Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. Ecology and Ecophysiology of Subantarctic Campbell Island Megaherbs A thesis presented in partial fulfilment of the requirements for the degree of Masters of Science in Ecology at Massey University, Palmerston North, New Zealand. Vivienne Joy Nicholls 2000 Abstract The megaherb growth form is not common in the New Zealand flora; yet it is a distinctive feature of the flora of New Zealand' s subantarctic islands, such as Campbell Island (52°33'S, 169°09'E). It occurs in four genera: Pleurophyllum, Stilbocarpa, Bulbinella and Anisotome. Their unusually large form and striking colourful flowers have long been commented as possible relics of a more widespread flora or possibly more recently dispersed and adapted to the present conditions. This research focussed on how well they are adapted to their environment and how adaptable they may be to novel conditions using ecophysiological methods. The breeding system of Pleurophyllum was also looked at as an example of the development of reproductive systems in an isolated island environment. Biomass allocation patterns were investigated in two species, A. latifolia and P. speciosum in order to determine whether this growth form was extraordinary compared with other herbaceous perennials. Six shoots of each species were harvested on the island, and sorted into components (leaf, stem, rhizome, reproductive) and dried back on the mainland. Leaf:stem ratios were found to be higher in both species than other perennials. This may be a response to ensure adequate resource harvesting i.e. light, water, nutrients, in an exposed, competitive environment. Gas exchange was studied in different light and temperature regimes using a growth cabinet in order to determine light and temperature tolerance and possible optima. Nine plants each of subantarctic Pleurophyllum criniferum, Anisotome latifolia, Stilbocarpa polaris (three plants only), New Zealand subalpine Ranunculus lyallii, and Chatham Island Myosotidium hortensia were exposed to three temperatures (7°C, l 7°C, 25°C) and four light levels (0, 150, 380 and 950 µmol m-2 s-1). P. criniferum had the fastest photosynthetic rate (of 8.6 µmol m-2 s-1 ), followed by R. lyallii (7.8 µmol m-2 s-1), A latifolia and M. hortensia (both with 4.6 µmol m-2 f 1 ) and S. polaris (2.1 µmol m-2 s- 1 ) (averaged over all light and temperature levels). All species had higher photosynthetic rates at the highest light level. P. criniferum did not appear to be inhibited by the high temperatures while A latifolia did. Respiration rates increased with temperature with A. latifolia having the highest rate followed by M. hortensia, P. criniferum, R. lyallii and S. polaris. These results iii suggest that although the subantarctic megaherbs appear to be well adapted to the low light and cool temperature regime of their environment, they may be more plastic to environmental change at low ranges, especially temperature, than expected. Experiments were carried out in situ on Campbell Island to ascertain the effect of environmental perturbations, using carbohydrate, chlorophyll, and growth analysis of leaves on P. speciosum, P. hookeri and P. criniferum, Bulbinella rossii, and A. latifolia. The rnicroenvironment around each plant was altered by combinations of reducing light, increasing shelter, increasing nutrients, altering photoperiod and increasing temperature over the course of eight weeks. Most of the treatments had very little effect on the carbohydrate pools of the species suggesting that environments were not limiting growth on the island. However leaf growth in P. criniferum increased in reduced light as did leaf growth in B. rossii, suggesting etiolation (sensitivity to light). A. latifolia showed an increase in carbohydrate pools with increase in temperature, compared with P. speciosum. Contrary to expectations increased night length also had a small positive effect on growth. Chlorophyll content remained unaffected by treatment but differed significantly between species (ranging from 56 mg g' 1 in A. latifolia to 149 mg g-1 in B. rossii). These results suggest that the megaherbs are quite plastic in their responses, relatively unaffected by environmental perturbations. Megaherbs may be pollen limited in an environment that might be considered unfavourable to insects. 