1). In the ROBERTS, S.W. , MILLER, P.e. & VALAMANESH, A. 1981 . Comparative field water relations of four co­ following South African herbaria (BLFU, BOL, RUH, occurring Chaparral shrub species. Oecologia 48: 360-363. PEU), specimens of this species were mislabelled as C. RYCROFT, H.B. 1955. The effect of riparian vegetation on bonariensis or C. canadensis. It is a very common species water-loss from an irrigation furrow at lonkershoek. Jl S. that grows in vast areas along roadsides in Transvaal , Afr. For. Assoc. 26: 80-88. Orange Free State, eastern Cape, southern and western