SOME ASPECTS OP THE WJffiER RELATIONS OP W O MAHOGANY SPECIES A thesis submitted by Gladys Dodoo, B.Sc, (Hons) in part fulfilment of the requirements for the M.Sc. Degree of the University of Ghana. Deco?Wê 5̂71 From; Botany Department University of Ghana . University of Ghana http://ugspace.ug.edu.gh G183168 University of Ghana http://ugspace.ug.edu.gh I hereby declare that the work presented in this thesis •Some Aspects of the Water Relations of Two Mahogany Species* was done entirely by me in the Department of Botany, University of Ghana, Legon, from September 1970 to November 1971. No part or parts of this thesis has been submitted for a degree elsewhere. o o (GLADYS D0D00) University of Ghana http://ugspace.ug.edu.gh ABSTRACT The general distribution, growth form and economic importance of Kha%a senega!sir (Bear) A. Juss. and Khaya ivorensis A. Chev. are described. Some aspects of the water relations of seedlings of these two species, the former a savanna species and the latter a forest species were studied with the view of ascertaining whether moisture plays an important role in determining the pattern of their distribution. &rowth of seedlings under four soil watering regimes namely - -0.3 (A), -0,4 (B), -0.8 (C), and -4.5 (D) bars; and under culture solution and culture solution to which polyethylene glycol was added to give the following osmotic potentials (bars): -0.3 (A), -2.8 (B), -5.3 (C), aid -10.3 (D) was studied. The experiment was done in the g r e e n h o u s e . G-rowth of K. S e n e g a lensis was more sensitive to moderate moisture stress and less sensitive to high moisture stress. K, ivorensis on the other hand showed less sensitivity to moderate stress but high sen­ sitivity to severe stress. However when soil moisture stress was -0.3 bars, growth of K. ivorensis was very poor. This was attributed to a reduction in root permeability due to poor aeration as a result of more permanent near saturation of the soil. Studies of diurnal patterns of plant water status were carried out by examining leaf relative water oontent, (R.W.C.) leaf water potential (L.W.P.) and stem diameter variation, in relation to soil moisture stress. R.W.C. was overall higher in seedlings of K. senegalensis than in those of K. ivorensis. L.W.P. on the other hand was lower for IC. senegalensis than for K. ivorensis seedlings. Stem shrinkage decreased with decrease of soil moisture content from 100 to about 50%, field capacity. The decrease in K. senegalensis was greater than in K. ivorensis-. However at 27% field capacity, shrinkage in K. senegalensis was consistently reduced more than in K. ivorensis. This may indicate better water conservation by the former species. Transpiration was also studied in relation to the soil moisture treatments, both in the greenhouse aid in the research room, the latter being a semi-controlled environment where temperature, relative humidity and light intensity were precisely known. The transpiration of seedlings growing in osmotic solution was also studied in the greenhouse, employing stresses of -0.3 and -10.3 bars. Transpiration generally decreased with moisture stress in the root medium. In the research room transpiration of K. senegalensis was higher than that of IC. ivorensis under all soil treatments. In the greenhouse, however, similar higher transpiration rates were recorded for K. senegalensis seedlings than for K. ivorensis seedlings when stress was from -0.3 to -0.8 bars, but at severer stresses (-4.5 snd -10.3 bars) transpiration of K. senegalensis was reduced more than that of K. ivorensis. University of Ghana http://ugspace.ug.edu.gh Infiltration of leaves of K. senegalensis and K. ivorensis with mixtures of liquid paraffin and commercial Kerosene showed that stomatal conductivity of K. senesalensis leaves was greater then that of K. ivorensis leaves at low moisture stress. At more severe stress (-4.5 bars) conductivity of leaves of both species was low. The desorption curve for K. senegalensis seedlings was above that for K. ivorensis seedlings. Curves of leaves for adult trees of both species were however similar. K. senegalensis leaves oould tolerate desiccation better than K. ivorensis leaves. Stomata of K. senega!ensis leaves close at higher relative water content values than those of K. ivorensis leaves. The results are discussed in relation to the distribution of these species and to the general ecological problem of the control of plant dis­ tribution between forest and savanna in West Africa. University of Ghana http://ugspace.ug.edu.gh CONTENTS Page ABSTRACT •*■ INTRODUCTION ... ... ... 1 CHAPTER I j Experimental species, materials, and general methods ... ... 11 CHAPTER II s Growth in relation to moisture stress in the root medium ... ... 25 CHAPTER IH: The diurnal pattern of plant water status 74 CHAPTER IV : Transpiration in relation to moisture stress in the root medium ... 104 CHAPTER V : Tissue water relations ... 139 CHAPTER VI j General Discussion ..« ... 165 SUMMARY ' ... ... ... 174 APPENDIX ... ... ... 181 ACKNOWLEDGEMENT ... ... 217 REFERENCES ... ... ... 218 University of Ghana http://ugspace.ug.edu.gh INTRODUCTION General In general, the natural pattern of distribution of plants in any area is largely controlled by environmental factors. The main physical factors of the environment - temperature, light, moisture, wind and soil, frequently set limits to the occurrence of species, both on a geographical and on an ecological scale (Raunkiaer, 1934; Daubenmire, 1947; Hopkins, 1965)* The importance of any of these factors in a given area is detemined both by its general mean level and by the range between its minimum and maximum levels. Temperature is the primary factor responsible for the main vegetational belts of the world. In many geographical regions it is also the overriding factor controlling the pattern of plant m distribution on an ecological scale. In West Africa, however, where daily and seasonal variations iai temperature are usually small (see Lawson, 1966), this factor is less important in controlling plant distribution. There is rather a general belief that the main division of West African vegetation into forest and savanna is causally related to environmental moisture conditions (see for example Taylor, 1952; Keay, 1959). Hopkins (1965) and Lawson (1966) in their general accounts of vTest African ecology point out that in this region rainfall University of Ghana http://ugspace.ug.edu.gh 2 varies greatly between areas, in amount and seasonal distribution* Over 80 ins. (200 cm) of rain falls annually in most parts of the forest areas. This anount decreases progressively inland from the coast until within the savanna areas there is usually less then 45 in. (114 cm) of rain per annum. Rainfall in the forest regions is distributed in such a way that there are not more than two months with less than one inch (2.5 cm) of rain. On the other hand the savanna regions may have up to five or six months with less than 1 inch (2,5 cm) of rain. The balance between precipitation and potential evaporation is perhaps a more meaningful criterion than amount er distribution of rain for assessing water availability for plant growth, From a simple survey of this balance for Africa, Davies and Robinson (1969) reached the tentative conclusion that the forest-savanna boundary, at least in West Africa, coincides reasonably well with the -200 mm isopleth, where although there is no excess of precipita­ tion over potential evaporation for iiie year as a whole, there is some amount of water stored in the soil which could support growth of forest species. Thus it is clear that savanna vegetation type is associated with a dzy climate, hence the general belief that savanna plants are more drought-adapted than plants of the forest regions. Comparative experimental work to test this belief is however largely lacking so that there is little knowledge of the University of Ghana http://ugspace.ug.edu.gh 3 mechanisms by which plants from these two regions are adapted to environmental moisture conditions. Mechanisms by which plants adapt themselves to drought have been much studied and reviewed (see for example Parker, 1956, 1968; Stocker, 1956, I960; Iljin, 1957; Oppenheimer, i960, 1968; and Kozlowski, 1968). Oppenheimer (i960) points out that these mechanisms generally fall under three main heads; plants may be drought escapers, drought evaders or drought endurers. Drought escapers apparently have no adaptation to withstand drought. They normally complete their life cycle before the unfavourable period. Unlike the escapers, drought evaders survive under drought by preventing the development of internal water stress under a given degree of external stress. This they do by efficient control of transpiration and efficient water absorption. As has been pointed out by several workers (for example Sullivan and Levitt, 1959; Oppenheimer, I960; and Parker, I968), numerous adaptive morphological characters are associated with the phenomenon of drought evasion. In contrast to drought escapers or evaders, drought endurers are able to cany on their normal life activities under drought even when there is internal stress. Most studies of plant adaptation to drought have been concerned with identifying which of these mechanisms is paramount in any given situation, and with understanding details of plant attributes which make drought resistance possible. Such studies University of Ghana http://ugspace.ug.edu.gh gain in value when they contribute towards understanding the causal factors of the pattern of plant distribution in any area. To achieve such an understanding is the general objective of the present study. The Problem The general paucity of experimental investigations on the water relations of West African plants has already been mentioned. The few studies conducted in this field (for example Lawson and Jenik, 1967; Okali, 1971a), were mostly concerned mth mechanisms of adaptation to environmental stress by seme species on the Accra Plains of Ghana. Although vegetation on the Accra Plains is dissimilar to Guinea savanna in many respects (Wills, 1962; Hall and Jenik, 1968), nevertheless the Accra Plains Lawson (I966)^indicates that ^ ishow'. many interesting ecological factors at work which may throw light on processes in other types of savanna. The vegetation here consists of short grass and tree- thickets which often occur in scattered clumps (Wills, 1962; Aubreville, 1959} Boughey/"1957)j described the Accra Plains as a kind of steppe. As in true savanna regions, the climate over the plains is relatively dry. Less than 30 in. (76 cm) of rain falls annually, (Lawson, 1966; Lawson and ‘•enik, 1967). Thus studies of mechanisms of plant adaptation to Uiis habitat may be useful in understanding the general nature of the control of plant life by University of Ghana http://ugspace.ug.edu.gh 5 environmental moisture in West Africa. For example, Lawson and Janik (1967) in their study of the interrelations of microclimate and vegetation on the Accra Plains, compared transpiration rates (by the rapid weighing technique) and other features of species on the wind-ward and leeward sides of a thicket. In general they found that species on the windward side transpired less and had more xeromorphic features than those on the leeward side, thus demonstrating tiiat microclimate (and here particularly desiccating winds) may exert considerable influence on vegetation. Okali. (1971a) further