15 plants each of P. criniferum (discoid capitulum), P. speciosum (rayed capitulum) and P. criniferum x speciosum (rayed capitulum) were randomly chosen and four treatments - control, bagged and hand-outcrossed, bagged and hand-selfed, and bagged (no assisted pollination) - individually applied to four scapes on each plant. Capitula were collected at the end of the season and proportion of seed set analysed in the laboratory. P. speciosum, the most colourful of the genus, is an obligate out-crosser (i.e. self-incompatible). The less colourful P. criniferum is autonomously self-compatible, and the hybrid, while being able to set seed autonomously, sets more seed when outcrossed. These results indicate that these plants are not pollen or pollinator limited. Observations showed that small midges were the most active on these plants, although their efficacy was not examined. Their activity was significantly affected by wind. A selection of breeding systems and the existence of a hybrid suggests a capacity for recombination of genetic material and potential for adaptive radiation of species. iv This study shows that the subantarctic megaherbs are well adapted to their environment. Their apparent plasticity in physiological responses to environmental, and their range of breeding systems, also indicates that they may be more adaptable to novel environments than previously considered. However, whether they are relicts of an ancient, more widespread flora, or whether they have evolved more recently in situ remains unresolved. V Acknowledgements I am grateful for the support of Massey University who provided financial assistance in this venture, via the Research and Graduate Funds, Professor S. Rumball (Science Dean) and the Ecology Group Development Fund. For logistical support I would like to thank Southland Depa11ment of Conservation, in particular Carol West, without whose cooperation this project would not have been possible. My thanks also go to Rodney and Shirley Russ of Southern Heritage Expeditions for their haulage and hospitality on the high seas. Island companions enhanced the experience and I thank: Geoff Walls and Colin Meurk for making excellent travelling companions around the Southern Ocean; Mike Joy for hi s assembly line structural creations, bread baking, and wonderful photography; Peter Moore for support and advice in my moments of scientific angst; and Bron, Georgie, Jenny, Kerri­ Anne and Brent for the laughs, the whisky and the haggis-hurling. On dry land I had assistance from many and I thank: my mother (Rae), Noeline Warr, and Ecology technician Cathy for their sewing skill s; Ecology engineer Jens for patience in the face of last minute deadlines, not to mention a ready ear for a yarn; Ecology technicians Carol and Hayden for endless hours of drying and weighing samples; Ian Henderson for statistical enlightenment in the face of a blank one; Liz Grant for map creation ; Ian Andrew for voluntary Diptera identification; Otago University Botanical Department and Plant Science, Institute of Natural Resources, Massey University for the loan of expensive measuring equipment; Ian Flux who collected data from Campbell Island while I stayed on terra firma; Lance Currie and the Fertiliser and Lime Research Centre, Institute of Natural Resources, Massey University for soil analysis; Crop and Food Research, Invermay, for the use of their growth cabinet, and in particular Graeme Parmenter, who assisted with this activity so obligingly, and Alison Evans for plant advice and support. And, lastly, but of course not least, my supervisors: co-supervisor Dr Alastair Robertson who guided the pollination study and provided statistical advice, editing and an open door, and Dr Jill Rapson, who got the whole project off the ground, and assisted throughout with extreme vi patience; thank you for enabling me to experience first hand some of New Zealand's most amazing flora (and lets not forget the fauna . . .. ). The rare opportunity of visiting these southern islands is particularly appreciated and I am extremely grateful for everyone's contributions to this project. For assistance with my soul I'd like to thank: my family therapists, Ma, Steve, and Geoff, who, provided good food (the odd fish), support and a sympathetic ear; my gardening therapists, Colin Mahy and Sue Pendergast, who provided a weekly breath of fresh air, gardeners' philosophy and date scones (on Wednesdays); my musical therapists, Adie Leng, Wal Robinson, Catherine Widdup, Ann Andrews and Colin Caddick who, as Callanish, offered a welcome interruption of music and good humour (on Monday nights); and my coffee therapists: Halema Flannagan, Grant Blackwell, Mike Joy, Al Hewit, Elissa Cameron, Jam Godfrey and Sarah Treadgold, Wayne Linklater, Alastair Robertson, Suzanne Bassett, who assisted with caffeine levels, inappropriate diversions and inane, yet stimulating conversation. To all these good folk, I thank you for the experience. vii TABLE OF CONTENTS ABSTRACT .... . . . .. . ... ..... .... . .... . .... . . . .. ... ... . ... . .. . ... . . . . .... . .. .. . .. .. .. .. . ..... . .. .iii ACKNOWLEDGEMENTS ..... ... .... . .... . . .. .............. .. .... . .... . .......... . ..... .. ... vi CHAPTER ONE: INTRODUCTION .................................................................................... 