UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of light conditions from patchy natural canopies on the growth and morphology of white clover… Marcuvitz, Sheldon 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-953624.pdf [ 4.29MB ]
JSON: 831-1.0088213.json
JSON-LD: 831-1.0088213-ld.json
RDF/XML (Pretty): 831-1.0088213-rdf.xml
RDF/JSON: 831-1.0088213-rdf.json
Turtle: 831-1.0088213-turtle.txt
N-Triples: 831-1.0088213-rdf-ntriples.txt
Original Record: 831-1.0088213-source.json
Full Text

Full Text

THE EFFECTS OF LIGHT CONDITIONS FROM PATCHY NATURAL CANOPIES ON THE GROWTH AND MORPHOLOGY OF WHITE CLOVER CLONES  by SHELDON MARCUVITZ B.A., Reed College, 1985  A THESIS SUBMITTED IN PARTIAL FULFiLLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Botany)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA September 1994 © Sheldon Marcuvitz, 1994  the requirements for an advanced In presenting this thesis in partial fulfilment of agree that the Library shall make it degree at the University of British Columbia, I er agree that permission for extensive freely available for reference and study. I furth oses may be granted by the head of my copying of this thesis for scholarly purp tives. It is understood that copying or department or by his or her representa not be allowed without my written publication of this thesis for financial gain shall permission.  (Signature)  Department of  /?lzr(f  The University of British Columbia Vancouver, Canada  Date  DE.6 (2)88)  ),1?’1’1  11  ABSTRACT In a community such as a pasture, the success of an individual plant might be affected by neighboring plants. One influence of neighbors on the environment of a plant is alteration of light conditions. Canopies in natural environments have only recently been recognized as heterogeneous in this regard, and responses of plants to alterations in light conditions under heterogeneous canopies are poorly understood. In this investigation, heterogeneity in canopy conditions experienced by white clover (Trifolium repens L.) in a pasture was acknowledged. Three natural canopy configurations were used to examine whether light conditions established by neighbors alter the growth and morphology of white clover clones. Live grass neighbors were used and at the same time, over one clover plant, patches of different light quality and/or quantity were provided. In the first set of experiments, light reflected from grass neighbors was provided simultaneously with direct light. There were. no consistent effects on white clover growth and morphology, but there was evidence of phototropic movement of plant structures, which became located in positions that might minimize the effects of grasses on more permanent features of the clones. In the second set of experiments, shade from three different species of grass was presented to different clover clones for parts of each day, with full sun around noon. The canopies reduced overall growth and branching of clones, while increasing length of, and biomass allocation to, petioles. Lolium perenne had different effects compared to Holcus lanatus or Dactylis glomerata, but between the latter two species, no differences were detected. In the final set of experiments, two interconnected portions of white clover clones, the apical and basal regions of a primary stolon, were subjected to local canopies. Basal region response was examined for independence from apical region conditions and vice  111  versa. Basal regions responded to apical conditions only when they were themselves shaded, while apical regions responded to basal conditions regardless of their local illumination. The ways in which plants respond to neighboring vegetation are complex, and difficulty exists in interpreting plant morphology in terms that are ecologically relevant. If we could identify advantages to particular strategies or responses, and then selectively control plant performance, we might be able to improve our use of plants through more efficient production and management schemes.  iv  TABLE OF CONTENTS  Abstract  ii  Table of Contents  iv  List of Tables  vi  List of Figures  vii  List of Plates  viii  Acknowledgement  ix  Chapter One. GENERAL INTRODUCTION  1  Chapter Two. GENERAL METHODS  28  2.1. Species used  29  2.2. Modular canopy design  37  2.3. Experimental chambers  44  2.4. Measurement of white clover  48  2.5. Data analysis  52  2.6. Microclimatic measurements  54  Chapter Three.  DOES LIGHT REFLECTED FROM NEIGHBORS (NORTHERN CANOPY) AFFECT THE GROWTH AND MORPHOLOGY OF WHITE CLOVER CLONES?  3.1. Introduction  59  3.2. Methods  60  3.3. Results  65  3.4. Discussion  70  V  Chapter Four. THE EFFECTS OF PARTIAL SHADE FROM THREE SPECIES OF GRASS NEIGHBORS ON THE GROWTH AND MORPHOLOGY OF WHITE CLOVER CLONES 4.1. Introduction  90  4.2. Methods  92  4.3. Results  94  4.4. Discussion  103  Chapter Five. IS THERE AN EFFECT OF REMOTE CANOPY CONDITIONS ON THE GROWTH AND MORPHOLOGY OF LOCAL REGIONS OF WHITE CLOVER CLONES?  Chapter Six  5.1. Introduction  128  5.2. Methods  130  5.3. Results  134  5.4. Microclimatic measurements  137  5.5. Summary and discussion  139  GENERAL DISCUSSION  155  Literature Cited  164  vi LIST OF TABLES  Table  3.1  77  3.2  81  3.3  81  3.4  82  3.5  82  3.6  83,84  3.7  86  4.1  115  4.2  115  4.3  116  4.4  116  4.5  120  4.6  121  4.7  122  4.8  124  4.9  125  4.10  126  4.11  127  5.1  146  5.2  149  5.3  150  5.4  151  5.5  152  5.6  154  vii LIST OF FIGURES  Figure  1.1  27  2.1  56  2.2  56  2.3  57,58  3.1  80  3.2  85  3.3  87  3.4  87  4.1  111  4.2  112, 113  4.3  114  4.4  117, 118  4.5  119  4.6  123  5.1  153  viii  LIST OF PLATES  Plate  1. 78 2  88  3  109  4  144  5  147  ix ACKNOWLEDGEMENTS  I’d like to thank my supervisor, Roy Turkington, for the help on many aspects of this thesis. Thanks need to be given to the many friends who offered help at the critical final harvests of the experiments. Pat Harrison deserves thanks, also, for the many calls to Plant Operations, trying to get help in repairing aspects of the greenhouse. Carole greatly helped me wrap this up. I’d also like to thank all of the people who helped me maintain a perspective on this thesis, whether through agricultural and botanical discussions, or through discussions over beer at the Grad Center.  1 Chapter One GENERAL INTRODUCTION 1.1.  BACKGROUND ON Trifolium repens L.  1.1.1. IMPORTANCE OF TifiS PASTURE LEGUME Plants experience different problems capturing resources than do animals. Plants are stationary and can not move around on the scale that animals do. However, plants have remarkable abilities to modify their growth and morphology in order to capture resources. Plants with creeping growth habits, i.e., clonal plants with plagiotropic stem growth and rhizomatous and stoloniferous plants, have been called “foragers” as they wander through heterogeneous resource patches (Bell, 1984).  One of the most  agronomically important of these foraging plants is white clover (Trifolium repens L.), a pasture legume of mild temperate regions (Fig. 1.1). As a nitrogen fixing perennial, clover can continuously supply a grass sward with up to 60kg nitrogen/ha/yr and thereby indirectly provide protein grazing animals. White clover is primarily grown in mild temperate regions, where it can overwinter with enough viable stolon material to continue growth in the next season (Harris et al., 1983). Even in these areas, its sensitivity to winter conditions makes its performance unpredictable from a farmer’s perspective, and its persistence requires pasture management in the form of controlled stocking or mowing (Frame and Newbould, 1986). With increasingly difficult economic times, unpredictability makes its use somewhat problematic, although at the same time the increased use of legumes is warranted to reduce the use of artificial nitrogen fertilizer on farmland. It seems logical, then, that a better physiological understanding of the interactions between white clover and other plants growing in a mixed sward should increase our ability to use white clover efficiently in pastures. The biology, ecology and agronomy of the species have been  2 reviewed by Erith (1924); Ahlgren and Fuelleman (1950); Chestnutt and Lowe (1970); Burdon (1983); Turkington and Burdon (1983); and Frame and Newbould (1986).  1.1.2. MANAGEMENT AND INTERACTIONS WITH NEIGHBORS In most plant communities it is the proximity of neighboring plants that largely determine an individual’s success.  Resources acquired by one plant are no longer  available to others. Because different species have slightly different patterns of resource use, and because resources are patchily distributed, multispecies communities can persist (Grime, 1979; Tilman, 1988), and the composition of multispecies mixtures changes under different environmental conditions (e.g. Pickett and Bazzaz, 1976). In pastures, the management regime often precisely determines the composition of species mixtures (Jones, 1933). For example, white clover persists because grazing or mowing maintains the sward at a height which prevents its elimination (Frame and Newbould, 1986). In taller grass, white clover growth is greatly inhibited, and it contributes little to the total yield of the pasture. The importance of white clover as a forage crop has led to management regimes that strive to maintain a particular proportion of it in a pasture. As white clover grows through a sward, it encounters heterogeneity from two sources, and it responds phenotypically to this heterogeneity. First, it encounters various species of plants, particularly grasses. The ability to respond in a specific manner to different species of grass (Chestnutt and Lowe, 1970) is a maj or factor contributing to white clover persistence in pastures (Bulow-Olsen et al., 1984). Secondly, it encounters environmental variability caused by local soil conditions, dung and urine patches, mole hills, and death of grass clumps (Parish, 1987). Turkington and Harper (1979), Burdon (1980), and Aarssen and Turkington (1985) have argued that the first component (the neighbors) has a much greater influence on success than the second component (abiotic factors). To continue growth and gathering of resources, the plant may need to modify its phenotype as it continually encounters the patchiness in its environment.  3  White clover shows rather consistent patterns of growth when associated with particular pasture conditions. When growing in a patch in the absence of neighbors, white clover will grow large, have more branches and have short petioles. These and various other characteristics of white clover such as the percentage of nodes with branches, internode length, stolon growth rate, and leaf size show consistent patterns of response when growing with different grass species (Chestnutt and Lowe, 1970; Turkington, 1979; Solangaarachchi, 1985). Turkington et al. (1991) compared two halves (branches on either side of a primary stolon) of the same white clover plant when the two halves were growing with monocultures of different grass species. They showed that white clover had many small modules when growing with Agrostis capillaris L. (bent grass) fewer, larger modules with longer internodes when growing with Holcus lanatus L. (velvet grass), and large modules on short internodes when growing with Lolium perenne L. (ryegrass). In general, white clover is found to be more productive when growing in association with ryegrass than with any other companion grass (Turkington and Burdon, 1983). While the different responses of white clover to grasses were quite marked, it is not clear what component of the physical environment the grasses alter, although it has been variously attributed to physical prevention of rooting (Solangaarachchi, 1985), soil microbial populations (Turkington et al., 1988), or light quantity and quality (Solangaarachchi, 1985; Solangaarachchi and Harper, 1987; Thompson and Harper, 1988; Thompson, 1993a). Besides differences induced by different species of grasses, different genotypes of a single species, ryegrass, induced different morphologies in a particular white clover clone (Hill, 1977; Aarssen and Turkington, 1985). Mixtures of white clover and ryegrass taken from the same site of origin can result in higher yields than mixtures of plants derived from different sites of origin (Turkington and Harper, 1979; Frame and  4 Newbould, 1986). This can result in widely different yields of white clover in apparently similar white clover/ryegrass mixtures. Evans and Turkington (1988) showed that the morphological effects induced in white clover by different grasses can persist in a common garden, in the absence of neighbors, for at least four months, but disappeared within two years. This carry-over effect indicates that a particular morphology may be “programmed” into white clover plants by their biotic environment, and internal controls might restrain phenoptypic changes during short-term changes in conditions. It also demonstrates that the conditions under which a white clover plant previously grew are important in determining that plant’s response to a newly encountered stimulus.  1.1.3. CLONAL GROWTH  PHENOTYPIC PLASTICITY As a clonal plant, white clover has a modular design whereby each module, or  ramet, has the potential for independent existence and proliferation. This clonal growth habit enables lateral spread, placing daughter ramets at a distance from the parent (Cook, 1983). The ramets are constructed of resource-acquiring structures (leaves, roots) and spacing-associated structures (internodes). The selective placement of resource-acquiring structures through differential elongation and branching of stolons in response to environmental heterogeneity is interpreted as “foraging behavior” in plants, and this term has gained wide acceptance in recent years (e.g. Sutherland and Stillman, 1988; Silvertown and Gordon, 1989; Hutchings and Mogie, 1990; Grime, 1994; Hutchings and de Kroon, 1994). The dynamics of white clover in a sward are determined largely by vegetative growth, rather than recruitment from seeds (Turkington et al., 1979; Chapman, 1983).  5 The ability of white clover to persist in pastures shows that it is capable of gathering sufficient resources even in these heterogeneous and competitive environments. Arguably, one of the main characteristics conferring this success is its phenotypic plasticity, the ability to make morphological and physiological adjustments. (Bradshaw, 1965;  Schlichting, 1986; Sultan, 1987; Grime, 1994; Hutchings and de Kroon, 1994).  The “form” that a plant takes under particular conditions might allow for its most efficient use of resources (Lovell and Lovell, 1985). Indeed, flexibility in morphology and physiology is a character under genetic control, and thus is expected to be affected by selection and evolutionary change (Schlichting 1986; Thompson, 1991; Bell and Lechowicz, 1994). In white clover, plasticity in response to phosphorus supply was recently observed to vary among genotypes, indicating that plasticity is a trait that can be bred (Caradus et al., 1993). White clover provides a good system in which to study the inheritance of plasticity. In a well-studied permanent pasture in North Wales (around 100 years since establishment), an individual white clover plant may co-exist with a single neighboring grass species for several generations because grass patches may be up to lOOm 2 (Turkington and Harper, 1979; Thorhallsdotir, 1983; Turkington and Mehrhoff, 1986). In a younger permanent pasture (50 years since establishment) in British Columbia, a more complex mosaic of neighboring grass species, with patches up to 1 m , means that a 2 growing white clover may experience many different grass neighborhoods in a season (Aarssen and Turkington, 1985; Evans, 1986; Parish, 1987). It has been suggested that the genotypes of white clover persisting in the permanent pasture in N. Wales have been screened (presumably through initial mortality of genotypes arriving as seeds, or, since there is occasional seed-set, through generational-selection) for their ability to adopt a form allowing for continued existence with one particular species of grass (Turkington and Harper, 1979; Gliddon and Trathan, 1985). In the B.C. pasture with a more fine grained mosaic of neighborhoods, the same type of screening could be expected to occur,  6 except that it would be for the ability to be flexible in a heterogeneous environment, because particular morphologies are more productive with particular grass species (Turkington and Mehrhoff, 1990).  PHYSIOLOGICAL INTEGRATION The extent to which connected ramets of a plant respond as an integrated unit, as  opposed to independent units, has been debated for some time (reviewed by Pitelka and Ashmun, 1985; Marshall, 1990; Hutchings and de Kroon, 1994). In some clonal species, the connections between ramets decay soon after establishment, while in others, the connections may persist throughout a growing season or for many years, with the ramets remaining attached and perhaps more or less physiologically integrated, either as a whole plant, or in groups of ramets defined as “integrated physiological units” (IPU’s sensu Watson, 1986). There seems to be variation in the extent of integration depending on the species examined, the architecture of the plant, and the conditions to which plants are subject. It appears that physiological integration can occur through carbohydrate (Price and Hutchings, 1992; Chapman et al., 1992a; Landa et al., 1992), water (Alpert, 1990; Evans and Whitney, 1992; Price et al., 1992) or mineral (Alpert, 1991; Marshall and Anderson-Taylor, 1992; Caradus et al., 1993) transfer, and possibly through signals, hormonal (Dong, 1993), or physiological (Hartnett and Bazzaz, 1983), transmitted between interconnected regions, although still no clear picture exists (Hutchings and de Kroon, 1994). It seems that many plants have the ability to regulate their sharing of resources between ramets, rather than simply sharing through diffusion from higher to lower concentrations (Caraco and Kelly, 1991), but the effects of physiological integration on an ecological and evolutionary scale remain largely unknown. There have been many arguments put forth for the advantages of each plant’s particular strategy (Pitelka and Ashmun, 1985; Caraco and Kelly, 1991; Hutchings and  7 Price, 1993). 1 Most of the attention in this area has focused on clonal, rhizomatous or stoloniferous plants and their ability to alleviate the effects of locally poor resource conditions through physiological integration. In many. of the species tested, ramets growing in locally poor conditions were supported by interconnected ramets growing in more favorable conditions (Hartnett and Bazzaz, 1983, in Solidago canadensis L.; Slade and Hutchings, 1987a, in Glechoma hederacea L.; Lau and Young, 1988, in Lycopodium flabelliforme (Fernald); Evans, 1991, in Hydrocotyle banariensis (Lam.); Alpert, 1991 in Fragaria chiloensis (L.) Duchesne; Evans and Whitney, 1992 in Hydrocotyle banariensis). It appears that temporary growth in poor conditions can be sustained through support from ramets growing in better conditions, and in some plants this may be a sound strategy, since by this time, an investment in stolon growth and ramet production has already been made by the plant. It would be advantageous to support the poorer ramets for a while, either to allow their escape from the poor conditions or to allow better mobilization of carbohydrate and nutrients back to the richer portion of the plant. Another possible effect of physiological integration between ramets is to intensify the local responses of ramets, i.e. ramets under poorer conditions would perhaps do worse, and ramets under richer conditions would do better than they would have done otherwise without this integration (Abrahamson et al., 1991; Hutchings and Price, 1993). This effect of physiological integration was observed in Lolium perenne in response to locally severe shading (Ong and Marshall, 1979), to some extent in nitrogen-limited ramets of Fragaria chiloensis to shared (translocated) nitrogen (Alpert, 1991), nutrient-starved  1 In this context, ‘advantages’ of one growth strategy over another implies a comparison, and the comparisons being made are not always clear. Many authors seem to be trying to synthesize a summary about the global strategies of plants regardless of the community structure. With respect to physiological integration, there is often no distinction made between comparisons at this level and where individuals in one population vary in the degree to which they integrate or fail to integrate local conditions. Caution needs to be exercised, since comparisons made between different species (unless having very similar growth form and found in very similar environments) have little relevance to their relative success or strategies, even in the same community. Comparisons of this type should be restricted to different individuals in a population or to comparisons between ancestral and modem lineages where relative success can be judged.  8 Solidago canadensis (Abrahamson et al., 1991), and in Carex bigelowii (Torr. Ex Schweinitz) to chronically carbon-poor tillers (Jonsdottir and Callaghan, 1989). Physiological integration between ramets, however, is not always detected. There are examples of a lack of integration in response to locally depleted light conditions, in Lamiastrum galeobdolon (L.) Ehrend. and Polatschek. (Dong, 1993) and in response to local competition from grass in Glechoma hederacea (Hutchings and Price, 1993). This may depend largely on the characteristics measured, since in Kemball et al. (1992) physiological integration in white clover due to local shading was not detected in morphological measurements, but radiolabelled 14 C export from the shaded branch was reduced. The designs of many of these experiments examining physiological integration make use of the concept that the phenotype (e.g., petiole and internode lengths, stolon branching, physiological functioning) of localized portions of several clones growing under identical conditions, would be indistinguishable if there were no dependence on connected ramets living in remote conditions. Phenotypic variation detectable between corresponding portions growing under identical conditions is interpreted as evidence for the existence of physiological integration. One requirement of this type of analysis, is that the portions of the plant are otherwise independent, i.e. the conditions experienced by one portion do not directly affect the remote portion, and it is unclear whether in many of the previous experiments this requirement was met.  Experiments involving local  differences in light conditions may need to pay special attention to this (see below). In many of the experiments described above, the locally-poor conditions were imposed artificially and abruptly by the researchers. Under natural conditions resources are likely to vary in a much more continuous way with different intensities of gradients in different places. So it might be expected that the response of the portion of a plant growing in rich conditions to connected ramets would differ depending on the degree to  9 which the locally-poor ramets are depleted. If this were so, the use of artificially depleted conditions might evoke responses quite different from those in the field. The use of more naturally-depleted conditions would greatly contribute to our understanding of inter ramet integration in the field. In a field situation, also, locally-poor conditions are more likely to develop gradually, with the poorer ramets remaining under nearly ideal conditions, at least for a while. For example, as a clonal plant approaches lower resource patches, e.g. lower light patches, a gradient might become established within the plant, where the ramets nearer the patch of low resource are under slightly poorer conditions than the rest of the plant. Under these conditions it might be advantageous for the richer ramets with a more rapid growth rate to limit their support of these poorer ramets with a slower growth rate. There has been little experimental work enabling an evaluation of the plausibility or generality of this scenario under natural conditions, although suggestions by de Kroon and Schieving (1990), Abrahamson et al. (1991), and Hutchings and Price (1993) indicate that in these situations, a lack of integration could lead to greater foraging efficiency. In a heterogeneous environment, it would also be advantageous to avoid these lower resource patches in the first place, and it remains unknown how the ability to integrate conditions and respond accordingly might be involved in an early avoidance response which may prevent resources from actually being limited. One of the potential benefits of physiological integration in white clover is that it permits new daughter ramets to be supported through their establishment phase by receiving photoassimilate from the parent or from older established ramets (Harvey, 1970; Ryle et al., 1981). In white clover, there are usually several branches growing as one physically interconnected unit (IPU). Characteristic sinks for carbohydrate produced in photosynthesis are stolon apices, developing buds, expanding leaves, root apices, root nodules, and flowers (Harvey, 1970, 1979; Chapman et al., 1992a,b). On a single stolon, it was suggested that all leaves acted as sources for all of the sinks, indicating that bidirectional transport was possible in these plants. As plants aged, however, there was  10 reduced transport into older portions from the newer growth (Harvey, 1970). Developing secondary stolons continued to import carbohydrate even after producing several of their own leaves, indicating a close connection between the main stolon and daughter ramets produced on branches.  The same pattern was observed by Kemball et al. (1992)  regardless of the illumination status of the branches themselves. They did notice, however, a reduction in the export of assimilate from a labelled branch if it was shaded. Chapman et al. (1991a,b) showed that carbohydrate exported from active leaves went primarily to nearby sinks, but labelled 14 C was found in distant sinks, in both basipetal and acropetal directions from the leaf. Their observations led them to conclude that relative sink strength as well as the distance to the sinks determine the patterns of carbohydrate movement in white clover.  They also observed that reserves of  carbohydrate could be mobilized in response to stress such as defoliation, which eliminates much of the active carbohydrate production. Phosphorus translocation in white clover has been recently studied (Chapman and Hay, 1993), and these authors concluded that the level of phosphorus transported from sources varied depending on the overall phosphorus supply. The strongest sink for phosphorus appeared to be the branch arising at the node of an actively fed root. Continued translocation from main stolon roots to branch ramets allows the branches to enhance their development of leaves while relying on the parent stolon for mineral nutrition before the ramets on the branch stolon were able to initiate their own roots. Chapman and Hay (1993) concluded that phosphorus translocation patterns were not restrained to acropetal movement, as was previously suggested, but rather, were determined by the structure of the plants and the location with respect to principal sinks. In an examination of intra-plant integration under slightly more natural conditions, Turkington et al. (1991) grew clover plants with ramets arising on opposite sides of the primary stolon directed to grow in different neighborhoods, formed by different species of grass. The morphology of individual ramets responded to their  11 immediately neighboring grass independently, while the rate of new ramet production was constant over the entire plant, even though this differed under different neighborpairs. This suggests that the responses of some phenotypic characters vary locally, whereas others are integrated within the whole plant. In this case, the selective plasticity of some characteristics enables white clover to explore preferentially, favoring resourcerich patches but continuing to produce new ramets, regardless of local conditions. It was also suggested that since the development of individual ramets (leaves, internodes) depended on the acquisitions of resources locally, and that the development of an apex requires resources from several interconnected ramets (Harvey, 1979; Newton, 1986), their data agree with what is known about the nutrition of ramets and apices in white clover. In summary, patterns of translocation within the plant provide detailed description of short-term transport in plants, but information on the long-term effects on the fitness and survival of ramets is needed when speculating about the ecological importance of physiological integration. In addition, experimental conditions unlike those experienced in the field might yield information on mechanisms involved in establishment and maintenance of integration within a plant, but might do little to increase understanding of its ecological importance.  1.2.  MECHANISMS INVOLVED IN PLANT/PLANT INTERACTIONS  1.2.1. IRRADIATION AND PHOTOMORPHOGENESIS The extent of spatial heterogeneity in pastures, shown to exist for both biotic and abiotic factors, is reflected in the heterogeneous structure of the pasture canopy. Arguably, some of the most important changes associated with different canopies are in the light conditions. A clonal plant’s ability to respond to changes in surroundings may be essential to its growth and survival within a heterogeneous canopy. The effects of  12 canopies found in the field on light conditions are very complex, and until recently speculation about the eco-physiological effects of light on plants was difficult, since the majority of studies were conducted under relatively uniform and artificial light conditions (e.g. using cheese-cloth, colored filters, or brief pulses of monochromatic light). Following is (i) a description of what is known about the effects of light beneath relatively uniform canopies on the growth of plants, and (ii) the consequences of heterogeneity, or patchiness, in a canopy such as a pasture, on the light conditions at (a) one location on a plant, (b) over interconnected portions of a clonal plant, and (c) throughout the life of a plant.  HOMOGENEOUS CANOPIES Above the canopy surface in pastures and other plant communities, irradiance  comes from the sun, either directly in the sun’s beam, or indirectly from the clear sky. Diffuse radiation from the sky has a much lower photon flux density (PFD) than direct radiation from the sun, and it has a higher proportion of blue wavelengths (Smith, 1982). The spectral energy distribution (SED), or quality, of the light from above a plant in an open situation depends on the proportion of diffuse to direct light. This changes with solar angle and aspect as well as proximity to opaque objects such as rocks or tree trunks (Stoutjesdijk, 1974; Smith, 1982). Under cloudy conditions, diffuse light from clouds may be homogeneous or mixed with direct light from the sun’s beam and diffuse light from clear sky. Below the cover of a plant canopy, much larger changes in the SED occur than above a canopy, even under varying cloud conditions (Holmes, 1981; Smith, 1982). Extrapolation from absorption characteristics of chlorophyll largely describe the spectral changes that occur beneath canopies of plants. There is great attenuation of blue and red wavelengths, as well as most of the energy in other visible wavelengths, but there is far  13 less attenuation of wavelengths longer than 700 nm (Holmes and Smith, 1977a; MacClellan and Frankland, 1985). Light quality in natural systems has most often been characterized by the ratio of photons in the red (R: 655-665 nm) region (greatly attenuated by vegetation) to the farred (FR: 725-735 nm) region (attenuated little by vegetation).  Phytochrome, a  photoreversible pigment having absorption maxima at 660 nm and at 730 nm, is thought to be the pigment responsible for plant response to variations in light quality formed by natural canopies (Smith, 1982; Kasperbauer, 1987; Casal and Smith, 1989a; Smith and Whitelam, 1990; Ballare, 1994).  For reviews of the photochemical properties of  phytochrome, see Kendrick and Kronenberg (1986), Furuya (1987), and Attridge (1990). Above a canopy the R:FR is usually between 1.1 and 1.3, depending on solar elevation, cloud conditions, and water vapor content (Smith, 1982; MacLellan and Frankland, 1985; Lee and Downum, 1991), and this is beyond the range of values that establish differences in the phytochrome photoequilibrium (Smith, 1982). Measurement of light quality beneath natural plant canopies often shows the R:FR ranging from 0.1 to 0.5, and this is within the range where slight differences will cause large changes in the photoequilibrium between the two forms of phytochrome, and hence, alter its activity (Smith and Holmes, 1977).  Short-term monitoring of the R:FR beneath a Corylus avellana (hazelnut)  canopy, however, showed that the R:FR fluctuated between 0.5 and 1.1 many times during a 20-second period, though this was bimodal and most often close to 0.5 (Woodward, 1983).  The bimodal distribution suggests that sunlight penetrates the  canopy intermittently as sunflecks, and brief periods of direct illumination can have profound effects on growth (Morgan and Smith, 1978; Baldocchi and Collineau, 1994; Pearcy et al., 1994). The patterns of growth observed under low R:FR conditions, such as beneath a relatively uniform closed canopy (e.g., beneath a forest canopy or beneath stands of many  14 cultivated crops) are consistent with postulated shade-avoidance responses (Grime, 1981). Etiolation of seedlings emerging from the soil is accentuated by a low R:FR. This promotes elongation towards the surface of the canopy in plants with orthotropic shoot orientation. Internode and petiole extension are also promoted by low R:FR light, branching is inhibited, and biomass partitioning to leaves, stems and roots is altered (Kasperbauer, 1971, 1987; Tucker and Mansfield, 1972; Field and Jackson, 1975; Holmes and Smith, 1977; Morgan, 1981; Casal and Kendrick, 1993). These responses are generally weaker in plants from closed-canopy habitats than in plants from more open habitats (Morgan and Smith, 1979).  In plants with plagiotropic shoot orientation,  responses to uniform low R:FR conditions are similar and include increased petiole elongation and decreased branching in white clover (Solangaarachchi, 1987; Thompson and Harper, 1988; Thompson, 1 993a), increased internode and petiole length, increased leaf area and decreased branching in Lamiastrum galeobdolon (Dong, 1993), and reduced leaf weight ratio (leaf weight to overall plant weight) and increased stem and petiole weight ratios in Veronica sp. (Dale and Causton, 1992). In white clover, conflicting results have been obtained on the effects of low R:FR on internode length, and this is apparently due to the differences in PFD from 400-700 nm (photosynthetic photon flux density, PPFD) in these experiments (Thompson, 1993a). All of the above responses presumably serve to increase the chances that the plant will continue to intercept light efficiently (Ballare, 1994). In several of the more recent experiments, fresh or living plant material was used to filter the light and create low R:FR conditions. Use of these natural filters precludes the need for assumptions about responsible wavelengths. Other wavelengths have been shown to elicit similar responses in plants. There have been suggestions that decreases in UV-B can lead to increased internode elongation (Barnes et al., 1990; Ballare et al., 1991). There is also a specific bluefUV-A photoreceptor which can detect decreases in these wavelengths and promote internode elongation in seedlings (Gaba and Black, 1979) and in fully de-etiolated soybean plants (Britz, 1990). Distinct  15 groups of wavelengths may interact, e.g., blue and FR in photomorphogenic and phototropic responses (Drumm-Herrel and Mohr, 1984; Ballare et al., 1992; Janoudi and Poff, 1992) and green and FR on night-closure of Albizzia julibrissin leaflets (Tanada, 1982). Work on this topic is scarce, so it remains largely unknown how groups of different wavelengths might interact in promoting plant responses to vegetation shade. However, correlations between growth responses under natural closed canopies and those under controlled conditions known to elicit phytochrome responses, such as a brief pulse of red followed by far-red, have led to reasonable confidence implicating phytochrome in shade-avoidance responses under uniform canopy situations.  