studied the water relations of seme of the woody species on the plains. Using mainly detached leaves or leaf discs taken from plants growing in the field, he compared transpiration rates and the relationships between leaf water content, leaf water potential, stomatal closure and tissue damage. His results showed that the species studied, exhibited several mechanians of adaptation to drought. Neither of the above studies examined the response of whole plants (e.g. growth) to drought, and although they demonstrate the possession of adaptive features by species in a diy area, they do not provide an adequate basis for testing the role of moisture in deteimining the pattern of plant distribution between forest and savanna, because they examined species adapted to one "type of climate alone. A more satisfactory approach for such a test would be to compare the water relations of forest and savanna species. University of Ghana http://ugspace.ug.edu.gh 6 There is little information in the literature on the water relations of West African forest species. The only suitable data which permit such a comparison to be made were recently given by Hopkins (l970a,b). Prom a study of species in forest and savanna sites of the Olokemeji Forest Reserve in Nigeria, Hopkins was able to show that for forest species leaf water may be 'severely limiting at the severest part of the dry season*; he could not demonstrate similar limita­ tion for the savanna species although he attributed this to probable effects of leaf age in masking iiie relation between environmental moisture and leaf water status. Hopkins' study examined, only one aspect of the water relations of forest and savanna jpecies and this limits its usefulness as a basis for generalization on the probable role of moisture in controlling plant distribution in this part of the world. Richards (1952) draws attention to the existence in West Africa of certain tree genera whose component species are restricted to sharply contrasting habitats. He cites the genera Khava and Lophira as examples, noting that the foimer genus is represented by K. ivorensis and K. anthotheca in the Wet Evergreen forest, by randif o 1 iola in the Dry Forest and by K. Senegalensis in the savanna. Lophira is similarly represented by L. alata in the forest and L. lanceolata in savanna. To these examples may be added such genera as Afronaosia. Paniellia and Cussonia which have both forest and savanna species. If environmental moisture is indeed the University of Ghana http://ugspace.ug.edu.gh overriding factor causing the restriction of these closely related species to contrasting habitats, it should be possible to demonstrate this by comparing the water relations of any two such species of the same genus - the one from forest and the other from savanna. The object of the study described in this thesis was to examine experimentally the water relations of two such species, Khava senegalensis (Des*) A. Juss, and K. ivorensis A. Chev, commonly called Dry Zone Mahogany and African Mahogany respectively, to find possible differences 'between them which might help in understanding the factors responsible for their restriction to contrasting habitats. The two species are not only taxonomically closely related, they are also as is more fully described at a later stage, similar in growth habit, phenology and reproductive biology. Differences between them could therefore be expected to have arisen more as an adaptation to their respective habitats than would be the case if the two species were dissimilar in the above four respects. If environmental moisture plays a large part in determining the restriction of these two species to different habitats, K. senegalensis would be.expected to show greater adaptation to drought than &. ivorensis. Approach to the problem Several investigators of plant water relations have approached such studies with different objectives. Some investigators have University of Ghana http://ugspace.ug.edu.gh 8 studied single species (for example Weatherley, 1950, 1951, 1965; Gates, 1955a,b; Hutter and Sands, 1958; Aspinall, 1965; Klepper, 1969); others have compared the responses of a group of species to similar environmental conditions (for example Slatyer, 1955, 1957’Dj I960; McKell, Perrier and Stebbins, I960; M.S. Jarvis, 1963; Jarvis and Jarvis, 1963a,b,c and e, 1965; Connor and Tunstall, 1968). The objective when single species have been studied has usually been to understand the processes through which environmental moisture affects plant growth. When several species have been studied together, the implied aim has been to understand the comparative effects of environmental moisture with the view 60 explaining ecological situations. The latter is the case in the present study. The value of comparative water relations studies in contributing to an understanding of the causal factors of plant distribution is illustrated by the work of many authors. Thus McKell, Perrier and sub . ^lomerata Stebbins (i960) compared two^species of Daqtvlis and concluded that the sub species the restriction of Iasi f.ani r.a to mesic habitats and i>. Judaic a to xeric habitats in Israel might be due to differences in the water demand of the two species. Similarly, investigations by M.S. Jarvis (1963) on factors which limit the distribution of Saxifrage liypnoifles and Prunus padus to upland areas of North western'parts of Great Britain and Filipendula vulgaris and Thelycrania sanguinea to more low land parts to the University of Ghana http://ugspace.ug.edu.gh 9 South east showed that this could be attributed to differences in soil moisture in these areas. Sullivan and Levitt (1959) found from their work that a possible explanation for the restriction of Querous palustris to moist areas and Q. rubra to upland areas in Missouri might be the difference in drought resistance of the two species. Kychnovska and Kvet (1963) accounted for the distribution of Festuoa domini. Corynephorus canesoens and Helichrysum arenarium in Czechoslovakia on the basis of their water relations. Bannister (1964) also related differences in distribution of Calluna vulgaris. Erica cinerea and E. tetralix to the water relations of these species. The distribution of various species of Artemisia in arid and ssmi arid communities in Kazakhstan has also been related to differences in the internal water balance of these species (Sveshnikova, 1965). Irrespective of the objective investigators have sought to understand plant water relations by observing growth and transpiration responses of the species to environmental moisture stress (for example Gates, 1955&; Rutter and Sands, 1958; M.S. Jarvis, 1963; Jarvis and Jarvis, 1963a,b,c), by studying the water relations of tissues through, for example, measurements of the relation between water content and water potential of tissue (Slatyer, i960, 1962c; Connor and Tunstall, 1968), the relation between water content and stomatal closure (Bannister, 1964, Jarvis and Jarvis, 1963d,e), and the relation between tissue water content and tissue damage University of Ghana http://ugspace.ug.edu.gh (desiccation tolerance) (Sullivan and Levitt, 1959; Noy-JSeir and Ginzburg, 1969b; Okali, 1971a). Some investigators have followed changes in stem diameter as a measure of variation in plant water status (Kozlowski, 1967; Ogigirigi, ICozlowski and Sasaki, 1970). The present study represents an attempt to combine several of these approaches. Following an analysis of the growth responses of the experimental species to varying moisture status in the root medium, internal water balance over most of the day for plants growing in wet or dry soil was compared. In order to understand the growth responses and diurnal patterns of plant water status more clearly, transpiration and seme aspects of tissue water relations of the experimental species were then studied. These experiments were carried out in the laboratory and mostly with potted plants as has been previously done by several workers such as Slatyer (1957b, 196l), Rutter and Sands (1958), McKell, Perrier and Stebbins (i960), Brix (I962), Jarvis and Jarvis (1963a,b,c,d,e, 1965)» and Lawlor (1969). There are clear difficulties in attempting to use laboratory studies in unc.erstending natural distribution of plants, but evidence in the literature, suggests that when these studies are comparative, individual differences revealed could be of ecological significance (cf. McKell, Perrier and Stebbins i960, M.S. Jarvis, 1963). 10University of Ghana http://ugspace.ug.edu.gh 11 CHAPTER I EXPERIMENTAL SPECIES, MATERIALS AND GENERAL METHODS ✓ 1.1. The Experimental Species The species chosen for this study - Khaya Senegalensis (Desr) A. Juss. and Khava ivorensis A. Chev. are both trees belonging to the family Meliaceae. They are both widely distributed throughout West Africa. Irvine (l96l) notes that K. senegalensis extends from Senegambia to Cameroons, Sudan and Uganda while K. ivorensis extends from the Ivory Coast to Gabon. Hutchinson, Dalziel and Keay (1954) however, indicate that K. ivorensis could be found as far south as o -OCabinda, that is between latitudes 4 and 6 south of the equator and o olongitudes 10 and 13 east of the Greenwich Meridian. Throughout their geographical range, the two species occur in distinctly different habitats. Richards' (1952) observation on this has already been mentioned. Hutchinson, Dalziel and Keay (1954) also point out tiiat K. senegalensis is a savanna tree but grows especially by streams. Taylor (i960) describes K. ivorensis as occurring throughout the high forest zone of West Africa. In his book 'Woody Plants of Ghana1 Irvine (1961) indicates that K. sene^alensis is found in savanna and fringing forest, particularly in low lying places besides streams aid K. ivorensis in deciduous and evergreen forests. The contrast in the habitats of the experimental species appears thus to be well docunented. University of Ghana http://ugspace.ug.edu.gh 12 The two species appear to be dissimilar in their soil preferences. K. seneaalensis prefers good alluvial soil but not swampy soils, while K. ivorensis is known to favour moist valley soil and apparently can also survive considerable flooding (Kinloch, 1945). Data to be presented later in the present study suggest that tolerance to flooding is limited in K. ivorensis seedlings. Taylor (i960) further points out that K. ivorensis occurs on heavy or rich alluvial soil with good drainage near water courses and damp areas. The growth form and general morphology of the two species have been adequately described by Lely (1925), Dalziel (1957) and Irvine (1961). K. S e n e g a lensis grows up to 30 m in height and three meters in girth. It has no buttresses. The baric is grey and scaly and the slash is red. Unlike K. senega!ensis. K. ivorensis grows up to 60 m in height and 4.5 m in girth. It is strongly buttressed and has a tall cylindrical bole up to 27 ra above the buttress. The bark is ashy white to brownish black. The slash is crimson. The crowns of adult trees of both species are dense with branches bearing alternate pariplnnate leaves about 25 cm long with 4 - 8 opposite or subopposite, exsiipulate, oblong or oblong elliptic leaflets. Examination of K. senefealensis indicates that each leaflet is about 10 cm long and * shortly acuminate at the apex. Each leaflet has a petiolule which is about one cm long. Each leaflet of K. ivorensis on the other hand is about 14 cm long and 