1 M EGAHERBS .. . .... ... . ..... . . ....... . ... . . . .. ..... . . . .. . .. .. . . . .. . ... ... . ... . ....... .. .... . . 1 N EW Z EALAND S UBANTARCTIC ISLANDS AND FLORA . ... . . . .. . ... . .... . ..... . . . ...... l THE MEGAHERB GROWTH FORM .... . ......... . .. . ........... . ... . .... . .. .. .. .. ...... .... 4 F LORAL BIOLOGY .. ... . ... .. ....... ...... . .. . . . . ........... . ... . .... . .. . ...... ... . . . ... ..... 4 THE PLASTICITY OF THE M EGAHERBS .. . .. . . . . . . . . . ..... . ... . . .. ... ... .. .. . . . . .. ... . . . . . . 5 GROWING INTEREST IN M EGAHERBS . ........ . ... . . .. .. . ....... ... ..... ... ......... . .... . 6 AIM .... .. ... . . .. .. ... ... ... ... . ... .... ... .. .... .. .......... .... .. ....... ...................... 6 RE FE REN CES ............................................................................... 7 CHAPTER TWO: BIOMASS ALLOCATION IN SUBANTARCTIC ISLAND MEGAHERBS, PLEUROPHYLLUM SPECIOSUM (ASTERACEAE) AND ANISOTOME LAT/FOL/A (APIACEAE) ................................. 11 INTRODUCTION ..................................................................................................... 11 METHODS ................................................................................................................. 12 RESULTS ................................................................................................................... 12 DISCUSSION ............................................................................................................. 14 STRATEGIES OFANISOTOME A D PLEUROPHYLLUM .. .. .. .............. . .... . ...... ....... ...... . ... . 14 EVOLUTION OF THE MEGAHERB FORM .... .... .... .......... . .. .......... .. .......... .. ..................... 14 CONCLUSIONS ............................................................................... 15 ACKNOWLEDGEMENTS ......... .. ........................ .. .. ... .. ..... .. .............................. . .... .. ..... 15 RE FE REN CES ................................................................................. 15 viii CHAPTER THREE: POLLINATION IN THE SUBANTARCTIC: PLEUROPHYLLUM.18 ABSTRACT ............................................................................................................... 18 INTRODUCTION ..................................................................................................... 19 STUDY ORGANISMS ................................................................................................... 20 STUDY AREA ....... ......... . ................ . ..... ... .. ....... ........ ... . .. . ........... .. ............................. 22 METHODS ................................................................................................................. 23 BREEDING SYSTEMS ....................... . ................................. . ....................................... 23 POLLINATOR OBSERVATIONS ................................................................................. ... 24 DATA ANALYSIS ................................................................... .............. ...... ..... ........... 24 RESULTS ................................................................................................................... 26 BREEDING SYSTEMS ................................................................................................. 26 POLLINATOR OBSERVATIONS ................................................................................ ... . 26 DISCUSSION ............................................................................................................. 30 SELF-INCOMPATIBILITY ..... ... ............................... ... .................................................. 30 POLLINATION ............................................................................................................ 31 RE FE REN CES .......................................................................................................... 34 CHAPTER FOUR: IMPACT OF IN SITU ENVIRONMENTAL MANIPULATIONS ON MEGAHERB CARBOHYDRATE, CHLOROPHYLL AND GROWTH LEVELS ................................................................................................... 37 ABSTRACT ......... ... .. ................... .. .... ..... ........ .......... . ...... ............. .......... . ................... 37 INTRODUCTION ....................................... .............................................................. 39 METHODS ................................................................................................................. 43 CON GENERIC EXPERIMENT ....................... ...... ..... ...... .... . ... .... ... ........ .. .. .... ... .. ... ... . .... 45 "Shade" .......... ............................ .. .............. .. .................................... ... ...... .. ........ 