HETEROGENEOUS CANOPIES Most plant communities present a very complex canopy structure that is far from  uniform. Besides intermittent penetration of sunlight through small spaces between leaves, as detected by Woodward (1983) beneath a complete canopy, there may be larger gaps that allow penetration of sunlight for longer durations. These larger sunflecks can supply a substantial portion of the energy for a planCs photosynthesis (Morgan and Smith, 1978; Baldocchi and Collineau, 1994; Pearcy et al., 1994). A plant within a sunfleck will also receive light from the clear sky that has been transmitted through leaves and in many cases, light that has been reflected off plants forming the boundary of the sunfleck (Anderson and Miller, 1974; Bazzaz and Wayne, 1994). In general, plants beneath a canopy that is not an entirely uniform surface of leaves will receive direct light from the sky or the sun and light that has been transmitted through and reflected off leaves. In recent years several studies (e.g. Casal and Smith, 1988; Woitzik and Mohr, 1988; Rice and Bazzaz, 1989; Ballare et al., 1989, 1990, 1992a; Novoplansky et al., 1990; Casal and Kendrick, 1993; Dong, 1993) have reported progress toward unraveling more complicated questions about how light affects plants growing in the field. These reports indicate how recent approaches have taken a more sophisticated and more  16 accurate view of a plant’s light environment, acknowledging, for example, the existence of heterogeneity, which has several consequences for the way a plant responds to its canopy (Ballare, 1994). In particular, this type of heterogeneity causes different inputs forming the global light environment to be present at one location on a plant, and causes conditions at a given time to be different at two interconnected portions of the plant. Each of these will change throughout a day and can also change in a predictable way over time through the length of a growing season. This provides some theoretical background for hypothesizing how the particular light conditions created by grass neighbors might influence the morphology of white clover.  Consequences at one location within the plant In a community such as a pasture, light received at one location will contain a variable proportion of transmitted, reflected and unfiltered light over time due to biotic (location, identity, and conditions of neighboring vegetation) and abiotic (diurnal and seasonal solar movement, cloud cover) factors. In addition, points in close proximity can have widely different proportions of light from each source at any given time. Both spatial and temporal heterogeneity mean that even though plants might receive light that has a low R:FR, they are likely to receive some unfiltered light as well. Within a gap, a plant may experience these conditions even in direct sunlight. In this case, the source of low R:FR light would be from reflection off upright leaves of neighboring plants (Stern and Donald, 1962; Wooley, 1971; Anderson and Miller, 1974; Bazzaz and Wayne, 1994). Reflection can create strong additions of FR wavelengths to a plant’s overall light regime if the sun’s direct beam is the source of the reflected light (Ballare et al., 1987; Kasperbauer, 1987; Ballare et al., 1989; 1992a; Smith et al., 1990; Casal and Kendrick, 1993). Kasperbauer (1987) suggests that  “...  light reflected from the leaves may be even  more important [than light transmitted through leaves] in adaptation of a plant to competition from other plants.” The increase in FR that is associated with reflection, it is  17 now argued widely, provides an indication of the proximity to neighbors and, hence, an indication of the potential for competition, before competition for the light resource is experienced. The responses to this increase in FR are consistent with shade-avoidance strategies in orthotropic plants, i.e. increased internode elongation and decreased branching. The designs of many recent experiments have achieved a goal stated by Smith (1982) as necessary to further our Understanding of the function of phytochrome in the field. Namely, they have provided differences in the R:FR at irradiation levels that are commonly experienced by many plants growing in the field. There have as yet been few studies satisfying this criterion in plagiotropic plants, and therefore, the effects of reflection from neighbors on plagiotropic plants remain largely unknown. An exception to this is the work by Novoplansky et al. (1990). They found that seedlings of Portulaca oleracea L. preferentially developed branches away from other seedlings, even though they were widely enough spaced to receive nearly full illumination.  Plants also  responded similarly to reflecting barriers that simulated neighbors. These results suggest that plagiotropic plants can respond to increased FR in the proximity of neighboring plants and can selectively control placement of branches to avoid growing towards neighbors.  Consequences at more than one location within the plant The horizontal growth form of white clover presents a situation where heterogeneity in the light environment may occur throughout interconnected parts of a single plant. The extent to which ramets of clonal plants respond independently to local light conditions, as opposed to coordinating their response to whole-plant conditions, has only recently begun to be studied and is considered here (see also above section Each of the following studies is based on the idea that corresponding parts of separate clones can be compared when the interconnected portions of each clone are experiencing different environments.  Hartnett and Bazzaz (1983) examined the rates of  18 photosynthesis, growth and survival of ramets of Solidago canadensis connected through a common parental node to sibling ramets experiencing different conditions. If connected to a ramet that received shade, the rate of photosynthesis in the target ramet was higher than if it was connected to another fully-illuminated ramet. They suggested that a greater assimilate demand on the fully-illuminated ramet was incurred by the shaded ramet, and these ramets responded with a greater maximum rate of photosynthes 2 is. Growth and survival measurements also indicated that shaded ramets were supported with assimilates from the other, fully-illuminated, ramet. These results suggest that translocation could occur in both basipetal and acropetal directions between ramets sharing a common parental node. They hypothesized that in Solidago canadensis physiological integration could alleviate the effects of patch-specific limitations of light, and therefore increase the genet’s survival in a heterogeneous environment. Slade and Hutchings (1987b) showed that unshaded ramets of Glechoma hederacea formed internodes the length of which were not dependent on the light conditions experienced by other parts of the stolon. However, shaded ramets formed shorter internodes if they were connected to basal ramets that were unshaded. This only occurred if the unshaded basal ramets were within two or three nodes of the shaded ramets in question. This suggests that there is some translocation of the effect of unshading in the acropetal direction for at least two nodes in this species. This is consistent with the patterns of labelled assimilate transport observed in this species (Price et aL, 1992). Dong (1993) did not observe the same effect in Lamiastrum galeobdolon, where shaded ramets apical of unshaded basal ramets showed no dependence on the conditions experienced by basal ramets.  Instead, basal ramets  produced shorter internodes if the apical region was under high light conditions. No  However, in this experiment, the conditions used to impose shade on one sibling ramet (a shade tent) 2 could have directly interfered with the light environment under which the other sibling ramet was grown, even though it did not receive its own shade ‘tent’. The details of the experiment are not clear, but it was likely that these “fully-illuminated” ramets (the ones which were connected to shaded sibling raniets) grew under lower illumination than the controls, which were connected to sibling ramets which received no shade “tent”, and this could have altered the way these ramets responded to the photosynthesis measurements.  19 basipetally-transported effect on internode length was seen if the basal ramets were highly illuminated and the apex was shaded. Dong (1993) suggests that this was an effect of hormones, rather than an effect of assimilate transport, because the apex is traditionally considered to be a strong carbohydrate sink (Pitelka and Ashmun, 1985; Marshall, 1990). In white clover, Harvey (1970) found that localized shading of mature leaves reduced to nil the small amount of imported assimilate they would otherwise have received. Young, expanding leaves also imported less when locally shaded, but shading of stolon apices had no effect on their import of assimilate. Solangaarachchi (1985) grew apices of white clover into different neighborhoods of grasses while basal portions remained in the first neighborhood.  She showed that growth in the basal portion was inhibited by a  neighborhood of grass around the stolon apex. She suggested that light conditions experienced by the apical region were responsible for the response of basal branches. Using the same technique on a smaller scale (with three connected ramets) Newton (1986) observed and concluded the same, suggesting that there may be basipetal translocation of the effects of shading which occurs only around the apical portion of a stolon. Thompson (1993b) observed that supplemental Red light (without increasing PPFD) on one node of white clover plants, using light emitting diodes (LED), affected the branching and petiole length of ramets produced acropetally, and this effect was reduced as the distance from the treated node increased. This suggests that the response to R:FR can be communicated acropetally along a stolon in white clover. However, the diodes were illuminated 24 hours per day, and the effect of nighttime supplementation of PPFD could not be determined, nor could the direct effects of scattering of the Red light within and around the plants, especially at night. Kemball et al. (1992) shaded only one basal branch on white clover clones. They found that there was no detectable growth response in other portions of the plant. Shading of the basal branch also did not change its import of 14 C labelled assimilate from the plant’s main stolon. However, labelling of a basal leaf on shaded basal branches indicated an increase in its acropetal transport of  20 assimilate towards its stolon apex and a decrease of assimilate in basal portions of that stolon and in the main stolon of the plant. This indibates that under unshaded conditions, there was both acropetal and basipetal transport of assimilate throughout the plant, and that export of assimilate from a single branch was reduced by localized shading. In summary, patterns of photoassimilate movement within several clonal plants have been observed to change due to local light conditions. However, the directions in which factors allowing for physiological integration (e.g. photoassimilate, hormones, etc.) move do not seem to be consistent, depending on the particular plant (perhaps due to structural limitations (Price et al., 1992)) and on the particular localized conditions. In several cases, ramets under poorer light conditions received support from better-illuminated ramets. In these experiments low fluence rates were often established with artificial filters, so these were examinations of the effects of locally low fluence rates. The effects of locally altered R:FR conditions remains to be examined.  The local differences  associated with neighboring plants may involve changes to R:FR with little change to PPFD. It could be beneficial to the plant to increase growth in the region with a high R:FR and limit growth in the low R:FR region, tperhaps hrough carbohydrate support from the low R:FR region. A requirement in these types of studies is that the separate environments created over two (or more) portions of one plant need to be completely isolated to enable the conclusion that intra-plant integration had an effect on the growth of ramets. In none of the published accounts of the experiments described above is the reader assured that the conditions imposed around one portion of the clonal plant in question do not interfere directly with the conditions experienced by the interconnected portion. The physical arrangement of “shade tents”, and the necessary proximity to the interconnected ramets, makes this requirement logistically difficult to satisfy, yet it is necessary for inferring the existence of intra-plant integration.  Without this it remains uncertain how much of the  response in proximal portions was due to conditions around the apex, and how much was  21 due to each ramet’s own local environment, which was itself altered by treatments on stolon apices. This requirement might be particularly stringent in experiments involving manipulations of light conditions, however, it can not be ignored in any experiments requiring independent manipulations of two portions of one plant.  Temporal variation in light conditions Great complexity is added to understanding photomorphogenesis in the field when considering the temporal variation in the light environment. Plant response in the field may be closely linked to changes in the light conditions occurring over many different time scales from seconds to the life of the plant, a topic recently reviewed by Baldocchi and Collineau (1994), Pearcy et al. (1994), and Pearcy and Sims (1994). McLaren and Smith (1978) showed that petioles of Rumex obtusifolius (L.) elongated more if the R:FR decreased over time than if the R:FR remained unchanged at the lowest value throughout. A decrease in R:FR (without much decrease in PPFD) under natural conditions would describe a canopy which was once open, but is now beginning to close, and elongation of partially-shaded plants could result in them overtopping the canopy. On the other hand, in a canopy that already presents a low R:FR from early in growth, there may be little chance for a plant beneath it to overtop the canopy. This suggests that an increase in reflection off vegetation over time could be involved in white clover shade avoidance, and there might be an ecological advantage, since it would indicate decreasing distance from neighbors or increasing height of neighbors, potential sources of PPFD limitation. Pearcy and Sims (1994) provide several examples of plant responses to changes in PPFD over time, changes largely between quite low levels (as in shade) and high levels (as in the open).  There are consistently observed effects on leaf  characteristics, and they suggest that many of these changes might be associated with acclimation to other environmental stresses that are concomitant with the new environment.  22 In summary, the photomorphogenic responses of plants in the field are only relatively well studied with respect to complete shading by other plants, a condition that might not be very common, especially in communities with heterogeneous canopy structures such as pastures. In clonal plants under highly variable canopy conditions, one could expect photomorphogenic responses to be expressed as horizontal movement of stolons through a sward or as lengthening of petioles to ensure placement of leaves in high light positions (Boiler and Nosberger, 1985).  Recent recognition of the  heterogeneity in many communities has led to several investigations of shade-avoidance responses in many orthotropic plants with, for example, investigations of the role of reflection off neighboring plants, although this has not been attempted on plagiotropic plants such as white clover. There have also been many recent investigations of the effects of gradients in light conditions over interconnected portions of one plant, mainly in cional, plagiotropic plants. There seems to be consistent evidence that there is some type of intra-plant integration of the response to spatial heterogeneity, although the responses are widely varied. Many of the experiments fail to meet the requirement of effectively isolating the plant neighborhoods. In addition, many of the conditions used to investigate photomorphogenesis in plants have been quite artificial, and extrapolation to plant responses in the field remains difficult.  The response of plants to temporal  heterogeneity is even less well understood. 1.2.2. OTHER  MECHANISMS  INVOLVED  IN  PLANT/PLANT  INTERACTIONS In a plant community such as a pasture, neighbors have other less obvious influences on the nature of a white clover plant’s environment. Since the objective of this thesis is to examine the influences of natural light conditions on the growth and form of white clover, these are only briefly discussed here and reference is made to recent reviews addressing this topic. The various conditions associated with gaps of different size in vegetation and at different locations within a gap are reviewed by Bazzaz and  23 Wayne (1994). Above ground, air circulation patterns change with the proximity of neighbors, perhaps increasing 02, water vapor, and ethylene, and depleting CO . The 2 effects of many of these influences on plant growth have been reviewed recently in a special issue of Plant, Cell, and Environment (13:7, 1990) entitled “Sensing the Environment”. Below ground, many precise microenvironmental conditions may be associated with particular grass neighbors including depth of roots, affinity of roots for water and nutrients, and soil microbe populations, all of which may influence the growth of associated white clover plants (Turkington et al., 1988; Stark, 1994; Caldwell, 1994). Accumulation of dead material within a Holcus lanatus or Agrostis capillaris sward may actually prevent white clover stolon contact with the soil, and this is likely to reduce rooting and branching from nodes (Chapman, 1983). These factors surely interact with each other and with light conditions in the field, and the responses to neighbors in the field are undoubtedly complex.  1.3.  OBJECTIVES When considering through what mechanisms neighbors affect white clover  morphology, it may at first be beneficial to simplify the environment and only allow possible interactions through alterations in the light conditions. This study represents a departure from traditional studies of the effects of light conditions on plants, which generally have used artificial conditions.  Providing a SED that is altered by live  vegetation more accurately mimics light conditions in the field, and assumptions about responsible wavelengths need not be made. This study investigates the importance of alterations in the light conditions as a possible mechanism responsible for neighbor detection, rather than being a study of the mechanisms involved in detecting light changes (e.g., phytochrome or other receptors, signal transduction, molecular basis for response, etc.). Three different canopy arrangements are used, all with separation below ground, which recognize the existence of patchiness in the above-ground environment. Each of these puts target white clover clones in a different arrangement with respect to  24 the canopy of grass neighbors, mimicking a field situation where there is a gap in which a white clover plant resides, either wholly or partially.  These arrangements create  contrasting light conditions dictated by the arrangement of the canopy and the position of the white clover clones with respect to the canopy and the sun. 1.3.1. NORTHERN CANOPY: REFLECTION It has been shown in several orthotropic plants that neighbors can be detected through changes in the R:FR of light impinging in the horizontal direction, and this allows them to detect a signal of impending competition before there is photosynthetic limitation (Ballare et al., 1987, 1988, 1989, 1990; Smith et al., 1990). The responses of plants in these experiments were similar to the responses under complete vegetation shading, e.g. longer internodes and reduced branching. In white clover, the responses to reflection are also expected to be qualitatively similar to responses under complete shade, e.g. increased petiole length and reduced branching (Solangaarachchi and Harper, 1987; Thompson and Harper, 1988; Thompson, 1993a). In plagiotropic plants, the possibility that the placement of ramets horizontally, in the plane of the growing surface, is altered through reflection from neighbors, has been suggested for Portulaca oleracea seedlings (Novoplansky et al. 1990; Novoplansky, 1991), however the effects of reflection have not been examined in white clover. In five very similar experiments, the effects of light reflected from grass neighbors on the growth and morphology of white clover clones was investigated (Chapter 3). In each of these experiments, a canopy of grass, Dactylis glomerata (orchard grass), was placed to the north of target white clover clones. In this arrangement, the target plant received direct light from the sun and reflected light from the grass barriers. In the first four experiments, control clones grew in front of a control “canopy” of pots with potting-medium only, and in the fifth experiment, control clones grew in front of a barrier of bleached grass. By describing clonal growth in response to reflected light, these experiments are intended to establish the sensitivity of white clover to potential signals of impending competition from neighboring grasses.  25  1.3.2. EASTERN AND WESTERN CANOPY: PARTIAL SHADE In one experiment the morphologies produced by a single clover clone in response to different species of grass forming the borders of a standard gap are described (Chapter 4). In this experiment, target white clover clones grew inside a corridor bordered to the east and west by Dactylis glomerata, Lolium perenne, Holcus lanatus, or control (empty) pots. In these arrangements, light was transmitted through and reflected off the grasses casting shade in a natural pattern. In the middle of the day, direct light was also received from the sun by the clover clones.  This experiment provides a description of  phytochrome-mediated responses to light altered by leaf-filters, rather than cellophane filters or monochromatic light, under conditions where plants also receive some direct unfiltered light (unlike the homogeneous leaf filters used in Solangaarachchi and Harper, 1987; Thompson and Harper, 1988; Thompson, 1993a).  It is possible that these  photomorphogenic responses are quite sensitive to natural changes in light brought about by different species of neighbor, and that gaps within different species of grass are recognized as different from white clover’s perspective. 1.3.3. SOUTHERN CANOPY OVER PART OF THE PLANT:  LOCALIZED  SHADE A series of experiments was designed to investigate the extent of integration between portions of a clover clone when subject to a natural canopy placed to the south of localized portions of target white clover clones (Chapter 5). The extent and direction of integration was determined by examining the effects of a localized apical or basal-region grass canopy on the growth and morphology of the remainder of the plant. These allowed the following questions to be investigated: Do the conditions experienced by the apical region affect the response of the basal region in these clones? Do the conditions experienced by the basal region influence the response of the apical region? These questions were both investigated when the region under examination (not receiving the  26 variable local canopy) was behind both open and grass canopies. These experiments allowed an assessment of the importance of physiological integration in determining the growth and morphology of white clover clones growing under different localized naturalcanopy conditions. 1.3.4. SUMMARY OF OBJECTIVES 1. Northern canopy, reflected light: to determine if light reflected from neighboring grasses influences the growth and morphology of clover clones otherwise growing under open conditions. 2. Eastern and western canopy, partial shade: to determine if shade cast from neighboring grasses influences clover growth and morphology when direct light is also received for part of the day, and to determine if the effects of shade cast in this fashion from different species of grass differ. 3.  Southern canopy over parts of a single plant,  localized shade: to  determine if and under what canopy conditions a localized region of a clover clone responds to shade experienced by a remote region of the clone.  10 CM  stolon contains Figure 1.1. Mature stolon of white clover (Trifolium repens L.). This , flower, or branch a either ing several ramets, each consisting of a leaf, the axil contain bud, and the root.  28 Chapter Two GENERAL METHODS  Three sets of experiments were designed to assess the effects of reflected, partial shade, and localized shade conditions on the growth and form of white clover. These experiments have many aspects in common, e.g., the species used, the modular canopy of grasses which formed the treatments, the experimental chambers, and the techniques for measuring and evaluating the temperature, the spectral energy distribution (SED), and the clover clones. To avoid excessive overlap in Methods, the characteristics in common will be described in this chapter. The particular characteristics which distinguish each set of experiments, e.g., the arrangement of the canopy, the details of the growth and measurement of clover, and the particular data analyses, are also described in this chapter, but specific methods used for the individual experiments within each set are described in their respective chapters. The designs of the canopies used in these experiments allowed clover clones to receive light conditions that mimicked different arrangements of neighbors growing in field while preventing below-ground interaction. Each of these canopy arrangements emphasized a different influence of the canopy on light conditions.  In the first  experiment, canopy modules were arranged to the north of target clover clones (Fig. 2.1) to emphasize reflected light. In the second experiment canopy modules were arranged to the east and west of clover clones (Fig. 2.2) creating a standard gap and allowing shade from different species to be received along with some direct light. In the third set of experiments, two portions of clover clones (apical and basal along a primary stolon) were isolated but remained connected, and each received a canopy treatment independent of the interconnected portion (Fig. 2.3a,b,c,d). The canopy modules were arranged to the south and east (over apical regions) or south and west (over basal regions) of clover  29 clones creating local shade conditions similar to a uniform sward, with little or no direct light.  2.1.  SPECIES USED  2.1.1. WHITE CLOVER  COLLECTION AND PROPAGATION OF MATERIAL All of the white clover material used in these experiments was obtained from a  single clone collected approximately six months earlier from an old permanent pasture (sown in 1939) in Aldergrove, B.C. described in Aarssen and Turkington (1985). This pasture is considerably well studied and the clover persists largely through vegetative propagation (Parish, 1987; Evans and Turkington, 1988). Clover clones from this pasture might possess a high degree of phenotypic plasticity because of their ability to persist in the long term among a complex mosaic of grass neighbors, an argument detailed by Turkington and Mehrhoff (1990). These clones would therefore make good test subjects for examining responses to varying above-ground canopy conditions. The clone which was selected possessed three qualities that made it desirable over several other clones derived from this pasture; i)  it was vigorous and healthy growing in pots in the  greenhouse, ii) its phenotype was observed to vary quite markedly with illumination conditions in the greenhouse, and iii) it was a relatively large-leaved clone, whose measurable morphological characters can vary to a greater degree than smaller-leaved clones (Caradus and Chapman, 1991). The use of a single clone makes these experiments examples of the use of a phytometer to measure the environment by mimicking the response of a plant growing within it. This type of experiment can then be conducted using several genetically distinct clover plants to establish the level of variation in plasticity that is genetically-based (Bell and Lechowicz, 1994).  30 In September 1988 several cuttings of white clover from this stock material were rooted and transplanted into 25x50x8 cm flats containing Fison’s Sunshine Mix #1, a nutrient-rich soilless potting mix. The flats of clover were irrigated as needed on a constant liquid-feed program (i.e. fertilized during at least 3 out of every 4 irrigations) at lg/l of 20-20-20, N-P-K to at least 10% beyond their point of saturation. These were grown in a 18/18°C day/night greenhouse at the University of British Columbia southcampus field station. The clover was kept free from pests with regular use of airborne insecticides (Malathion, Ambush, and Sevin), an airborne fungicide (chlorothalonil), and a soil insecticide (Diazanon granular) at recommended rates. When the flats had a dense population of clover, cuttings were taken from the ends of the longest stolons. Taking cuttings from the ends of stolons eliminated potential age-dependent responses by the clover to the imposed treatments. Six of the cuttings were transplanted into each of eight new flats containing fresh potting medium. Thus, the material was sub-cloned periodically, and when grown in this fashion, the morphology of this stock material, e.g., leaf size, internode length, and stolon diameter, remained relatively consistent. This method of subdividing and planting in fresh media ensured a consistent supply of healthy, uniform stock material from which cuttings could be taken for use in the experiments.  PREPARATION FOR EXPERIMENTS It was important to start the experiments with uniform cuttings, because early  variation in size could lead to large differences later, perhaps altering the effects of the grass. There was always a phase after the stock material was subdivided and transplanted into new containers, when each flat of clover had several rapidly-expanding stolons advancing along the potting-medium surface. This was the stage at which cuttings were taken for use in the experiments; before the flats became overly crowded, and when stolon apices from the same order of branching could be selected. At this stage there  31 were nearly always advancing apices from established stolons containing several intemodes which made ideal material for propagating and for growing in the experiments. Stolon tips from the stock material were selected for uniformity and excised before the newest node had rooted in the potting medium. These cuttings consisted of the newest expanded leaf and the region apical to it, and a small root initial at the node. The cuttings were placed horizontally on 3.0 cm x 3.0 cm x 1.5 cm rock-wool cubes, with plastic-coated wire hooks placed over the stolon at the point where the expanded leaf originated. These were placed under shade on a heated mist-propagation bench for 5-7 days and then transferred to the greenhouse where they were allowed to grow into the rock-wool cubes. The rooted clones were placed in a flat with drainage holes inside a flat without holes, so that they could be bathed in a nutrient solution (20-20-20; NPK lgIl) which conveniently could be changed every second day. The cuttings were allowed to acclimate thus for 7 more days. An excess of visually uniform cuttings were then selected and transplanted into the appropriate containers to be used for each experiment, either 30 x 15 x 8 cm deep, for the partial shade experiment, or 50 cm x 25 cm x 8 for all others. One cutting was placed near the end of each flat, with the stolon pointing straight down the middle of the flat. All flats were filled with the same potting mix, Fison’s Sunshine Mix #1. The transplanted clones were allowed to root-in in a common area of the greenhouse (for 4-8 days) with their position rerandomized every second day. The plants were continued on the same constant liquid feed program described above for clover stock material, and were always being irrigated whenever the driest flat required. This allowed plants to be grown more uniformly and reproducibly. The clones were also treated with pesticides (airborne insecticides as above, Diazanon granular watered-in to kill fungus-gnat larvae, and Captan fungicide sprayed on the potting-medium surface to prevent algal growth), which began the bi-weekly program which was to continue throughout the experiment.  32 After 4-8 days, the length of the stolon, the number of ramets, and the length of the newest petiole were measured on each clone. These clones were then selected for uniformity, randomly assigned a treatment, and then placed into the experimental compartments. The clones were allowed to acclimate in the experimental compartments for 2-4 days (reflection and partial shade experiments) or 18-40 days (localized shade experiments) before the treatments were imposed.  TREATMENT DURING THE EXPERIMENTS During all experiments clover clones continued on the constant liquid feed and bi  weekly pesticide programs described above. During the reflection and partial shade experiments, the position of clover clones was rerandomized every four days. In the reflection experiments, to accomplish the rerandomization without compromising the experimental conditions, clover clones were moved only after all of the canopy modules had been removed from the compartments. The clones were then moved to the appropriate new compartments, and the canopy modules were replaced. This way, control clones were prevented from being close to neighboring grasses while being rerandomized, and therefore, they never received pulses of FR which could have had an effect on their morphology. During the partial shade experiment, a canopy of each type was prepared but remained vacant (without clover clones). After each rerandomizatioñ a different canopy became the spare. This system was used so that the grasses could be clipped (see below) when the clover was removed during rerandomization. First, each spare aisle was clipped, after which clover clones from the same treatment in a different block were moved into this aisle, vacating the next aisle to be clipped. This procedure was followed until the final aisle of each species was vacated and the grasses clipped. During the localized shade experiments, the clover clones were fixed into position and were not rerandomized. In these experiments, the two regions of clover clones always received irrigation and pesticide treatment at the same time.  33  2.1.2. GRASSES  COLLECTION AND PROPAGATION OF MATERIAL The grass species chosen to be neighbors of the clover are species that co-occur  with it in pastures of British Columbia. Dactylis glomerata (Dactylis), Holcus lanatus (Holcus), and Lolium perenne (Lolium) are found growing largely in monospecific clumps or in patches also containing clover (Aarssen and Turkington, 1985). Patches of these grass species offer varying degrees of competitiveness to white clover (Haynes, 1980; Frame and Newbould, 1986; Evans and Turkington, 1988). Seed of the three species was obtained from Buckerfields Seed Co, Vancouver, and sown with 50-100 seeds per pot in approximately one hundred 800 cm 3 (10-cm) pots containing a mixture of 1/3 peat, 1/3 sand, and 1/3 perlite. The seeds were covered lightly with sand and placed in the heated greenhouse at the University of British Columbia south-campus field station in September 1988. To produce material suitable for the experiments, it was important to grow these grasses under nearly-ideal conditions. It was necessary for the pots of grass to (i) alter light conditions comparably to grasses in the field, (ii) be easily reproducible from the grower’s standpoint, and (iii) form a canopy that was uniform and quantifiable. Under these conditions, replicates of treatments in individual experiments would more closely resemble each other, and an experiment could be repeated using almost identical canopy conditions. Plants growing in containers in a greenhouse generally find conditions relatively resource-poor compared to the locations in which they normally grow in the field, where they often have greater access to light and a larger rooting-space than in a greenhouse. Examples of these three species of grass were available in fields outside the greenhouse. It was only under extremely rich conditions in the greenhouse, i.e. growing with heavy fertilization and in spaced rows allowing one side full exposure, that these  34 canopy modules closely mimicked grasses growing in the field and, hence, would alter the light conditions similarly. The consistency to be gained by growing canopy modules in this fashion came from each species’ ability to self-thin under rich conditions. This was accomplished by growing the grasses in a relatively open situation with consistent fertilization allowing each species to reach its own natural density. The grasses were irrigated as needed on a constant liquid-feed program (i.e. fertilized during at least 3 out of every 4 irrigations to at least 10% beyond saturation) at lg/l of 20-20-20, N-P-K. These were placed on benches in the greenhouse in spaced rows grouped by species, in order to encourage uniformity within each species. The pots were kept free from pests with a regular program using airborne insecticides (Ambush, Malathion, and Sevin), airborne fungicides (chlorothalonil, Maneb, and benomyl), a soilapplied insecticide (Diazanon granular), and a fungicide soil-drench (chlorothalonil) at the recommended rates. The soil was also limed (1 gm/pot) twice yearly. Twice per week beginning early in growth, the grasses were clipped to encourage sturdy growth and tillering and to fill the pots with lateral growth. Material overhanging the bounds of each pot was also clipped, helping to maintain a tidy, self-supporting canopy which had the same lateral dimensions as the pot. Progressively, the grasses were allowed to grow taller before being clipped, until they formed a sturdy, square “hedge”, lOxlOx2O cm tall. They were then clipped weekly to this height. During this time the pots were allowed to fill and thin without interference, except for occasional removal of the thatch layer before it became a significant feature of the canopy.  