7 cm wide with well University of Ghana http://ugspace.ug.edu.gh defined drip-tips, (see appendix 1, Plate i), smooth, acuminate and usually rounded at the base with a short petiolule about 1.5 cm long. Observations on seedlings in the greenhouse, on herbarium specimens from trees of different ages as well as on trees in the Botanical Garden, Legon, show ,that the well defined drip-tips of K. ivorensis leaflets decrease in length as the tree grows older. In general mature leaves of K. senegalensis are comparatively much smaller, with a thick net of veins, than those of K. ivorensis. This is probably a xeromorphic feature (cf. Lawson and Jenik, 1967)• Detailed examination, of the leaves, made on epidermal prints and on microtome sections revealed that stomata occur mostly on the lower surfaces of leaves in both species; those of K. senegalensis being more numerous (Table l). There are very few stomata (less than l/mm ) on the upper surfaces of the leaves of both species. Metcalfe end Chalk's (1950) statement that stomata occur only on the lower surface in leaves of Khava species probably therefore does not necessarily apply to seedlings. The stomatal frequencies obtained for the two species are comparable to figures quoted by Yanney-Wi 1 son (19.63) for three species on the Accra Plains: Baphia pubescens. Yemonia senegalensis and Pluggea virosa. Yanney-V/ilson considers frequencies more than 502/rm to be an indication of xeromorpby in leaves. On this basis the leaves of both species could be assumed to have some xeromorphic features. Metcalfe and Chalk (1950) note that the palisade mesophyll in Khaya species is more than one layer. University of Ghana http://ugspace.ug.edu.gh Table 1 s e e d l i n g Anatomical features of^leaves of Khava senegalensis and K. ivorensis 14 Leaf anatomical character K, senegalensis K. ivorensis Mean thickness of leaf ( ji) 822 508 Mean thickness of cuticle ( ju) 14 13 Length of stomata ( ju.) 14-20 11-18 Stomatal frequency (mm ) 660 580 No. of palisade mesophyll layers 1 1 Mean height of palisade cell ( p) 272 104 Mean width of palisade cell ( ji) 46 38 Thickness of spongy mesophyll (p) 347 324 Mean diameter of spongy mesophyll cell ( p) 56.5 61.0 Species University of Ghana http://ugspace.ug.edu.gh 15 This has been found to be true of leaves from adult trees of K. senega!easis and K. Ivorensis; they have two or three layers of palisade. However mature leaves of seedlings were found to have only a single layer of palisade (Pig. l) with those of K. senegalensis being much deeper (about 270 microns) than those of K. ivorensis (about 104 microns). In their natural habitats, K. senegalensis and K. ivorensis are never deciduous (Kinloch, 1945; Taylor, i960), Kinlock (1945) however indicates that in the drier areas of the high forest zone where K. ivorensis is seldom found, the tree is deciduous for a short period at the beginning of the dry season. The similarity between the two species extends to tiieir reproductive biology. The flowering season for both ^ecies is July to January. Flowers are arranged in axillary panicles. Fruits of both species are capsules. They ripen and open to release their seeds between February to May. The fruit of K. senegalensis is about 6 cm in diameter and generally four-valved, while that for K. ivorensis is about 9 can in diameter and opens by five valves. Seeds of the two species are brown and winged. Measurements show that the Seeds of K. senegalensis are about 1.8 cm long and 2.4 cm broad and are oblong-elliptic in shape, while that of K. ivorensis are about 2.3 cm long and 3.7 cm broad; they are oblong to triangular in shape. The seed of K. senegalensis is comparatively much thicker with an average dry weight (without testa) of about 16 mg while that of K. ivorensis weighs about 11 mg. Seeds University of Ghana http://ugspace.ug.edu.gh FIG-. 1 CAMERA. LUCIM DRAWINGS OF EXAMPLES OF TRANSVERSE SECTIONS OF LEAVES OF KBAYA SMEGALENSIS (KS) iND K. IVORMSIS (Kl) Seedlings * LEFT (THM L E A V E S 5 7 ^ ® ' (THICK LEAVES)! University of Ghana http://ugspace.ug.edu.gh C U T IC L E UPPER EPIDERM IS P A L IS A D E M ESO PHVLL SPONGY MESOPHYLL LOW ER EPIDERM IS University of Ghana http://ugspace.ug.edu.gh 17 of both species are short lived. Observations in the present study show that gemination, in both species, generally starts eleven days after sowing* by the 21st day almost all the viable seeds geminate. Often K. senegalensis seeds start germinating before K. ivorensis. especially when fresh. Germination is hypogeal for both species. The shoot of K. senegalensis is green and that of K. ivorensis is red and slender. The first two leaves of both species are simple and opposite and are borne about 8 - 10 cm above the soil. The areas of these leaves when about two months old are respectively about 14.0 and 1 1.5 sq. cm for K. senegalensis and K. ivorensis. The next few leaves are also simple but alternate. These are followed by a few unifoliolate leaves. Compound leaves develop later, normally after two months of gemination. The first few sets are trifaliolate,, later anes are imp a rip inna t e with about 4 - 9 leaflets which are opposite. Leaflets of K. ivorensis are much larger than those of K. senegalensis (see Appendix l,Pla£e ). -One year old seedlings of both species have both paripinnate and impari- pinnate leaves. At emergence seedlings of K. senegalensis are much larger than those of K. ivorensis and by the age of two months, the height of the seedlings is about 13.0 cm with 6 - 9 leaves while that of K. ivorensis is about 10.0 cm with about 4 - 6 leaves. The larger initial size is presumably due to the larger seed size of K. senegalensis (see p. 1 5). This might be advantageous for quicker establishment. University of Ghana http://ugspace.ug.edu.gh 18 The two species are not only of ecological interest; they are also economically important species. The first exploitation of Mahogany species for timber for export begaa in about 1891 (Taylor 1952). It has since remained one of the most important species in the timber economy in this country. Economic reports of the Forestry Department indicate that in 1969 (JAnon) the log volume output of K. ivorensis. whose trade name is Dubini, from the forest reserves was about one million cubic feet (28000 cu.m). Over 19 million logs from 34 timber species valued at about 24. million new cedis were exported from January to December, 19^9, and out of this about 1.5 million cu. ft. (£2000 cu.m) valued at about 2.2 million new cedis was K, ivorensis. Further reports in 1970 indicate that Mahogany yield from forest reserves in the high forest zone, was about 1.3 million cu. £t.(36000 cu.m) out of which one million was K. ivorensis. K. senegalensis being of smaller size is of less importance. The first export of the species as timber was made from tiie Gambia over a century ago. Although an excellent timber, it is seldom now exported from Ghana because of its small size and weight. Local uses of these two Mahogany species include their conversion to charcoal for fuel. Annual reports of the Forest Froducfss Research Institute here (1968 and 1969) emphasise that because of the general increase in industrial consumption of wood, it is necessary to establish plantations of the more important timber species of which the mahoganies are certainly leading University of Ghana http://ugspace.ug.edu.gh examples. It is clear that successful establishment of such plantations can be enhanced by the availability of adequate information on the ecology of the timber species that are considered as important. Some of the results from the present study might contribute towards providing such information. In particular, the findings from this study might assist in the choice of combination of management practices if it is desired to extend the habitat ranges of these two species with respect to environmental moisture conditions. 1.2, Experimental Plants Seedlings of K. senegalensis and K. ivorensis were used for -this study except in one instance when certain determinations were made on adult trees growing in the Botanical Garden. The main advantage of using seedlings is that they are more convenient to handle in laboratory work. Although the seedling phase represents only & small fraction of the entire life of a plant, distributional patterns of plants in nature are often governed by factors affecting seedlings. (Moore, 1926; Daubenmire, 1943; Jarvis and Jarvis, 1963a). Seedlings of K. senegalensis for experiments between October 1970 and April 1971 were raised from seed supplied by the Forest Products Research Institute, Kumasi, Ghana. The seeds had been collected from Koforidua in May 1970. Seedlings for later studies 19University of Ghana http://ugspace.ug.edu.gh 20 were raised from seeds collected in March 1971> from tiie grounds of the University Campus, Legon where this species lines "the main University avenue. Seedlings of K. ivorensis were raised from seed also obtained from the Forest Product Research Institute, Kumasi. The seeds had been collected from Asenanyo > Forest reserve, in Ghana, also in May 1970. For seme experiments larger seedlings were used to supply sufficient leaf material for investigations of tissue water relations or to provide stems large enough to enable observations on diameter changes to be made. Some of these seedlings were less than one year old while others were just over one year old. At the time of experiment, these seedlings were between 50 and 80 can tall in both species. 1.3. Growth media Plants were grown on soil or in culture solution to which polyethylene glycol was added when it was desired to vary the water potential of the root medium. Both media have advantages and disadvan­ tages for the study of plant water relations. Several investigators (Sands and Rutter, 1959; Jarvis and Jarvis, 19&3&,% Assinall,. 