45 "Wind" .. ............................... ...... .. ... ... ...... .... ....... ..... .......... .. ........ ..... ..... ..... ........ 45 "Nutrients" ........................... ... ... ..... ........ .......... ....... ... ......... ................... ... ... ...... 45 ADAPTABILITY EXPERIMENT ...... ... .. ... ............................... . .................... . ................. 47 "Cloche" .............. .. ... .. .... .. ............ .. ............ ..... .............. ... .......... ..... .... ............... 47 "Night" .. .......... ..... .. .......... ... ...... ....... .... ... ...... ..... ........ ... .. .. ...... ................. ........... 48 ENVIRONMENTAL DATA .. ... .. .. ....... ... .. . ... . ........... ..... .... .... .......... .................. .. ........ .. . . 50 TEST FOR SAMPLING EFFECT ........... .... .. . ............. .. .............. . .. ... ........................ .... .... 50 CHO ANALYSIS METHODS .. . ............................................................. . .... .. ...... ........... 51 CHLOROPHYLL ..... .. ......... .. .. ... .. .... ....... ..... .... .. . ........ ... ............ .. ... ..... ... .. .. .......... ....... 53 ix STATISTICAL ANALYSIS .... ...... ... .......... .. ... ..... . ............ ..... ...... ... ......... .... . ...... .. ...... .... 54 RESULTS ............................................................... .................................................... 55 ENVIRONMENTAL CONDffiONS .................... . ..... .. ........ .. .... .. .... . .......... ... .. ... . .. . .......... 55 Rainfall .... .. ..... ....... ..... .... .................... ..... .. ........... .. ...... ...... .. ......... ...... ... ............ . 55 Temperature ....... .... .... ... ..... ....... .. .... ...... .... ........... ...... ..... .... .. .. ........ ... ..... ............ 55 Light .... .... ............. ... ......... .... ..... ... ..... ...... ................ ... ......... ..... .... ... ........... ... ....... 60 Wind .......... .. ...... .. ... ... ... .... ......... .... ......... .. ... ..... .. ....... ...... ..... ... ............. ... ...... ....... 61 Soils .................................................. ...................................... .. ....... ........... ......... 61 Repeated sampling test ...... ....... .... ......... .. ... ... ......... ....... .. ..... ................. ...... ... ... .. 62 CON GENERIC EXPERIMENT .... ...... . .... . ... . ................................................................... 63 Phenology .. ..... ... ... ........................... ... .... ... ... .......... .. ...... ..... ....... ...... ...... .... ......... 63 Growth rates ..... .. ...... ..................... ........ .. .. .... ..... ...... ........................................... 63 Chlorophyll ................... ........ ... ... ................... .. ... .... ............... .... .......................... 65 Carbohydrates ......... .... .... ... ........ ..... ..... .... ... ........ .... ....... ... .. ....... ..... .. ...... ........ .... 66 ADAPT ABILITY EXPERIMENT .... .. ... .. ... .. . . .... . .... ... . .. .......... ......... .... .. ....... . ......... ....... .. 67 Phenology .... ... .... ..... ........................................ .... ....................... ......................... 67 Growth Rates .... .. ..... .... ........ ...... ....................... ... ........................ ..... ................... 68 Chlorophyll ...... .. ... .... ... ... ...... ... ... ........ ... ... .. ... .. .. ...... .... ........... ... .. .... ....... ........ ..... 68 Carbohydrates .... ............. ... .......... ...... ... ..... ... ....... .. .. .. ............. ........ .................... 69 CHO content ... ............................................................... ........ ...... ... .......... .... ...... . 70 DISCUSSION ........................ ...... .......... ........ .... ............ .. ........................ .. ........ ........ . 73 LIGHT EFFECTS ON MEGAHERBS ... . ........ ...... .. .. .. . .................................. . ............... ... . 73 PHOTOPERIOD EFFECTS ........ ...... .................. ........ ... . .. ..... ............... .... ................. .. .... 7 5 N UTRIENT EFFECTS .................................................................................... .. ... .. ........ 7 5 TEMPERATURE EFFECTS .. ................ .. . ...................... ...................... ....... . ...... ............ . 7 6 SHELTER EFFECTS .... ..... .. ....... .......... .. ... .. ........... . .. .................... .... .... ... .. .... ............... 77 SEASONAL EFFECTS .... .... . .... .. ..... .... . ....................................................... ... ............. .. 78 PHENOLOGY .. . .. . .. .. ..... ... ......... .. .. .... . ...... .. .... .. .... ........ ......... ... ..... .. . ... ............... . ..... . .. 80 CONCLUSION .............................................. ................................. ................... ........ 82 RE FE REN CES .......................................................................................................... 83 CHAPTER FIVE: PHOTOSYNTHESIS AND RESPIRATION OF FIVE MEGAHERB SPECIES IN A RANGE OF LIGHT AND TEMPERATURE REG IMES ................................................................................................ 87 ABSTRACT ............................................................................................................... 87 INTRODUCTION ................................ ..................................................................... 88 METHODS ................................................................................................................. 90 X RESULTS ............................................................................................... .................... 92 PHOTOSYNTHESIS ...... .. ... ..... .. .. . ... ... ..... . ............. . .. . ... ...... . ...... .. ... .... .. . . ....... . ......... ..... 92 RESPIRATION .. . ....... .. . .... ... ....... .. .. . ...... ................... ................ .. . .... .... . .. .. .... . ... ...... . .... 96 DISCUSSION ............................................................................................................. 98 LIGHT ·· · ····· ···· ··· ··· ·· ····· ··· ·· · ··· ·· · · ·········· ·· ·· · ·· ···· · · ······························ ·· ········· ·· · · · · ···· ······· 98 T EMPERATURE ....... ... . ................... . ...... ............. .. . ..... . ... .... .. ... .. .. .. ...... ... . .. . ... .... ... .... .. 99 CONCLUSION .......... .............................................................................................. 101 RE FE REN CES ........................................................................................................ 102 CHAPTER SIX: DISCUSSION ........................................................................................ 105 P HYLOGENETIC C ON SID ERA TI ONS ................................. .................... .. ...... . .... . .. .. ... 106 T HE M EGAHERB F ORM ......... . .......... .... ........ . ... .... .. ....... .............. ... .... . . . .. ............... . 107 PHENOLOGICAL RESPONSES ... . ... . ... . .. . .......... . ............... ............... ...... . ...... . . . ... . ....... 110 E COPHYSIOLOGICAL RESPONSES ................ . . . ... .......... . ... .... .. .... .. . .......... .. .. . ............ 112 FOLIAR P ANELS············· ·· ··· ······· ·· ·· ···· · ·· ··· ·· ·· · · ·········· ··· · ·· ······ ·· ····· · ······ · ··· · ···· · ··· · ···· · ·· 114 C ULTIVATION ....... .... ... ............................................ .................. ...... . .............. . ....... 11 6 FUTURE W ORK .. ............................................................. ..... . . .................. ............... 116 CONCLUSION ............................................................... ......................................... 117 REFERENCES ........................................................................................................ 119 xi CHAPTER ONE Introduction Megaherbs 'Megaherbs', 'pachycauls' (Mabberly, 1979) and 'giant rosettes' (Hedberg, 1964) all describe plants with little or no arborescent growth, and with a limited number of growing points. They tend to have a 'stocky' habit i.e. having short internodes, large leaves and large rhizome. Examples of these large plants occur in the Hawaiian islands as silverswords, Argyroxiphium (Goldstein et al., 1989) in the slightly more dendroid forms in Kenya as Lobelia and Dendrosenecio (Schulze et al., 1985) and as Espeletia (Monasterio and Sarmiento, 1991) in the Andes. The species mentioned inhabit environments where extremes of temperature occur on a diurnal basis and irradiances are high. Research into their physiology and ecology has shown a wide range of adaptations to cope with these conditions including pubescence ( e.g. Meinzer and Goldstein, 1985), rosette form ( e.g. Hedburg, 1965), leaf orientation ( e.g. Melcher et al., 1994) and large pith (Goldstein et al., 1984). While diurnal temperatures are more extreme in the alpine tropics comparisons have been made with this climate and that of the New Zealand subalpine and subantarctic (Troll, 1960; Mark et al., 2000), and with the plant life forms found there (e.g. Hedberg, 1964). One representative occurs on the Kerguelen Islands: the Kerguelen cabbage, Pringlea antiscorbutica (Aubert et al., 1999). In the subantarctic islands of New Zealand, the growth form is a common element of the vegetation and is represented by four genera. Here it is a herbaceous growth form of typically large size (up to 1 m high), with a