Eventually, the stock of grass material for the experiments, consisting of these three species of grass growing as uniform hedges above the pots, could be maintained indefinitely. Each pot of grass formed a pillar-shaped canopy that could be arranged with others like it in a standard, repeatable fashion. These pots became the canopy-forming  35 modules that were used to present the desired above-ground canopy conditions to clover clones. Canopies that were used in control treatments were designed to present clover clones with no influence from grass, though otherwise be the same. These “empty” canopies were formed by the same pots filled with the same potting medium but containing no grasses, and in one reflection experiment control pots contained bleached grass (see below). Control pots were treated in the same manner as the pots containing live grass; i.e. they were irrigated and treated with pesticides as if grasses were growing them. In the experiments, they were placed in the same arrangement as the grass-canopy modules, creating a “canopy” of empty pots (or bleached grass).  PREPARATION FOR EXPERIMENTS An excess of pots of all the types needed for each experiment was arranged with  species in rows 10cm apart, on benches in the greenhouse. The orientation of the pots was kept consistent, so the sides next to neighboring pots in the rows would remain the same. These also would be the sides that were next to neighboring pots of grass in the experiments, on the inside of a row of canopy modules. Of the two remaining sides on the grass pillars, one consistently received the irrigation, which came from a tight shower spray and matted down some lower leaves and thatch. The remaining side was directed toward the south wherever possible. This south face of the pillars of grass most closely resembled the appearance of a clump of grass growing in the field. This was the side that faced clover clones during the experiments. Throughout this time the average tiller density in the pots of grass remained between 40-150 per pot (4000-15,000 per 2 m ) , the leaf density remained between 80-300 per pot, and the estimated LAI remained between ca. 2 and 5. These pots of grass resembled clumps growing in the field, and altering the SED similarly, could present clover clones with light conditions found in a sward while  36 preventing interaction below-ground.  The control pots were placed in the same  arrangement and treated identically before the experiments.  TREATMENT DURING THE EXPERIMENTS A two-fold excess of uniform canopy modules (either pots of grass or control-  canopy pots) was prepared for the experiments to ensure uniformity in the canopies throughout the experiments. It was uncertain as to how consistent the grass canopy modules would remain when growing in the experimental treatments. With an excess, pots could be rotated in and out of the experimental chambers as needed, ensuring that a healthy, field-like, grass-canopy could be presented to each clover clone receiving that treatment. During the experiments all excess canopy modules remained in similar rows outside of the treatments, and these were irrigated and sprayed the same as modules in the experiment. The canopy modules were fed throughout the experiment on the same constant liquid feed program described above, with irrigation occurring as soon as any pots began to dry out. The same pesticides as above were applied to grasses and control pots, but since there was an excess, this could always take place outside the experimental compartments.  In the reflection experiments, the pots forming the canopies were  replaced with pots freshly clipped to 20 cm every 4 days, at the same time that the position of clover clones was rerandomized. In the partial shade experiment, the pots of grass were clipped to 15 cm every 4 days in the rows when clover clones were removed. In localized shade experiments pots of grass were clipped every 4 days to 20 cm, outside of the experimental compartments. The edges of each pot were clipped at the same time ensuring that leaves of grass canopies did not overhang the flat of clover, thus confining the canopies’ effects to the hedge of grass forming the standard canopy, and preventing grass contact with clover leaves. Clipping of grasses always took place at a distance from clover clones, preventing cut leaves, motion, etc. from affecting them.  37  2.2.  MODULAR  CANOPY  DESIGN--CONSTRUCTION  OF  NEIGHBORHOODS An individual treatment in all experiments consisted of a single cutting of the clover clone rooted in its own container and a canopy formed by either a hedge of live grasses growing in pots (treatments) or control pots. The different canopy conditions used in each experiment can be described by the species, density, and orientation of the canopy-forming modules with respect to the clover clones.  For controls in these  experiments, pots containing potting-medium only were placed in the same arrangement as pots with grass, creating an “empty canopy”. Thus, the differences between controls and experimental treatments were confined to the zone where the grass plants were located, restricting to above-ground the possible mechanisms used for detection of these canopies by clover clones. It was important to design the neighborhoods so the grass hedges were prevented from interacting with the clover clones through leaf contact with clover foliage or potting-medium, splashing from the potting medium, and wicking of drainage water. These and other potential modes of interaction had to be minimized to ensure that the effect of the canopy was primarily due to alterations in the light conditions. Care was taken to avoid these, by treating the different types of canopy used in each experiment identically. For example, water, fertilizer, and pesticides were applied equally to the regions occupied by the canopy pots regardless of the canopy design or the species of grass used. All pots used for the canopies were positioned at or above the level of the flat containing clover. If the pots had been lower, grass leaves in experimental treatments would have covered the sides of clover flats, while in controls the sides of clover flats remained exposed. This location assured that the canopy modules would insulate the soil  38 similarly regardless of its contents, since in all cases the sides of each flat of clover were equally exposed. In addition, it was important for grasses to be high up towards the zenith in order for the canopy to have the desired effects for the experiments. As mentioned, the species chosen to form the canopy could be varied, and in the partial shade experiment each species could produce its own distinct neighborhood influences. Pots of different grass species consistently differed in many parameters including tiller and leaf density, leaf width, pubescence, and leaf angle, as they do in the field (Burdon, 1982; Burdon and Harper, 1982). These canopy modules, however, were considered to be equivalent, well-defined, and reproducible because of the similar conditions under which they were grown. Since each pot (within species as well as between) determined its own density (through self-thinning), the final outcome of all the species (and pots within species) would be determined by the amount of resources provided, and this can be duplicated. A reliable way of providing equal resources to different plants is to saturate them. The constant liquid-feed program and spaced rows, both before and during the experiments, were designed to accomplish this. The density of modules used to form the canopy could be varied, and this would produce a distinct neighborhood. The traditional definition of density as plants per m 2 does not adequately describe conditions in these experiments where below-ground interaction is prevented. A one or two-dimensional measurement does not adequately describe density, even with below-ground interaction, because it does not describe the restrictions to the space which provides resources (Ross and Harper, 1982). For the above-ground applications in these experiments, density can be described better by quantifying the amount of sky obscured, or the amount of the global-radiation hemisphere that is occupied, using an angle, wedge, or percentage of a hemisphere. Using this definition density is determined by both the distance and height of neighboring plants.  39 For example, tall grass at a greater distance would cause the same amount of sky obstruction as shorter grass that was closer. Although a short and very close canopy arrangement would obscure the sky the same as a tall, distant one, the two canopies could differ in factors such as temperature, humidity, and other airborne factors. The further away the canopies are located, in addition, the less these proximity factors are involved, while changes to the light conditions will still occur. To focus the differences between canopies on alterations to the light environment rather than to these other proximity factors, the distance to the canopy was maximized and circulating fans were used to increase air flow. To maximize the distance, the height of the canopy also had to be maximized to present the desired amount of sky obstruction. The height of the canopies was constrained by the height the grasses could grow and still be uniform in the pots and by the maximum depth of the subcanopy of pots, which could be elevated slightly above the clover soil level before it influenced a large a proportion of the global radiation hemisphere itself. The specific distance and height used in each type of experiment is described below. The orientation of canopy modules with respect to the sun was the most important factor determining the light conditions experienced by clover clones. The specific placement of the canopies in each experiment (Figs. 2.1, 2.2, 2.3) created a particular light environment. If the orientation with respect to the sun stayed the same, two canopies of the same species at a constant height and distance would create the same light conditions. This feature was utilized so that replicates of each treatment could be created. Following is a description of the orientations used in the three types of experiments.  2.2.1. REFLECTION (NORTHERN CANOPY) Grass canopies used in these experiments (Chapter 3) were designed to provide clover clones with a signal of impending interspecific interaction largely through  40 reflection from neighboring grasses, without limiting the available PPFD. A solid barrier of live grasses was used to reflect light with an SED, and in a pattern, that far more accurately simulates field conditions than artificial filters. Experimental clover clones were divided into two groups, both to be grown under high levels of PPFD, with one group receiving additional light reflected from live grass neighbors. Experimental treatments consisted of a row of five Daclylis canopy modules (described above) placed in an east-west orientation 15cm to the north of, and totally separated from, target clover plants, which were arranged with the primary shoot growing parallel to the row of canopy modules (right side of Fig. 2.1). Controls were arranged with clover clones growing in front of pots with no grasses and containing only the growing medium, a soilless pottingmix (left side of Fig. 2.1). In this arrangement the two sets of clover clones both received direct light from the south, while experimental clones received, in addition, reflected light from the row of grasses to the north. This design simulates a field situation where a clover stolon is growing within a large gap but is near grass neighbors at the northern edge. In two of the experiments in this set (Exps. 1, 2; Table 3.1), each set of canopy modules and controls were separated from the corresponding target clover clone with a transparent Plexiglas sheet (Fig. 2. ib). This was an attempt to minimize the differences in temperature, humidity, 2 C0 or other factors created by the different canopies (grass or , control) beyond what the air-circulation would accomplish. The transparent barriers would keep environmental conditions more uniform between treatments, while allowing light signals to pass between grass canopies or controls and clover. In this arrangement, the influence on light conditions that grass canopies had on target clover plants would be emphasized even further.  41  2.2.2. PARTIAL SHADE (PARTIAL CANOPY) In this set of experiments (Chapter 4) canopy modules of Lolium, Holcus, Dactylis, and controls were positioned to the east and west of clover clones (Fig. 2.2). This arrangement resulted in the clovers and grasses being physically separated both above and below ground, preventing root competition yet allowing interference with regard to light. The orientation of canopy modules allowed light to be transmitted through the grass canopies while also allowing a substantial amount of unaltered light (as well as reflected light) to reach clover clones. This is a pattern experienced by a clover plant in the field, when at different times of the day, a plant might receive unequal light conditions. The gap created directly over the clover clones allowed a period of direct sun during the middle of the day, and on overcast days, a patch of high R:FR light would be experienced while receiving light filtered by grasses. This mimics the field situation where a clover stolon is in a small gap, but is near grass neighbors on at least two sides. Each treatment consisted of two north-south rows of canopy modules with two flats containing clover clones placed in the same orientation (north-south) between the rows. Each row of canopy modules was 100 cm long, consisting of 10 individual pots placed side-by-side, the rows containing grass forming a uniform continuous hedge. The spacing of the rows created an aisle of standard width, and because all grasses were clipped to 15 cm above the pot (or 17 cm above clover ground level), these rows formed an aisle of standard height. The dimensions chosen (15 cm wide x 17 cm tall) created a gap that allowed equal exposure to the sky from any point on a line in the middle of the floor of the gap. From any of the points on this line, a “corridor” of sky running north to south, directly overhead, would never be obstructed. The primary stolon of each clover clone was oriented so it grew from north to south on the line down the middle of each tray, beginning at a point near the north end of  42 each flat. This kept each primary stolon apex under relatively constant conditions throughout the experiment (because it generally grew straight down this line) and assured that the environment was symmetrical from the point of view of the clones, i.e. having conditions imposed equally from the left or from the right.  2.2.3. LOCALIZED SHADE (LOCALIZED CANOPY) In the localized shade experiments (Chapter 5) grass canopy modules were located to the south of clones which were divided into two portions (Fig. 2.3). These canopies provided shade in a natural pattern, with sunflecks and natural scattering through the canopies of grass while maintaining below-ground separation. The shade was imposed on one or both portions of a clone, independently, by physically separating the two neighborhoods with opaque barriers (see below). The intent was to present two regions of a clover clone with distinct conditions, formed by shade that mimics neighbors growing in the field. This allows an evaluation of the independence of the responses of the two regions of the clover, by testing the conditions experienced by one region for an effect on the other region of the clone. The designs of the neighborhoods used in these experiments mimic a field situation where a single clover stolon has portions experiencing different environmental conditions. In the field, as the apex of a stolon advances into a gap it experiences high light conditions, while the basal portion of that stolon might remain within a clump of grass. The reverse is also common in the field, where a stolon apex invades a clump of grass from a gap and experiences shade, while the basal portion remains in the gap. In the field, the two portions clearly do not experience their own conditions independently. The design of these experiments, with separation between the two neighborhoods, allows an evaluation of the independence of two regions of a clone under different sets of conditions. A response in the remote portion of a plant would have to come from an  43 intra-plant signal rather than from the direct effects of the distant canopy, effects which could not be disentangled without isolation of the two regions. Clover clones were divided between basal and apical regions of a primary stolon, and all four combinations of two canopies (formed by grass or control canopy modules) and two regions (apical or basal) were used in the course of four experiments. Only two arrangements were used in each experiment, due to limitations of space and the need to grow replicates of each treatment. In the first two experiments, the effect of apical-region shading on the growth of basal regions of clover clones is investigated, first with the basal regions themselves in the open (Fig. 2.3a), and second, with the basal regions shaded by a grass canopy (Fig. 2.3b). In these experiments, the manipulation is done only to the apical regions of clones, and the response of only basal regions is examined. Differences in the basal regions of clones having their apical regions manipulated differently, is taken as evidence of physiological integration within the plant. In the second two localized canopy experiments, the effects of a basal-region canopy on the growth of apical regions of clones is examined, first with the apical regions in the open, (Fig. 2.3c) and second, with the apical regions shaded by a grass canopy (Fig. 2.3d). In these experiments, it is the response of apical regions, themselves growing under identical conditions, which determine the existence of physiological integration throughout the plant. To summarize, in three sets of experiments, three different types of neighborhoods were designed from pots of grass or control pots and flats of clover clones. Each of these presented a unique environment which was established by the presence of grass in the canopy, and these all simulate conditions white clover could experience in the field. The canopies of grass, however, could cause no corresponding below-ground interaction, so the responses of clover clones could only have been due to  44 detection of changes in above-ground conditions, perhaps primarily through changes in light conditions.  2.3.  EXPERIMENTAL CHAMBERS  -  DETAILS OF A ‘BOX’ AND FLAT OF  CLOVER In each experiment, regardless of type, it was necessary to isolate treatments in the greenhouse to prevent the canopy of one treatment from influencing the conditions of other clover clones. A difficulty in experiments of this type is ensuring that each canopy only exerts influence on its neighboring clover clone. The distance needed to effectively isolate different treatments of this type without an opaque barrier is unknown, and possibly quite large. Space was limited in the greenhouse, and compartment walls allowed treatments to be adjacent, maximizing the number of replicate treatments that could be grown. These “boxes” were designed within practical constraints, e.g. cost, transportability, reusability and are described below. Each treatment, consisting of a canopy and a single clone of clover (see above), was placed within a standard wooden compartment (box) designed to isolate the treatments and to standardize conditions. These boxes ensured that the light conditions experienced by a single clover clone were affected only by grasses immediately neighboring that clone. The main features that these boxes had in common were the opaque walls and the uniformity imposed through the height of the walls and their orientation in the same direction. The specific design of the boxes used in each type of experiment is described in more detail below. Walls were always high enough to maintain isolation to the point where the canopy of an adjacent treatment could have no effect on any other clover clone. However, the height of the walls was limited due to their blocking of the sun and sky, and the corresponding reduction in the quantity of light.  It was important to grow clover  45 clones with quantities of light that were typical of the field, in order to be able to apply clover response to the field. The orientation of the boxes remained consistent throughout each experiment. Within each experiment, since all boxes were arranged in the same direction and had walls of the same height, the sun tracked identically from any particular coordinate within the boxes. In all cases they were aligned so that they were open (i.e. widest exposure) to the south. This allowed as much light as possible into the boxes and allowed the sun to track symmetrically through the boxes, spending an equal time left and right of center.  2.3.1. REFLECTION The boxes used in reflection experiments were designed so that the grass canopies added reflected light without impeding the photosynthetic light available for clover clones. The boxes had three walls, as compared to a complete box having six; they were open to the top, bottom, and south (Fig. 2. ic; Plate 1). The side walls (east, west) were higher towards the north where the grass canopy modules were situated, although the rest of the walls were tall enough to obscure clover clones from grass or clover in adjacent boxes. The north wall was the same height as the higher portion of the side walls. This extra height towards the north prevented the sun (or sky) from appearing through the canopy, limiting light transmission through the canopies, and reducing the PPFD available from behind control canopies. The only light which could originate from behind the canopies would be the reflection off the inside walls of the boxes, and this could be reduced to a minimum. This design allowed as much of the available light as possible into the box, increasing PPFD (towards field level) and maximizing the exposure of grass canopies. The walls of the boxes were painted with a matte black finish (CIL indoor latex matte-black) to reduce reflection within the boxes. It was especially important to reduce  46 the amount of reflection from the northern wall of the box, behind the barriers, so that experimental treatments, having grass canopies which obscured the wall, contained roughly the same amount of PPFD as controls, which did not obscure the (upper portion of the) walls.  In this arrangement then, experimental and control clover clones received  equal light from the south, however, experimental clones in addition received reflected light from the grass canopies to the north, and control clones in addition received reflected light from the matte-black-painted northern wall of the box.  2.3.2. PARTIAL SHADE Experimental compartments used in this experiment were also designed to prevent light passage between treatments and allow equal exposure to the sky. These were constructed from separate wooden boards (120cm long x 10cm high x 1cm wide) painted with matte-black paint (Cit indoor latex matte-black) and placed N-S on stands to form east and west walls (Fig. 2.2; Plate 3). The height of the stands could be adjusted, and they were placed at a standard height (30 cm) in each box, high enough to never allow a grass leaf from one treatment to be exposed to a neighboring treatment. Black darkroom curtains were draped downward from the boards, so the height of the wall could be extended while maintaining isolation down to the level of the pots. These boards with curtains made opaque walls which restricted the differences in light conditions experienced by clover clones to the canopy present in its own compartment. The treatments were placed adjacent so that one of these opaque walls served as a wall in two adjoining treatments. Twenty compartments were thus created, and with the walls placed 45 cm apart, each compartment had enough space behind the rows of canopy modules to allow access to the “rear” of the pots. This space allowed water to be applied to the grass on the side away from clover clones, creating minimal alteration of the light conditions by matting of grass and minimal splashing off the canopy modules. The space also allowed the grasses to grow within the treatments without experiencing too much shade  47 from the barriers, which may have caused yellowing and loss of leaves from the pots of grass. As in reflection experiments, the walls were matte-black to reduce reflection from the barriers, because this would focus the differences between treatments to influences of the grass rather than influences of the wooden barrier. With dark rather than bright walls, control clones, which had no grasses obscuring the walls, would receive a more similar PPFD to the experimental clones, which have grasses obscuring the walls.  2.3.3. LOCALIZED SHADE In localized shade experiments the standard wooden compartments (boxes) in which each clone grew were each two-chambered, which allowed the independent manipulation of light conditions experienced by the two inter-connected parts of each clone.  In these experiments, localized portions of experimental clover clones were  subjected to shade from a grass canopy. To isolate the direct effects of this canopy to one portion of the plant, an opaque wall was present between the two portions of each plant, and the second portion would then occupy a separate chamber. The boxes used were the same as the ones used in reflection experiments (Fig. 2.1; Plate 4). These boxes originally had been constructed in pairs, i.e. two boxes were always side-by-side, sharing the middle wall (this made them easier to construct and transport). In these experiments, a single two-chambered box was used for each clover clone (Fig.2.3). An extension to the lower part of the middle wall, made of 2 mm-thick high-impact polystyrene, was installed, dividing the two chambers just after the plants were placed with their primary stolon apices extending into the second chamber. This lower-wall addition was necessary to fully isolate the two neighborhoods right down to soil level while allowing the primary stolon through. A small gap (2mm wide x 5mm high) was cut into the bottom edge of this wall at the point where the primary stolon  48 crossed, so the wall could be pressed into the soil around the stolon until it fit snugly into the gap. Thick white tape was placed around the top of the gap, so that light transmission would be blocked without applying damaging pressure from the top of the gap down on the clover stolon. Light conditions in the two chambers could then be independently manipulated. In an attempt to increase similarity to the field, all of the walls within the chambers used for localized shade experiments were painted white, which would increase the overall illumination.  It was thought that with maximally-reflecting barriers,  conditions within the boxes would have remained more similar to field light conditions. These treatments, therefore, would be indicative of a more common situation in the field, i.e. where there are better and worse light-resource patches due to neighboring grass, rather than having patches with light present and light absent. In addition, with white walls inside the boxes, the advantage that open neighborhoods have over shaded ones increases, perhaps promoting separation of the responses to the different localized canopies.  2.4.  MEASUREMENT OF WHITE CLOVER The piece of stolon that was taken from the stock material always continued its  development throughout the experiments, and this was considered the primary stolon. It developed secondary stolon branches at nearly every node. In nearly all experiments, these secondary stolons also branched to produce tertiary stolons. A single clover ramet in these experiments, consisted of a leaf, the root from its node, the axillary bud in the node, and the internode just basal to the leaf (Fig. 1.1). The plants’ total number of ramets considered only ramets with the leaf petiole  >  2.0cm,  which was approximately the same time that leaf lamina unfolded, while the total length of stolon incorporated all stolon length, right up to each apex. A young leaf was  49 considered to be part of a new ramet when its petiole length reached 2.0 cm, otherwise it was considered as part of the expanding apex. A branch stolon was considered to have been produced when the axillary bud produced its first leaf petiole > 2.0 cm. In localized shade experiments, since it is only the response of one region of the plant that is necessary to determine the existence of intra-plant integration, only the measurements made on the appropriate region in each experiment are presented (in two experiments it is the apical region, in the other two it is the basal region). Clover clones were measured in detail at the final harvest of each experiment as well as monitored over time in all of the experiments. At the final harvest of each experiment, the length, the number of ramets, and the number of branches on each stolon were recorded. Also, five fully expanded petioles, leaf lamina, and internodes were recorded from the primary stolon of each plant, and in the partial shade experiment, in four clones from each treatment, all petiole lengths and leaf areas were recorded. From these the following variables were measured or calculated: Total stolon length Total number of ramets Total number of branches Primary stolon-% of nodes branching Secondary stolon-% of nodes branching Whole plant-% of nodes branching Primary stolon-age to first branch (number of nodes more distal without branches) Secondary stolons-age to first branch Secondary stolons-mean internode length Tertiary stolons-mean internode length Whole plant-mean internode length Mean primary stolon internode length (n=5) Mean primary petiole length (n=5) Mean primary leaf lamina area (n=5) Whole plant-leaf area Whole plant-mean petiole length Branch stolon length ratio (ratio of branch stolon length to whole plant length) Branch ramet number ratio (ratio of branch ramet number to whole plant length) At the final harvest of each experiment, plants were divided into leaf lamina, petioles and stolons on each particular order of branching (primary, secondary, and tertiary) and dried at 60°C for 96 hours. These measurements allowed the determination of the following:  50  Total above ground dry weight Petiole weight ratio (percentage allocation to petioles, i.e. petiole dry weight/whole plant dry weight) Stolon weight ratio (percentage allocation to stolons) Leaf lamina weight ratio (percentage allocation to leaf lamina) Branch weight ratio (percentage allocation to branches Ratio of petiole/stolon weight Mean weight per ramet Whole plant-stolon specific length (length per unit weight) Whole plant-petiole specific length Whole plant-leaf lamina specific area In the reflection experiments, in addition, the newest five ramets on primary stolons of clones were harvested, measured (leaf area, petiole length, and stolon length) and dried separately from the rest of the primary stolon. In partial-canopy experiments, all leaf lamina, petioles, and stolon internodes on the primary stolon were measured and dried. From these measurements of length or area and the corresponding weights, the following were calculated: Primary stolon-specific stolon length (length per unit weight) Primary stolon-specific petiole length (length per unit weight) Primary stolon-specific leaf area (area per unit weight) In the localized shade experiments, only the stolon length associated with a particular weight was known, so only the specific stolon lengths (whole plant and primary stolon) were determined. Measurements over time did not involve destructive sampling, so only the length of stolon, the number of ramets, and the number of branches could be recorded over time. From these the following were calculated: Mean absolute growth rate (AGR) of stolon length Mean absolute growth rate (AGR) of ramet number Mean absolute growth rate (AGR) of the number of branches Mean relative growth rate (RGR) of stolon length (change in natural logarithm of stolon length over time) Mean relative growth rate (RGR) of ramet number Mean relative growth rate (RGR) of number of branches Ratio of primary stolon length growth to whole plant stolon length growth Ratio of primary stolon ramet number growth to whole plant ramet number growth Ratio of branch number growth to primary stolon length growth  51  In two of the reflection experiments (#‘s 2,5; Table 3.1) and in all localized shade experiments, the length of expanding petioles was monitored over the 8 days from the petiole first appearing. When recording stolon length and number, it was not difficult to measure the length of the newest petiole and a few of the next older petioles. When measurements of this type were made frequently enough (every 2 days) the growth rate could be determined for these petioles. The lengths of successive petioles on each plant were then grouped according to the number of days since first appearing, 2-4, 4-6, and 68 days, and the growth rates during these intervals were compared between treatments, disregarding any ontogenetic effects on petiole growth rate. In reflection and partial shade experiments, the position of leaves on the primary stolon was sampled on overcast days during the experiments, and the following were determined: Height of leaf above potting-medium surface Lateral displacement from the stolon (excluding leaves that crossed the stolon) The side on which the leaf originated The side on which the leaf lamina was currently positioned The angle normal to the lamina as viewed from the basal portion of the clones (elevation angle). In reflection experiments, the lateral displacement of the primary stolon apex was also recorded. This was a measurement of the displacement north/south from the center of the flat indicating if the stolon grew straight down the middle of the flat. In the localized shade experiment, the height of the highest leaf lamina and the number of leaves crossing the primary stolon was also recorded.  2.5.  DATA ANALYSES The specific methods for data analysis depended on the type of data collected, and  this differed slightly between experiments. In many of the experiments, a few plants, which were inhibited by some uncontrolled factor, were visually detectable as smaller,  52 slower-growing plants,. and these differed from the overall group mean by at least 4 standard deviations in at least one of the variables examined (95% confidence interval n=6, Glantz, 1987). The experimental group. of clones was always compared to the control group after these plants (if any) were removed. The remaining plants were used to compute the treatment-specific means and standard deviations for each of these variables. Homogeneity of variance between groups was tested using a Bartlett test and if  p<0.05, the data were transformed as follows:  10 (of size measurements), arcsine (of log  proportions), and square-root (of internode length and age to branch). If after such transformations the data were still heteroscedastic, a non-parametric test was used. Means were compared with Student’s t-tests, or ANOVA and Tukey HSD when multiple comparisons were to be made (Systat, 1990), and non-parametric Mann-Whitney or Kruskal-Wallis rank-sum tests were used when the variances of the groups to be compared were significantly different (Bartlett test p<O.O5). All tests used were twotailed. In reflection experiments, the following ANOVA’s were conducted: Leaf position data were compared using ANOVA model: Data  =  Constant  +  Side of plant  +  Treatment  +  (Side of plant x Treatment)  +  error. Data collected at several intervals during one experiment were analyzed using a repeated measures ANOVA model.: Data = Constant + Interval + Treatment + (Interval x Treatment)  +  error.  