1965), who have used soil as a rooting medium point out that the main disadvantage is that it is not possible to know the exact moisture stress to which roots are subjected, since water potential at the root-soil interface changes continuously with transpiration by the plants and as roots grow into new regions of soil. Exact knowledge University of Ghana http://ugspace.ug.edu.gh of soil moisture stress at the root—soil interface is however important since this factor appears to he the main soil characteristic controlling soil water availability for plant growth (Slatyer, 1967). The use of well stirred solutions containing osmotic substrates such as sodium chloride (Slatyer, 1961; Nieman, 1962), msnnitol (Gingrich and Russell, 1956; Slatyer, 1961, Jarvis and Jarvis, 1963c,d) and polyethylene glycols (Largerwerff, Ogata and Eagle, 1961; Jarvis and Jarvis, 1963d, 1965; Kaul, 1966; Kauftaann, 1968; Lawlor, 1969, 1970), attempts to overcome this limitation with soil. But there are also disadvantages in using aqueous solutions of osmotic substrates. The first is that the water potential will comprise only an osmotic component whereas soil water potential is made up of both osmotic and matric components, although there is evidence (Wadleigh and Ayers, 1945) that this difference may have little significance for plant growth response. Another disadvantage is that some osmotic substances tend to be absorbed by the roots, thus decreasing plant water potential. Jor example, Slatyer (1961) found that sucrose, sodium chloride and potassium nitrate were readily diffusible into the roots of tomato. It has however been found by Lawlor (1970) that undamaged roots have low permeability to polyethylene glycols of molecular weights 1000, 4000 and 20,000. Some of these substances, when absorbed, are also toxic to the plant (Largerwerff et al, I96I; Jarvis end Jarvis, 1963d; Macklon and Weatherley, 1965; Leshem, 1966; Lawlor, 1970). In.spite of these disadvantages, both soil and osmotica have been 21University of Ghana http://ugspace.ug.edu.gh 22 successfully employed, by most of the authors cited a ;ove, to study several aspects of plant response to moisture stress in the root medium, hence their use in the present study, 1,4, Culturing technique Seeds for soil experiments were sown in wooden boxes (51 x 45 x 10 cm) containing John Innes Potting Compost II which was obtained from the Botanical Garden, Legon. Seeds for culture solution experiments were sown in plastic bowls (about 35 cm in diameter and 14 cm deep) containing vermiculite, The containers were placed on a platform in the greenhouse. Gemination normally took place 11 to 21 days after sowing. Germinated seedlings in soil were allowed to grow for one month before they were transplanted. Those in vermiculite were allowed to grow only for three weeks before transplanting. The shorter time here was found necessary as it became more difficult to free the roots from the vermiculite Mien seedlings were allowed to grow for a longer period. At the time of transplanting into the respective media seedlings were mostly at the three to five leaf stage (K. Senegalensis^ or the two to three leaf stage (K. ivorensis). John Innes Potting Compost II was also used for subsequent growth of seedlings in soil eiiperiments. This medium allows good aeration while holding sufficient moisture for plant growth. As prepared by the Botanical Garden the mineral nutrient status of the soil was found to be adequate for growth of University of Ghana http://ugspace.ug.edu.gh the experimental seedlings and there were no harmful organisms since preparation included heat sterilization. A bulk quantity of this soil was obtained at the beginning of the experiments. Soil for each experiment was taken from this bulk to maintain uniformity of growing medium. This also meant that characteristics determined on samples from this bulk were applicable to soil used in each experiment. For culture solution experiments modified half strength Amon and Hoagland (1940) (see Hewitt 1952) nutrient solution was used. This culture solution was recommended by Dr N.I.C. Nwachuku of this department who had found it suitable for growth of tree seedlings. Seedlings were grown in polyethylene containers. These containers were preferred to porous clay pots because of their lightness and also because they prevented evaporation from their sides. Containers for the soil experiments -were 12 cm in diameter with a capacity of about one litre. Five drainage holes were bored at the base of each bucket and over these was placed a snail nylon mesh to prevent loss of soil particles. The containers were filled with soil to about 2.0 cm below the rim so that each held 1.4 - 1.5 kilogram soil at field capacity. There was one seedling in each container. Containers for culture solution experiments were 13.0 an in diameter with a capacity of about 1.1 litres. They were covered with black paper to prevent light reaching ihe solution in order to limit algal growth. Bach container held about 700 ml of culture solution, and was covered by a lid through which seven holes had been made for insertion of seedlings University of Ghana http://ugspace.ug.edu.gh and an aerator. Larger contsiners were used for culture solution experiments in which transpiration was measured. These were about 15 cm in diameter and about 2.5 litres capacity. 1.5. Experimental growth room Most of the experiments described in this study were carried out in a greenhouse in the Botany Department, Legon. Thermohygrographs were installed in the greenhouse to record temperature and relative humidity. Midday temperatures in the greenhouse before the actual experimental work were found to be too high (about 40°C); it was therefore found necessary to shade the house with palm leaves. /On certain occasions, Piche evaporimeters were installed in the greenhouse to record evaporation. Throughout the experimental period temperatures in the greenhouse never exceeded 36°Cs these normally occurred between 11.00 and 14.00 hours depending on the sunniness of the day. Minimum values were obtained at night and these never fell below 18°C. Minimum relative humidities were around 45 to 50$: these also were obtained between 11.00 and 14.00. Some transpiration experiments were carried out in the Research room of tiie Botany Department, Legon, where, through air conditioning, temperature and humidity were controlled slightly more precisely than in the greenhouse. Because of the near constant temperature maintained in this room, determinations requiring more precise temperature control were also carried out here. 24University of Ghana http://ugspace.ug.edu.gh 25 CHAPTER H v GROWTH IN RELATION TO MOISTUEE STRESS IN THE ROOT MEDIUM 2.1. Introduction. Normal plant growth is sensitive to both too much or too little water. Hunt (l95l), Walker (1962), Hosner and Leaf (1962), Gaertner (1963) aid Mueller — Dombois (19&4), observed that excess water is detrimental to most species presumably through poor aeration and the consequent reduction in root permeability (Kramer, 1956; Mees and WeatSserley, 1957). Several workers, however, indicate that water deficiency is even more important than excess water in limiting plant growth (Slavik, 1965; Kozlowski, 1968), Thus Marsden (1950, cited by Kozlowski, 1958) observed that drought in New England had been responsible for great reduction in growth of trees, poor fruit development, early leaf fall, dieback, thin foliage, transplanting failures, sunscorch and premature death of trees. In temperate climates, although duration of the growing season is controlled mainly by photoperiod and temperature, water deficiency during the growing season also has influence on growth of forest trees (Zahner, 1968). In tropical regions* the dry season is usually the least favourable period for plant growth, although the relative importance of moisture as against other environmental factors (for example, photoperiod and low night temperatures, Njoku, 1959J 1964; Longman, 1969) is little understood. University of Ghana http://ugspace.ug.edu.gh 26 Reduction of growth results because of the effects water deficit existing in the root medium has on internal water balance of plants. Plant growth is related directly to the internal water balance of plant (Kramer, IS63; Slatyer, 1967, 1970). In nature plant water status is controlled by the degree of soil moisture stress and the diurnal lag of absorption behind transpiration (Kramer, 1962; Slatyer, 1970). These are in turn controlled by environmental factors such as temperature, solar radiation, humidity and wind, the effects of which are modified by features of the plant which tend to reduce transpiration (e.g. leaf characters) or enhance absorption, (e.g. efficient root system). During the day transpiration usually exceeds absorption, internal water deficits develop, leading to a steep gradient in water potential between plant and soil. At night when transpiration is low or negligible this gradient promotes the absorption of water into the plant so that the internal water deficit is reduced or even eliminated (Slatyer 1967). The extent to which deficits are eliminated each night depends greatly on soil moisture status. When this is high deficits in plant are rapidly eliminated; but when the soil is dxy the overnight improvement in plant water status is limited. Thus the moisture status of the root-medium effectively sets the upper limit of plant water status at the beginning of each day, and this controls plant growth. It is probably for this reason that the effect on plant growth of water University of Ghana http://ugspace.ug.edu.gh stress in the root medium has been studied by a large number of investigators, such as Sates (l955$» Slatyer (1957b), Sands and Sutter (1959), Jarvis aid Jarvis (1963a, 1965), M.S. Jarvis (1963), Aspinall (19^5) and Lawlor (1969). Plant growth is the ultimate expression of the various processes that occur in the plant. It reflects the interaction of environment and plant genotype (cf Lirxmoore and Millington, 1971) and is known to be directly affected by water stress (Wardlaw, 1966). If environmental moisture has different effects on the two species in the present study, because of any adaptive features they possess, this should be reflected in their growth response to soil moisture stress. The experiments described in this chapter were carried out to compare the growth responses of the two species, K. senegalensis and X. ivorensis. to increases in moisture stress in the root mediun, both when grown on soil and in culture solution. 2.2. Growth response to moisture stress in soil. Seedlings which had been transplanted singly onto soil in 12 cm diameter containers (see p. 23 Chapter One) were allowed to grow until they were well established and their roots thoroughly permeated the soil mass. During this establishment period, which lasted about four weeks from transplanting, the soil was kept moist to reduce the development of stress in the plants. 27University of Ghana http://ugspace.ug.edu.gh 28 Seedlings of each species were divided into five matched groups (I, II, III, IV and V) of eight seedlings each, matching being based on leaf number and partly on height. This was done to reduce possible errors which might arise from differences in initial size of plants. Pour of the matched groups of seedlings (II - v) were assigned, each to one of four soil moisture treatments, designated A, B, C and D. Soil moisture level was controlled by pot weighing. Several workers (e.g. Sands and Rutter, 1959) have pointed out that in an experiment such as this soil moisture status is best expressed in terns of moisture tension since this indicates the force with which the soil holds water, and hence the ease with which the plant can obtain water from the soil. Therefore, in order to relate pot weights to soil moisture tension the moisture characteristic of the experimental soil was deteimined. The pressure plate apparatus was used to establish the relation between soil moisture content (as percent dzy weight) and soil water potential (bars) for the range 0 to -1.0 bars. A pressure membrane apparatus was used for the range -1.0 to -11.0 bars. It was not possible because of technical failure in the apparatus to extend the determination to -15.0 bars, which is generally accepted as the permanent wilting point for most soils (Black, 1957; Russel, 1961). The extension of the curve (Pig. 2) to this range was done by extrapolation. The pressure apparatus, as used here measures only the metric potential of the soil University of Ghana http://ugspace.ug.edu.gh 29 whereas total water potential of the soil includes an osmotic component (cf. Jarvis ana Jarvis, 1963a). No attempt was made here to correct for this osmotic component as was done by Jarvis and Jarvis (loc. cit.) since there was no good reason to expect this to be high (cf. Slatyer, 1967). Because the apparatus used was loaned frcm the Soil Science Department, Faculty of Agriculture, Legon, it was not possible to complete determination of the soil moisture characteristic before the experiment was begun. Therefore soil moisture levels in the various treatments were controlled simply by pot weighing throughout the experiment. Conversion of soil moisture contents to the equivalent potentials was done at the end of the experiment, A heavy duty top-pan Mettler balance (Mettler P3) was used for pot weighing so that weights could be obtained to an accuracy of +1.0 gram. The moisture content of the soil at field capacity, that is after standing in water to become thoroughly saturated and allowing the pots to drain for 24 hours was found to be 23.0+1.6$! of the soil dry weight. The pot weight at this point was taken to be 100 percent moisture status. Treatment A consisted of soil kept at near this point by daily addition of water. In the other treatments, the soil, initially at field capacity, was allowed to dry out respectively to 75, 50 and 25% of the weight at field capacity before rewatering. Soil water potentials (bars) corresponding to these weights were subsequently read from the curve in Pig. 2 and were as followsi University of Ghana http://ugspace.ug.edu.gh PI&. 