brightly coloured floral display and stout rhizome and occurs in the fo llowing species: Pleurophyllum speciosum, P. criniferum, P. hookeri, P. criniferum x speciosum, Bulbinella rossii, Anisotome latifolia, A. antipoda, A. latifolia x antipoda, A. acutifolia and Stilbocarpa polaris, all of which are endemic to the Southern Ocean. New Zealand Subantarctic Islands and Flora The subantarctic islands are a group of exposed islands lying in the Southern Ocean between 47°40'S and 52°38'S latitude and incorporating the Bounty Islands, The Snares, Antipodes Islands, Auckland Islands and Campbell Island (Figure 1). Their weather is characterised by cool but relatively stable temperatures, strong winds, a high degree of cloud cover and frequent precipitation giving a high degree of humidity (De Lisle, 1965). 1 800 km ~.. Chatham Islands The Snares · · Bounty Islands ------+-------+------..---+-Antipodes Islands -- 50° ~ Auckland Is lands Campbell Island 160° 170° 180° Figure 1: The subantarctic islands in relation to New Zealand . Campbell Island lies 700km south of New Zealand. The vegetation of the subantarctic islands has been a source of great interest and centre of research since the expeditions of the early botanists (Hooker, 1844; Buchanan, 1883; Cockayne, 1903) and this continues to the present day (e.g. Godley, 1982; Meurk et al. , 1994). The reason for such interest lies in the contrasting features between these southern islands and the flora of the mainland, leading to much speculation and debate concerning their origin and development (e.g. Wardle, 1978; Godley, 1979; Lloyd, 1985). The megaherb growth form is considered unusual on the mainland, where the flora is characterised by diminutive flower form and pale colour and the herbs are generally small. Only a few examples of megaherbs occur outside the subantarctic: Ranunculus lyallii, Anisotome lyallii and Aciphylla species on the main islands and Stilbocarpa 2 robusta and S. lyallii on Stewart Island, and the Chatham Island forget-me-not, Myosotidium hortensia. There are several areas of intrigue surrounding the megaherb in the subantarctic context. Of interest is the reason for the abundance of this growth form in the subantarctic flora (and conversely its absence in the mainland flora), and how well suited it is to a subantarctic environment - i.e. one that is very windy, cool, often overcast and wet. Climate change, isolation and the shifting zone of the Antarctic Convergence (Walton, 1984) may all have influenced the evolution of the flora of these islands. This growth form may represent relicts of a more widespread pre-glacial flora (Godley, 1975), which disappeared from the mainland during less suitable conditions, the islands being possible remnants of a larger subcontinental landmass (Gressitt et al., 1964). Alternatively, they may have dispersed there from neighbouring land masses and evolved in situ (Lloyd, 1982). This study was based on Campbell Island, 52°33 'S , 169°09E, the southernmost of the New Zealand islands, situated 700 km south of Bluff, New Zealand (Figure 2). It receives a mean annual rainfall of 1360 mm distributed throughout the year (a mean of 265 annual raindays) and a mean wind speed of 32 km hou( 1 (9.2 m s-1). The mean annual temperature is 6.9°C while the mean monthly maximum temperature for the warmest month is 15.9°C and mean minimum for the coldest month is -2.7°C. Although sunny, bright days can occur the island is usually overcast receiving less than 1 hour of sunshine a day for 215 days of the year [De Lisle, 1965]) and only a mean of 659 hours of sunshine a year (NZ Meteorological Service, 1983). The vegetation of Campbell Island, predominantly consists of Dracophyllum shrubland with mixture of Myrsine and Coprosma, and tussock.land of Chionochloa antarctica and Paa litorosa. Megaherb meadows are scattered over various areas of the island consisting of Pleurophyllum, Anisotome and Stilbocarpa and other species (for a more detailed vegetation description of the island see Meurk and Given (1990) and Meurk et al., [1994]). Megaherb species have quickly recolonised these areas from their previously restricted habitat of rocky ledges, with the removal of sheep (Meurk, 1982). 3 The development of the megaherb form in what is the equivalent of a subalpine/alpine environment (Bliss, 1979) is of interest. Altitude generally governs plant life forms by imposing increasingly harsh regimes of desiccation and temperature (Wardle, 1974). As temperatures decrease with altitude canopy height generally becomes lower allowing the prostrate plants to exploit the relatively warmer temperatures at ground level (Wardle, 1974). 