Some of the measurements in reflection experiments were available from several different experiments, and the treatment effects were analyzed between as well as within experiments: Data = Constant + Experiment + Treatment + (Experiment x Treatment)  +  error.  53 Leaf position data that were available across several experiments, were analyzed using ANOVA model: Data  =  Constant  Treatment)  +  +  Experiment  +  Side of plant  +  Treatment  (Experiment x Side of plant x Treatment)  +  +  (Side of plant x  error.  In partial shade experiments, differences among variables were examined with a one-way ANOVA using all treatments (3 species of grass and the control) as the effect using the model: Data=Constant + Treatment + error. Tukey HSD multiple-comparison tests were then conducted to determine which treatments had different effects. In the localized shade experiments, one or two of the largest plants reached the edges of the flat well before the others, and when this occurred, an equal number of plants from each treatment was harvested early. For the remaining plants, means of finalharvest measurements were compared using Student’s t-tests if the variances between groups were not significantly different (Bartlett test p>O.O ) and the observations were 5 approximately normally distributed. If these conditions were not met, non-parametric Mann-Whitney or Wilcoxon signed-rank tests were used instead. In cases where the pairing of individuals allowed, Wilcoxon signed-rank tests included the pairs harvested early. For the measurements made over time, the early-harvested pairs were excluded from the analyses. These repeated-measures ANOVA’s were conducted using the model: Data = Constant + Interval + Treatment + (Interval x Treatment)  2.6.  +  error.  MICROCLIMATIC MEASUREMENTS While detailed measurements of microclimate were not attempted, prudent use of  crude measurements was used to describe these conditions. The temperature and SED were measured under different conditions. As resources permitted, measurements were  54 made under extremes of conditions, under various weather conditions, and at different times of the day and of the season. This would allow better comparison with other experiments, and it would describe the extremes that clover clones had to deal with. These data were either collected directly in the experiments or under simulations of the experiments which could provide more standardized conditions than were possible during the experiment itself. The measurements were not intended to quantify differences between treatments, but merely to establish if there was a difference, and if so, its direction.  2.6.1. SPECTRAL ENERGY DISTRIBUTION Simulations of the experiments constructed in summer 1991 were used to produce spectroradiometric data representative of these experiments.  A LICOR LI- 1800  Spectroradiometer (Lincoln, Nebraska) with a cosine-corrected sensor on a fibre-optic cable was used to collect and analyze spectral information. Scans were made at several locations within one compartment, under simulations of treatment and controls. All scans recorded were means of three consecutive scans. These data were collected outdoors, since in the greenhouse the length of time it took to scan from 300 to 1100 nm in each simulation corresponded with changes in ambient light, e.g., the sun moving behind a support beam.  Some measurements also were made in the greenhouse to compare  outdoor and greenhouse light conditions. Spectroradiometric scans were made under uniform sky conditions, i.e., clear sky or overcast, at various times of the day including solar noon, morning and afternoon. Using arrangements identical to the experiment, pots of grass and pots of soil could be alternated without disturbing the sensor to produce two spectra, differing only in the presence of live grass. Any significant differences in the spectra were considered as indicative of spectral differences between the two treatments of the experiment. The R:FR ratio and PPFD were calculated from these scans.  55 2.6.2. TEMPERATURE Temperature was monitored in several experiments by simultaneously recording in the different treatments.  Resistors calibrated for temperature were placed in  experimental compartments, either taped to the wall of the compartment with the resistor supported 3cm away from the wall (used when the sun was not able to heat the resistor itself), in a small white tent 3cm above the potting-medium surface, or buried 1cm below the surface. The sensors were mounted in place and left to stabilize. The temperature was then recorded by moving the monitor (ohmmeter calibrated for temperature) between treatments while leaving the sensors in place. This was repeated a minimum of three times as rapidly as possible, and these were averaged. This was done at representative times (e.g. near noon, early evening, late evening), throughout the course of each set of experiments to give a general picture of the temperature regime in all treatments.  (a)  Pots of neighboring grass Pots of soil with no grass Flats with clover  ‘‘‘‘‘I’ll.—  N  I I  >1 I  >1  Black plywood  (b)  partition Plexiglas partition  iiiiiiiii•i  I  >1 I  (c)  >1  Scale:  .—  10 cn  60  Fig. 2.1. Treatments used in four reflection experiments in 1990 showing the arrangement of the modular grass canopy. (a) shows the design without using Plexiglas dividers between the regions occupied by the grass canopies and the clover clones, while (b) shows the design with a Plexiglas barrier. (c) is an enlarged side-view of (b).  •  w w  ii  •rn•  •iii•  .uJ • . • I  •Pots of neighboring grass Pots of soil with  no grass Flats with clover Black plywood partitions Scaie: 10cm  II  Figure 2.2. Design used in partial shade experiment. Canopy modules were pure stands of either Dactylis glomerata, Holcus lanatus, or Lolium perenne, and pots of potting-medium only were used as the control. Clover clones grew for 25 days.  S j  Pots of neighboring grass Pots of soil with  no grass Flats with clover  N  Black plyvood partitions Scele:  ,—i  10cm  I  I  111111  AVB  RI 1R 11111111 Neighborhood 1  Neighborhood 2  RI  1111551 Neighborhood 1  OPEN-OPEN  Neighborhood 2  OPEN- CLOSED  Figure 2.3a. Design of localized canopy experiment where canopy variation is presented to apical regions of clover clones, while basal regions remain in the open. Growth of basal portion in A: Neighborhood 1 was compared to growth of the same portion in B: Neighborhood I.  I 11111  I  •  n  5151111 Neighborhood 1  Neighborhood 2  CLOSED -OPEN  I  •  S 55555  Neighborhood I  Neighborhood 2  CLOSED CLOSED -  Figure 2.3b. Design of localized canopy experiment where canopy variation is presented to apical regions of clover clones, while basal regions receive closed canopies. Growth of basal portion in A: Neighborhood 1 was compared to growth of the same portion in B: Neighborhood 1.  ‘  I  11111 I  II  1R  •RiiII Neighborhood 1  Neighborhood 2  CLOSED-OPEN  ftI Lilili Neighborhood 1  Neighborhood 2  OPEN-OPEN  Figure 2.3c Design of localized canopy experiment where canopy yariation is presented to basal regions of clover clones, while apical regions remain in the open. Growth of apical portion in A: Neighborhood 2 was compared to growth of the same portion in B: Neighborhood 2.  I I•I I I I  •  •  Neighborhood 1  )  • • Neighborhood 2  CLOSED-CLOSED  El El  liii Neighborhood I  • ••I Neighborhood 2  OPEN -CLOSED  Figure 2.3d. Design of localized canopy experiment where canopy variation is presented to basal regions of clover clones, while apical regions receive a closed canopy. Growth of apical portion in A: Neighborhood 2 was compared to growth of the same portion in B: Neighborhood 2.  59 Chapter Three DOES LIGHT REFLECTED FROM NEIGHBORS (NORTHERN CANOPY) AFFECT THE GROWTH AND MORPHOLOGY OF WHITE CLOVER CLONES?  3.1.  INTRODUCTION The importance of the ability to sense and respond to neighboring plants before  there is resource limitation has been recognized recently (see Chapter 1). There is convincing evidence that neighboring plants can be detected through changes in the R:FR of light before there is photosynthetic limitation by neighbors (Ballare et al., 1987, 1988, 1989, 1990; Novoplansky et al., 1990; Smith, 1990).  These changes in the light  environment are primarily through additions of FR light from reflection off neighboring plants. The ability to successfully avoid photosynthetic limitation by neighbors may allow a plant to achieve success in a competitive environment, and this might be a trait that is under genetic control, hence, a population can be under selective pressure from its habitat to improve in this regard (Ballare et al., 1992b; Ballare, 1994; Bell and Lechowicz, 1994).  The responses of plants to photosynthetic changes in the light environment brought about by neighboring grasses are relatively well studied (see Chapter 1). These responses, which typically include increased internode length, reduced branching, greater specific leaf area, and increased biomass allocation to stems, largely match the responses to artificially imposed conditions that establish differences in both the quantity and quality of light.  The responses in orthotropic plants to neighbors through non  photosynthetic changes in the light environment have included increased internode length, reduced branching, and a greater proportion of biomass allocated to stem (Ballare, 1987, 1988; Kasperbauer, 1987; Kasperbauer and Hunt, 1992). In plagiotropic plants, the  60 responses to non-photosynthetic changes in the light environment are less well studied, but can include differential horizontal placement of ramets in response to reflection from neighbors (Novoplansky et al., 1990; Novoplansky, 1991). In white clover this has not as yet been studied, although its response to neighboring vegetation through changes in the overall light environment include reduced branching and alteration of petiole and internode elongation (Solangaarachchi and Harper, 1987; Thompson and Harper, 1988; Thompson, 1993). In these studies, the R:FR along with the PPFD were altered by neighboring plants. It seems that the responses to alterations in R:FR depend on the available PPFD. At relatively high PPFD (at levels typical of light shade from neighbors in the field) internode and petiole elongation are increased, and branching is reduced. However, the responses to changes in the R:FR at PPFD levels typical of relatively open conditions or when unfiltered light is also received, has not been studied in white clover or other plagiotropic plants. For example, it is unknown if the reception of light reflected from grass neighbors (lowered R:FR at high PPFD) has an influence on white clover. Using natural vegetation to alter the light conditions, the hypothesis could be tested that light reflected off grass neighbors alters the growth and morphology of white clover clones.  3.2  METHODS Five experiments were designed to assess the effects of light reflected from  grasses on the growth and form of white clover clones (Table 3.1). The experiments have unique features, but they are otherwise similar in many aspects. For example, the location (indoors or outdoors) and the time of year created different background conditions, so the experiments are not true repetitions, but many of the primary techniques are similar, such as the treatment of the plant material, the design of the canopies and barriers (see Fig. 2.1; Plate 1), the type of data collected, and the data analysis (Chapter 2). In four of the experiments, all conducted in 1990 (Exps. 1-4, Table  61 3.1), whole clones were exposed to light reflected from neighbors, while, in the fifth, conducted in 1992 (Exp. 5, Table 3.1), it was only the apical region of clover clones that was exposed to reflected light, basal regions remaining untreated. In the fifth experiment also, control “canopies” were formed by hedges of bleached grass, rather than pots with only potting-medium, so the apical regions in these plants received reflection from nonliving neighbors.  These and other distinguishing features are described here as a  supplement to the descriptions in Chapter 2.  Preparation of plant material Grasses and clover clones were prepared as described in Chapter 2.1. The species of grass chosen for these experiments was Dactylis glomerata (Dactylis), and the grass was clipped to 20cm above the pot, 22cm above the clover medium surface. In all of the experiments, an excess of these and control pots was prepared.  In the first four  experiments, control pots contained only potting medium making a “canopy” of “empty” pots, while in the fifth, control pots contained bleached grass forming a hedge similar to live grasses. The bleached-grass control pots were formed by spraying paraquat (0.01mg active ingredient per pot) on live Dactylis canopy modules. After a few weeks, these had completely yellowed, and along with a set of the same Daclylis  which was left  unsprayed, made up the control and canopy modules used in this experiment.  Clones of white clover were taken from stock material (Chapter 2.1.1). The cuttings were then selected for uniformity from an excess, placed into experimental compartments and allowed to acclimate, in the first four experiments, for a further 2-4 days before treatments began, and in the fifth experiment, for 43 days before treatments began. In the latter experiment, unlike the earlier four, only the apical region (of a primary stolon) of clones received light reflected from neighbors. For there to be a substantial amount of basal-region growth in these clones, extra time spent before  62 applying the treatment was needed. During this period, the position of clones was rerandomized every four days. To begin this experiment, the clones were repositioned with only the apical-most ramet protruding into the second neighborhood, where grass canopies were positioned two days later. The experimental chambers used in this experiment were similar to the ones used in localized shade experiments (Chapter 2.3.3) except that they were painted black in the region where grass canopies were situated and white in front (to the south) of this. Eight replicate clones were used in each treatment (grass canopies or controls) in the first four experiments, and six replicates were used in the fifth experiment. Treatments were randomly assigned to compartments, and this was changed every four days at the same time that clones were rerandomized.  Arrangement of canopy modules The arrangements of canopy modules used in these experiments are described in Chapter 2.2.1. In Experiment #5 (Table 3.1), bleached grasses were used instead of control pots with potting-medium only.  Treatment of plant material Clover clones during the experiments received treatment as described in Chapter In Experiments #2-4 (Table 3.1), the orientation of the clover clones was rotated 1800 at the same time as rerandomizing (every 4 days). This was done to avoid the asymmetry which developed in all clones including controls in preliminary experiments and Experiment #1. This type of asymmetry, which inhibited growth on the northern side of all clones, could have prevented reflection from this direction from having any additional effect.on clover clones. By rotating the clones, the darkening from the the northern side of the experimental compartments would come from no particular direction, nor would the reflection from neighboring grasses. In addition, a description of the treatment of clover clones in Experiment #5 is needed. In this experiment, a small apical  63 portion of the primary stolon was the only region of the clones that received reflected light. These portions were fixed in location once treatments began, so they could not be rerandomized during the experiment. The two portions of each clone were treated as one, both receiving irrigation and pesticide applications at the same time. Grass canopy modules and controls were treated as described in Chapter  Data collection Techniques for spectroradiometric measurement are described generally in Chapter 2.6.1., and are detailed further here. Under as many different ambient conditions as possible, simulations of the reflection experiments were arranged outside the greenhouse. One experimental compartment with a flat of potting-medium and canopyforming pots (Dactylis or controls) was placed on a table at the same height as the benches in the greenhouse. The sensor was placed horizontally in the middle of the flat, in place of a clover plant, and this was mounted securely so it could not move between measurements.  Then, the only changes that were made between spectroradiometer  measurements were the type of canopy modules used, and the measurement made with control canopy modules present became the standard to which the, other scans were compared.  All combinations of canopy type and Plexiglas presence were placed  individually within the compartment, and the global radiation (300-1100 nm) under each arrangement was measured. Next, a clover plant was placed into position, and the sensor was remounted on the potting-medium surface, directly in the shade of one of the clover clon&s primary leaves. This was a mid-sized clone (the same as the one used in all experiments) with several large primary leaves and few secondary branches. The canopy was then rearranged while the sensor remained in the same position. Measurements were also made with the sensor mounted vertically 7 cm above the potting medium (with no clover clone present), facing the northern wall of the compartment.  This greatly  enhanced the emphasis of horizontally impinging radiation, and would allow the sensor to  64 detect conditions similar to those experienced by upright clover structures such as petioles (Ballare et al., 1987). The different canopy arrangements were then placed in the compartment and measured as above. In sets of scans made at a later date using similar techniques, bleached grass plants were used as the controls, and several of the same arrangements as above were used.  Temperature was monitored using techniques described in Chapter 2.6.2. An outdoor reflection experiment (Exp. 2; Table 3.1) was used to determine the effects of Plexiglas on the temperature just above the potting-medium surface (sensor 3cm above). This was recorded around 1640 hrs (PDT) under a clear sky with a 8-15 kph wind by first recording with the Plexiglas present, and then again after the Plexiglas was removed. Indoors, (in Exp.3; Table 3.1) two sensors were mounted in each compartment, one 3cm above the potting-medium surface and another 15cm above, mounted 3cm away from the eastern side wall of the compartment. From these locations the temperature was recorded just before noon (solar time) and several hours after noon (around 1830 hrs PDT), after the sun ceased to directly affect the plants or the compartments.  The collection of data on clover clones is described in Chapter 2.4. This addition describes the techniques for measuring primary leaf position in these experiments. Leaf position measurements were taken on overcast days at least once during each experiment. These were made near solar noon and always when the plants were relatively turgid. At these times, the position of the oldest two of the newest three primary leaves was described in terms of: (i) height of lamina above potting-medium surface, (ii) lateral distance from the lamina to the stolon, (iii) the side on which it originated, (iv) the side on which its lamina was presently positioned, and (v) the angle normal to the larnina as viewed from the basal portion of the clones (i.e. looking east or west) giving the angle with respect to north/south. While the height of leaves was easy to record within 0.5 cm,  65 and it was easy to tell whether or not a leaf was on the same side of the stolon on which it originated, the leaf lamina angle was determined visually with a protractor and pencil, and therefore this measurement was only accurate to ca.  150.  3.2.5. DATA ANALYSIS These are described in Chapter 2.5 3.3.  RESULTS  Light Measurements Spectroradiometric measurements made under simulations of the experiments indicate that there was a detectable increase in FR wavelengths in treatments with grass barriers (Fig. 3. la,b,d,e). These differences in reflected FR light led to decreases in the global R:FR, as detected by the horizontally-positioned sensor which was either in full sunlight or in the shade of a clover leaf (Table 3.2) The presence of a grass canopy increased the PPFD observed in the open but not beneath a clover leaf. Beneath the cover of a clover leaf, the increase in FR due to neighboring grasses is noticeable both in terms of absolute increase and the resulting decrease in R:FR (Fig. 3. ib, Table 3.2). Plexiglas had no observable effect on light quality, but the quantity of PPFD was increased in the experimental compartments with Plexiglas (Fig. 3. ic, Table 3.2).  With the sensor placed directly in front of the canopy of empty pots or live grass and oriented vertically facing the canopy, the relative influence of the canopy over the light conditions detected is increased. Fig. 3. le and Table 3.3 describe the changes in the light scattered horizontally, due to the presence of live grass neighbors.  These  measurements indicate a strong increase in FR and resulting decrease in the R:FR due to neighboring grasses. There was also a noticeable increase in PPFD in the horizontal direction due to a canopy of live grass. Readings from the vertical-orientation of the  66 sensor (Fig. 3.1 a, Table 3.2) indicate that this led to an overall increase in PPFD of less than 2%. Horizontally-scattered radiation from bleached grass (Fig. 3. le, Table 3.3) was higher in R:FR, similar to controls using empty pots, and was higher in PPFD, similar to live grass canopies. Global (horizontally-detected) PPFD, however, was approximately 10% higher in treatments with bleached controls than treatments with live grass (Fig. 3.ld), whereas in controls using empty pots, PPFD was <2% higher than in treatments with live grass (Fig. 3.la).  Spectroradiometric measurements also were made under several other ambient conditions (e.g. different time of day, sky conditions, etc.). The results (data not shown) follow a similar pattern, i.e. a decrease in R:FR and a slight increase in PPFD was detected in the presence of live grass neighbors. Measurements were also made using two other species of grass, Holcus lanatus and Lolium perenne (also used in Chapter 4), to see how the influence of Dactylis compared to other common neighbors of white clover. These indicate that while the influence of the grasses on the light conditions is qualitatively similar, Dactylis is a relatively strong FR-reflecting grass (data not shown).  Temperature Experiments conducted outdoors clearly experienced lower temperatures than experiments conducted inside the greenhouse (Tables 3.4, 3.5). Under all conditions, the temperatures measured within the experimental compartments were elevated over the ambient greenhouse or outside air temperature. While outdoors with a slight breeze (8-10 kph) Plexiglas warmed the area occupied by the clover clones, it did so equally in both treatments. The increase over ambient temperature was more or less equal in both treatments under all conditions measured, except late in the afternoon inside the greenhouse, where, after the influence of the sun’s direct beam no longer directly affected the compartments, the air in treatments having grass barriers was warmer (Table 3.5).  67 Under overcast conditions, both indoors and outdoors, no difference between treatments was detected.  Clover growth and morphology In the five reflection experiments, there were few detectable differences between clover clones growing next to barriers of live grass and control clones in measurements made at the final harvest of each experiment (Table 3.6). Measurements differing significantly in one particular experiment generally did not show a significant difference in the other experiments.  The ANOVA on these measurements using all five  experiments, reveals that mean primary leaf area was the only measurement that differed significantly between treatments over all experiments (Fig. 3.2). It appears that clones next to live grass had a larger mean primary leaf area than clones next to empty pots. In this ANOVA, also, there was a significant “experiment” effect, showing that measurements differed between experiments, but there was no significant “treatment x experiment” interaction, which was common in the other measurements.  Despite this consistent lack of response for so many characters across experiments, a few observations are worth noting. No differences between clones developed in the two experiments without Plexiglas. The experiment in which most of the significant differences developed was the one in which whole plants were oriented in the same direction (growing east to west) for the entire experiment. In this experiment (Exp. 1), clover clones in front of live grass grew somewhat larger with longer stolons, more above-ground biomass, larger primary leaves, and longer internodes on their secondary branches. In the other experiment in which clones’ orientation remained constant (Exp. 5), the neighbor-treatment was given only to the apical region of the plants, and this region also produced longer stolons with more ramets and a reduced percentage allocation of biomass to leaf lamina. In all five experiments, there was no  68 evidence of increased allocation specifically to petioles, nor of differences in specific length of petioles and stolons, or the specific area of leaf lamina. None of the branching measurements indicated any significant differences in branching between treatments in any of the experiments. Mean internode lengths did not differ consistently, and petiole measurements indicated no significant differences in length. Petiole growth rates in Exp. 5, measured at 2-day intervals during the twelve days after appearing, also indicated no differences between treatments (data not shown). In addition, neither the mean AGR (analyzed by repeated-measures ANOVA) nor the mean RGR throughout the whole experiment (in stolon length, ramet number, or number of branches) differed between treatments, even in the experiments where differences in size developed in the experiment (data not shown).  A summary of the variation in measurements made on primary leaf position during each experiment is shown in Table 3.7. In Exp. 1, w,here the clones’ orientation was not rotated, the height at which leaves were displayed varied significantly between the two sides of the primary stolon on which the leaf could have originated. Primary stolon leaves were displayed at a higher elevation if the leaf was toward the north, regardless of treatment. Leaves towards the grass neighbors were displayed at a higher elevation when the clones’ orientation was reversed, in Exp. 3 and 4, i.e. leaves that originated and were situated on the side of the primary stolon towards the north, were displayed at a higher elevation above the soil than leaves on the side away from grasses (Fig. 3.3; Plate 2). This occurred at consecutive measurements, when a different side of the clone was toward the grass. In control clones, the leaves towards the south were often at least as high (or higher) than leaves toward the wall.  In one of the experiments where the orientation of the clover clones was rotated (Exp. 4), the lateral distance of primary stolon leaf lamina from the stolon was reduced in  69 clones growing in front of grass barriers on both sides of the plant (Fig. 3.4). This indicates that these leaves on clover clones in front of grass were kept at a greater distance from the canopy, and that the leaves retained this posture when they were away from the barrier.  A similar effect on lateral placement of primary stolon leaves,  especially if the leaves were currently towards grass, was noticeable but not statistically significant in the other experiments where clover orientation was rotated.  In Exp. 1, where clover orientation was not changed, all plants regardless of treatment had several primary leaves cross the primary stolon from north to south and very few cross from south to north (Yates’-corrected Chi-square=5.45, p<O.O5). In one of the experiments where clones were rotated, there was a greater number of leaves crossing the primary stolon from north to south, regardless of treatment. Observations of leaves crossing the stolon over consecutive periods in one experiment, when different leaves originated towards the north, show that a leaf that was once crossing the stolon from north to south, after rotation will often re-cross the stolon and reside on the other side of the stolon after it is rotated (usually within one day). Although the number and direction of leaves crossing the stolon seemed to be unaffected by the canopy of grass to the north in the first four experiments, primary stolon leaves on clones growing in front of live grass in Exp. 5 crossed the stolon significantly more frequently from north to south than on clones in front of bleached grass (Yates’-corrected Chi-square=4.7, p<O.O5). In this experiment, the orientation of the clover clones was unchanged throughout the experiment.  There was no evidence that the angle at which the primary stolon leaf lamina was displayed was affected by its side of origin on a plant or by a grass canopy, in any of the experiments.  70 The above data also support the visual evidence of the leaves and primary stolon rolling away from the north, and this happened more or less regardless of treatment and regardless of which side of the clone was toward the north, i.e., in experiments with clonal orientation reversed, the stolon rolled in the opposite direction after each rotation. In one of the experiments where the clones’ orientation was not reversed (Exp. 5), there was a statistically significant deflection of the stolon apices towards the south, regardless of treatment (i.e., the mean distance south from the middle of the flat was significantly different from 0 and always positive; n=12,  p<0.05,  Sign test, Systat, 1990), and there  was no evidence that this differed between treatments. There was no asymmetry detected in the secondary branch growth in these clones under this condition. In the experiments where the clones had their orientation reversed, there was no significant deflection of the primary stolon apices from the middle of the flat in any of the treatments. In these clones, the older portions of the primary stolon appeared to zig-zag down the flat, indicating that while stolon apices generally proceeded down the middle of the flat, there was tendency to orient the stolon tip away from the northern barrier, regardless of its design.  3.4.  DISCUSSION The microclimatic measurements indicate that the presence of grass neighbors to  the north under controlled conditions influenced the microclimate, including the light conditions, in ways detectable to instruments, although differences between these measurements were unreliable predictors of the response in nearby clover.  The  measurements of light conditions on the surface of an open flat (Figs. 3.la, d, e) are representative of the conditions in the experimental chambers before considering the influence of the clover itself. These measurements also would be indicative of conditions present early in the experiment (when clover clones were small) as well as of conditions above the clover leaf canopy. Measurements made under the shade of a clover leaf (Fig.  71 3. ib) are more indicative of the conditions experienced by a clover stolon or petiole below the cover of leaf lamina. In the shade of a leaf, the increase in FR due to neighboring grasses is perhaps more noticeable to the clover, since the resulting decrease in R:FR (Table 3.2) is in the range where there may be more sensitive phytochrome photomorphogenic responses (Smith and Holmes, 1977; Morgan, 1981). However, no effect on petiole or internode extension was observed in these experiments.  The vertical orientation of the spectroradiometer sensor enhances the contribution of horizontally-scattered radiation to the measured SED, and this would be more representative of the light conditions received by vertical structures such as clover petioles (Ballare, 1987). With this sensor orientation, it was clear that there was an increase in the horizontally-scattered FR due to live neighboring grasses. There was also an increase in the PPFD in front of neighbors when compared to “empty” controls. In this northern orientation, a grass canopy such as this may provide conflicting signals to a clover plant. On the one hand, it provides low R:FR light, while at the same time provides additional PPFD, even though the overall increase in global (sensor horizontal) PPFD was small (<2%) (Table 3.2).  Temperatures taken at representative times can give only a partial description of the temperature regimes under the different treatments.  Few differences between  treatments were detected, so the effects of Plexiglas with respect to decreasing between treatment temperature variation in the region occupied by clover clones could not be determined. The difference detected late on a clear day (Table 3.5) suggest that there was a difference in the way grass barriers retained and emitted heat after the sun went below the “horizon”. A Plexiglas barrier was not in place, and therefore its effect under these conditions could not be tested. Under late-afternoon clear conditions, the northern and eastern edges of a gap might retain more heat (in the soil and surrounding air) under  72 calm conditions (Bazzaz and Wayne, 1994). There is also likely to be long-wave radiation emanating from the leaves of live grasses at this time of day (Smith, 1982). Increases in temperature have been reported to increase stolon extension rate, leaf size, and petiole length in white clover (Turkington and Burdon, 1983; Frame and Newbould, 1986), and in these experiments the presence of inconsistent temperature regimes was likely, perhaps causing the increased stolon extension observed in experiments 1 and 5. Humidity differences are likely to be reflected in these temperature measurements. In the presence of grass, there will be a slight increase in water vapor pressure in the air near the grass neighbors, especially under relatively calm conditions. The higher water-vapor pressure at similar temperatures (as detected at mid-day) would give the air in these treatments a smaller vapor pressure deficit (higher relative humidity), and this could promote growth in white clover when under low-humidity conditions, e.g. bright sun or inside the heated greenhouse (Frame and Newbould, 1986). There were times when clover clones were not completely turgid, and after application of water to the floors and radiators in the greenhouse, clover clones would rapidly regain turgidity. A slight increase in local humidity in treatments with neighboring grass could have reduced the time spent in these water deficits, and hence could have increased clonal growth as seen in experiments 1 and 5. Another response to increased humidity which might have been observed in the clones was increased leaf size, however most responses to increased humidity are dependent on other factors, e.g. light, temperature, water status of roots (Grantz, 1990). The higher temperature reading found late in the day in front of live grass also could have been caused by increased latent heat, a characteristic of more humid air.  