2 THE RELATIONSHIP BSTWE3J SOIL WATER POTENTIAL (BARS) AND 3DIL MOISTURE OONTENT (% DRY ’3EEGHT). EOS THE EXPERIMENTAL SOIL, JOHN INNES POTTING- COMPOST H. (see also Appendix 2, table, i for statistical treatment op the DATA). University of Ghana http://ugspace.ug.edu.gh S O IL W A T E R P O T E N T IA L (B A R S ) 30 F I G . 2 University of Ghana http://ugspace.ug.edu.gh Treatment A (-0.3 bars), Treatment B (-0.4 bars), Treatment C (-0.6 bars) and Treatment D (-4.5 bars). This method of soil moisture control has been successfully used by previous workers (see for example Slatyer, 1957b). The experiment was carried out in the greenhouse. Over the experimental period (13 November 1970 to 12 January 1971) average temperatures were around 28°C at 09.00 hrs and 33°C at 15.00 hrs; relative humidity was around 60jFa (09.00 hrs) or (15.00 hrs). The general light intensity within the greenhouse was compared from time to time with light intensity outside. Two matched selenium barrier layer photometers (BEL Lightmaster Model 18) were used for this purpose. Internal lighting was about kOfo of that outside. The roof of the greenhouse was fairly heavily shaded to reduce mid-day temperatures (see>. p. 24 Chapter i). The experiment was carried out on a centre bench of the greenhouse; a randomized-block design was adopted. Randomization was based on a table of random numbers (Fisher and Yates, 1957). It was found necessary to re-randomize the groups and plant pots weekly because environmental conditions within the greenhouse varied appreciably from one end to the other. In order to examine as fully as possible how soil moisture stress might operate to control growth in the experimental plants, the growth analysis technique (Gregory, 1917) was used. The remaining group of seedlings, group I, was harvested at the beginning of the experiment. Length and breadth 31 University of Ghana http://ugspace.ug.edu.gh 32 of the leaves of each seedling were measured. The leaves were then blue printed on 'Ozalid’ paper and then planimetered to obtain their areas. From the regression of planimetered area on length x breadth (l x b) of these leaves (see .Appendix 2Sgraph I), leaf area of the unharvested plants could subsequently be derived by simple measure­ ments of length and breadth. Dry weights of leaves, stem and roots of the harvested plants were determined separately by oven-diying at 90 - 100°C for 24 hrs. The unharvested seedlings were allowed to grow on for approximately nine weeks. Periodic measurements of seedling height and leaf area were made over this interval. During this period K. senegalensis and K. ivorensis seedlings in Treatment JR received* overall, approximately 26 and 24- watering cycles respectively, while those of treatments C and D received 12 and 6 cycles respectively for both species. Pots in treatment D did not often reach the predetermined weights when wilting took place, therefore the criterion of watering in this group was the apparent inability of vdlted leaves to recover at sunrise. At the end of the nine weeks the seedlings were harvested; leaf areas and dry weights were determined as for the initial harvest. The initial weights of these seedlings were obtained from their initial leaf areas by using the relation between leaf areas and plant dry weight obtained from plants in the critical harvest (see Appendix 2, graph II)* University of Ghana http://ugspace.ug.edu.gh 33 (b) .Analysis of growth From the total diy weights (W-̂ and and leaf areas (^^ and, A2) at times ^ and tg, relative growth rates (R.G.R.), net assimilation rates (N.A.H.) end mean leaf area ratios (M.L.A.R.) were calculated as follows: R.C-.R. (g/a/wk) = loge - loge t2 " *1 N.A.R. (g/m2/wk) = (W2 - W^) (loge A2 - loge A ^ A2 “ ̂ ^ 2 “ V or * (»2 - V (ii) 2 M»L.A.R. (cm /g) = 2 + A 2 W1 W 2 Fisher (1920) showed that the above equation for R.G.R. is valid irrespective of the relationship between dry weight and time. The equation for IT.A.R and M.L.A.R. are however subject to certain limitations. Thus Williams (1946) emphasized that the use of the above equations for N.A.R. and M.L.A.R. is valid only when the total plant weight (W) is linearly related to the total leaf area (A) over the experimental period. In this experiment there were harvests on only University of Ghana http://ugspace.ug.edu.gh 34 two occasions (initial and final harvests) it was not possible, therefore to know the exact relation between W and A over -the period. Cocmbe (i960) has suggested that *en the relation between A ^ ^ is small (less than 2.0) the difference between the two methods of calculating 1T.A.R. is small. In the above experiment it m s found that for most of the treatments A^/A^ was less than two; the exceptions are for K. ivorensis in treatments B and D. N.A.R. was generally calculated by equation (i) except when A^ - A^ was zero when equation was used. The above equation for the M.L.A.R. was also used to sinplify the arithmetic. Distribution of dry matter between the main organs - root, stem and leaf - was compared between treatments, at the final harvest by calculating the ratios of dry weights of these organs to total plant dry weight, thus obtaining root-weight ratio (R.W.R.), stem weight ratio (S.W.R.) and leaf weight ratio (L.YiT.R.). These were expressed as percentages of total dry weight. In addition, root- shoot ratios (ly’s) and specific leaf areas (leaf area/leaf diy weight) (S.L.&.), were calculated. To bring out the comparison between the species more clearly all these parameters were further expressed as percentages of the values in treatment A. (c) Results During the experimental period it was observed that some of the leaves of K. ivorensis in Treatment A became chlorotic. Towards iiie end of the ej;perimental period, one of these seedlings University of Ghana http://ugspace.ug.edu.gh shed most of its leaves* "This seedling was not included in any of the final calculations. It was also observed that the length of the watering to drying-out cycles differed from pot to pot within the same treatment; the bigger seedlings with larger leaf areas took a shorter time to get to the predetermined weights. The tensions stated for each treatment are therefore averages of the tensions experienced by all the seedlings in that treatment before rewatering took place. Changes in height and leaf area of the seedlings over the experimental period are shown in Pigs. 3 and 4 respectively. Initial and final heights, dry weights and leaf areas, per plant together with the least significant differences (L.S.D., P = 0.05) between these values are presented in Table 2. A breakdown of the dry weights into leaf, stem and root fractions is given in Appendix 2, Table ii. In Table 3, growth functions and mean leaf area ratios derived from the primary data are similarly presented, together with leaf area ratios and ratios indicating the distribution of dry matter within and between the main organs at final harvest. To aid comparison, the derived data are presented again in Fig.4 and Table 4, as percentages of the values in Treatment A. K. Senegalensis seedlings were by far larger than K. ivorensis seedlings, both at the beginning and at the end of the experimental period. The larger initial size is presumably partly due to the larger seed size of K. senegalensis (see p. 15). This difference is reflected in the initial and final dry weights of the seedlings. The growth 35 University of Ghana http://ugspace.ug.edu.gh 36 Table 2 unary data for the analysis of growth of Khaya senegalensis and ivorensis seedlings under varying soil moisture treatment: A (-0.3); (-0.4); C, (-0.8); D, (4.5) bars; t and t^ indicate initial and lal harvest occasions respectively. K. senegalensis ait indices A Treatments B C D least sign diff. (P =0.05) m height per plant (cm) t 14.0 13.9 14.0 11.7 3.2 *2 16.5 16.7 14.6 14.6 3.7 sn total diy weight per ait (g) *1 0.751 0.829 0.876 0.813 0.097 *2 4.100 4.500 3.200 3.000 1.940 sn leaf area per ant (cm^) 114 125 103 114 52 *2 225 246 179 174 73 (continued next page) University of Ghana http://ugspace.ug.edu.gh 37 (Table 2 continued) K. ivorensis lant indices A Treatments B C D least sign. diff. (P=0.05) ean height per plant (cm) t̂ 9.7 10.5 10.6 9.6 2.2 *2 12.6 13.3 13.1 12.7 2.3 ean total dry weight per lant (g) *1 0.460 0.410 0.430 0.370 0.130 *2 1.710 2.150 1.980 1.710 0.600 [eaj^leaf area per plant t1 100 87 93 78 31 t2 169 189 182 153 31 University of Ghana http://ugspace.ug.edu.gh rates obtained for the two species are in general comparable to values obtained by Jarvis and Jarvis (1963a) for tree seedlings (aspen, pine and spruce) in a similar experiment using soil moisture tensions ranging from 0 to 4 atmospheres, and to values found by Ampofo (1969) for Afronaosia elata. a West African rain forest tree; but they are lower than most of the values found by Okali (1971b) for seme West African forest-tree seedlings in full day light, presumably partly because of the relatively heavy shade {kO per cent of full day light) employed in the present study. These comparisons suggest that, apart fran the shade, conditions of growth were relatively good in the present experiment. (i) Height growth At the beginning of the experiment mean heights of K. senegalensis ■ seedlings assigned to Treatments A, B and C were similar but significantly different from those of seedlings for Treatment D (Pig. 3)• This initial difference -was maintained and possibly increased by treatment, over the experimental period. K. ivorensis seedlings assigned to Treatments B and C were similar in height, but significantly taller ihan seedlings for Treatments A and D, which were themselves similar -when the experiment was started, Again, these differences appear to have been maintained, though slightly reduced, over the experimental period. For K. senegalensis. height of seedlings in Treatments A, B and C increased rapidly at the beginning of the experiment but slowed down by the sixteenth day. Seedlings in University of Ghana http://ugspace.ug.edu.gh M E M HEIGHT OF KHAYA SENEGALMSIS (KS) JND K. rVORENSIS (Kl) SEEDLINGS UNDER VARIOUS SOU MOISTURE TREATMENTS (BARS). Tlffi&mENTS: A (-0.3), B (-0.4), C (-0.8) M D D (-4.5). FIG. 3 University of Ghana http://ugspace.ug.edu.gh M E A N H E IG H T (C M .) 39University of Ghana http://ugspace.ug.edu.gh k0 Treatment D showed a slow rate of height increase throughout the whole period. TJnlike seedlings of K. senegalensis those of K. ivorensis showed a more uniform rate of change in height throughout the experimental period. Height growth was more or less linear in this case and there was no striking difference in rate between treatments. However, in both species, there was a tendency for seedlings in Treatment S to be taller than seedlings in all other treatments. This and the slow growth of K. senegalensis in Treatment D were the only observable effects of the treatments applied on height growth. (ii) Changes in leaf area. As with height, total l.eaf area per seedling of K. senegalensis was generally greater than that per seedling of K. ivorensis. both at the beginning and at the end of the experimental period (Fig. For both species initial leaf areas of seedlings assigned to the various treatments were more similar than were heights, and, by the end of the experiment, differences between seedlings in the various treatments were still not significant. This presumably was partly due to the large variability in leaf area of plants within any one treatment. In spite of this, a tendency for more rapid rate of leaf area increase, particularly up to Day 32, is indicated for K. senegalensis seedlings in Treatments A and B. A more uniform rate of leaf area increase is also shown by K. ivorensis seedlings. As vath height growth, seedlings in Treatment B, for both species, tended to have higher leaf areas. University of Ghana http://ugspace.ug.edu.gh PIG-, 4 LEAP AREA OP KHAYA SMEGALBNSIS (KS) JND K. IVORMSIS (Kl) SEEDLINGS UNDER VARIOUS SOIL MOISTURE TREATMENTS (BAHS). TREATIMTS: A (-0.3) 5 B (-0.4) j C (-0.8), D (-4.5). THE BROKEN LINES INDICATE ACCIDENTAL LOSS OP LEAVES WHEN THE GREENHOUSE WAS SPRAYED TO CONTROL INSECTS. University of Ghana http://ugspace.ug.edu.gh L E A F A R E A ( C M 2 ) a FIG. 4 University of Ghana http://ugspace.ug.edu.gh 42 (iii) Relative growth rate. In general the relative growth rates of the two species (Table 3) appeared not to be markedly different. They ranged from 0.14 to 0.18 s/g/wk in It. senegalensis and from 0.15 to approximately 0.19 g/g/wk in K. ivorensis. The data presented in Pig.5 suggest that K. senegalensis was more sensitive to the diy treatment than K. ivorensis: its relative growth rate decreased steadily with soil diyness so that in the driest treatment (D) the rate was significantly reduced to 11% of that in treatment A. The differences between seedlings of this species in Treatments A, B and C were however not significant. In / terms of growth rates, the response of K. ivorensis seedlings can be considered as that to a wet treatment (Treatment A) and a dry treatment (Treatments C and D) (cf. Jarvis and Jarvis, 1963a). Relative growth rate was highest in Treatment B and tended to be reduced, though not significantly in either wetter or drier soil. The lcyest growth rate occurred in Treatment A, and although the rate in Treatment D was slightly higher than in C the difference is not significant. (iv) Het assimilation rate. Net assimilation rates obtained for K. senegalensis were much higher than those for K. ivorensis (Table 3). Like the relative growth rates, maximum net assimilation rate for K. senegalensis (22.6 g/m /wk) occurred in Treatment A and the minimum (15.0 g/m2/wk) in Treatment D, The difference between these two values is significant. University of Ghana http://ugspace.ug.edu.gh 43 Table 3 Derived growth data and ratios indicating distribution of dry matter between main organs for Khaya senegalepsis and K. ivorensis seedlings under varying soil moisture treatments; A (-0.3); B (-0.4); C (-0.8); D (-4.5) bars. K. senegalensis Plant indices A Treatments B C D least sign* dSttfifi. (P = 0.05) Relative growth rate (g/g/wk) 0.183 0.179 0.170 0.140 0.033 Net assimilation rate (g/m2/wk) 22.6 19.4 21.9 15.0 6.9 Mean leaf area ratio (cm2/g) 103.6 108.5 106.1 107.6 17.7 p Leaf area ratio (cm /g) 62.1 64.8 62.5 60.5 23.9 Specific leaf area (cm/g) 201.5 219.0 220.2 200.8 59.0 Leaf weight ratio {%) 31.7 29.4 28.6 29.2 2.8 Stem weight ratio {%) 25.0 25.5 24.7 24.2 3.2 Soot weight ratio (j?) 44* 1 45.3 46.3 46.6 6.2 Root shoot ratio (%) 80.0 81.1 87.7 87.5 13.9 $ continued next .'.page) University of Ghana http://ugspace.ug.edu.gh Table 3 (continued) 44 K. ivorensis Plant indices A Treatments B C D least sign, tiff*. (P = 0.05) Relative growth rate ; (g/g/wk) 0.150 0.189 0.165 0.171 0.060 Net assimilation rate (g/m^/wk) 11.3 14.9 13.3 13.5 3.0 Mean leaf* area ratio (cm2/g) 154.2 149.8 151.7 148.4 17.7 Leaf area ratio (cm /g) 93.0 88.0 91.3 88.4 17.7 Specific leaf area (cm2/g) 240.7 217.0 231.6 226.9 34.2 Leaf weight ratio (%) 59.5 40.6 37.7 58.3 5.9 Stem weight ratio {%) 26.7 24.1 24.8 25.1 3.5 Root weight ratio (%) 37.0 35.3 37.2 36.6 9.9 Root shoot ratio 55.2 55.1 59.7 58.6 11.4 University of Ghana http://ugspace.ug.edu.gh THE RELATION BEOTE® RELATIVE GROWTH RATE (R.&.R.), NET ASSIMILATION BATE (N.A.R.) OR MEAN LEAF A R M RATIO (M.L.A.R.) AND SOIL MOISTURE POTENTIAL (BARS), FOR KHAYA SME&ALENSIS ( & ) AND K. IVORMSIS ( 0 ) SEEDL3N&S. THE EATA ARE EXPRESSED AS PERCENTAGES OP THE VALDES IN TREATMENT A. 2TG-. 5 University of Ghana http://ugspace.ug.edu.gh M E A N L E A F N E T A S S I M I L A T l O N A R E A R ATI O R ATE (o/o) (o/o) (NAR) University of Ghana http://ugspace.ug.edu.gh RELATIVE GROWTH RATE <%) University of Ghana http://ugspace.ug.edu.gh 46 There was, however, no significant difference between either of these values and those obtained for seedlings in Treatments B and C. In K. ivorensis. the maximum net assimilation rate (14.9 g/m /wk) was obtained for seedlings in treatment B and the minimum (11.3 s/m /wk) for plants in treatment A, Unlike the situation with K. senesalensis no significant treatment effect was found between K. ivorensis seedlings. (v) Mean leaf area ratio. In all the treatments, mean leaf area ratios of K. ivorensis seedlings were greater than those of K. seneftalensis seedlings. The differences between treatments within the species were however not significant (Table 3 and Fig. 5). (vi) Leaf area ratio, specific leaf area and leaf weight ratio at final harvest. K. ivorensis had higher leaf weight ratio and specific leaf area ratio than K. senegalensis so that leaf area ratios at the second harvest were also higher in the former species (Table 3). Within the species, the treatments appeared not to have had any significant effect on any of these parameters, (vii) R&gfc weight ratio, root-shoot - ratio and stem-weiaht ratio. For both species the treatments apparently had no significant effect on any of these ratios although root weight ratio tended to increase with soil dryness. K, senefcalensis had proportionately more dry matter present in the root than had K. ivorensis. while the reverse University of Ghana http://ugspace.ug.edu.gh 47 Ratios indicating distribution of diy matter between main organs at final harvest, for Khaya senegalensis and K. ivorensis seedlings grown under various soil moisture treatments — A , (0.3)} B (-0.4) C (-0.8); D (-4.5) bars. The data is expressed as percentages of values in Treatment A. Table 4 K. senegalensis Plant indices A Treatments B C D Leaf area ratio 100 104 100 97 Specific leaf area 100 109 109 100 Leaf weight ratio 100 93 90 92 Stem weight ratio 100 102 99 97 Root weight ratio 100 103 105 106 Root shoot ratio 100 101 110 109 K. ivorensis Plant indices A Treatments B c D Leaf area ratio 100 95 98 95 Specific leaf area 100 90 96 94 Leaf weight ratio 100 103 95 97 Stem weight ratio 100 90 93 94 Root weight ratio 100 95 101 99 Root shoot ratio 100 100 108 106 University of Ghana http://ugspace.ug.edu.gh is true for the proportion of diy matter in the leaf (Table 3)» Similarly higher V s was obtained for K. senegalensis than for K. ivorensis seedlings. The ratio of stem weight to total dry weight was similar in both species. The greater proportion of root in K. senegalensis may be of adaptive significance, although, in the present experiment its advantage was not apparent. (a) Discussion The view of VeihSeyer and Hendrickson (1950) that soil moisture stress is equally available for plant grov/th within the available range (0 - 15 atmospheres) has been actively contested (Richards and Wadleigh, 1952; Slatyer 1957a; Stanhill, 1957)* According to Slatyer (19&7), this view was based primarily on field experiments with deep-rooted tree crops. It is now more generally accepted that growth may vary with soil moisture' status within the available range., (Richards and Wadleigh, 1952). In the present study, where potted seedlings were used, same reduction in growth of the two experimental species was expected, but the results presented show that any such reduction was small. Relative growth rate of K. senegalensis generally decreased with increasing soil moisture stress, so that there was a significant reduction for plants in Treatment D (-4.5 bars) as compared with those in Treatment A (-0.3 bars). In K. ivorensis however the rate vt&8 apparently not significantly affected by the treatments although there was a tendency for growth to be depressed in the wettest treatment (A). 48University of Ghana http://ugspace.ug.edu.gh In general net assimilation rate was lower in K. ivorensis than in K. senegalensis but this was compensated by higher mean leaf area ratios in the former species so that relative growth rates were very similar for both species. Mean leaf area ratios for both species showed very little response to the treatments so that reduction in relative growth rate of K. senegalensis iB response to dry soil was due to reduction in net assimilation rate; similarly the tendency for reduced relative growth rate in response to wet soil in K. ivorensis can be accounted for by reduction in net assimilation rate. Jarvis and Jarvis (1963a) obtained a similar relationship between relative growth rate and net assimilation rate in their response to soil moisture status with the species they studied. The apparent low growth rate for K. ivorensis in Treatment A, the wettest treatment, could be attributed to poor aeration. Kramer (l9§f») and Meesand Weatherley (1957) have pointed out that poor aeration in the rooting medium reduces root permeability sufficiently to cause large reduction in water uptake which may limit growth. Jarvis and Jarvis (1963a) observed a reduction in growth of aspen, birch and spruce in the wettest of the series of water regimes they practised in their study. They also suggested that poor aeration was responsible for low growth rates in response to wet soil. In the present study growth of one of the species, K. senegalensis. was however not reduced by Treatment A. The generally larger size of the seedlings and the amount of root they had may have caused rapid transpiration per plant and hence rapid soil moisture depletion, so that any excess water, which might have caused prolonged water logging after each watering operation, was rapidly removed. Transpiration data 49University of Ghana http://ugspace.ug.edu.gh 50 to be presented later support this s suggestion. The growth of K. senegalensis was significantly reduced by soil moisture stress of about 4*5 bars while that of K. ivorensis was not significantly reduced over this isnge. Sands and Rutter (1959) reviewed examples of tensions which have been found to reduce growth significantly in plants and came to the co&clusion that