The measurements of clover clones themselves should indicate whether conditions with neighbors to the north were different from conditions in control treatments. Few consistent differences arose in the morphological characteristics of  73 clones measured in different experiments conducted at different times of the year, in different locations, with and without Plexiglas. Only one variable compared across all experiments varied significantly due to a treatment effect, mean leaf area. This normally is seen as a response to leaf shade (Smith, 1982) or to a reduction in light quantity (Morgan and Smith, 1981; Caradus and Chapman, 1991; Dong, 1993), while in this case it could not have been caused by shade, but perhaps by another alteration to the light conditions, e.g., R:FR (Solangaarachchi and Harper, 1987), or an effect of humidity. The lack of response in petiole length to the substantial increase in horizontally-impinging FR, is surprising because it is known that petioles of white clover can respond quite sensitively to R:FR under uniform natural-canopy conditions (Solangaarachchi and Harper, 1987; Thompson, 1989; Thompson, 1993a). It seems that for this response in white clover, light with high PPFD and R:FR might override the effects of low PPFD and R:FR light. One effect of the northern grass barriers seemed to be an increase to clone’s size in front of grass neighbors, which appeared in experiments where the clones’ orientation was not reversed (Exp. 1, 5; Table 3.6). However, there was no detectable effect on internode length or branching patterns in any of the experiments, indicating that there was little if any selective placement of ramets, i.e. no foraging was exhibited. Selective placement of ramets in white clover can be affected by physical displacement of the stolon in addition to changes in internodes and branching. The newest node or two on each stolon does not root for several days after appearing, and the potential exists for lateral movement before these nodes root. Phototropic bending away from strong FR sources has been reported in shoots of orthotropic plants (Ballare et al., 1992), and the direction in which plagiotropic shoots of Portulaca oleracea seedlings grow is affected by the direction of a strong FR source (Novoplansky et al., 1990). In the present experiment, it appears that the primary stolon in white clover was affected in this way by neighbors, and in these clones this had a stronger effect on ramet placement than different internode extension or branching patterns  74  Foraging in white clover can also be considered with respect to placement of resource-gathering organs, leaf lamina, as leaves grow, e.g., through differences in petiole elongation, and in the shorter term, through the proximal positioning of the leaf lamina at any given time. There was evidence that the positioning of leaf lamina could change over time, even after the petioles have stopped elongating. Under the experimental conditions, leaves were allowed to adopt their own position. In experiments where the clone’s orientation was reversed, data on leaf position indicate that leaves can adopt a position suitable for brief periods of growth. Changes occurring at four-day intervals can be “tracked” by these leaves, and this phenomenon was affected by the presence of live grasses to the north. In the field, clover leaves are often intermingled with neighboring grasses (Haynes, 1980), not allowing them as much flexibility in positioning as seen here. These results suggest that in a gap, clover leaves are able to alter their position daily (and perhaps during the day) using phototropic responses to differences in FR.  In clones having their orientation reversed every four days (Exp.2, 3, 4), there seemed to be fewer differences between clones under the different treatments than in the experiments where the clones’ orientation was not reversed (Exp. 1, 5). It appears that light reflected from neighbors that does not consistently originate in a particular direction might cause little response in white clover. This suggests that there may be some sidedependent effects that develop when one side of the plant remains in closer proximity to neighbors. It has been suggested that in some plants this might be accomplished through an integrated system able to compare conditions between interconnected portions and “choose” between them (Novoplansky, 1991). Solangaarachchi and Harper (1989) found that clover clones growing near neighboring clones produced greater growth on the side away from the neighbors than the corresponding side of an isolated clone. This would suggest the possibility of asymmetric growth in experiments with unrotated clones. In  75 one of these two experiments (Exp. 5), no asymmetry was found, so even though clones in front of grass neighbors produced more stolon length in this experiment, stolons on both sides of the clone were larger. In the other experiment where clones were not rotated (Exp. 1), asymmetry was found, but the side of clones towards the barriers was the larger, regardless of treatment. The asymmetry in these clones could have been due to the Plexiglas barriers. The presence of Plexiglas in these chambers increased the PPFD in full sun by more than 10% (Table 3.2), provided that the reflection of the sun was cast onto the detector. During the time the experiment was conducted, ending just before the summer solstice, the reflection from the sun in the 30cm-tall Plexiglas sheet would reach approximately 13 cm out onto the flat containing clover clones, or approximately half-way. This would have created the conditions where branches to the north were under greater PPFD for most of the day. Reversing the orientation of the clones eliminates the possibility of an effect developing from the interaction between one side receiving a particular condition and the barrier itself. For example, if the side that always remained near the neighbors displayed leaves higher, this could alter the way the clone responded, e.g., by reducing mutual shading in clover (shading of northern clover leaves by southern ones would be reduced). In clones that were rotated, the number of significantly different morphological characteristics dropped to almost nil. The effects on primary stolon lateral displacement were similar, this being eliminated when the orientation of the plants was periodically reversed. In clones that were not rotated, and perhaps clover in the field with uni-directional reflection from neighbors, the combination of greater stolon extension along with lateral movement of stolon apices may increase the speed with which a plant diminishes the effects of neighboring plants. Perhaps in clones that were reversed, placement of ramets was unaffected, since there was no directionality to the signal of neighbors and hence, no clear signal of where they could best be avoided. Finally, if clones under different treatments in experiments without Plexiglas barriers (Exps. 3, 4, 5) had more obvious differences than experiments with Plexiglas  76 (Exps. 1, 2), the suggestion might have arisen that factors other than light were influencing clover clones in front of neighboring grasses. Experiments without Plexiglas showed few detectable differences between clones (Table 3.6), while more differences were detected in experiments with Plexiglas, so no such suggestion about the effects of grass neighbors can arise from these data.  In conclusion, under this canopy configuration, possible interactions through the radiation environment were confined largely to reflection off the neighboring grasses, and it appears that this potential signal of neighboring grasses, whether directional or not, initiates little change in the growth and morphology of clover clones. Other than mean primary leaf lamina area (n=5 leaves/plant), it appears that few morphological characters were affected by the presence of neighboring grasses. It appears that the grass neighbors were detected by the clover clones however, because leaf and stolon apex positions were altered by neighboring grasses.  The movement of these structures, presumably  phototropic, placed them in positions where the effects of neighbors was reduced, and this could have been involved in the lack of further morphological response in the clones.  Initially, in the presence of neighbors, behavioral responses such as these may alleviate the need for morphological changes. Perhaps, if these initial responses are inadequate, then morphological modifications might be expected. Variable sensitivity in either of these responses to neighbors could regulate an individual’s response to heterogeneous conditions, and a wide range of these may be present in a population.  77  Table 3.1. Summary of five reflection experiments conducted at different times in 1990 and 1992. Neighbors used  Plexiglas  Location  Time  Exp. 1  Dacrylis glomerata empty pots  yes  indoor  Spring, 1990  Exp. 2  Dactylis glomerata empty pots  yes  outdoor  Autumn, 1990  Exp. 3  Dacrylis glomerata empty pots  no  indoor  Summer, 1990  Exp. 4  Dactylis glomerata empty pots  no  outdoor  Summer, 1990  Exp. 5  Dactylis glomerata bleached grass  no  indoor  Spring, 1992  78  Plate 1. Layout of reflection experiments: (top) Experiment 2 near the beginning of the experiment; (bottom) Experiment 4 near the final harvest.  77  :  ‘I  -  1  Ir A,  I  1V  I  -  -  Jdj r  Grass present  Grass present  Control  Control  70--  (a)  -  4 r\A  &fl-  .I  /  II  a,  a, z  C C  3_fl-  2.0--  a,  ‘S  —  1.0—  2.0—-  1.0-0i  0.04(A)  5(10  (.00  700  400  500  600  700  without Plexiglas  Grass present  with Plexiglas 7_li  -  (c)  :: :::  Control-Bleached grass  3.0—  -  (d)  --y  ‘  1.00.5-  1.1)0. 0  0.0-  I, 5130  ‘0  600  70(1  -  300  400  500  600  700  800  900  1000  Wavelength (rim)  Control-Empty pots  Grass present 1.81.8-  (e)  Control-Bleached grass  A 1.4— 1:1,..  1.2—  I  1.0-  I,  •jlj  ii—’.— I “ I  •  A  0.80.6—  1  0.40.2-  300  400  500  600  700  800  900  1000  110  Wavelength (rim)  Fig. 3.1. Spectroradiometric measurements made using spectroradiometer (LICOR, Lincoln, Nebraska) under simula a LI-1800, portable tions near solar noon. Each measurement was the average of three consec of the experiments utive scans. (a) the effect of a live reflecting barrier of Dactylis glomerata measu same as (a) but measured beneath a clover leaf, (c) the effect red on an open flat, (b) treatment, (d) the effect of live reflecting barrier of Dactylis of Plexiglas in a control go?nerata compared to a control using bleached grass, and (e) sensor vertical to detect horizontally-impinging light.  110  81  Table 3.2. R:FR (quantum ratio 655-665:725-735) and PPFD (i.tmolm s-’) derived from 2 spectroradiometric scans (using LI-1800, LICOR, Lincoln, Nebraska) made at solar noon on a clear day. Scans were made (i) on the surface of an open flat (containing potting medium but no clover) with Plexiglas, (ii) on the surface of an open flat without Plexiglas, and (iii) on the surface of a flat in the shade of a clover leaf without Plexiglas. Each measurement was the average of three consecutive scans. (Uncertainties: R:FR<0.01, PPFD<5) Canopy type  Conditions Open, no Plexiglas  Open, Plexiglas  Under clover leaf, no Plexiglas  R:FR Control empty pots  1.20  1.20  0.55  Dactylis glomerata  1.12  1.14  0.46  PPFD Control empty pots  1542  1723  280  Dactylisgiomerata  1561  1739  273  -  -  Table 3.3. R:FR (quantum ratio 655-665:725-735) and PPFD (!i.molm s’) derived from 2 spectroradiometric scans (using LI- 1800, LICOR Lincoln, Nebraska) with the sensor vertical, facing the canopy to the north, to detect horizontally-scattered radiation. Canopy type R:FR  PPFD  Dactylis glomerata  0.42  106  Control bleached grass  0.94  140  Control empty pots  1.19  54  -  -  82 Table 3.4. Mean (±SE) of temperature measurements (°C) made outdoors in reflection experiments with and without Plexiglas (Wind speed 8-10 kph, clear sky, air temp. 13.5 OC). Each individual compartment was measured three times and averaged before determining the mean for all treatments of that type. Shown below are the significance levels for the ANOVA effects of canopy type, Plexiglas, and their interaction. nsnot significant at p<O.05. Canopy type  with Plexiglas  without Plexiglas  Dactylis glomerata  18.1 (0.3)  16.0 (0.5)  Control  17.6 (0.4)  16.2 (0.4)  ANOVA results: Canopy type ns Plexiglas 05 . 0 p< Interaction ns  Table 3.5. Mean (±SE) of temperature (OC) measurements made under different conditions in the greenhouse. Measurements were made on a clear day around noon (greenhouse temperature 26.0°C) or late in the afternoon, after the sun no longer directly affected the plants (greenhouse temperature 23.0°C). Shown below the column heading is the significance of the difference between means of each measurement made under the two treatments. ns=not significant at p<O.05.  Noon Canopy type  Late PM  Soil surface (ns)  Air (ns)  Soil surface (ns)  Air (p<O.05)  Dactylis glomerata  32.0 (0.7)  28.5 (0.5)  23.4 (0.4)  22.5 (0.3)  Control  32.9 (1.3)  28.2 (0.4)  22.6 (0.4)  20.3 (0.3)  83 Table 3.6. Summary of differences between morphological measurements of clover clones made in five reflection experiments. If significant, the relative size of the measurement between treatments is shown. C=Control, D=Dactylis glomerata, * p<005, ** p<O.Ol, ns = not significant. Exp. 1 Indoor, with Plexiglas  Exp. 2 Outdoor, with Plexiglas  Exp. 3 Indoor, without Plexiglas  Exp.4 Outdoor, without Plexiglas  (n=5)  C (n=7) D (n=6)  (n=7)  (n=7)  ns  ns  ns  *  (D>C)  *  (D>C)  Variable Total stolon length  *  (D>C)  Exp. 5 Indoor, Control is bleached grass C (n=5) D (n=6)  Total number of ramets  ns  ns  ns  ns  Total number of branches  ns  ns  ns  ns  ns  ns  ns  ns  ns  Total above-ground dry weight  *  (D>C)  Mean weight/ramet  ns  ns  ns  ns  ns  Petiole weight ratio  ns  ns  ns  ns  ns  Stolon weight. ratio  ns  ns  ns  ns  ns  Leaf lamina weight. ratio  ns  ns  ns  ns  Ratio: petiole/ stolon weight  ns  ns  ns  ns  ns  Mean primary petiole length  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  (D>C)  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  Mean primary leaf lamina area Mean primary stolon internode length Ratio: primary petiole/ stolon length Secondary stolons- internode length  *  **  (D>C)  Tertiary stolons- internode length  ns  Whole plant- internode length  ns  *  (C>D) ns  *  (D<C)  84 Table 3.6. cont.  Variable  Exp. 1  Exp. 2  Exp. 3  Exp. 4  Exp. 5  Primary stolon- specific leaf area  ns  ns  ns  ns  115  Primary stolon-- specific petiole length Primary stolon--specific stolon length Secondary stolons-ratio: new petiole/ new internode Primary stolon-% of nodes branching Secondary stolons-% of nodes branching Whole plant-% of nodes branching  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  flS  ns  *  (D>C)  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  Primary stolon- age to first branch  ns  ns  ns  ns  flS  Secondary stolons- age to first branch Branch weight ratio  ns  ns  ns  ns  ns  ns  ns  ns  ns  ns  Branch stolon length ratio  ns  ns  ns  ns  ns  Branch ramet number ratio  ns  ns  ns  ns  ns  Mean RGR-total stolon length  ns  ns  ns  ns  ns  Mean RGR-total ramet number  ns  ns  ns  ns  ns  Mean RGR-total number of branches  ns  ns  ns  ns  ns  85  12.0  • Control Dactylis  11.0 Mean primary leaf area (cmxcm)  10.0 9.0 8.0 7.0 6.0 1  2  3 Experiment #  4  5  Fig. 3.2. Mean (±1 SE) of five primary leaves per white clover plant under the two treatments (reflecting bathers of live Dactylis glomerata (Dactylis) or control pots (Control)) for the five reflection experiments. Shown below is a summary of the ANOVA of mean primary leaf area by treatment (Neighbor) and experiment.  Source of variation  Sum of Squares  Degrees of freedom  Probability  138.0  4  0.000  Neighbor  2.3  1  0.039  Exp. X Neighbor  1.0  4  0.750  26.3  52  Experiment  error  86  Table 3.7. Summary of ANOVAs of white clover primary leaf position measurements made on both sides of the primary stolon during reflection experiments. Treatments had neighbors of either live Dactylis glomerata (D) or controls (C). The significance of the various effects (T=treatment, S=side of plant, i.e. towards grass or away from grass) is shown: *= p<O.O5, ns=not significant.  Primary leaf position measurement Height of leaf  Bxp. 1  Exp. 2  Exp. 3  Exp. 4  Exp. 5  ns ns ns  ns ns  ns ns  ns ns  *  *  T S TxS  ns  Lateral distance from stolon (excluding leaves that crossed stolon)  T S TxS  ns ns ns  115  fl5  *  ns ns  ns ns  ns ns  Number of leaves crossing stolon (using Yatescorrected Chi square)  T 5 TxS  ns  ns  *  *  fl5  *  ns  ns ns ns  ns  *  ns  ns  ns  Lamina angle  ns T ns S TxS ns  ns ns ns  ns ns ns  ns ns ns  ns ns ns  *  ns  ns ns ns  87  5.0  4.0 Height of primary leaves (cm) 3.0 2.0 1.0 0.0  -  -  -  • Towards D Away  -  -  -  Dactylis  Control  Fig. 3.3. Mean (1± SE) height of white clover primary leaves (Exp. 3) on two sides of the primary stolon, one towards the north (Towards), the other towards the south (Away). ANOVA results; “treatment” ns; “side” ns; “treatment x side” p.<0.05. (Dactylis Towards n=6, Away n=4; Control Towards n=6, Away n=7 leaves)  10.0  8.0  Lateral distance from stolon (cm)  • Towards D Away  6.0  4.0  2.0  0.0  Dactylis  Control  Fig. 3.4. Mean (±1 SE) lateral distance of white clover primary leaves (Exp. 4) on two sides of the primary stolon, one towards the reflecting wall (north), the other away from the wall (south). ANOVA results: “treatment”, p<O.05; “side”, ns; “treatment x side”, ns. (Dactylis Towards n=7 Away n=7; Control Towards n=7, Away n=7 leaves)  88  Plate 2. Side-view of clover (from east looking west) after the last four days spent with control pots (top) or live grasses (bottom) to the north (i.e., immediately to the right). Photographs were taken near noon on an overcast day.  90 Chapter Four THE EFFECTS OF PARTIAL SHADE FROM THREE SPECIES OF GRASS NEIGHBORS ON THE GROWTH AND MORPHOLOGY OF WifiTE CLOVER CLONES  4.1.  INTRODUCTION Interactions between a plant and its neighbors can occur through several  physiological systems simultaneously. Competition from neighbors is, by definition, ascribed to limiting resources either above- or below-ground. This experiment was designed to allow interaction only above-ground and to assess the effects of interference from neighboring grasses on the growth and form of a white clover clone. This experiment mimics a field situation where an individual clone or seedling of white clover experiences a small gap in the canopy. As a clover clone grows through a grass sward, it encounters a range of biotic and abiotic microenvironments. These patches of microhabitat are often imposed by neighboring grasses or by gaps within the sward. Such gaps can be created in many ways, e.g. patchy grass-growth, hoof-marks, non-uniform grazing, and deposition of dung pats (Parish, 1987). At different times of the day, different light conditions are experienced by plants in these gaps, depending on the location within the gap, the height of the neighbors, and the solar angle. In most situations when a clover plant is in a gap, the clover would have extended into the open area from a neighboring region having different conditions, and this is likely to affect its growth response in its new gap environment (Pitelka and Ashmun, 1985; Hutchings and de Kroon, 1994; and see Chapter Five). The experiment described in this Chapter examined the effects of shade from grass on clones of clover that were located entirely within gaps.  91 Attempts have been made to duplicate sward light conditions under controlled conditions in the greenhouse. For example, in studies by Solangaarachchi and Harper (1987) and Thompson and Harper (1988), target plants were arranged in such a way as to prevent root competition from neighbors. These target plants were shaded by canopies which were formed by healthy green leaves of neighbors that were either detached and floating in a transparent water tray above them (Solangaarachchi and Harper, 1987), or attached and held in position by wire frames (Thompson and Harper, 1988). A unique feature of the conditions imposed by these canopies was that, while preventing belowground interaction, they mimicked shade cast in the field, having the precise spectral qualities of natural conditions because it was being made by competitors’ leaves and contained many small sunflecks. This shade was presented uniformly, along with tiny sunflecks, obstructing the entire patch of available light above experimental clover plants. This shading pattern creates conditions which are more likely to be experienced by an understory plant in a dense woodland than a white clover plant in a pasture. The field conditions experienced by clover are commonly far more patchy.  In the present  experiment, shade from different species is presented in a more natural way, with clover clones continuing to receive large patches of unfiltered light. Gaps in a pasture will provide light with higher R:FR and PPFD than the light filtered through neighbors (Holmes and Smith, 1977a,b; McCartney and Unsworth, 1975). Under such a canopy, the overall R:FR would depend on the size of the available patch. In laboratory experiments, it has been shown that changes in the overall R:FR through the addition of incandescent light from individual bulbs can cause changes in plant response (Morgan and Smith, 1981). It is unknown how the reception of a distinct patch of higher R:FR and PPFD light modulates the response to patches of low R:FR and PPFD light. On sunny days, even within a tall sward, there are likely to be periods when direct sun is available for a duration longer than a large sunfleck (Parish, 1987). It is known that short periods of direct light can have profound morphological effects on  92 otherwise shaded plants under experimental conditions (Morgan and Smith, 1978; Pearcy et al., 1994), although it is unknown how this would modulate the response of clover to neighboring grasses. The present experiment was designed to present clover clones with shade from neighbors for only part of the day, to determine the effects of shade when some direct (unfiltered) light (much bigger than sunflecks) is also received by clover clones. In this study the different treatments were imposed by canopies of three different species of grass neighbors arranged in a standard design (see Fig. 2.2; Plate 3). This design created a gap among the grasses which presented all clover clones with equal exposure to direct light. The clones were placed in the same orientation within the gap. The design allowed different clover clones to receive shade from different species of neighbors for the same part of each day, while all clones continued to receive relatively high levels of PPFD. Thus the experiment investigates the effects of shading by different grass neighbors on the growth of white clover clones under conditions more typical of the field, primarily by presenting the various light stimuli in patches and allowing clones to receive PPFD typical of the level experienced by clover in the field. The clones under different grass canopies were compared to each other and to clones growing under no grass canopies, which served as controls for this experiment.  4.2.  METHODS  Preparation of plant material Clover clones and pots containing Dactylis, Holcus, and Lolium were prepared for the experiment as above (see Chapter 2). For this experiment, the grasses were clipped to 15 cm above the pot every four days. Due to natural variation in the density of leaves and tillers and in the architecture inherent in a canopy of each species, the precise nature of the shade cast by different grasses was expected to differ. The light transmitted and  93 reflected by these canopies was considered to be representative of the field, where different species of grass form clumps of varying density, SED, and competitiveness to white clover.  Arrangement of canopy modules Each treatment in this experiment consisted of two rows of canopy modules with two flats containing clover clones placed in the same orientation (north-south) between the rows (see Chapter 2.2.2, Fig. 2.2, Plate 3). Each aisle of standard width (15 cm wide x 17 cm tall) created a gap that allowed a “corridor” of sky running north to south, directly overhead, to never be obstructed. The amount of potential direct sunshine available to clover clones can be calculated from these dimensions and astronomical formulae for the position of the sun at given times on given days. This is shown in Fig. 4.1, and this was used to calculate the daily PPFD and average R:FR in the different treatments.  Treatment of plant material See Chapter for treatment of white clover clones, and Chapter for treatment of grass canopy modules.  Data collection and analysis Clover clones were measured as described in Chapter 2.4. Temperature and spectroradiometric measurements were also conducted as described in Chapter 2.6, with the following exception. Because the sun moved behind the walls of grass, casting a different pattern of shade and sunflecks, three such sets of scans were made at slightly different times during this period, in order to produce a better representation of the conditions. Analysis of variance between species could then be conducted between these measurements of PPFD and R:FR. In this experiment also, the daily PPFD and mean  94 daily R:FR could be calculated as a weighted sum or average, respectively, using the time spent at each level (from Fig. 4.1) as the weighting factor.  In addition to these  measurements, the density of grass tillers in pots of different species forming the canopies was determined shortly after the experiment. A transect was placed along a row of pots of each type at a standard height within the canopy of grass (2 cm below the clipping line). The number of different tillers touched by the transect was recorded, and this linear measurement was converted to the number of tillers per square centimeter. Also, once during the experiment, the height of regrowth in the four days between clippings was recorded for the three species of grass used to form the canopies. Data analysis for this experiment are described in Chapter 2.5.  4.3.  RESULTS  4.3.1. MICROCLIMATIC MEASUREMENTS  Spectral Energy Distribution Light measurements taken under simulations of the experiment indicate that there were large differences in the quantity and quality of light under the various types of canopy (Fig. 4.2). Data from these spectroradiometric scans indicate that all treatments received the highest PPFD and R:FR during the middle part of the day (Fig. 4.3, Table 4.1).  At times when sunlight was filtered by grasses, which occurred equally in the  morning and afternoon, the PPFD under the grass canopies fell to less than 14% of full sun (see “Morning, Afternoon” in Table 4.1). At the same time under grass canopies, the R:FR fell below 0.50, while the R:FR under control canopies remained above 1.1. At mid-day the PPFD and R:FR were slightly reduced by grass canopies (see “Morning, Afternoon” in Table 4.1). These effects were due to the increase in “blue” radiation in controls (more blue sky exposed) and the increase in radiation above 700 nm from  95 reflection in treatments with grass (Fig. 4.2c,d). With the sun less than 100 from the horizon, and blocked .by the wall of the experimental compartment, (see “Dawn, Dusk” in Table 4.1), grass canopies also reduced the PPFD and R:FR, although overall PPFD levels were quite small (<1% of full), and the time spent under this condition was limited (less than 2 hrs/day). At the canopy dimensions chosen, direct sunlight was available for approximately 4 hours during the middle of the day (at the ending date of the experiment) in treatments with grass canopies, while controls had 7 hours and 20 minutes of possible direct sunshine (Fig. 4.1). Using this and the above observations of PPFD, the clover clones under grass canopies received approximately 70% of the full total daily PPFD received by controls (Table 4.2). Although the effects of the different grass canopies on PPFD and R:FR were qualitatively similar, these observations indicate small differences among the grass species. In the mornings and afternoons when the sun was obstructed by either the east or west wall of the grass aisle respectively, Dactylis lowered the PPFD significantly more than canopies of Lolium but not more than canopies of Holcus (Table 4.1, Fig. 4.2c,d). Dactylis canopies also reduced the R:FR significantly more than canopies of both Holcus and Lolium at these times, and it appears that this was due to both an increase in PFD at wavelengths above 700 nm and a decrease below 700 nm (Fig. 4.2d). At times when the sun was within 100 of the horizon (see “Dawn, Dusk” in Table 4.1), Holcus canopies reduced the PPFD and R:FR more than canopies of the other two species, while Lolium canopies reduced the PPFD more than Dacrylis canopies.  Temperature Temperature measurements made under various conditions indicate no differences between treatments, except in the afternoon on sunny days (Table 4.3). Control clones experienced temperatures 1 20 C higher than clones under grass canopies at these times (and perhaps in the morning when sun was blocked). These were times when direct  96 sunlight would reach the space occupied by control clover clones, while shade was cast through grass canopies. The weather during the experiment was frequently sunny, so controls would have experienced elevated temperatures on several occasions.  The  relative humidity was likely to have been higher on these occasions under grass canopies due to the lower temperatures and proximity of the grass. At mid-day in the sun, the humidity could also have been higher, since the temperatures experienced at this time were quite similar. The same is true of overcast conditions, when the temperatures observed under different canopies were not significantly different, however there was still likely to be an elevation of water-vapor content due to the proximity of live grass.  4.3.2. GRASS CANOPY MEASUREMENTS Canopies of Lolium had a lower tiller density than the other two grass canopies (Table 4.4).  The density of leaves would have been 2-3 times higher than the density of.  tillers, as each species had approximately 2-3 leaves per tiller (data not shown). Since the width of leaves of these three species were quite different (approximately 4.0, 3.0, 2.Omm:Dactylis, Holcus, Lolium respectively), the leaf area clearly differed, with Dactylis having the most, followed by Holcus and then Lolium. Leaf growth in the four days between clippings, was the greatest in Dactylis, then Holcus and Lolium (Table 4.4). This observation again indicates that Dactylis glomerata canopies had the greatest leaf area followed by Holcus lanatus and then Lolium perenne.  4.3.3. CLOVER MEASUREMENTS  Yield All but one of the 32 clones (four treatments X eight replicates=32 plants) grew for the duration of this experiment without flowering. The one that flowered was a control clone, growing in the open, and it was excluded from any further data analysis.  97 The remainder of the plants were vigorous and grew with their primary stolon approximately down the center of the flat. By the end of the experiment, clones growing in control treatments (n=7) had produced a mean (±1 SE) of 48.7 (±1.6) cm of stolon, 35.0 (±1.0) ramets, 9.9 (±0.6) branches, and a biomass (when dried at 60° C for 96 hours) of 1.85 (±0.09) g. Each of these determinants of yield was substantially reduced by partial shading from all three species of grass  (p<O.Ol; Figs. 4.4a, b, c, d).  Differences in these yield variables among grass treatments were not as obvious as the differences between grass and control treatments. The Tukey multiple pair-wise comparisons indicate that, while there was no evidence of different amounts of reduction in the total stolon length and dry weight between different grass species, the different grass canopies caused different amounts of reduction in the number of ramets and branches produced by clover clones. Canopies of Holcus reduced the number of ramets (Fig. 4.4b) significantly more than canopies of Lolium, and canopies of Dactylis and Holcus reduced the number of branches significantly more than canopies of Lolium. There was no evidence of significant differences in any of the yield measurements made on clones under canopies of Holcus and Dactylis. However, the effects under canopies formed by Dact-ylis were always intermediate between the other two grasses. Although the effects of different grass canopies on clover stolon length and biomass were not significantly different, these values suggest the same trend, with canopies of Holcus having the largest effect, while those of Lolium had the smallest effect (Figs. 4.4 a,c).  Growth Every clone, regardless of canopy treatment, developed a highly symmetrical branching pattern with branches growing from successive nodes. Measurements made  98 over time indicate a regular and predictable increase in the total length of stolon, the number of ramets, and the number of branches (Fig. 4.5). The total number of ramets and total stolon length increased more or less exponentially in all treatments (Figs. 4.5a, b). In both measurements, the three grass canopies induced a significantly slower mean relative growth rate (i.e. the change in natural logarithm of the measurement per day, RGR) than control canopies, but under different grass canopies these were not significantly different. While there is a strong indication of an exponential increase in the number of branches on control clones (Fig. 4.5c; R =.994), the number of branches on clones under grass canopies increased much 2 more linearly 2 (R = .995, .999, .996; Dactylis, Holcus, Lolium).  The increase in the  absolute rate of appearance of branches (as is necessary for exponential increase), was seen somewhat in clones under control canopies before Day 25, when branches from secondary stolons first appeared. This early exponential increase in control clones must have been due to an increasing rate of appearance of new branches from the primary stolon. In the clover clones under grass canopies, secondary stolons rarely branched (see below), so it was only the primary stolon that produced branches. In these clones the rate of branch production (entirely by the primary stolon) was more or less constant, producing one new branch approximately every 3.6 days. Results of the measurements made over a two-day period during the experiment, where the primary petiole and internode on the expanding ramet of primary stolons were monitored, are shown in Table 4.5. While there was no indication of different extension rates for petioles in these clones, control clones and clones under Lolium had more rapid internode extension than clones under Dactylis and Holcus canopies. On clones under Holcus and Dactylis over the two-day period, the expanding petiole produced 3 to 4 times as much length as the expanding internode, while in control clones and clones under  99 Lolium, the expanding petiole produced less than twice as much length as the internode in  the same period. The above measurements of clover growth support the observations of final yield, which suggest that while all of the grass canopies restricted growth relative to controls, clones under Lolium were inhibited least by the canopy. Short-term evaluation of the rates of primary internode and petiole extension also indicate that clones grown under canopies of Lolium more closely resembled control clones with rapid internode and slow petiole extension, than clones grown under Holcus or Dactylis. Similar to final harvest measurements, the effects of Daclylis and Holcus canopies were not distinguishable.  Morphological measurements There are significant differences in several measurements describing biomass partitioning and organ dimensions (Tables 4.6, 4.7 and Fig 4.6). The most consistent differences are between clones under grass canopies and clones under control canopies, although there are some differences between clones under the canopies of the different grass species. Under all canopies approximately half of the clones’ biomass was allocated to leaf lamina (Fig. 4.6b; Table 4.6), although this was reduced under canopies of grass. On control clones the remaining half of the biomass was shared approximately equally by petioles and stolons, while on clones under grass, petioles retained approximately 60% of the remaining biomass, and stolons, 40%. Clones under Lolium had a significantly greater proportion of their remaining biomass in stolons than clones under the other two grasses, and a smaller percentage of biomass in petioles than clones under Dactylis. The ratio of petiole to stolon weight in clones under Lolium was significantly lower than clones under the other two grasses (Table 4.7). These large differences in biomass  100 partitioning between aerial structures were achieved at a relatively constant mean weight/ramet in clones under the different canopies (Table 4.7). Significant differences in internode length were observed only on the secondary stolons (first-order branches) of different clones, where controls developed longer internodes than clones under all of the grass canopies (Table 4.7). Although the mean internode length on secondary stolons were shorter on clones under grass, the mean petiole length for the whole clone (n=4; Table 4.7), and the ratio of petiole to internode length were significantly higher on clones under grass canopies than on control clones (Table 4.7), as was the case when considering only primary stolon structures (data not shown). The specific length (or specific area) of the three aerial primary ramet structures is shown in (Table 4.6). All clones under grass canopies produced more primary leaf area per unit weight than control clones, while clones under different species of grass did not differ. Primary stolons had longer specific length under Dactylis and Holcus than under controls, and primary stolons under Dactylis were longer on the same weight as under Lolium. While there were clear differences in mean petiole length due to partial shade, there was no indication of significant differences in the length per unit weight of petioles. Grass canopies caused an increase in biomass allocation to petioles and a decrease to stolons (Fig. 4.6).  