for most species considerable restriction of growth was caused by moisture tensions of about 0.5 atmospheres. In their own work wilii PiBus svlStestris. they found diy weight production by first year seedlings to be sensitive to this tension. Jarvis and Jarvis (1963a) however found that maximum growth of birch, aspen, spruce and pine occurred in soils at this tension but that diy weights of the conifers were reduced by about 335̂ in soils dried to 1 .7 atmospheres and those of the engiospeims by about 10 - 20^ in soils at 2 atmospheres. M.S. Jarvis (1963) also observed that measurable growth (as assessed by leaf development) of Saxifraga hvpnoides and ffilipendula vulgaris ceased at tensions of 0.5 and 2.Q atmospheres respectively. Although she could not reach a definite conclusion it appeared that for the woody species, Prunus padus and Thelyorania sanauinea. growth ceased at soil moisture tension above 2 - 3 atmospheres. In contrast to the above results indicating low soil moisture tensions for the cessation of growth, Wadleigh and G-auch (1948) studying the effect of increasing total soil moisture stress on the daily elongation of cotton leaves, found that elongation ceased at a narrow range of stress values, close to 15 atmospheres. Slatyer (1957^) also observed that for privet and cotton increases in dry matter ceased at University of Ghana http://ugspace.ug.edu.gh 51 dawn values of leaf water status corresponding to the permanent wilting point for these ^ecies. It thus appears that a variable range (low or high) of soil moisture stress has been found by different workers to reduce plant growth. Some plants may be sensitive to low moisture tensions while others are sensitive to high tensions. The growth response described here indicate that K. senegalensis. a savanna species was more sensitive to moderate soil moisture stress than K. ivorensis which is a forest species. This result was not expected. M.S. Jarvis (1963) has however observed that high desiccation resistance and ability to grow well at moderate moisture stresses may be inversely correlated in seme species. Thus a plant with high sensitivity to small reduction in water potential may still possess strong drought resistance, because other features which promote drought resistance e.g. rapid / stomatal closure at relatively low moisture stress or slow rate of water loss per unit decrease in water potential of tissues, may be disadvantageous for growth. Stomatal closure increases carbon dioxide diffusion resistance and this could limit growth. Data to be presented later show that K. senegalensis possesses some drought resistant features. Interpretation of differences which arise in plant growth response to soil moisture stress is complicated by the difficulty of knowing the precise moisture stress experienced by plants, particularly at the root-soil interface, throughout an experimental period. 3h the present case it is possible that the correct stress which existed at the root—soil interface has not been adequately described. Since the same type of soil was used University of Ghana http://ugspace.ug.edu.gh and the two species transpire at different rates (see later), the stress experienced at the root surface may have been different (cf. Jarvis and Jarvis, 1963a), because of different rates of depletion of soil moisture in the root zone. The more rapidly transpiring plant could be expected to have experienced greater stress than the plant with a lower transpiration rate. Evidence to be given later indicate that K. senegalensis generally transpires more rapidly than K. ivorensis. It is possible therefore that the significant reduction in growth shown above for senegalensis was in response to more severe water stress than occurred in K. ivorensis or than was detected by weighing of the whole soil mass. The differences in growth of the two species might also have arisen because of variation in the length of time seedlings took to reach the predetermined weights before watering took place, so that the method of taking the mean soil moisture tensions of the individual seedlings in a treatment to represent the tension experienced by the group (see p. 35 ) was probably inadequate. M.S. Jarvis (1963), in order to obtain estimates of the amounts of effective soil moisture tensions experienced by the seedlings she worked with, summed up the areas beneath the drying out curves she obtained for seedlings in each treatment and expressed the results in percentage atmosphere-days. In this way the length of the drying out cycle was accounted for in each treatment. Jarvis (1963) also adopted a similar method for estimating effective soil moisture tensions in his experiments. In the present experiment however no attempt was made to account for the length of the 52University of Ghana http://ugspace.ug.edu.gh drying out cycle,* therefore, differences in the growth results which are due to this factor are not known. In addition to the above complications, Owen aid Watson (1956) indicate that experimental comparison of plant growth response to soil moisture stress is further confounded by the effect of re-watering on plants which had been subjected to drought. They observe that re­ watering could cause temporarily greater increases in dry matter production in plants which have been subjected to prolonged drought as compared with irrigated plants that had never been subjected to severe water stress. For the reasons outlined above it became necessary to examine further the growth responses of the two species to moisture stress in the root medium, this time using a root medium in which the water potential could be more precisely controlled; hence the water culture experiment which is described in the section that follows. By using a culture solution technique it was also possible to extend the severity of stress applied^ to see whether the pattern of response is similar under both high and low stress conditions. As a further improvement upon the soil experiment it was thought desirable to grow the two species in the same container so that there would be no doubt about the similarity of the water potential experienced by the two species at the root surface. This condition however introduces inter-species competition, but there was no good reason to expect large interference effects. 53 University of Ghana http://ugspace.ug.edu.gh 54 2.3. Srowtl re as onse 'to-.moistaitrsu, atre a a in culture solution. a« Method Seedlings for this experiment were raised from seed sown in vermiculite (see p. 22). They were transplanted into half-strength Amon and Hoagland culture solution (see below) held in two large (38.0 cm diameter) polyethylene containers. The solutions were not sterilized as had been done by previous workers (for example Jarvis and Jarvis, 1963d, 1965). The solutions were however aerated continuously, and the seedlings were allowed to grow in this condition until they were required for experiment. This took about four T®eks from transplanting. The seedlings were then divided into matched lots, corresponding to the number of pots to be used in the experiment. Matching was based on lsgf number alone; seedling height was not used for this purpose because at this stage seedling stems (particularly in K. ivorensis) were too short. After matching, the seedlings were carefully inserted through holes made in the lids of the experimental pots. They were supported with cotton v s o o l. In order to ensure that the growth medium was the same for the two species in each treatment, each experimental pot carried three seedlings of each species so that there were six seedlings per pot. The experimental pots each held 700 ml half-strength Amon and Hoagland solution aerated continuously. The seedlings were allowed to grow for a further period of two weeks before the experiiaent was started. It was necessary after each transplanting operation, to University of Ghana http://ugspace.ug.edu.gh allow the seedlings to grow for some t ime so that any root damage caused by the transplanting operation could heal, Lawlor (1970) has shown that uptake of polyethylene glycols by plant roots is reduced if roots are undamaged. The total length of time allowed for recovery from any root damage in this experiment was six weeks - four weeks after the initial transfer from veimiculite and two weeks after transfer from large pre-treatment containers to experimental pots. Preliminary experiments, by growing the seedlings in different strengths of Amon and Hoagland (1943), (see Hewitt 1952) solution, showed that the half-strength concentration was optimum for growth of the seedlings. The basic culture medium used consisted ofj 55 Ca(N03)24H20 0.354 g HgSO^HgG 0.247 g KNO^ 0.404 g * V ° 4 0.136 g N\ H2P04 0.231 g "^Microelements 0 .5 ml Distilled water 1000 ml •i(The composition of tine microelements solution was: MnClg.W^O, 1.810 g/lj CuSO^.SHgO, 0.080 g/L; ZnSO^HgO, 0.220 g/L; H^BO^, 2.860 g/L; and (NH^MOyO^.i^O, 0.124 g/L). The seedlings were subjected to four levels of osmotic potential in the root medium: A (control, -0.3 bars), B (-2.8 bars), C (-5.3 bars) and D (-10.3 bars). Treatment A consisted of plants grown in the culture solution alone; the quoted osmotic potential for this medium is derived University of Ghana http://ugspace.ug.edu.gh 56 from the literature (Hewitt, 1952) for half-strength of solutions of identical composition. The solutions in Treatments B, C and D were made respectively by dissolving 100, 150 and 256 g of polyethylene glycol (molecular weight i»JOOO) in a litre of the half-strength culture solution. These weights were derived from the curve of Lawlor (1970, 3?ig. l) relating concentration to osmotic potential for polyethylene glycol of the same molecular weight. The osmotic potentials quoted , I are -therefore only approximate. Because of peculiarities in behaviour of polyethylene glycols (cf. Lagerweft et al, 1961), there was no ready means of measuring osmotic potentials of these solutions directly. Two containers were assigned to each treatment. The volume of solution in each container was maintained at 700 ml by daily addition of distilled water. The solutions were renewed twice over the experimental period of 21 days. The first renewal took place seven days after 12ie experiment had started and the second was ten days after the first renewal. This experiment also was carried out in the greenhouse. Over the experimental period (12 July 1971 to 2 August, 1971) average temperatures were around 26°C at 09.00 hrs and 30°C at 15.00 hrs; relative humidity was around 75/i (09.00 hrs) or 60% (15.00 hrs). The sky was never over­ cast during this period} this made' it impossible for light intensity measurements to be made as was done for the soil experiment. The light intensity might be lower then that encountered in the soil experiment. In all, growth conditions were comparatively less severe for this experiment than during the soil experiment. As in the previous case University of Ghana http://ugspace.ug.edu.gh 57 the experiment m s carried out on the centre bench of the greenhouse and a randomized block-design was adopted. At the beginning of the growth period, ten seedlings of K. senegalensis and nine of K. ivorensis were harvested. At this time alsoj marks were made with white paint a few millimeters above the pot lid, on the stems of the plants allowed to grow on. These marks served as the base line from which height measurements were subsequently made. Height