The functional size of stolons (length) seems to have been  maintained through increases in the specific length. The measurements of biomass partitioning and organ dimensions also suggest the same trend as yield and growth measurements, with the morphology of clones under Lolium canopies intermediate between control clones and clones under the other two grasses.  101 Branching Table 4.8 contains several measurements of clover clonal branching. The number of new branches, or apices, produced was substantially decreased by grass canopies, while branch production by clones under Lolium canopies was intermediate between clones under control canopies and clones under the other two grasses. The primary stolons on all clones regardless of treatment produced roughly the same proportion of nodes with branches, and the age of the node having most recently branched on this stolon did not differ between treatments. Secondary stolons, on the other hand, had produced branches by the end of the experiment on control clones, whereas secondary stolons on clones under grass canopies scarcely branched. This led to a greater overall branching percentage on control clones over all others but to no differences in branching percentage between clones under different grass canopies. Branching in white clover can also be described by determining the proportion of the whole plant (in terms of dry weight, stolon length, number of ramets) found in branch structures as opposed to on the primary stolon (branch weight, length, and ramet number ratios; Table 4.8). The higher this value, the weaker the suppression of branches by the primary stolon. On control clones branching described in this manner was increased, with a greater percentage of each plant’s total stolon length, ramet number, and biomass accounted for by secondary branches. Again, Lolium canopies produced clones that were intermediate between control clones and clones under Holcus or Dactylis canopies. Since branch growth was followed over time, the strength of the suppression of secondary length and ramet production could be determined at specific times during the experiment (Table 4.9). The proportion of each plant’s growth allocated to the primary stolon declined throughout the experiment. There was greater suppression of branch length growth in clones under grass than in controls, and the level of suppression in clones under Lolium was intermediate between control clones and clones under Dactylis  102 or Holcus.  Even though suppression of branch length growth occurs under grass  canopies, there was no clear indication of suppression in branch ramet production under grass canopies.  Leaf Position Many leaves that originated on one side of a stolon were displayed in their final position on the other side of the stolon (Table 4.10). In control clones fewer primary leaves crossed the primary stolon than in clones under grass canopies. The height at which primary leaves were displayed during the experiment was also greater on clones under grass canopies than in controls (Table 4.10). Even though the mean primary petiole lengths were greater under grass canopies, the primary leaves on these clones were displayed at a steeper angle from their point of origin on the stolon, at a height that approached the height of the grass neighbors.  Summary Clones under the different canopies had visually distinct forms, ranging from the most branched with many ramets and thick leaves in controls, to fewer branches and fewer ramets in Lolium treatments, to very few branches, few ramets, and a large allocation to petioles in clones under Dactylis and Holcus. Table 4.11 shows a summary of the analyses of variance of the effects of the three grass canopies on several plant measurements. This indicates the measurements in which clones were the most distinct, and also indicates which clones differed among the groups under the three grass canopies. Measurements concerned with branching, petiole vs. stolon production, and clonal size in terms of ramet number and dry weight, differed the most between clones under canopies formed by different species of grass. In nearly all measurements, clones grown under canopies of Lolium were intermediate between clones shaded by Dactylis or Holcus and control clones. While  103 clones under Lolium were occasionally different from clones under Dactylis, they were more consistently different from clones under Holcus, and in none of these measurements were clones under canopies of Holcus different from clones under canopies of Daclylis.  4.4.  DISCUSSION The effects on morphology observed in clover clones under partial shade in this  experiment (reduced yield, increased petiole length, decreased branching) are generally consistent with the effects of complete shade at a wide range of daily PPFD (Solangaarachchi and Harper, 1987; Thompson and Harper, 1988; Thompson, 1993). Therefore, the effects of spending part of the day in full sunlight does not entirely alleviate the influence of shade received at other times during the day. This supports what is known about white clover’s demands for light (Haynes, 1980; Frame and Newbould, 1986) and suggests that small portions of the day with shade have an effect on morphology, illustrating the strength of the “shade avoidance” responses (Grime, 1981). The species of grass neighbor forming the small gap in this experiment also had a variable influence on the growth of clover clones within, suggesting that clover clones in the field may be able to adjust their phenotypes to the neighboring vegetation under partial-shade conditions and in the absence of below-ground interaction. The responses to neighboring grasses in the field are likely to involve responses to similar alterations in the light conditions, i.e. to shade interrupted by periods of direct sun, rather than to consistent day-long shading. The selective placement of ramets (foraging) in clones under natural partial shade in this experiment involved changes in ramet production, branching, and partitioning of biomass, rather than through modification of internode extension. Differences were also evident in resource-gathering structures, leaf lamina and petioles, suggesting that both types of modifications are adopted by white clover in response to partial shade from their neighbors.  Clones beneath Lolium canopies were nearly always intermediate in  104 phenotype between control clones and clones under the other two grass canopies, which suggests that many of the same changes in morphology can occur in response to relatively small changes in conditions due to different species of neighbors. Clones beneath all grass canopies produced internodes of similar length to control clones, however they were lighter per unit length (higher specific length), and in this regard stolon extension was increased in shaded clones. Leaf lamina size and specific area were also increased beneath canopies of grass. Petioles, on the other hand, were longer with proportionately more biomass (equal specific length), and they accounted for a greater proportion of total dry weight in clones under partial shade. Stolons and leaf lamina in these clones accounted for a lower percentage of total biomass, suggesting that while extension of stolons, larnina, and petiole were maximized, petiole extension may have had the highest priority in these clones. Among the different grass treatments, although there was a visual indication of longer petioles when observing clones grown under Holcus and Dactylis next to clones grown under Lolium, this was not statistically significant. The proportion of biomass that went into petioles, and the ratio of petiole to stolon biomass, however, was greater under Dactylis and Holcus, indicating more intense demands of this type under canopies of Dactylis and Holcus than Lolium. Branching in clones under grass was greatly limited, while the primary stolons had similar length and ramet number as in control clones (data not shown).  The  proportion of the plant accounted for by branches (Table 4.8), and the observations of growth of branches (compared to the primary stolon, in Table 4.9), describes the amount of branch inhibition caused by the primary axis, and under grass canopies a greater level of suppression led to more linear extension. While the restriction of length growth to primary stolons was greater in clones under grass at all three intervals, measurements of ramet number show no such suppression in branch ramet number production. Foraging in this regard does not appear to be conducted through modulation in the rate of  105 appearance of branch ramets, but rather through the suppression of growth in length of branch internodes. In support of this, secondary internodes were found to be shorter in clones under grass (Table 4.7). Among grass canopies, this type of foraging difference is most clearly seen between clones under Holcus and clones under Lolium canopies. Canopy modules of Lolium developed a lower tiller density than the other two grasses, even though grown under the same fertilizer and spacing conditions. However, the observed effects on temperature and light conditions beneath canopies of Lolium and Holcus were nearly indistinguishable. The effects of Daclylis and Holcus canopies on the clones growing beneath them were very similar (Holcus seemed to have a stronger effect), yet for much of the afternoon, Dactylis lowered the R:FR significantly more than canopies of Holcus. The effect of these reductions in R:FR beneath Dactylis could have been alleviated or offset by conditions at some other part of the day, or perhaps by some other (unmeasured) feature of the canopies. The conditions observed at Dawn, Dusk (Table 4.1), indicate that Holcus had the largest effect on PPFD and R:FR at this time, and this is consistent with the observed effects of Holcus canopies on clonal morphology. Light conditions at the end of the day, especially R:FR, can have large effects on plant morphology (Kasperbauer, 1971; Smith, 1982; Casal and Smith, 1989b), and this may account for the strong response beneath Holcus canopies. In Thompson and Harper (1988), growth of clover clones also was affected more substantially beneath canopies of Holcus than canopies of Lolium, and this occurred at a constant PPFD, while the shade differed in R:FR. In the present experiment, shade cast from these canopies reduced PPFD and R:FR equally for a large majority of the day. The difference between these effects of the canopies on the light measurements might have been caused by the presence of leaf pubescence, which, it has been suggested, increases its reflection of red light (Thompson and Harper, 1988). With only small gaps between leaves in a canopy of Holcus, as in Thompson and Harper (1988), there would be no  106 source of light rich in red wavelengths to be reflected from leaf hairs. With a relatively large gap in the canopy, as in the present experiment, and the sun shaded, there would still be a source of light relatively rich in red light (unfiltered diffuse sky or clouds) available for reflection from the leaf hairs, and the R:FR detected beneath “equally” dense shade would be increased. Since all grasses were clipped to 15cm for all spectroradiometric measurements, the growth of grasses that took place between clippings in the experiment is not reflected in these measurements. Each centimeter of additional height in the grasses had the potential to interfere with sunlight for several minutes each morning and afternoon, and the differences in height growth between species were significant (Table 4.4). While Holcus grew taller than Lolium, and this may explain the differences between growth under these two canopies, Dactylis grew significantly taller than Holcus between clippings, yet there was little difference in the resulting clover morphology. From the data on light conditions, grass tiller densities, leaf widths, and height growth between clippings, it was expected that canopies of Lolium would be the least influential, that its effects would resemble the effects of Holcus canopies more closely than Dactylis canopies, and that Dactylis canopies would be the most influential. There seems to be a quality of Holcus canopies (perhaps its effects on end-of-day R:FR) that gives it a relatively strong influence for its size. This result may reflect the strong competitive ability of Holcus observed in the field towards many of its neighboring species (Haynes, 1980; Solangaarachchi, 1985; Turkington et al., 1991). There is another possible explanation for the strong effect of Holcus on the growth of clover in these experiments. Between clippings (every four days in the present experiment), grass grew into the space formed within the corridor of canopy modules. Little intermingling with clover leaves occurred, but small differences in the width of the corridor could have affected the conditions experienced by clover clones within. This  107 temporary encroachment was not measured, but it appeared that Holcus canopies extended most obviously into this area between clippings. So, even though clipped to the same dimensions every four days, its effective size may have been larger than Dactylis canopies, which primarily grew upwards. Regrowth between clipping beneath Holcus canopies may have had an influence on the morphologies observed in Thompson (1987) (7 days) and Thompson and Harper (1988) (3 days), as well as in the present experiment. The vigorous lateral growth of Holcus lanatus has been attributed with a role in its success in the field (Solangaarachchi, 1985). The foraging expressed in clones under partial natural shade in this experiment involved changes in ramet placement through changes in ramet production, branching, and partitioning of biomass, and modification of resource-gathering structures through changes in biomass partitioning, leaf lamina and petioles.  A similar pattern of  modifications to phenotype occurred under Lolium canopies when compared to clones under Dactylis and Holcus, which may suggest that many of the same changes to morphology occur in response to slight alterations in conditions due to different species of neighboring grass. Finally, many of the characters shown to respond most sensitively to these small differences in partial-shade conditions might be likely to respond to subtle changes in the light conditions, e.g., increased FR from reflection off neighboring grass (Chapter 3). It appears that a particular phenotype in white clover can be developed in response to the species of neighbor growing around a gap. Success of clover in the field might rely on its ability to alter its morphology when growing with different grasses (Evans and Turkington, 1988). An appropriate morphology might be necessary for the successful capture of resources, allowing persistence under particular conditions. These results show that the different species of grass can induce different morphologies in  108 clover clones under standard, partially-shaded conditions. This stresses the importance in plants of responding to relatively small differences in natural light conditions.  109  Plate 3. Layout of partial shade experiment showing canopy modules, flats of clover, and opaque barriers.  111  waflof grass  47.6 17cm  pot of soil  vaflof grass  15cm  pot of soil  10cm  66.2  flatof clover  10cm 15cm  80 70 a.)  I 50  — — — —  0  Elevation needed-grass Elevation needed-control  40 Sun-April 1 Sun-May 1  30 20 10 0 0  1  2 3 4 Hours from mid-day  5  6  Figure. 4.1. above) The canopy dimensions used in the partial shade experiment. In this cross-section from point A in the middle of the flat of clover, the angle perpendicular to the wall of grass is 66.2 degrees (vertical: 17.0cm, horizontal: 7.5cm). Using the fact that the further south along the top of the wall that the sun crosses it, the lower the elevation angle needed, the potential hours of sunshine during the experiment can be estimated for grass “corridors” and controls (perpendicular angle to experimental compartment wall 38.7 degrees; vertical: 20.0cm, horizontal; 25cm, not shown). below) are the elevation angles needed to clear the grass barriers and experimental compartment walls depending on the time of day (time angle) at which the sun would cross, and the plots of the sun’s elevation by the time of day, near the starting (April 1) and ending points (May 1) of the experiment.  112.  D. glomerata  (a)  -  6.0—  5.0—  jS/\ J’  -.  _.‘  f  Control  /‘___  4.O-  0  L. perenne  /__  .  —  .  H. lanatus  I Vi  a)  D  o o  ‘  q  I  ‘7  3.0  :H  I  400  500  600  700  (b) 6.0  3.0—  2.0—  1.0  0.0— 650  I 670  I  690  710  730  750  Wavelength (nm)  Fig. 4.2. Spectroradiometric measurements made on a clear day at solar noon (a) and (b), and three hours after solar noon (c) and (d) under simulations of the experiment. Note that (b) is a closer the red and far-red wavelengths in (a), and (d) is a closer view of the red and far-red wavelengths in (c).  view of  D. glomerata (c) H. lanatus 4.0 L. perenne Control  3.0— C)  a 2 ‘4--  0  .,  E 0  1.0-  .  400  500  600  700  (d)  3.0—  20—  650  670  590  710  730  750  Wavelength (nm)  Fig. 4.2 Spectroradiometric measurements made on a clear day at solar noon (a) and (b), and three hours after solar noon (c) and (d) under simulations of the experiment. Note that (b) is a closer view of the red and far-red wavelengths in (a), and (d) is a closer view of the red and far-red wavelengths in (c).  114  “  1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 0  CDHL CDHL CDHL MORNING MID-DAY AFTERNOON  (b)  1.2 1.0  R:FR  0.8 0.6 0.4 0.2 0.0  Figure 4.3. Photosynthetic photon flux density (PPFD) and the red:far-red ratio (665665:725-735; R:FR) during three periods on a clear day under Controls (C), Dactylis glomerata (D), Holcus lanatus (H), and Lolium perenne (L). Morning and afternoon values are derived from 3 sets of scans (330-1100 nm, using a LICOR LI-1800 portable spectroradiometer) taken at slightly different times one afternoon, hence the error bars (±1 SE). Means at these times with different letters above are significantly different (P<0.05; ANOVA with Tukey multiple comparisons. Times corresponding to these periods can be calculated from Figure 4.1. Treatments with grass canopies received approximately 4 hours of sunshine around mid-day and 1 hour 40 minutes of leaf-filtered shade each morning and afternoon.  115 Table 4.1. Photosynthetic photon flux density (PPFD) as a percentage of Control, and the red:far-red ratio (R:FR, quantum ratio of 665-665:725-735) under four different canopies. Measurements were taken on a sunny day at solar noon (Mid-day), 3 hours after solar noon (Morning, Afternoon), and <1 hour from sunset (Dawn, Dusk) using a LICOR LI- 1800 portable spectroradiometer. Different letters preceding “Morning, Afternoon” PPFD and R:FR indicate that the measurements under the different canopies were significantly different (ANOVA with Tukey multiple comparisons) Time of day Canopy type  Morning, Afternoon  Mid-day  Dawn, Dusk  Dactylis glomerata  PPFD R:FR  a 12.8 a 0.35  98.2 1.08  40.6 0.92  Holcus lanatus  PPFD R:FR  a,b 13.1 b 0.45  99.8 1.09  20.3 0.82  Loliumperenne  PPFD R:FR  b 14.0 b 0.48  99.1 1.10  30.4 0.91  Control  PPFD R:FR  c 100 c 1.11  100 1.20  100 1.26  Table 4.2. Total photosynthetic photon flux density per day, and this as a percentage of Control on a theoretical sunny day calculated from the time spent at “Morning, Afternoon” and “Mid-day” levels (from Fig. 4.1). This is a weighted sum (PPFD) or weighted average (R:FR) from the time spent at each of the observed levels. PPFD/day (molIm /day) 2 and as percentage of control  Mean daily R:FR  Dactylis glomerata  (18.9) 68.9%  0.75  Holcus lanatus  (19.2) 70.1%  0.80  Lolium perenne  (19.2) 70.0%  0.82  Control  (27.4) 100%  1.16  Canopy type  116  Table 4.3. Mean temperatures under the different canopies at two different times on a sunny day, and on an overcast day. Means in each column that are preceded by a different letter are significantly different (p<O.05; ANOVA with Tukey multiple comparisons). PDT = Pacific Daylight Time.  Canopy type  Temp (°C) 1320 PDT on a sunny day (±1 SE)  Temp (°C) 1600 PDT on a sunny day  Temp (°C) 1300 PDT on an overcast day  Daclylis glomerata (n=4)  a 24.4 (0.3)  a 21.0 (0.3)  a 20.7 (0.3)  Holcus lanatus (n=4)  a 24.4 (0.3)  a 20.8 (0.2)  a 20.2 (0.3)  Loliumperenne (n=4)  a 24.8 (0.2)  a 21.5 (0.2)  a 20.4 (0.2)  Control (n=4)  a 25.0 (0.3)  b 22.6 (0.3)  a 20.8 (0.2)  Table 4.4. The density of grass tillers and the height attained between four-day clippings. Grasses were grown in 10cm pots and used to form the canopies in this experiment. Means in each column that are preceded by a different letter are significantly different (p<O.O5; ANOVA with Tukey multiple comparisons among grasses)  Grass tiller density ) 2 (tillers/cm  Grass height above clipping (cm)  Dactylis (n=8)  a 1.19 (0.14)  a 4.8 (0.2)  Holcus (n=8)  a 0.97 (0.07)  b 3.5 (0.2)  Lolium (n=8)  b 0.68 (0.04)  C  Canopy type  Control (n=7)  0.00  2.4 (0.2) 0.0  117  (a)  60  50 45 40 Total stolon 35 length (cm) 30 25 20 15 10 5 0 Control  Dactylis  Holcus  Lolium  Neighbor  40  a  b,c  b  C  (b) 30 Number of ramets  25 20 15’ 10’ 5. Control  Dactylis  Holcus  Lolium  Neighbor  Figure 4.4 (a,b). Mean (±1 SE) yield at final harvest in (a) total stolon length, (b) total number of ramets, (c) total above ground dry weight, and (d) total number of branches under four different canopies: Control, i.e. empty pots (n=7), Dactylis glomerata (n=8), Holcus lanatus (n=8), and Lolium perenne (n=8). Bars with different letters above are significantly different (p’<0.O5; ANOVA with Tukey multiple comparisons).  118  (c)  2.2 2.0 1.8 1.6 Total aboveground dry 1 4 wieght (gm) 1.2 1.0• O.8 O.6 0.4 0.2 0.0  a  b  b  b  Holcus  Lolium  -  Control Dactylis  Neighbor  12 (d) ii10 9. Number of 8: branches  -  a  b  b  c  1  -  5 43 2 1 0Control, Dactylis Holcus Neighbor  Figure 4.4 (c,d)  Lolium  lag (a)  55 50 45 40  Total stolon length (cm)  Control  35  • Dactylis • Holcus o Lolium 15  (b)  05  10  o  10  15  20  25  20  25  40 35 30  Number of 25 ramets  1;  I....,  0 5  I  15  12.0 11.0 10.0 9.0 8.0 Number 7.0 of branches  (c)  3.0 2.0  0  5  10  15  20  25  30  Days  Figure 4.5.  Increase of mean  (±1 SE)  in (a)  total  number of branches, in clover clones grown under  the  best-fitting  exponential  number in grass treatment.  curve  (y=alObX)  stolon  the  length, (b)  number of ramets, (c)  four different treatments.  or linear regression  (y=ax+b)  Shown is branch  for  120  Table 4.5. Mean (±1 SE) of primary petiole and internode growth and their ratio in clones under four different canopies. Over a two-day period growth of the expanding portion of primary stolons was monitored. Within each column, means preceded by a different letter are significantly different (p<O.O5; ANOVA with Tukey multiple comparisons).  Canopy type  Primary petiole growth (cm/day)  Primary internode growth Petiole! (cm/day) internode growth  Dactylis glomerata (n=8)  a 1.55 (0.22)  a,bO.61 (0.08)  a,b2.90 (0.69)  Holcus lanatus (n=8)  a 1.57 (0.19)  a 0.45 (0.08)  a 3.85 (0.82)  Loliumperenne (n8)  a 1.18 (0.20)  b 0.88 (0.12)  b 1.62 (0.36)  Control (n=7)  a 1.35 (0.16)  b 0.76 (0.07)  b 1.92 (0.35)  121  Table 4.6. Mean (±1 SE), describing the proportion of total biomass allocated to each white clover structure, primary ramet mean leaf area or petiole and internode lengths, and the specific area or length of each part on primary stolons of white clover under four different canopies (D=Dacrylis glomerata, H=Holcus lanatus, L=Lolium perenne, C=control). Values in a column subset preceded by a different letter are significantly different (p<O.05; ANOVA with Tukey multiple comparisons).  Plant character  % of total plant dry weight allocated to character  Canopy type  Primary ramet mean  Leaf lamina area  D H L C  a 48.3 a,b49.3 a 49.0 b 50.6  (0.3) (0.7) (0.3) (0.1)  2 a 12.9 (0.2) cm a 13.3 (0.6) a,b12.2 (0.5) b 11.0 (0.3)  Leaf petiole length  D H L C  a 31.4 a,b30.7 b 29.4 c 25.4  (0.2) (0.5) (0.3) (0.2)  a a a b  Internode length  D H L C  a a b c  (0.3) (0.6) (0.3) (0.2)  a 2.41 (0.05) mm a 2.40 (0.09) a 2.35 (0.10) b 2.28 (0.05)  20.3 20.0 21.7 24.1  13.1 13.3 12.4 10.8  (0.1)cm (0.3) (0.4) (0.3)  Primary ramet specific leaf area lmg) or 2 (mm specific length (mm/mg) a 3.34 a,b 3.23 a,b 3.23 b 2.85  (0.05) (0.08) (0.11) (0.13)  a a a a  5.29 5.31 5.46 5.28  (0.11) (0.19) (0.16) (0.25)  a 1.31 a,b 1.25 b,c 1.19 c 1.10  (0.03) (0.03) (0.03) (0.03)  122  Table 4.7. Means (±1 SE) of morphological measurements of clover clones grown under four different canopies. Values of each measurement preceded by a different letter are significantly different (p<O.05; ANOVA with Tukey multiple comparisons).  Canopy type Plant character Primary stoloninternode length (mm)  Dactylis glomerata  Holcus 1 anatus  Lolium perenne  Control  a 24.1 (0.5)  a 24.0 (0.9)  a 23.5 (1.0)  a 22.8 (0.5)  a 9.6 (0.4)  a 9.0 (0.6)  a 10.4 (0.6)  b 12.3 (0.4)  Wholeplantinternode length (mm)  a 14.4(0.3)  a 14.3(0.6)  al4.4(0.6)  a 13.9(0.3)  Mean weight per ramet (mg)  a 54.5 (1.8)  a 57.1 (3.0)  a 54.9 (2.5)  a 52.9 (2.1)  Petiole weight! stolon weight  a 1.55 (0.03)  a 1.54 (0.05)  b 1.36 (0.03)  c 1.05 (0.02)  Whole plant-mean petiole length! seconday stolon internode length (n=4)  a 8.0 (0.2)  a 7.8 (0.3)  a 7.6 (0.1)  b 6.2 (0.1)  Secondary stoloninternode length (mm)  123  (a)  Mean dry weight (gm)  1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0  b a  a  a Leaf lamina 0 Petiole Stolon  ab b  Control  Dactylis  Holcus  Lolium  Neighbor  (b)  Percentage of total biomass  55 50 45 40 35 30 25 20 15 10 5 0  b Leaf lamina Petiole Stolon  Control  Dactylis  Holcus  Lolium  Neighbor  Figure 4.6. Mean (±1 SE) of (a) dry weight per plant of the three above-ground plant structures and (b) the percentage of above-ground biomass allocated to each of these structures under the four different canopies: Control, i.e. empty pots (n=7), Dactylis glomerata (n=8), Holcus lanatus (n=8), and Lolium perenne (n=8). Means of the same structure with a different letter above are significantly different (p<O.OS; ANOVA with Tukey multiple comparisons).  124  Table 4.8. Means (±1 SE) of branching measurements made on clover clones grown under four different canopies. Values of each measurement preceded by a different letter are significantly different (p<O.05; ANOVA with Tukey multiple comparisons).  Canopy type Dactylis glomerata  Holcus lanatus  Lolium perenne  Control  Number of new branches produced  a 5.50 (0.19)  a 5.38 (0.18)  b 6.38 (0.32)  c 9.86 (0.55)  Primary stolonbranching (%)  a 67.7 (2.2)  a 69.4(2.1)  a 71.4(1.8)  a 71.0(2.1)  a 2.6 (1.4)  b 16.4 (1.2)  Plant character  Secondary stolonbranching (%) Whole plantbranching (%)  0  (-)  0  (-)  a 22.6 (0.8)  a 24.6 (0.7)  a 24.8 (0.7)  b32.3 (0.3)  Branch length ratio  a,b 43.9 (1.9)  a 40.3 (1.4)  b 48.1 (1.4)  c 58.5 (0.6)  Branch ramet number ratio  a,b 66.3 (1.5)  a 64.4 (0.8)  b 68.2 (1.0)  c74.6 (0.5)  a 49.5 (1.4)  a 46.5 (1.3)  b 53.6 (0.9)  c 61.5 (0.8)  Branch weight ratio  125  Table 4.9. Percentage of new growth in total stolon length and total ramet number (mean, 1 SE) that occurred on the primary stolon, during three consecutive intervals, under four different canopies. Values at each interval preceded a different letter are significantly different (p.<O.O5; ANOVA with Tukey multiple comparisons).  Canopy type Dactylis glomerata (n=8)  Interval (Period-days) 1 (6-12)  Percentage of Percentage of all new growth all new growth by primary by primary stolon (length) stolon (number) a,b38.3 (2.6)  a 36.5 (6.3)  Holcus lanatus (n8)  a 45.6 (2.6)  a 41.7 (3.2)  Loliumperenne (n=8)  b 36.6 (2.7)  a 35.8 (4.4)  Control (n=7)  c27.7 (2.6)  a 44.6 (4.1)  Daclylis glomerata (n=8)  2 (12-18)  a,b17.7 (1.2)  a 22.3  (2.0)  Holcus lanatus (n=8)  a 21.4 (1.8)  a 26.2  (3.3)  Loliumperenne (n=8)  b 14.8 (1.1)  a 21.7  (1.7)  Control (n7)  c 10.1 (0.7)  a 21.7  (0.7)  a,b 6.6 (0.6)  a 18.7  (1.8)  Holcus lanatus (n=8)  a  7.5 (0.4)  a 18.6  (2.1)  Loliumperenne (n8)  b  6.3 (0.3)  a 15.4  (1.6)  Control (n=7)  c  4.1 (0.1)  a 14.1  (1.4)  Daclylis glomerata (n=8)  3 (18-25)  126  Table 4.10. Measurements describing the fmal position of primary leaves (mean, 1 SE). A primary leaf crossed the stolon if its lamina was displayed entirely on the side opposite its origination from the stolon. The height of these leaves was determined by the maximum height of the highest primary lamina on each plant. Means preceded by a different letter are significantly different (p<O.O5; ANOVA with Tukey multiple comparisons).  Mean # of primary leaves crossing stolon  Mean height of highest primary leaves  Dacrylis glomerata (n8)  a 3.6 (0.8)  a 14.1 (1.1)  Holcus lanatus (n8)  a 3.6 (1.4)  a 15.3 (0.8)  Loliumperenne (n=8)  a 3.6 (1.4)  a 14.8 (1.4)  Control (n=7)  b 1.1 (0.8)  b 8.8 (1.6)  Canopy type  127 Table 4.11. Summary of analyses of variance (ANOVA model: Variable = Constant + Treatment + residual) using grass canopies (3 species: D=Daciylis glomerata, H=Holcus lanatus, L=Lolium perenne) as the treatments. (n=8) for all treatments. Comparisons of means having t tests (Tukey) with p<0.05  % sum of squares between groups  significance  Branch weight ratio  45.4  0.002  H<L  Petiole/stolon weight  44.1  0.002  D,H>L  Petiole weight ratio  42.9  0.003  D,H>L  Branch length ratio  36.9  0.008  H<L  Number of branches  32.8  0.015  D,H<L  Stolon weight ratio  31.9  0.018  H<L  Number of ramets  29.4  0.026  H<L  Primary stolon specific length  26.8  0.03 8  D<L  Branch number ratio  21.0  0.084  Whole plant-stolon specific length  19.8  0.097  Total stolon length  19.5  0.103  Total biomass  16.0  0.161  Secondary stolon internode length  13.3  0.222  Leaf lamina weight ratio  12.0  0.260  Mean primary leaf area  11.4  0.279  Mean primary petiole length  9.1  0.365  Primary leaf specific area  6.4  0.501  Primary petiole specific weight  3.0  0.732  Primary stolon internode length  1.4  0.857  Whole plant internode length  1.0  0.988  Variable  128 Chapter Five IS THERE AN EFFECT OF REMOTE CANOPY CONDITIONS ON THE GROWTH AND MORPHOLOGY OF LOCAL REGIONS OF WHITE CLOVER CLONES?  5.1.  INTRODUCTION Clonal plants are widespread in many of our natural and agricultural ecosystems,  and this has led to a great increase in studies on how plants respond to heterogeneous conditions around interconnected portions of a single plant. In the field, the extending apex of a clover stolon occupying a gap in the canopy may reach the edge of the gap and begin extending into the sward. In these two adjacent “patches”, the apical region will be experiencing conditions typical of a dense sward, while the rest of the plant remains under a far more open canopy.  In a white clover plant under such a two-patch  environment, it may be advantageous for support from basal regions to be reduced, to limit investment in the apex and focus new growth in the resource-rich patch. On the other hand, support from basal regions under these conditions may be increased in order to reduce the effects of more intense apical-region competition. The strategies which can be adopted by clonal plants are numerous and have been discussed in Chapter 1. Under patchy conditions, an ability to regulate the extent and direction of integration under different conditions might improve a clover plant’s ability to efficiently exploit the environment. These are traits which could be selected in white clover if under genetic control, and/or could develop in the clones and remain “programmed” into their growth patterns (Evans and Turkington, 1988). To establish if these patterns of within plant integration of ramets are important in white clover, it first would be useful to determine if, without directly influencing the remote portion of the clone itself, conditions existing in one patch can influence the growth of portions in an adjacent patch. It is useful at first to use “large” patch-differences which are defined so by being near the  129 extremes of what the plant might encounter in the field. If no effect is found due to connection with a distant portion of the plant that is manipulated, then the response of a local portion of a plant in this two-region system would be a local response. A great complication to studying this in the natural habitat of many plants is that the conditions existing over one part of the plant are rarely independent of the conditions existing over another part. For example, if one part of a plant is next to very tall neighbors, while another is farther away, the influence of the tall neighbors is likely to be experienced directly by the portion farther away. Under experimental conditions, it is often difficult to alter the conditions experienced by one portion of a plant without directly altering the conditions of the interconnected portion. Unless this is satisfied in studies of this type, how much of the observed effect is due to local conditions and how much is due to an influence from the conditions on the other portion cannot be determined. It is only when the two portions of a plant remain connected but in isolated environments, that these two can be disentangled. Experiments of this design using white clover need to pay special attention to isolation of the regions of the plant due to the relatively short distances between interconnected ramets, and this precaution has not always been observed (e.g. Newton, 1986; Solangaarachchi, 1987; Turkington et al., 1991). White clover has relatively short connections between ramets when compared to other clonal plants used in similar studies, such as Fragaria chiloensis, Glechoma hederacea, and Lamiastrum goleobdolon, which have longer stolons between ramets. A series of experiments was designed to determine if there was a response in local portions of white clover clones to the conditions experienced by interconnected portions. The portions of the plant manipulated and examined for response correspond to the fieldpatch size observed in a pasture with a relatively fine-grained mosaic of species (e.g., permanent pastures in B.C., Aarssen and Turkington, 1985). These experiments mimic field situations where an individual white clover plant experiences different conditions  130 over an interconnected length of one stolon. Are the conditions responded to in isolation, or is there evidence that the particular conditions experienced by the distant portion of the plant can invoke a local response; i.e., are they integrated? If integrated, what types of patterns are common; are ramets under poorer conditions supported by ramets under better conditions, or are ramets under better conditions supported by ramets under poorer conditions? Do these patterns change under different conditions? A better understanding of these questions should increase our ability to maintain and manage the natural habitats of clonal plants and to use them efficiently in production systems. The objective of the following experiments is to determine if and under what canopy conditions a localized region of a clover clone responds to shading experienced by a remote region of the clone. These experiments subjected clones to two canopy patches along the length of the primary stolon. The dependence of basal-region growth on apical-region conditions was examined by manipulating the neighborhood occupied by the apical region, using shade from natural neighbors, and examining for response in the basal region of the clone, (i) under open conditions, and (ii) under naturally shaded conditions. The dependence of apical-region growth on basal-region conditions was examined by manipulating the neighborhood occupied by the basal region with shade from neighbors and examining the response in the apical region of the clone, i) under open conditions, and ii) under shaded conditions. A lack of integration within the plant under all or some of these conditions may indicate if and when physiological integration plays a major role in determining white clover morphology and foraging behavior under patchy canopy conditions.  5.2.  METHODS Canopy modules and clover clones were prepared for these experiments as described  in Chapter 2.1).  