and leaf area measurements were made at intervals during the experimental period. Leaf area and dry weights of leaves, stem and roots were determined for "the harvested plants as was done in the soil experiment. Similar determinations were made at the final harvest. As in the soil experiment, the results were examined by the growth analysis technique, b. Results Wilting was observed frequently in almost all the seedlings of the two species in Treatment D during the experiment. Occasional wilting was also observed in most of K. ivorensis seedlings in Treatment C. The results of the height and leaf area measurements are expressed in graphs (Pigs. 6 and 7 respectively). Table 5 shows the primary growth data together with the least significant differences (L.S.D, P = 0.05), while Table 6 presents the derived growth data and ratios indicating distribution of dry matter between the main organs for both K. senegalensis and K. ivorensis. also with the L.S.D. (P = 0.05). University of Ghana http://ugspace.ug.edu.gh 58 Table 5 Primary data for the analysis of growth of Khaya senegalensis and Kg. ivorensis seedlings in culture solutions of different osmotic potentials (bars)s A (-0.3);® (“2.8); C (-5*3); D (-10.3); and t„ indicate initial and final harvest occasions respectively. Treatment A consisted of culture solution alone, B, C and D consisted of culture solution plus polyethylene glycol. K. senegalensis Plant indices A Treatments B C D least sign diff. (P, 0.05) Mean height per plant (cm) 15.4 16.3 15.8 14.3 2.3 *2 16.6 17.3 16.2 14.4 2.2 Mean total dry weight per plant (g) . T. 0*359 0.406 0.459 0.418 0 .11 7 *2 0.921 0.781 0.743 0.547 Q.223 Mean leaf area per plant (cm2) t 42.1 47.8 53.1 48.5 12.9 t 2 90.2 64.6 62.6 61.4 17.9 K. ivorensis Plant indices A Treatment B C D least sign diff. (P = 0.05) Mean height per plant (cm) t 10.1 12.4 10.4 9.7 2.3 t 2 11.3 13.3 11.0 9.9 2.0 Mean total dry weight per plant (g) t 0.205 0.205 0.209 0.201 0.104 t 2 0.370 0.312 0.202 0.236 0.176 Mean leaf area per plant (cm2) t]_ *2 32.1 61.0 31.9 45.7 32.5 35.1 31.1 34.1 16.4 25.2 University of Ghana http://ugspace.ug.edu.gh 59 In appendix 2, Table iii, is shown the breakdown of the total dry weights at the initial and final harvests into the weights for the main plant organs. Table 7 and Pig. 8 show the derived growth data and the other ratios presented as percentages of the values in Treatment A, Total dry weights of both species decreased with decreasing solution osmotic potential with those of K. senegalensis decreasing less than those of K. ivorensis. The relative growth rates and net assimila­ tion rates of K. senegalensis in this experiment were comparatively higher (maximum values are 0*31 s/gjweek and 29.2 g/m /wk respectively) than in the soil experiment, K. ivorensis on the other hand showed lower values here (maximum values being about 0,18 g/g/wk and 1 1 .3 g/m /wk respectively) than in the soil experiment. In spite of these lower values for K. ivorensis. the rates on the Tiiole suggest that growth conditions were again relatively good, (i) Height growth J)’or both species changes in seedling heights over the experimental period were small (Pig,6), The intial heights of seedlings assigned to the various treatments were variable with those of Treatment D being the lowest (Table 5)• Differences between treatments were not significant for K. senegalensis seedlings while for K. ivorensis mean seedling height in Treatment B vsas observed to be significantly different from that of seedlings in i reatxaent D. idl’ferences between these and the other treatmen were however not significant. Changes in seedling heights over the experimental period both within and between the species appeared University of Ghana http://ugspace.ug.edu.gh MEM HEIGHT OF KHAYA SESEGALENSIS (KS) M D K. IVOHMSIS (El) SEEDLINGS GR0V3N IN CULTURE SOLUTEJNS OP DIFFERENT OSMOTIC POTE'TTIALS (BARS). TREATMENTS: A (-0.3); B (-2.8); C (-5.3); D (-10.3). TREAMENT A, CONSISTED OP CULTURE SOLUTION ALONE} B, C 1 D D CONSISTED OP CULTURE SOLUTION PLUS POLYETHYLENE GLYCOL. PIG. 6 University of Ghana http://ugspace.ug.edu.gh |o* i University of Ghana http://ugspace.ug.edu.gh '0* 1 University of Ghana http://ugspace.ug.edu.gh 61 similar. There was however a significant difference between heights of seedlings in Treatments B end D for K. senegalensis at the end of the experiment. The difference observed for K. ivorensis seedlings in Treatments B and D could be attributed to the initial difference between these seedlings. There was no marked difference in rate of height increase between seedlings under these treatments. (ii) Changes in leaf area Initial leaf area of seedlings assigned to the various treatments were similar within the species (Table 5). Leaf area of K. senegalensis seedlings were generally larger than those of K. ivorensis. Leaf development in both species was stimulated by Treatment A, so that -there was a rapid increase in leaf area throughout the experimental period (Fig. 6). This effect was comparatively more pronounced in it. senegalensis than in K. ivorensis. Leaf area increase was however affected when stress was imposed so that at final harvest (see Table 5) leaf area in both species, for plants in Treatment A was greater than that for plants in the other treatments. The difference between Treatments A and B plants of K. ivorensis was however not significant; that between plants in Treatments B, C and D was also not significant for either species. (iii) Relative yxowth rate Relative growth rate decreased significantly with decreasing osmotxc potential of the root medium (Table 6). Negative rates of growth were recorded for some seedlings of K. ivorensis in Treatments C University of Ghana http://ugspace.ug.edu.gh FIG. 7 LEAF AHEA OF KHAYA SMEGALENSIS (ivS) AND K. IVORMSIS (Kt) SEEDLINGS GROWN IN CULTURE SOLUTIONS OP DIFFEBMT OSMOTIC POTH'TTIALS (BARS). TREATMENTS A (-0.3); B (-2.8); C (-5.3); D (-10.3). TREATMENT A CONSISTED OP CULTURE SOLUTION ALONE: B, C AND D, CONSISTED OP CULTURE SOLUTION PLUS POLYETHYLENE GLYCOL. University of Ghana http://ugspace.ug.edu.gh L E A F A R E A (C M * ) 62 FIG. 7 University of Ghana http://ugspace.ug.edu.gh 63 Derived growth data and ratios indicating distribution of dxy matter between main organs at final harvest, for Khaya senegalensis and K. ivorensis seedlings grown in culture solutions of different osmotic potentials (bars). A (-0.3); B (-2.8); C (-5.3), D (-10.3). Treatment A consisted of culture solution alone; B, C and D consisted of culture solution plus polyethylene glycol. K. senegalensis Table 6 Plant indices A Treatment B C D least sign diff. (P = 0.05) Relative growth rate (g/g/wk) 0.310 0.212 0.163 0.083 0.009 Net assimilation rate (g/m /wk) 29.2 23.5 16.8 9.1 6.7 lean leaf area ratio (cm /g) 107.8 100.9 100.9 106.6 14.6 Leaf area ratio (cm2/g) 98.0 84.0 86.0 97.0 14.6 Specific leaf area (cm /g) 225.0 221.0 215.0 210.0 28.2 Leaf weight ratio {%) 44.0 38.0 40.0 46.0 4.5 Stem weight ratio (/?) 24.0 28.0 29.0 27.0 4.5 Soot weight ratio {%) 32.0 34.0 31.0 28.0 7.3 Root shoot ratio 02) 51.0 53.0 47.0 40.0 11.3 (continued next page) University of Ghana http://ugspace.ug.edu.gh 64 Table 6 (continued) K. ivorensis Plant indices A Treatments 0 B C D least sign cliff. (P = 0.05) Relative growth rate (a/a/;*) 0.184 0.134 0.025 0.045 0.08 Net assimilation rate (g/m /wk) 11.3 9.3 1 .8 3.3 5.4 Mean leaf* area ratio (cm2/g) 157.7 151.0 159.9 157.5 39.6 Leaf area ratio (cm2/g) 171.0 146.0 165.0 160.0 39.6 Specific leaf area (cm2/s) 311.0 318.0 320.0 317.0 71.1 Leaf weight ratio (%) 55.0 46.0 52.0 50.0 4*9 Stem weight ratio {%) 25.0 31.0 29.0 25.0 5.1 Root weight ratio {%) 20.0 23.0 19.0 25.0 5.8 Hoot shoot ratio {%) 26.0 30.0 24.0 33.0 6.8 University of Ghana http://ugspace.ug.edu.gh 65 and D. Fig. 8 suggests that K. senegalensis was more sensitive to moderate moisture stress (up to about -2.8 bars) than K. ivorensis. so that around this point relative growth rate of K. senegalensis was significantly reduced to 68% of -that in Treatment A, vrtiile that of K. ivorensis was still about 73?® of the value in Treatment A. However K. ivorensis showed greater sensitivity to moisture stress beyond -2.8 bars so iiiat at 5 .3 bars the rate was significantly reduced to 12# of that in Treatment A while the rate for II. senegalensis was reduced only to 53%. For K. ivorensis the percentage was however slightly raised to 21& in Treatment D (-10.3 bars) but this was still slightly lower than the percentage reduction for K. senegalensis which was about 27% at this point. The slight increase for K. ivorensis was accounted for by one large value out of six in ttie sample. Without this value which was 0.149 &/g/iak, the mean relative growth rate for K. ivorensis at 10.3 bars would be 0.024 g/g/wk vhich is about 13% of the rate in Treatment A. (iv) Net assimilation rate Net assimilation rates (Table 6) were higher for K. senegalensis than for K. ivorensis. The responses of net assimilation rate to decreasing osmotic potential for both species followed the same tread (Fig. 8) as was encountered with relative growth rate. It is remarkable that there was no significant difference between the net assimilation rates of seedlings of K. ivorensis subjected to Treatments A (-0.3 bars), and B (-2.8 bars). This further indicates the tolerance of this species to low moisture stress (cf. p. 51 ). University of Ghana http://ugspace.ug.edu.gh THE RELATION BETWEEN RELATIVE GROWTH RATE (R.&.R.), NET ASSIMILATION RATE (N.A.R, ) OR MEAN LEAP AREA RflTIQ (M.L.A.R.) AND MEAN WATER POTENTIAL OE THE BOOT MEDIUM FOR KHAYA SME&AMSIS ( & ) AND K. ITORMSIS ( 0 ) SEEDLING-S. THE DATA ARE EXPRESSED AS PERCENTAGES OF THE VALUE M TREATMENT A. FIG-. 8 University of Ghana http://ugspace.ug.edu.gh ME AN LE AF NE T AS SI MI LA TI ON RE LA TI VE GR OW TH AR EA RA TI O RA TE RA TE (o /o ) (° /o ) (° /o ) 66 (RGR) (N AR) (MLAR) WATER POTENTIAL(BARS) University of Ghana http://ugspace.ug.edu.gh 67 (v) Mean leaf area ratio K. ivorensis had higher mean leaf area ratios than K. senegalensis (Table 6). Within the species, there appear to be no significant treatment effects on this ratio. Thus the reductions in relative growth rates could not, again, be attributed to changes in mean leaf area ratio. (vi) Leaf area ratio, specific leaf area and leaf weight ratio at final harvest The results presented in Table 6 shows that leaf area ratio, specific leaf area and leaf weight ratio, at final harvest, for Kg ivorensis were greater than those for K. senegalensis. The osmotic potentials imposed apparently had no significant effects on any of these parameters. However specific leaf area for K. senegalensis decreased slightly with decrease in osmotic potential. (vii) Root weight ratio, root-shoot ratio and stem weight ratio. For both species the treatment effect did not seem to have followed a particular trend (Table 6). However for K. senegalensis the difference between root/shoot ratios of Treatments A and D was significant. K. ivorensis had the greatest root development in Treatment D and poor development in Treatment C, but there appears to be no marked difference in the root weight ratios of Treatment A and B plants. Stem weight ratio was also not greatly affected by the stresses imposed on the seedlings of both species. University of Ghana http://ugspace.ug.edu.gh 68 Ratios indicating distribution of dry matter between main organs at final harvest, for Khaya