The arrangements of canopy modules used in these experiments are  described in Chapter 2.2.3 (and see Fig. 2.3, Plate 4). In these experiments, an extra length of time was required to prepare clones for the experiments. During this period, clover clones  131 were grown under uniform conditions in the experimental chambers without a canopy of grass, until they were large enough to have a primary stolon with several nodes (4-7) and a rapidly-expanding apex (19-31 days). This was done to ensure that there were patch-sized portions of clover clones available to experience local canopy conditions. To ensure that clones were of uniform size and developmental stage when encountering the second neighborhood, all plants were repositioned simultaneously at the start of each experiment, with only the apex of the primary stolon in the second chamber (see Fig 2.3). The exact position to which the plants were moved was determined by fixing the barrier at the node of the same numbered leaf in each clone (nodes 5 to 8 for the various experiments), and this always left the apical portions of each clone with at least one and no more than two ramets in the second neighborhood at the start of each experiment. In this arrangement then, it was the length of the segment of primary stolon in the two chambers that differed from clone to clone, rather than the number of branches or ramets. Six clones were randomly assigned to the two treatments used in each particular experiment. The methods of data collection and analysis in these experiments are described in Chapter 2.4, 2.5. Details relevant to each particular experiment are described below.  5.2.1. DOES APICAL-REGION SHADING AFFECT THE GROWTH AND MORPHOLOGY OF BASAL REGIONS OF CLOVER CLONES? (BASIPETAL TRANSFER OF RESPONSE)  Basal region open The canopies designed for this experiment allowed all clones to experience an open canopy around basal portions, while apical regions of experimental clones grew behind a canopy formed by live grass plants (see Fig. 2.3a). In controls, the apical region remained in the open for the duration of the experiment. All clones grew entirely under an open canopy for 18 days before being exposed to the second neighborhood. At this  132 time all clones were placed under the dividers between the two compartments at the same node on the primary stolon, leaving four basal nodes with branches, and allowing the newest one or two nodes to extend into the new treatment neighborhood. The clones remained like this for two days before the canopies were constructed in the appropriate neighborhoods. Clones grew for 19 days (April 20, 1991 to May 8, 1991) before being harvested.  Basal region shaded The canopies designed for this experiment allowed apical regions of different clones to experience different conditions, while basal regions of all clones grew behind a grass canopy (Fig. 2.3b). As above, in control treatments the apical region remained in the open for the duration of the experiment, while the apical region of the primary stolon in experimental clones grew in the shade of a grass canopy. As in all of the experiments reported in this chapter, all clones in this experiment grew entirely under an open canopy before encountering the second neighborhood. After 30 days the clones were large enough to be placed under the wall, leaving five basal nodes with branches and only the most apical node in the second neighborhood.  Two days later, the canopies were  constructed in the appropriate neighborhoods. Clones grew for 31 days (October 3, 1991 to November 3, 1991) before being harvested.  5.2.2. DOES BASAL-REGION SHADING AFFECT THE GROWTH AND MORPHOLOGY  OF APICAL REGIONS  OF  CLOVER  CLONES  ?  (ACROPETAL TRANSFER OF RESPONSE)  Apical region open The canopies designed for this experiment presented all clones with an open canopy over apical portions, while basal regions of experimental clones grew in the shade  133 of a grass canopy (Fig. 2.3c; Plate 4). In controls, the basal region remained in the open for the duration of the experiment. All clones grew entirely under an open canopy for 30 days before being repositioned, having six basal nodes with branches and their most apical node in the second neighborhood.  Two days later, canopies of grass were  assembled in front of the basal portions of experimental clones (Fig. 2.3c.) Clones grew for a further 51 days (November 21, 1991 to January 11, 1992) before being harvested.  Apical region shaded The canopies designed for this experiment allowed basal regions of clones to experience different conditions while all clones grew with apical regions in the shade of a grass canopy. Basal regions of experimental clones grew in the shade of a grass canopy (Fig. 2.3d). As above, in control treatments, the basal region remained in the open for the duration of the experiment. After 27 days of growth behind an open canopy, the clones were large enough to be placed with their apex in the new neighborhood, leaving six basal nodes with branches. Two days later the appropriate canopies were assembled. The clones grew for 18 days (August 6, 1991 to August 24, 1991) before being harvested.  5.3.  RESULTS  5.3.1. DOES APICAL-REGION SHADING AFFECT THE GROWTH AND MORPHOLOGY OF BASAL REGIONS OF CLOVER CLONES? (BASIPETAL TRANSFER OF RESPONSE) Since in this pair of experiments it is only the response of basal regions that is necessary for determining the existence of intra-plant integration, results of the measurements made on this region only are presented.  134 Basal region open Of the six clones under each treatment, one of the clones with the apical region under a closed canopy produced flowers from ramets in the basal region, and it was excluded from the analyses. There were no significant differences in any of the clonal measurements made on basal regions of the remainder of these plants (Table 5.1).  Basal region shaded In this experiment, one clover clone under each treatment, reached the edges of the flats well before the others, and they were harvested early and excluded from further analyses. A photograph of two clones near the final harvest of this experiment is shown in Plate 5. It appears that the clone below (apical region shaded) has a less compact design and appears to be slightly smaller. However, few significant differences in morphological measurements were detected between treatments.  There were no  significant differences in the final yield, branching measurements, or internode lengths between clones having their apical regions under different conditions (Table 5.2) The data suggest, however, that there might have been a slight improvement in yield due to apical regions that were unshaded. There was also a strong suggestion that biomass allocation patterns differed between the basal regions of these clones.  While the  allocation to leaf lamina was similar in the two treatments, of the remaining portion of above-ground biomass, a greater proportion was allocated to petioles rather than stolons in clones which had apical regions under a closed canopy (58.0 ±0.3% apical canopy closed, 56.5 ±0.5% apical canopy open;  p<0.05 Student s t-test). t  However, there was no  evidence of differences in final petiole or stolon internode lengths. The reduction in allocation to stolon weight in basal neighborhoods of clones with their apical region behind grass canopies, was also reflected in the mean dry weight of the portion of primary stolon that traversed this neighborhood (0.083 ±0.002 g vs. 0.100 ±0.006 g; p<O.05 Student’s t-test).  Although this structure had been formed prior to the the  135 initiation of the canopy treatments, by the final harvest of the experiment it was significantly lighter on clones which had their apical region shaded. There were no significant differences in repeated-measures ANOVAs on absolute or relative growth rates in stolon length, ramet number, number of branches or petiole length measured over time in this experiment. Apical region shading appears not to affect growth and morphology in basal regions of clones if basal regions themselves are beneath an open canopy. When basal regions are shaded there was a reduction in primary stolon weight and in the proportional allocation of biomass to stolons over petioles.  5.3.2. DOES BASAL-REGION SHADING AFFECT THE GROWTH AND MORPHOLOGY OF APICAL REGIONS OF CLOVER CLONES? (ACROPETAL TRANSFER OF RESPONSE) Since in these experimental designs it is only a different response by apical regions that is necessary for determining the existence of intra-plant integration, results of the measurements made on this region only are presented.  Apical region open In this experiment, one plant in each treatment reached the edges of the flat more than two weeks before the rest, and they were excluded from the analyses. This left five replicates in each treatment. There were no significant differences in any of the clonal measurements made except in the amount of branching that occurred in the apical neighborhoods. Table 5.3 shows that there was a significant reduction in the number of branches, the primary branching percentage, the age to first branch, and percentage of total ramets that were on branches in apical regions of clones which had their basal region shaded by a grass canopy. There were no significant differences in the growth rates of the portions of the plants measured over time.  136  Apical region shaded In this experiment two clones under each treatment reached the edges of the flat well before the rest of the clones, and they were excluded from most of the analyses. Of the remaining clones (n=4) there was a strong suggestion that a greater dry weight was produced in apical regions of clones which had their basal regions under open canopies (Table 5.4; Student’s t-test  p<0.063, Mann-Whitney; p<O.O24).  Also when the clones  harvested early (above) are paired and considered, apical regions of clones with basal regions unshaded had a larger dry weight than apical regions of clones having a shaded basal region (p<O.05; Wilcoxon paired-sign test). There were no significant differences in any other yield measurements. No differences were detected in branching, biomass partitioning, or internode lengths between the two treatments. Clones with basal regions shaded, however, produced longer stolons per unit weight in apical regions. Measurements of a more detailed nature in apical regions of these clones did not indicate any significant differences between final primary stolon length, but rather that this structure was heavier in clones with basal regions open (0.154 (±0.004) g vs. 0.118 (±0.004) g; p.<O.OO1 Student’s t-test). Regardless of local canopy conditions, the growth and morphology of apical regions of clones responded to basal region shade. When apical regions were in the open, several measurements describing the amount of branching were reduced when basal regions were shaded. When apical regions were shaded, there was a reduction in dry weight, especially in primary stolons, if basal regions were shaded. Even though the weight was reduced, the length of primary stolons was not different, indicating an increase in specific stolon length.  137 5.4.  MICROCLIMATIC MEASUREMENTS  5.4.1  TEMPERATURE MEASUREMENTS Measurements of temperature were made during one representative experiment  with localized canopies. For these readings two sensors were used, one 3 cm above the soil surface in a small white tent, the other buried 1cm below the soil surface. The temperature at each location was measured three times in rapid succession and averaged, and then the treatment-specific mean and standard deviation were calculated. The temperature was measured under different ambient conditions: i) in full sun near mid-day; ii) late in the afternoon on a clear day when direct sunlight was no longer present in the greenhouse; and iii) under overcast skies near mid-day. Table 5.5 shows the results of these measurements and the significance of the differences between treatments using Students t-tests. The observed reduction in temperature due to a grass canopy only occurred under clear conditions, and in the soil, the increase carried over until the evening, even though the soil temperature fell to near ambient by this time.  5.4.2  SPECTRORADIOMETRIC MEASUREMENTS Behind canopies similar to those used in the experiments, spectroradiometric data  was collected as described in Chapter 2. Light conditions were measured under clear skies near solar noon and again late in the afternoon after the sun’s direct influence was no longer present in the experimental compartment used for the simulation (more than 6 hours after solar noon). Scans were made at several locations behind grass and open canopies to give a better picture of the range of conditions available within one neighborhood. Results from noon scans are presented for three locations: 6.5 cm north of the canopy, i.e., 6.5 cm nearer to the canopy than the middle of the flat containing clover (Fig. 5.la); 13.0 cm to the north of the canopy, i.e., at the position occupied by the primary stolon of clover clones (Fig. 5.lb); and 18.5 cm to the north of the canopy, i.e.,  138 6.5 cm further from the canopy than the middle of the flat containing clover (Fig. 5. ic). The measurement taken nearest the canopy would be representative of conditions near secondary stolons growing towards the canopy (south), while the furthest position would be representative of conditions near secondary stolons which were growing away from the canopy (north). Due to the height that leaf lamina were elevated, the conditions observed at the surface of the flat would be similar to leaves that were closer to the canopy but located at a higher elevation, at a similar angle from the top of the canopy. The scans presented after the sun’s direct beam was no longer available (Fig. 5.1 d) were recorded at the midpoint of the flat, 13.0 cm to the north of the canopy, the position that the primary stolon of clover clones occupied. The R:FR and PPFD were calculated from these scans and are presented in Table 5.6. The PPFD increased as the sensor moved away from the canopy in both treatments, although the increase was more profound behind the canopy of grass. The R:FR followed the same pattern behind the canopy of grass, but remained stable behind the open canopy. Late in the day, after direct sun was no longer available, presence of the grass canopy continued to alter both measurements of the light conditions in the location occupied by clover primary stolons.  5.5.  SUMMARY AND DISCUSSION The measurements of temperature indicate that it was not only light conditions  that were altered by the presence of a grass canopy in these arrangements. The soil temperature remained elevated beneath the open canopies until late in the day, but it was likely that overnight, the temperatures equalized under the two canopies. The lower air temperature beneath the grass canopy would lead to higher relative humidity, especially when the presence of grass itself is likely to increase the amount of water vapor in the air. Also, the differences in soil temperature along with lower humidity, increased the speed with which the soil dried beneath open canopies, creating larger differences in the water  139 status of the clones under the two canopies than the air temperature and humidity alone would have caused. The differences in water status established within one clone with regions under two different canopies might have been involved in the ‘signal” indicating the presence or absence of a canopy around one portion of the plant, and this could be an important indicator of neighboring vegetation in the field. The canopies of grass reduced the PPFD to around 50% of full sun with a R:FR greater than 1.0, in the middle of the flat of clover, indicating that these conditions were probably not as severe as in many previous experiments applying localized shade (Hartnett and Bazzaz, 1983; Slade and Hutchings, 1987b; Solangaarachchi and Harper, 1987; Thompson and Harper, 1988; Davies and Evans, 1990). This also suggests that experimental conditions providing less than this level of PPFD and a drastically lower R:FR, e.g. <0.5, might exaggerate the effects of neighboring plants on the light conditions in a patchy community. In the present experiment, the grass canopies provided clover clones with a localized region where there was still a reasonable potential for growth, though not as great as in open regions. This type of heterogeneity is likely to be a much more common occurrence in the field, where there are numerous gaps, and where clover leaves are often at or near the top of the sward. In this clover clone, there appears to have been no functional dependence of basal portions on the conditions experienced by apical portions, when basal regions were under open canopies. In contrast to Solangaarachchi (1987), basal regions of these clones, which had apical regions extend into shaded conditions, did not respond to the manipulation of the apex. It remains uncertain whether this discrepancy is due to a difference in background conditions, e.g. different time of year, use of different clones, or to a difference in experimental techniques, where in the present experiment, no direct influence of the apical-region canopy on basal regions was allowed. This result agrees with others demonstrating independence in response between apical and basal portions of  140 white clover (Kemball et al., 1992), and other species (Slade and Hutchings, 1987b; Dong, 1993) There appears to have been some interdependence on apical conditions when basal regions themselves were under closed canopies. The increased allocation of biomass to petioles over stolons in basal regions of clones with apical regions shaded, is consistent with observations on whole plants when they are shaded (Thompson and Harper, 1988; and see Chapter 4; Results). It seems that this effect of shade on biomass partitioning in basal regions can be alleviated somewhat if the apical region only experiences an open canopy.  Some type of signal, perhaps hormonal, might be  transmitted from apical to basal regions because photosynthate transfer is generally not expected from an expanding apex to older basal branches (Harvey, 1970, 1979; Marshall, 1990). Since the portion of the stolon which was already formed prior to imposing the canopies had a smaller weight when apical regions were shaded, it seems likely that this portion of the stolon provided support for the growth of shaded apical regions (Davies and Evans, 1990; Chapman and Robson, 1992). Chapman and Robson (1992) also demonstrated that old stolon material can have a close link with the expanding apex for several nodes along the stolon, through mobilization of existing starch. Although there were no statistically significant differences in the stolon length, ramet number, total dry weight, and number of branches at the end of the experiment, a closer examination of the data shows that in the group with apical regions open, where the means of all these measurements are larger, there is a greater variance, due to one small plant. This and visual evidence tentatively supports the conclusion that a larger yield was produced in basal regions of these clones relative to clones with apical regions shaded. These observations are consistent with other observed patterns of carbon translocation in white clover, where there can be movement in the basipetal direction, but only under special conditions (Harvey, 1970; Ryle et al., 1981; Chapman et al., 1991). It  141 seems that translocation can occur in this direction (basipetal flow) from an apex to recently formed secondary branches and to defoliated or severely shaded ramets, provided that they are not too distant. In the present experiments, some support may have been received from apical regions, and it seemed to be stronger when basal regions themselves were shaded.  The effects of different background conditions over the  observed portion’s interdependence on the apical region conditions, need to be compared with caution however, since the experiments were not conducted simultaneously, and many differences existed between clones (age, size, pre-treatment conditions) and between the grass canopies used to shade apical regions at the different times. There is limited evidence that manipulations of the basal portions of clover clones altered the response of apical regions. This indicates that, to some extent, acropetal integration in a two-patch system such as this can occur in white clover. When apical regions were open, and perhaps more likely to produce branches themselves (Thompson and Harper, 1988; Solangaarachchi and Harper, 1987), the amount of branching that occurred was strongly dependent on the conditions experienced by the rest of the plant (Table 5.3).  In shaded apical regions, when little overall branch production was  observed, there was no enhancement of branching from interconnection with betterilluminated basal regions. With daughter ramets under sufficiently high resource levels, support from mother ramets may be beneficial to the entire plant, whereas with daughter ramets under low resource levels, integration may not be advantageous (Caraco and Kelly, 1991). They suggested that with a growth function that is convex-concave (i.e. increase in growth is small with increasing resources at low levels, while increase in growth is more rapid with increasing resources at high levels), and daughter ramets under sufficiently high resource levels, support from mother ramets is beneficial to the clone and hence should occur under these circumstances. This is in agreement with the observations of branching in the apical regions of the present clover clones. In the apical regions shaded by grass canopies, the increase in dry weight indicates that a benefit is  142 obtained through connection to unshaded basal regions. Under shaded conditions these apical regions can receive support from resource-rich basal regions and continue extension and investment of resources, perhaps at very little cost to these unshaded basal regions (see Results above). Shading of the basal regions seemed to affect apical regions of these clones regardless of their illumination condition, and this is supported by translocation patterns observed in white clover, where a stolon apex is a sink that draws from stolon as well as recent branch-produced carbohydrate (Harvey, 1970; Chapman et al., 1991; Chapman and Robson, 1992) Shading of older branches has been observed to decrease export to the main stolon (Kemball et al., 1992). This suggests that in white clover, shading of the basal portion of a stolon is responded to in a globalized way, regardless of the illumination status of the rest of the plant. This seems to emphasize the importance of acropetal transport in determining white clover clonal morphology. Support of the stolon apex regardless of its local conditions, serves to support and maintain growth and expansion of the clone into new environments, even if this currently occupies an inhospitable environment. The measurements on clover clones indicate that while most characteristics were determined locally, some responded globally to conditions experienced by other parts of the plant, in general agreement with Turkington et al. (1991). The extent of integration demonstrated within these clones, however was very limited, in agreement with Bulow Olsen et al. (1984) and Solangaarachchi and Harper (1989). This indicates that to a large degree, portions of one white clover plant in the field growing in two distinct patches respond in isolation to their local conditions. When integration between the portions of the plant occurred, it was generally seen as some form of support for portions of clones under less favorable conditions (i.e., a shaded basal region when the apical region was open, or a shaded apical region when the basal region was open). The prediction that  143 portions of a clone under poorer conditions would receive less from the rest of the plant (or, indeed, shunt some of its own resources to the better-illuminated portion) was not supported in any of these experiments. This also indicates that an alternative hypothesis, that an increase in branching and ramet production in high-light conditions would occur when the remainder of the plant is more carbon-limited, could not be supported in these clones (Hutchings and de Kroon, 1994).  144  Plate 4. Layout of localized shade experiments. This shows one experiment, where basal regions (left side of each two-chambered “box”) are given the variable (shaded by grass or open) canopy, while apical regions of all clones remained in the open. Morphological comparisons in this experiment are made between apical regions only.  146  Table 5.1. Mean (±1 SE) of measurements made at the final harvest on basal regions of clones with their apical regions under different conditions. In this experiment all basal regions were under an open canopy. Means were compared using Student’s t-test.  Plant measurement  Apical region shaded (n=5)  Apical region open (n=6)  Probability  Total stolon length (cm)  114.4 (7.7)  117.3 (8.9)  0.807  Total ramet number  95.0 (5.0)  96.8 (6.9)  0.910  Above-ground dry weight (g)  3.83 (0.34)  3.90 (0.33)  0.888  Number of branches  22.0 (0.7)  21.2 (0.9)  0.479  Branching percentage  67.0 (1.4)  66.2 (1.6)  0.7 13  2.7 (0.1)  2.7 (0.1)  0.954  Percentage of total ramets on branches  49.5 (2.0)  49.5 (2.9)  0.993  Percentage of total length on branches  65.4 (1.2)  66.4 (2.1)  0.692  % of aerial dry weight allocated to leaf lamina  48.2 (0.7)  48.6 (0.4)  0.588  % of aerial dry weight allocated to petioles  25.8 (0.5)  25.4 (0.5)  0.607  % of aerial dry weight allocated to stolons  26.1(0.3)  26.0 (0.6)  0.903  Petiole/stolon weight  0.82 (0.02)  0.83 (0.01)  0.7 17  0.127 (0.006)  0.127 (0.005)  0.990  1.19 (0.05)  1.21 (0.03)  0.784  Age to first branch (number of nodes)  Specific stolon length (cm/mg) Mean internode length (cm)  147  Plate 5. Two clones growing from right to left, and grown with their apical regions (left section) under different canopies; (top) open, (bottom) shaded by grass, while both basal regions (right section) were grown shaded by grass canopies. The morphological comparisons in this experiment were made between the two basal regions.  149  Table 5.2. Mean (±1 SE) of measurements made at the final harvest on basal regions of clones with their apical regions under different conditions. In this experiment all basal regions were shaded by a grass canopy. Means were compared using Student’s t-tests.  Plant measurement  Apical region shaded (n=5)  Apical region open (n=5)  Probability  Total stolon length (cm)  97.3 (3.7)  108.8 (8.5)  0.250  Total ramet number  110.6 (6.2)  120.8 (10.3)  0.421  2.100 (0.055)  2.380 (0.195)  0.195  Number of branches  20.8 (1.8)  24.0 (2.7)  0.355  Branching percentage  41.4 (3.0)  46.6 (4.3)  0.345  Age to first branch (number of nodes)  5.9 (0.3)  5.4 (0.3)  0.268  Percentage of total ramets on branches  54.3 (1.5)  56.8 (2.9)  0.469  Percentage of total length on branches  24.3 (1.4)  26.4 (3.3)  0.567  % of aerial dry weight allocated to leaf lamina  45.4 (0.3)  45.5 (0.2)  0.911  % of aerial dry weight allocated to petioles  31.6 (0.3)  30.8 (0.3)  0.073  % of aerial dry weight allocated to stolons  22.9 (0.2)  23.7 (0.3)  0.05 1  Petiole/stolon weight  1.17 (0.02)  1.09 (0.03)  0.074  0.2 10 (0.005)  0.202 (0.003)  0.154  0.88 (0.03)  0.91 (0.05)  0.674  Above-ground dry weight (g)  Specific stolon length (cmlmg) Mean internode length (cm)  150  Table 5.3. Mean (±1 SE) of measurements made at the final harvest on apical regions of clones with their basal regions under different conditions. In this experiment all apical regions were under an open canopy. Means were compared using Students t-tests.  Plant measurement  Basal region shaded (n=5)  Basal region open (n5)  Probability  Total stolon length (cm)  21.4 (1.5)  24.3 (2.0)  0.282  Total ramet number  39.6 (1.9)  45.6 (2.7)  0.110  0.413 (0.042)  0.469 (0.052)  0.400  5.6 (0.4)  7.6 (0.7)  0.035  48.2 (2.9)  65.4 (4.7)  0.014  6.0 (0.3)  4.0 (0.5)  0.013  Percentage of total ramets on branches  70.5 (1.2)  74.5 (0.7)  0.019  Percentage of total length on branches  32.4 (1.1)  35.4 (1.5)  0.145  % of aerial dry weight allocated to leaf lamina  47.6 (0.5)  47.9 (0.3)  0.650  % of aerial dry weight allocated to petioles  27.2 (0.8)  26.9 (0.7)  0.800  % of aerial dry weight allocated to stolons  29.3 (1.4)  29.5 (3.2)  0.950  Petiole/stolon weight  0.94 (0.06)  0.96 (0.13)  0.850  0.208 (0.006)  0.209 (0.004)  0.976  0.54 (0.04)  0.53 (0.06)  0.805  Above-ground dry weight (g) Number of branches Branching percentage Age to first branch (number of nodes)  Specific stolon length (cm/mg) Mean internode length (cm)  151  Table 5.4. Mean (±1 SE) of measurements made at the final harvest on apical regions of clones with their basal regions under different conditions. In this experiment all apical regions were shaded by a grass canopy. Means were compared using Student’s t-tests.  Plant measurement  Basal region shaded (n=4)  Basal region open (n4)  Probability  Total stolon length (cm)  38.6 (1.2)  43.7 (2.6)  0.124  Total ramet number  25.8 (1.8)  30.5 (3.0)  0.216  1.179 (0.041)  1.45 1 (0.112)  0.063  Number of branches  7.0 (0.6)  9.0 (1.2)  0.190  Branching percentage  30.1 (1.8)  34.5 (3.0)  0.25 1  Age to first branch (number of nodes)  3.0 (0.4)  Percentage of total ramets on branches  69.1 (2.6)  71.6 (1.9)  0.475  Percentage of total length on branches  50.7 (2.8)  52.3 (2.3)  0.679  % of aerial dry weight allocated to leaf lamina  45.1 (0.7)  45.3 (0.5)  0.762  % of aerial dry weight allocated to petioles  28.3 (0.7)  28.9 (0.3)  0.528  % of aerial dry weight allocated to stolons  26.6 (0.9)  25.8 (0.6)  0.494  Petiole/stolon weight  1.07 (0.06)  1.12 (0.04)  0.480  0.153 (0.005)  0.127 (0.002)  0.003  1.51 (0.06)  1.45 (0.06)  0.520  Above-ground dry weight (g)  Specific stolon length (cm/mg) Mean internode length (cm)  3.0  (-)  152  Table 5.5. Mean (±1 SE) of temperature (°C) measurements taken at two locations (Soil), 1cm below soil surface, and (Air), 2cm above the soil, under both canopy types under three different conditions. Shown is the significance of the difference between the pair of means taken under grass and open canopies at each location under each condition. (* p<O.O5, ** p<O.Ol, ns not significant (Student’s t-test)).  Soil  Air  Greenhouse temperature  Canopy type Open  Grass  Open  Grass  25.7 (0.2)  20.3 (0.3)**  25.9 (0.3)  20.8 (0.4)**  21.5  18.3 (0.2)  18.3 (0.1)ns  18.1 (0.1)  18.5 (0.2)ns  18.0  19.0 (0.2)  18.4 (0.2)*  18.8 (0.2)  18.6 (0.3)ns  18.0  Condition Full sun-midday (n=6) Overcast-midday (n=6) Clear sky-late PM (n=6)  153 Solar noon  Conti-ol  (a)  D. glomerata 6.0—  40  —  SIX)  600  700  Late afternoon  (d)  (b)  400  500  600  700  400  600  600  700  Wavelength (na)  (c)  400  500  600  700  Wavelength (nnn)  Fig. 5.1. Spectroradiometric measurements made on a clear day at solar noon, (a), (b), and (c), and four hours after solar noon, (d), under simulations of the experiment. Distance of sensor from grass (control) barrier: (a) 6.5 cm north, (b) 13.0 cm north, (c) 19.5 cm north, and (d) 13.0 cm north.  154  Table 5.6. PPFD 1 s- and R:FR under the two canopy types used in localized2 (.tmolm) shade experiments at two different times. At solar noon on a clear day, measurements were made at three different locations within one experimental compartment, and again six hours after noon at one location. This was after the sun’s beam had left the compartment. Data are from spectroradiometric scans shown in Figs. 5.la-d Noon Location of sensor 6.5 cm north  13.0 cm north  18.5 cm north  Canopy type Grass  PPFD R:FR  356 0.88  826 1.07  1711 1.13  Open  PPFD R:FR  1669 1.21  1738 1.22  1862 1.22  1800 hrs Grass  PPFD R:FR  268 0.87  Open  PPFD R:FR  399 1.30  155 Chapter Six GENERAL DISCUSSION  The series of experiments reported in this work were designed to investigate the relationship between white clover growth and its light environment. An attempt was made to recreate, in as natural a way as possible, the types of light conditions typically experienced by the growing plant in a pasture. As clover grows through a sward, its light environment will be altered by the identity and distance to its neighbors, primarily grasses. When clover is growing through a gap in the sward, it will receive direct sunlight as well as light both reflected from and transmitted through neighboring grasses. Continued growth through the gap will again bring the clover apex into close proximity to neighbors and the relative proportions of the different light sources will change. If the light conditions are important to its survival, then for continued survival in the sward, the clover must respond to these changes either by altering its growth patterns or form, although the two are intimately linked. Because white clover growth is plagiotropic, in the field different parts of the same clone will be exposed to different light conditions. An apex that has just invaded a patch of grass might remain connected to another part of the plant still growing in the gap. This adds complexity to the responses that can be expected from white clover in heterogeneous light environments. The results show that in reflection experiments (Chapter 3), except for mean primary stolon leaf area, growth and morphology of unshaded clover clones was largely unaffected by reflection from neighboring grasses. Movement of leaves through petiole repositioning was influenced by neighboring grasses, perhaps through phototropic response to increased FR from neighbors. This movement may have lessened the effects on morphological characters of reflection from neighbors.  In the partial shade  experiment (Chapter 4), the effects of spending part of the day in direct sun did not eliminate the effects of shade from neighboring grasses, indicating that white clover has a  156 low tolerance for shade even when in partial sunshine. The species of grass used to form the standard gap in this experiment had different effects on the clover clones growing within them, and the strength of the effects of different grass species descended from Holcus lanatus and Dactylis glomerata to Lolium perenne, however; this can be only partly attributed to their effects on the light environment. In localized shade experiments (Chapter 5), canopy conditions experienced by apical regions affected basal region response, but only in the experiment where the basal regions were shaded. Canopies experienced by basal regions affected the morphological responses of apical regions in both experiments, where apical regions received different canopies. While generally there was a reduction in growth due to interconnection with more poorly illuminated ramets, the changes in morphology associated with remote-region shading were not consistent.  The range of phenotypes produced by this single clone of white clover describes the plasticity with which a clone of clover is endowed, and this undoubtedly plays a role in its persistence in a heterogeneous community such as a pasture. The various canopy situations presented to this clone reflect some of the variability in the field, where white clover will receive light that is transmitted through and reflected from neighbors, as well as receive different light conditions over interconnected portions of the same plant. The responses to the various canopy arrangements indicate that the responses of white clover are quite complex and can not always be interpreted from an ecological perspective without making large assumptions about what specific changes in morphology do for the plant. By describing the changes to controlled semi-natural conditions we may create an understanding of what changes confer an advantage under particular field conditions, but this should also involve long-term studies to demonstrate an increase in fitness due to these changes.  157 Growth of the clones in treatments with neighbors to the north shows that the slight changes in R:FR associated with reflection from neighboring grasses do not cause modifications in the phenotype of the clover directly, but that neighbors are detected and may be responded to in ways that lead to little difference in observed morphology. Instead, in these clones, movable structures which could respond phototropically (new leaves, stolon apices) did so, perhaps in response to the directional nature of the added FR (Novoplansky et al., 1990; Ballare et al., 1992a). This may have strongly reduced the impact of neighboring plants on more permanent aspects of the clones’ morphology. Changes in physiological functioning might have also occurred rather than changes in outwardly-visible (and easily-measured) characteristics. With a slight change to one process, the plant may be able to maintain a constant morphology, and it may be this type of plasticity that allows plants to be quite similar under different conditions. The importance and prevalence of this type of modification to phenotype in determining plant success is unknown (Hutchings and de Kroon, 1994). Also, a lack of detection of significant differences in measurements between these groups, is not necessarily a strong indicator of a lack of real difference between them. If the power of a statistical test (Glantz, 1987) such as a t-test is weak, it will be difficult to detect a difference between means even if, in actual fact, there were one. With small sample sizes (e.g. <8) in t-tests, the ratio of the differences between means to the pooled standard deviation of the two groups needs to be greater than one for there to be a 50% chance of detecting a difference (at p<O.O5) when there really is one (Glantz, 1987). If this ratio is less than one, there is a smaller than 50% chance of detecting a difference when there really is one. Data from the partial-shade experiment (Chapter 4) indicates that many of the differences in means between pairs of clones were approximately the same magnitude as the pooled standard deviation. In the other experiments, where the effects of the treatments were expected to be more subtle, the differences between means are likely to be smaller with roughly the same magnitude to the pooled standard deviations.  This means that, for these  158 experiments, larger sample sizes would be needed to have at least a 50% chance of detecting differences when they they exist and are of this magnitude. The lack of morphological response in these experiments, therefore, may be due to the lack of sufficient replication. Changes in plant morphology do not only involve adaptations to current conditions, but also involve adaptations to assure continued access to sufficient resources. This may be a particularly important aspect of plant plasticity especially in communities with rapidly developing canopies such as in many of our agricultural systems (Ballare, 1994).  Plants in a monoculture, even if genetically identical, will have different  phenotypic attributes depending on the precise conditions at their particular location. For example, the plants near the center of the monoculture are often taller and less branched than plants near the edge. In fact, the shape of the canopy attained in such a monoculture is often “dome” shaped, with a regular increase in height towards the center. This suggests that each plant senses the height of its neighbors and responds quite accurately, only producing the needed increment of height for successful capture of light resources. The changes that neighbors cause are complex, and this phenomenon indicates the level of sensitivity with which plants can respond to their environment. Among other features, the existence of small changes in FR associated with these neighbors is likely to be responsible (Smith et al, 1990; Ballare, 1994). Earlier, it had been suggested that since light with a R:FR over 1.1 has little effect on phytochrome photoequilibrium, changes in this range in the natural environment will not affect plant growth through phytochrome (Smith and Holmes, 1977; Smith, 1982).  With the  recognition of heterogeneity in many plant canopies, a directionality to the various radiation inputs is likely to occur. This means that the effects of neighboring plants (increased FR or a low R:FR) are also often received in conjunction with both a high PPFD and R:FR. In some of these situations, the globally-detected R:FR might not drop  159 below 1.1 (Chapter 3, Table 3.2).  Horizontally-measured radiation under these  conditions is rich in FR with a R:FR that would alter the state of phytochrome in cells that receive this radiation (Ballare et al., 1987; Smith et al., 1990; Chapter 3). The radiation environment inside upright structures such as stems is remarkably sensitive to this horizontal radiation (Ballare et al., 1989; Mancinelli et al., 1992). In white clover, there was little permanent response to reflected FR radiation while also receiving unfiltered daylight-level radiation. The difference between these results and those using orthotropic plants such as Chenopodium sp. (lamb’s quarters), Sinapis alba (white mustard), and Cucumis sp. seedlings may have been due to the size differences between target plants. In the present experiment, clover clones grew quite large over the course of each experiment, and had several times as many leaves as the orthotropic plants used above. There would have been considerable amounts of FR in the region occupied by the petioles and stolons in control clones from the large number of clover leaves present. The older leaves of clover may have given the more recently formed leaves on all clones, including controls, a signal similar to neighboring grasses. Perhaps the response to reflection from neighbors in clover clones which are very small, would be easier to detect. The results from the reflection experiments suggest, however, that if a response is detected when a clover genet is small, these differences might not lead to apparent differences later in growth. White clover exhibited little shade tolerance under partial shade conditions. Defining a “shaded” environment as uniformly obstructed precludes the possibility of some direct (unfiltered) radiation being received by a plant. In this experiment, “partial shade” meant that sunlight was filtered in a natural pattern (allowing small sunflecks through) and that direct sun was also available. The clover clones responded to the shade, and quite sensitively, even though several hours were spent in the direct sun. The response to different species forming the edges of the standard gap varied in several regards, suggesting that clover clones can detect the species of grass they are neighboring  160 and respond accordingly, without interaction below ground. The types of morphological changes observed to occur from different species might have been able to predict the types of responses to more subtle changes in the light conditions. (However, none of these changes occurred consistently in response to treatments with light reflected off neighbors.)  The morphologies produced with only shoot interaction support the  observations of the effects on clover morphology when growing associated with the different species of grass in the field, with Holcus lanatus and Daclylis glomerata being among the most aggressive, and Lolium perenne being among the least aggressive (Chestnutt and Lowe, 1970; Solangaarachchi, 1985; Frame and Newbould, 1986). This suggests that light conditions associated with different neighbors are important in altering the phenotype of white clover, allowing it to grow, presumably, in an efficient manner under each of the particular conditions. It would be interesting to pre-condition plants by growing them with one species of neighbor, allowing it to adopt a suitable morphology, and then transfer them to neighborhoods with different species of grass, or different heights or densities of the same species. By comparing the long-term success between clones that came from different backgrounds, and between similarly pre-conditioned clones which went to different treatments, a possible fitness advantage of having a particular shoot morphology under each condition could be demonstrated. Under clear skies, more than 90% of the daily PPFD received by partially shaded clones in this experiment occurred during the middle period of the day, when direct sunlight was received. In further experiments, it would be interesting to vary the PPFD received in this period and monitor the change in response to equal shade at other times of the day. This may suggest a role for PPFD in modification of the response to low R:FR.  Under uniform natural shade conditions, PPFD and the R:FR would drop  simultaneously no matter what the particular PPFD used, so this can no.t be determined. It may also be interesting to give a range of shade regimes defined by varying the length of the shaded period(s) during the day. This may expose some threshold responses,  161 where below a particular length of time spent shaded, there is no response to shade. A range of plant species known to vary in their shade tolerance should be examined for this. Shade tolerance in some species might be exhibited by a range of critical lengths of shading among individuals, perhaps allowing selection for improved shade tolerance. By testing a range of genotypes for their response to partial shade given in conjunction with some unfiltered light, we may be able to more accurately predict responses to shade in the field, e.g., under higher planting density, growing in companion plantings. Surely the use of more realistic shade in our experimentation on plants should lead to more rapid accumulation of knowledge on how to manage them. The results of localized shade experiments suggest that physiological integration can occur in white clover in a patchy environment. The effects on the plant varied depending on the direction of the stimulus and the conditions under which the observed portions grew. Shading of the stolon apex did not influence the rest of the plant when it remained unshaded. This suggests that in the field, if the basal regions of a plant are in an opening while an apical region becomes shaded, the basal regions might continue growing at the same rate as it would had the apex remained unshaded. This demonstrates a localized response in a region under relatively good conditions, and this would maintain a steady production of ramets on branches. This might occur at the expense of the expanding apex which receives support from basal regions (Harvey, 1970; Chapman et al., 1991a; Chapter 5, Results). When the basal regions were shaded, more effects of shading the apical region shade were noticed, describing a more globalized response from apical region conditions on basal region response, and requiring some sort of basipetal transport.  The globalized response under these conditions suggest that when no  alternative is available, an apical region that is under poor conditions might begin drawing at cost from the rest of the plant through mobilization of stored carbohydrate and through effects on its morphology and physiology.  Also, unshaded apical regions  connected to shaded basal regions were inhibited from branching, perhaps promoting its  162 own escape from the area. It appears that an apical region can respond to stimuli (shading) affecting only basal portions, allowing it to increase its distance from the stimulus before the apex itself becomes shaded. It would be interesting to test if the apex increased its expansion solely due to a signal from neighbors (from reflection) received at basal regions rather than actual light-resource limitation. The R:FR could be varied over basal regions without altering PPFD, providing a signal of neighbor presence, while the response in apical regions is monitored. The same could be conducted examining the basal region response to apical region differences in R:FR, but the results above (Chapter 5) suggest that apical region conditions would be more likely to have an effect if the basal regions themselves were shaded. What is involved in determining a plant’s morphology to maximize current and continued access to sufficient resources is undoubtedly complex, and our ability to interpret changes in plants with respect to this is rather limited. The experiments conducted here indicate that even with relatively well understood effects of neighboring plants on the environment of a clone, the morphologies attained are difficult to interpret. Perhaps by describing a range of morphologies attained under particular conditions and long-term evaluation of the differences in fitness caused by these different morphologies, we might improve our understanding of what particular morphological changes can do for the plant. The goal of many studies of phenotypic plasticity is the ability to predict and improve the efficiency with which a crop plant exploits a particular controlled or natural environment. A genetic basis for this level of phenotypic plasticity would allow for selection of types based on the ability to perform to a desired standard. The importance of this aspect of potential crop improvement has only recently been realized, and the extent of phenotypic plasticity in many of our important crop plants is largely unknown.  163 There are many potential advantages of a better understanding of how plants respond to neighboring vegetation. Having the ability to predict if a particular crop will grow to satisfactory standards when in a new planting arrangement will clearly be beneficial. Increased knowledge of the behavior of plants in response to reflected light, for example, could increase the efficiency with which we use plants in our production facilities. For example, planting in double- and triple-rows of staggered height, rising from the south to north within each row running in an east-west direction, has the potential for increasing the amount of light intercepted and hence, crop production per unit area. The amount of increase needed in row spacing over single rows of the tallest crop would not be large enough to offset the benefit of 2-3 times the productivity per row. In order to successfully predict the yields of the shorter crops used in each multiple row, an understanding of these species’ responses to increased FR from light reflected off the taller companion plants is needed. A similar effect might develop in plantings on a southern aspect, with increasing elevation to the north and slight increasesin reflected light over level plantings. In some species the response to increased FR from reflection might affect production, whereas in others, it might not. It seems that breeding for an optimization of the response to FR reflected from neighbors (in this example perhaps eliminating the response), might increase our abilities to use more finely-tuned production techniques, which in the long run should have less negative impact on the global resources.  164  LITERATURE CITED Aarssen, L. W. and R. Turkington. 1985. Biotic specialization between neighboring genotypes in Lolium perenne and Trfolium repens from a permanent pasture. J. Ecol., 73: 605-614. Abrahamson, W.G., S.S. Anderson, and K.D. McCrea. 1991. Clonal integration: nutrient sharing between sister ramets of Solidago altissima (Compositae). Am. J. Bot., 78: 1508-1514. Ahigren, G.H. and R.F. Fuelleman. 1950. Ladino clover. Adv. Agron.., 2: 207-232. Alpert, P. 1990. Water sharing among ramets in a desert population of Distichlis spicata (Poaceae). Am. I. Bot., 77: 1648-165 1. Alpert, P. 1991. Nitrogen sharing among ramets increases clonal growth in Fragaria chiloensis. Ecology, 72: 69-80. Anderson, M. C. and E. E. Miller. 1974. Forest cover as a solar camera: penumbra effects in plant canopies. J. Appl. Ecol., 11: 691-698. Attridge, T.H. 1990. Light and Plant Responses. Edward Arnold, London. Baldocchi, D. and S. Collineau. 1994. The physical nature of solar radiation in heterogenous canopies: spatial and temporal attributes. In: Exploitation of Environmental Heterogeneity by Plants (Ed. by M.M. Caldwell and R.W. Pearcy), pp. 21-7 1. Academic Press, San Diego. Ballare, C. L., R. A. Sanchez, A. L. Scopel, J. I. Casal, and C. M. Ghersa. 1987. Early detection of neighbor plants by phytochrome perception of spectral changes in reflected sunlight. Plant, Cell, and Env., 10: 55 1-557. Ballare, C.L., R.A. Sanchez, A.L. Scopel, and C.M. Ghersa. 1988. Morphological responses of Datura ferox L. seedlings to the presence of neighbors. Their relationship with canopy microclimate. Oecologia, 76: 288-293. Ballare, C.L., A.L. Scopel, and R.A. Sanchez. 1989. Photomodulation of axis extension in sparse canopies. Role of stem in the perception of light quality signals of stand density. Plant Physiol., 89: 1324-1330. Ballare, C. L., A.L. Scopel, and R. A. Sanchez. 1990. Far-red radiation reflected from adjacent leaves: An early signal of competition in plant canopies. Science, 247: 329-332. Ballare, C.L., P.W. Barnes, and R.E. Kendrick. 1991. Photomorphogenic effects of UV B radiation on hypocotyl elongation in wild-type and stable-phytochrome-deficient mutant seedlings of cucumber. Physiol. Plant., 82: 652-658. Ballare, C.L., A.L. Scopel, S.R. Radosevich, and R.E. Kendrick. 1992a. Phytochrome mediated phototropism in de-etiolated seedlings. Occurrence and ecological significance. Plant Physiol., 100: 170-177.  165 Ballare, C.L., A.L. Scopel, R.A. Sanchez, and S.R. Radosevich. 1992b. Photomorphogenic processes in the agricultural environment. Photochem. and Photobiol., 56: 777-788. Ballare, C.L. 1994. Light gaps: Sensing the light opportunities in highly dynamic canopy environments. In: Exploitation of Environmental Heterogeneity by Plants (Ed. by M.M. Caidwell and R.W. Pearcy), pp. 73-110. Academic Press, San Diego. Barnes, P.W., S.D. Flint, and M.M. Caidwell. 1990. Morphological responses of crop and weed species of different growth forms to ultraviolet-B radiation. Am. J. Bot., 77: 1354-1360. Bazzaz, F.A. and P.M. Wayne. 1994. Coping with environmental heterogeneity: The physiological ecology of tree seedling regeneration across the gap/understory continuum. In: Exploitation of Environmental Heterogeneity by Plants (Ed. by M.M. Caidwell and R.W. Pearcy), pp. 349-390. Academic Press, San Diego. Bell, A.D. 1984. Dynamic morphology: A contribution to plant population ecology. In: Perspectives in Plant Population Ecology (Ed. by R. Dirzo and J. Sarukhan), pp. 48-65. Sinauer, Sunderland. Bell, G. and M.J. Lechowicz. 1994. Spatial heterogeneity at small scales and how plant respond to it. In: Exploitation of Environmental Heterogeneity by Plants (Ed. by M.M. Caidwell and R.W. Pearcy), pp. 391-414. Academic Press, San Diego. Boller, B.C. and J. Nosberger. 1985. Photosynthesis of white clover leaves as influenced by canopy position, leaf age, and temperature. Ann. Bot., 56: 19-27. Bradshaw, A.D. 1965. Evolutionary significance of phenotypic plasticity in plants. Adv. Genet., 13: 115-155. Britz, S.J. 1990. Photoregulation of root:shoot ratio in soybean seedlings. Photochem. and Photobiol., 52: 151-159. Bulow-Olsen, A., N.R. Sackville-Hamilton, and M. J. Hutchings. 1984. A study of growth form in genets of Trifolium repens L. as affected by intra- and interplant contacts. Oecologia, 61: 383-387. Burdon, J.J. 1980. Intra-specific diversity in a natural population of Trfolium repens. J. Ecol. 68: 717-735. Burdon, J.J. 1983. Biological flora of the British Isles: Trifolium repens. J. Ecol., 71: 307-3 30. Caldwell, M.M. 1994. Exploiting nutrients in fertile soil microsites. In: Exploitation of Environmental Heterogeneity by Plants (Ed. by M.M. Caldwell and R.W. Pearcy), pp. 325-347. Academic Press, San Diego. Caraco, T. and C. K. Kelly. 1991. On the adaptive value of physiological integration in clonal plants. Ecology, 72: 81-93. Caradus, J.R. and D.F. Chapman. 1991. Variability in stolon characteristics and response to shading in two cultivars of white clover (Trifolium repens L.). New Zealand J. of Agric. Res., 34: 239-247.  166  Caradus, J.R., M.J. M. Hay, A.D. Mackay, V.J. Thomas, J. Dunlop, M.G. Lambert, A.L. Hart, J. van den Bosch, and S. Wewala. 1993. Variation within white clover for phenotypic plasticity of morphological and yield related characters, induced by phosphorus supply. New Phytol., 123: 175-184. Casal J.J. and R.E. Kendrick. 1993. Impaired phytochrome-mediated shade-avoidance responses in the aurea mutant of tomato. Plant, Cell, and Env., 16: 703-7 10. Casal, J.J. and H. Smith. 1988. Persistent effects of changes in phytochrome status on internode growth in light-grown mustard: Occurrence, kinetics and locus of perception. Planta, 175: 214-220. Casal, J.J. and H. Smith. 1989a. The function, action and adaptive significance of phytochrome in light-grown plants. Plant, Cell, and Env., 12: 855-862. Casal, J.J. and H. Smith. 1989b. The “end-of-day” phytochrome control of internode elongation in mustard. Kinetics, interaction with the previous fluence rate and ecological implications. Plant, Cell, and Env., 12: 511-521. Chapman, D.F. 1983. Growth and demography of Trfolium repens stolons in grazed hill pastures. J. Appl. Ecol.,. 20: 597-608. Chapman, D.F. and M.J. Robson. 1992. The physiological role of old stolon material in white clover (Trifolium repens L.). New Phytol., 122: 53-62. Chapman, D.F. and M.J.M Hay. 1993. Translocation of phosphorus from nodal roots in two contrasting genotypes of white clover. Physiol. Plant.,. 89: 323-330. Chapman, D.F., M.J. Robson, and R.W. Snaydon. 1991a. The influence of leaf positions and defoliation on the assimilation and translocation of carbon in white clover. 1. Carbon distribution patterns. Ann. Bot.,. 67: 295-302. Chapman, D.F., M.J. Robson, and R.W. Snaydon. 199 lb. The influence of leaf positions and defoliation on the assimilation and translocation of carbon in white clover. 2. Quantitative carbon movement. Ann. Bot., 67: 303-308. Chapman, D.F., M.J. Robson, and R.W. Snaydon. 1992a. The carbon economy of clonal plants of Trfolium repens L. J. Exp. Bot., .43: 427-434. Chapman, D.F., M.J. Robson, and R.W. Snaydon. 1992b. Physiological integration in the clonal perennial herb Trfolium repens L. Oecologia 89: 338-347. Chestnutt, D.M.B. and J. Lowe. 1970. Agronomy of white clover/grass swards-review. Occas. Symp. Br. Grassi. Soc. 6: 191-213. Cook, R. E. 1983. Clonal plant populations. Am. Sci., 71: 244-253. Dale, M.P. and D.R. Causton. 1992. The ecophysiology of Veronica chamaedrys, V. montana, and V. officinalis. I. Light quality and light quantity. J. Ecol., 80: 483-492. Davies, A. and M.E. Evans. 1990. Axillary bud development in white clover in relation to defoliation and shading treatments. Ann. Bot., 66: 349-357.  167 Dong, M. 1993. Morphological plasticity of the clonal herb Lamiastrum galeobdolon (L.) in response to partial shading. New Phytol., 124: 29 1-300. Drumm-Herrel, H. and H. Mohr. 1984. Mode of coaction of phytochrome and blue light photoreceptor in control of hypocotyl elongation. Photochem. and Photobiol., 40: 261266. Erith, A.G. 1924. White clover (Trifolium repens L.) A monograph. Duckworth, London. Evans, J.P. 1991. The effect of resource integration on fitness related traits in a clonal dune perennial, Hydrocotyle bonariensis. Oecologia, 86: 268-275. Evans, R.C. 1986. Morphological variation in a biotically patchy environment: evidence from a pasture population of Trfolium repens L. M.Sc. thesis, The University of British Columbia, Vancouver, Canada. Evans, R.C. and R. Turkington. 1988. Maintenance of morphological variation in a biotically patchy environment. New Phytol., 109:369-376. Evans, J.P. and S. Whitney. 1992. Clonal integration across a salt gradient by a nonhalophyte, Hydrocotyle bonariensis (Apiaceae). Am. J. Bot., 79: 1344-1347. Field, R.J. and D.I. Jackson. 1975. Light effects on apical dominance. Ann. Bot., 39: 369-374. Frame, J. and P. Newbould. 1986. Agronomy of white clover. Adv. Agron., 40: 1-88. Furuya, M., ed. 1987. Phvtochrome and Photoregulation in Plants. Academic Press, Tokyo. Gaba, V. and M. Black. 1979. Two separate photoreceptors control hypocotyl growth in green seedlings. Nature (London), 278: 51-54. Glantz, S.A. 1987. Primer of Biostatistics. McGraw-Hill, New York. Gliddon, C. and P. Trathan. 1985. Interactions between white clover and perennial ryegrass in an old permanent pasture. In: Structure and Functioning of Plant Population 2. (Ed. by J. Haeck and J.W. Woldendorp), North Holland Publishing Co., Amsterdam. Grantz, D.A. 1990. Plant response to atmospheric humidity. Plant, Cell, and Env., 13: 667-679. Grime, J.P. 1979. Plant Strategies and Vegetation Processes. Wiley, Chichester. Grime, J. P. 1981. Plant strategies in shade, In: Plants and the Daylight Spectrum. (Ed. by H. Smith), pp. 159-186, Academic Press, London. Grime, J.P. 1994. The role of plasticity in exploiting environmental heterogeneity. In: Exploitation of Environmental Heterogenity by Plants (Ed. by M.M. Caldwell and R.W. Pearcy), pp. 1-19. Academic Press, San Diego. Harris, W., I. Rhodes, and S.S. Mee. 1983. Observations of environmental and genotypic influences on the overwintering of white clover. J. Appi. Ecol., 20: 609-624.  168  Hartnett, D. C. and F. A. Bazzaz. 1983. Physiological integration among intraclonal ramets in Solidago canadensis L. Ecology, 64:779-788. Harvey, H. J. 1970. Patterns of assimilate translocation in Trfolium repens. Occas. Symp. Br. Grassi. Soc. 6: 181-186. Harvey, H.J. 1979. The regulation of vegetative reproduction. Ph.D. thesis, University of Wales. Cardiff, Wales. Haynes, R.J. 1980. Competitive aspects of the grass-legume association. Adv. Agron., 33: 227-261. Hill, J. 1977. Plasticity of white clover grown in competition with perennial ryegrass. Welsh Plant Breeding Station, Annual Report (1976), Aberystwyth, 24-25. Holmes, M.G. 1981. Spectral distribution of radiation within plant canopies, In: Plants and the Daylight Spectrum, (Ed by H. Smith), pp. 147-158. Academic Press, London. Holmes, M.G. and H. Smith. 1977a. The function of phytochrome in the natural environment I. Characterization of daylight for studies in photomorphogenesis and photoperiodism. Photochem. and Photobiol., 25: 533-538. Holmes, M.G. and H. Smith. 1977b. The function of phytochrome in the natural environment IV. Light quality and plant development. Photochem. and Photobiol., 25: 551-557. Hutchings, M.J. and M. Mogie. 1990. The spatial structure of clonal plants: control and consequences. In: Clonal Growth in Plants: Regulation and Function (Ed. by J. van Groenendael and H. de Kroon), pp. 57-76. SPB Academic Publishing, The Hague. Hutchings, M.J. and E.A.C. Price. 1993. Does physiological integration enable clonal herbs to integrate the effects of environmental heterogeneity: Plant Species Biol., 8: 95105. Hutchings, M.J. and H. de Kroon. 1994. Foraging in plants: the role of morphological plasticity in resource acquisition. Advances in Ecological Research, 25:159-238. Janoudi, A. and K.L. Poff. 1992. Action spectrum for enhancement of phototropism by Arabidopsis thaliana seedlings. Photochem. and Photobiol., 56: 655-659. Jones, M. G. 1933. Grassland management and its influence on the sward. J. Royal. Agric. Soc., 94: 21-41. Jonsdottir, I.S. and T.V. Callaghan. 1988. Interrelationships between different generations of interconnected tillers of Carex bigelowii. Oikos, 52: 120-128. Kasperbauer, M. 1. 1971. Spectral distribution of light in a tobacco canopy and effects of end of day light quality on growth and development. Plant. Physiol., 47: 775-778. Kasperbauer, M.J. 1987. Far-red light reflection from green leaves and effects on phytochrome-mediated assimilate partitioning under field conditions. Plant Physiol., 85: 350-354.  169 Kasperbauer, M.J. and P.G. Hunt. 1992. Cotton seedling morphogenic responses to FRIR ratio reflected from different colored soils and soil covers. Photochem. and Photobiol., 56: 579-584. Kemball, W.D., M.J. Palmer, and C. Marshall. 1992. The effect of local shading and darkening on branch growth, development and survival in Trfo1ium repens and Galium aparine. Oilcos, 63: 366-375. Kendrick, R.E. and G.H.M. Kronenberg. 1986. Photomorphogenesis in Plants. Martinus Nijhoff, Dordrecht, The Netherlands. de Kroon, H. and F. Schieving. 1990. Resource partitioning in relation to clonal growth strategy. In: Clonal Growth in Plants: Regulation and Function (Ed. by J. van Groenendael and H. de Kroon), pp. 177-186. SPB Academic Publishing, The Hague. Landa, K., B. Benner, M.A. Watson, and J. Gartner. 1992. Physiological integration for carbon in mayapple (Podophyllum peltatum), a clonal perennial herb. Oikos, 63: 348-356. Lau, R.R. and D.R. Young. 1988. Influence of physiological integration on survivorship and water relations in a clonal herb. Ecology, 69: 215-219. Lee, D.W. and K.R. Downum. 1991. The spectral distribution of biologically active solar radiation at Miami, Florida, USA. Int. J. Biometerol., 35: 48-54. Lovell, P.H. and P.J. Lovell. 1985. The importance of plant form as a determining factor in competition and habitat exploitation. In: Studies on Plant Demography: a Festschrift for John L. Harper. (Ed. by J.L. Harper), pp. 209-22 1, Academic Press, London. McClaren, J.S. and H. Smith. 1978. Phytochrome control of the growth and development of Rumex obtusfo1ius under simulated canopy light environments. Plant, Cell, andEnv., 1:61-68. MacLellan, A. J. and B. Frankland. 1985. A simple field method for measuring light quality: seasonal changes in a temperate deciduous wood. Photochem. and Photobiol., 42: 689-695. Mancinelli, A.L., 0. Horbert, and G. Nikas. 1992. In vivo phytochrome-mediated perception of reflected light signals. Photochem. and Photobiol., 56:585-59 1. Marshall, C. 1990. Source-sink relationships of interconnected ramets. In: Clonal Growth in Plants: Regulation and Function (Ed. by J. van Groenendael and H. de Kroon), pp. 23-41. SPB Academic Publishing, The Hague. Marshall, C. and G. Anderson-Taylor. 1992. Mineral nutritional inter-relations amongst stolons and tiller ramets in Agrostis stolonifera L. New Phytologist, 122: 339-347. McCartney, H.A. and M.H. Unsworth. 1975. A spetroradiometer for measuring the spectral distribution of radiation in plant canopies. In: Light as an Ecological Factor, (Eds. G.C. Evans, R. Bainbridge, and 0. Rockham), pp. 565-574. Blackwell, Oxford. Morgan, D.C. 1981. Light quality and plant growth. In: Plants and the Daylight Spectrum, (Ed by H. Smith), pp. 205-22 1, Academic Press,.London.  170 Morgan, D.C. and H. Smith. 1978. Simulated sunflecks have large, rapid effects on plant stem extension. Nature, 273: 534-536 Morgan, D.C. and H. Smith. 1979. A systematic relationship between phytochrome controlled development and species habitat, for plants grown in simulated natural radiation. Planta, 145: 253-258. Morgan, D.C. and H. Smith. 1981. Control of development in Chenopodium album L. by shadelight: The effect of light quantity and light quality. New Phytol., 88: 239-248. Newton, P. C. D. 1986. The establishment, growth and fate of white clover plants. Ph.D. thesis, University College of North Wales, Bangor. Novoplansky, A., D. Cohen, and T. Sachs. 1990. How portulaca seedlings avoid their neighbors. Oecologia, 82: 490-493. Novoplansky, A. 1991. Developmental responses of portulaca seedlings to conflicting spectral signals. Oecologia, 88: 138-140. Ong, C. K. and C. Marshall. 1979. The growth and survival of severely-shaded tillers in Loliumperenne. Ann. Bot., 43: 147-155. Parish, R. 1987. The role of disturbance in permanent pastures. Ph.D. Thesis, University of British Columbia, Vancouver, Canada. Pearcy, R.W. and D.A. Sims. 1994. Photosynthetic acclimation to changing light environments: Scaling from the leaf to the whole plant. In: Exploitation of Environmental Heterogeneity by Plants (Ed. by M.M. Caidwell and R.W. Pearcy), pp. 145-174. Academic Press, San Diego. Pearcy, R.W., R.L. Chazdon, L.J. Gross, and K.A. Mott. 1994. Photosynthetic utilization of sunflecks: A temporally patchy resource on a time scale of seconds to minutes. In: Exploitation of Environmental Heterogeneity by Plants (Ed. by M.M. Caldwell and R.W. Pearcy), pp. 175-208. Academic Press, San Diego. Pickett, S.T.A. and F.A. Bazzaz. 1976. Divergence of two co-occurring successional species on a soil-moisture gradient. Ecology, 57: 169-176. Pitelka, L. F. and J. W. Ashmun. 1985. Physiology and integration of ramets in clonal plants, In: Population Biology and Evolution in Clonal Organisms, (Eds J.B.C. Jackson, L.W. Buss, and R.E. Cook), Yale University Press, New Haven. Price, E.A.C. and M.J. Hutchings. 1992. The causes and developmental effects of integration and independence between different parts of Glechom.a hederacea clones. Oiko, 63: 376-386. Price, E.A.C., C. Marshall, and M.J. Hutchings. 1992. Studies of growth in the clonal herb Glechoma hederacea I. Patterns of physiological integration. J. Ecol., 80:189-203. Rice, S.A. and F.A. Bazzaz. 1989. Growth consequences of plasticity of plant traits in response to light conditions. Oecologia, 78: 508-5 12. Ross, M.A. and J.L. Harper. 1972. Occupation of biological space during seedling establishment. I. Ecol., 60: 77-88.  171  Ryle, G.J.A., C.E. Powell, and A. J. Gordon. 1981. Patterns of 14-C-labelled assimilate partitioning in red and white clover during vegetative growth. Ann. Bot., 47: 505-5 14. Schlichting, C. 1986. The evolution of phenotypic plasticity in plants. Ann. Rev. Ecol. Syst., 17: 667-693. Silvertown, J. and D.M. Gordon. 1989. A framework for plant behavior. Ann. Rev. Ecol. Syst., 20: 349-366. Slade, A. J. and M. J. Hutchings. 1987a. An analysis of the costs and benefits of physiological integration between ramets in the clonal perennial herb Glechoma hederacea. Oecologia, 73: 425-431. Slade, A. J. and M. J. Hutchings. 1987b. The effects of light intensity on foraging in the clonal herb Glechoma hederacea. J. Ecol., 75: 639-650. Smith, H. 1982. Light quality, photoreception, and plant strategy. Ann. Rev. Plant. Phys. 33: 481-518. Smith, H. and M.G. Holmes. 1977. The function of phytochrome in the natural environment III. Measurement and calculation of phytochrome photoequilibrium. Photochem. and Photobiol., 25: 547-550. Smith, H. and G.C. Whitelam. 1990. Phytochrome, a family of photoreceptors with multiple physiological roles. Plant, Cell, and Env., 13: 695-707. Smith, H., J.J. Casal, and G.M. Jackson. 1990. Reflection signals and the perception by phytochrome of the proximity of neighboring vegetation. Plant, Cell, and Env., 13: 73-78. Solangaarachchi, S. M. 1985. The nature and control of branching pattern in white clover (Trfolium repens). Ph.D. thesis, University College of North Wales, Bangor. Solangaarachchi, S. M. and I. L. Harper. 1987. The effect of canopy filtered light on the growth of Trifolium repens. Oecologia, 72: 372-376. Stark, J.M. 1994. Causes of soil nutrient heterogeneity at different scales. In: Exploitation of Environmental Heterogenity by Plants (Ed. by M.M. Caidwell and R.W. Pearcy), pp. 21-7 1. Academic Press, San Diego. Stern, W.R. and C.M. Donald. 1962. Light relationships in grass-clover swards. Aust. J. Agr. Res., 13: 599-6 14. Stoutjesdijk, P. 1974. The open shade, an interesting microclimate. Acta Bot. Neerl., 23: 125-130. Sultan, S.E. 1987. Evolutionary implications of phenotypic plasticity in plants. Evol. Biol., 21: 127-178. Sutherland, W.J. and R.A. Stiliman. 1988. The foraging tactics of plants. Oikos, 52: 239-244. SYSTAT. 1992. Statistics, Version 5.0 Edition. Evanston, IL.  172 Tanada, T. 1982. Effect of far-red and green irradiation on the nyctinastic closure of Albizzia leaflets. Plant Physiol., 70: 90 1-904. Thompson, J.D. 1991. Phenotypic plasticity as a component of evolutionary change. Trends Ecol. Evol. 6: 246-249. Thompson, L. 1989. The influence of radiation quality on the behavior of grassland species. Ph.D. thesis, University College of North Wales, Bangor. Thompson, L. 1993 a. The influence of natural canopy density on the growth of white clover. Trifolium repens. Oilcos, 67: 321-324. Thompson, L. 1993b. The influence of the radiation environment around the node on morphogenesis and growth of white clover (Trfolium repens), Grass and For. Sci., 48: 271-278. Thompson, L. and J.L. Harper. 1988. The effect of grasses on the quality of transmitted radiation and its influence on the growth of white clover (Trifolium repens). Oecologia, 75: 343-347. Tilman, D. 1988. Plant Strategies and the Structure and Dynamics of Plant Communities. Princeton Univ. Press. Princeton, NJ. Tucker, D.J. and T. A. Mansfield. 1972. Effects of light quality on apical dominance in Xanthium strumarium and the associated changes in endogenous levels of abscissic acid. Planta, 102: 140-151. Turkington, R. and J.J. Burdon. 1983. Biology of Canadian weeds. 57. Trifolium repens L. Can. J. of Plant Science, 63: 243-266. Turkington, R. and J.L. Harper. 1979. The growth, distribution and neighbor relationships of Trifolium repens in a permanent pasture. IV. Fine-scale biotic differentiation. J. Ecol., 67: 245-254.. Turkington, R. 1979. Neighbor relationships in grass-legume communities. IV. Fine scale biotic differentiation. Can. J. Bot., 57: 2711-2716. Turkington, R.A. and L.A.. Mehrhoff. 1990. The role of competition in structuring pasture communities. In: Perspectives on Plant Competition (Ed. by J.B. Grace and D. Tilman), pp. 307-340. Academic Press, San Diego. Turkington, R. A., R. Sackville Hamilton, and C. Gliddon. 1991. Within-population variation in localized and integrated responses to Trifolium repens to biotically patchy environments. Oecologia, 86:183-192. Turkington, R., M.A. Calm, A. Vardy, and J.L. Harper. 1979. The growth, distribution and neighbor relationships of Trifolium repens in a permanent pasture. III. The establishment and growth of T. repens in natural and perturbed sites. J. Ecol., 67: 231243. Turkington, R., F. B. Holl, C. P. Chanway, and J.D. Thompson. 1988. The influence of microorganisms; particularly Rhizobium, on plant competition in grass-legume communities. In: Plant Population Ecology (Eds A.J. Davy, M.J. Hutchings, and A.R. Watkinson), pp. 343-366, Blackwell Scientific, Oxford.  173  Watson, M.A. 1986. Integrated physiological units in plants. Trends Ecol Evol., 1: 119123. Woitzik, F. and H. Mohr. 1988. Control of hypocotyl phototropism by phytochrome in a dicotyledonous seedling (Sesamum indicum L.). Plant, Cell, and Env., 11:653-661. Woodward, F. I. 1983. Instruments for the measurement of photosynthetically active radiation and red, far-red and blue light. J. Appl. Ecol., 20: 103-115. Wooley, J. T. 1971. Reflection and transmittance of light by leaves. Plant Physiol., 47: 656-662.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items