'Phytomicroclimates' are created around plants and are influenced by the growth form - leaf arrangement, size, shape, pubescence - and may allow an increase in plant temperature more than surrounding air aiding metabolic processes (Billings, 1974). As alpine plants have relatively high respiration rates (Billings, 1974), carbon balance - net gain of photosynthesis over respiration - is also crucial at these altitudes. The gain during the growing season must outweigh both the structural and winter maintenance costs (Wardle, 197 4) and so taller vegetation gives way to low woody shrubs which are in turn replaced by herbaceous plants as altitude increases. Alpine environments are more extreme than subantarctic environments with below freezing temperatures, possibility of desiccation, and possibility of high temperatures during the day. The subantarctic climate in comparison is more equable and temperate. While low temperatures and strong winds may not favour woody vegetation, the high humidity might allow for the larger leaf form of the megaherb. The Megaherb Growth Form A study investigating the 'megaherb' growth form may require some justification for the term megaherb and what might typify a megaherb. Biomass allocation patterns such as shoot:root or shoot:stem ratios may indicate a characteristic megaherb feature. In such plants one might expect a significantly large proportion of biomass attributed to foliage compared with other herbaceous plants. A study of tissue allocation patterns of P. speciosum and A. latifolia (Chapter 2) explores the possibility of extraordinary allocation patterns and may give some indication into the advantages this growth form has in the subantarctic environment. Floral Biology In general, islands are noted for their lack of zygomorphic, tubular and brightly coloured flowers (Carlquist, 1966) i.e. flowers that attract specialist pollinators such as birds, 4 butterflies, long-tongued bees. Instead the flowers are usually small, pale and simple, attracting a range of generalist pollinators, usually diptera, moths or beetles. This is especially so in the New Zealand environment (Lloyd, 1985). It is unusual then, given this trend, to find brightly coloured flowers on the subantarctic islands where, in accordance with the windy, cool, wet environment one would expect a high degree of autogamy, self-compatibility and wind pollination (Leppick, 1977). In Chapter 3 I investigate the pollination and breeding systems of the showiest genus, Pleurophyllum: P. criniferum and P. speciosum - and their hybrid - P. criniferum x speciosum. The Plasticity and Adaptability of the Megaherb Over time, after initial colonisation of islands, plants may undergo adaptive radiation (sometimes resulting from stochastic processes) (Barrett, 1996) to fit different niches offered by the new environment and climate. Alternatively, during the course of their evolution they may also suffer bottlenecks (Vitousek, 1988; Barrett, 1996) and consequently have smaller levels of genetic variation than mainland sister species (e.g. Barrett and Husband, 1990). This may actually reduce their ability to adapt to new environments . If genetic variation is limited it might be expected that these species are restricted in the range of environmental conditions i.e . cool temperatures, low light levels and high winds. If exposed to light regimes above their optima, photoinhibition (damage to the photosynthetic apparatus) can result (Bjorkman, 1981 a), while high temperatures might push respiration rates beyond economic levels (Bjorkman, 1981b). Under novel conditions such as controlled environments and environmental manipulation in situ, these possibilities may be examined using gas exchange measurements and carbohydrate analysis giving some insight into environmental optima and possible plasticity of the species (Chapters 4 and 5). Gas exchange studies have the advantage of providing a measure of a plant's immediate response to its environment (e.g. Komer and Diemer, 1987). Study of carbohydrate use on a daily basis, on the other hand, may give an indication of daily respiration loads on the plant, indicating the potential for growth. By using environmental manipulation in situ this type of study can be taken a step further to look at such factors as environmental optima and species plasticity. This may also aid the 5 understanding of how they survived past climate change and may survive it in the future as well as suggesting how they might perform in cultivation. Growing Interest in Megaherbs Latterly, the subantarctic islands have become a popular attraction for visitors fascinated by these far-flung islands and the abundant wildlife. With the increasing profile of the islands, one area of attention has focussed on the megaherbs. Cultivation of these species on the mainland has been attempted with mixed results. Anisotome, Stilbocarpa and Bulbinella survive (pers. comm. Alison Evans) but Pleurophyllum has proven the hardest to grow. Yet it is P. speciosum in particular that has the most appeal with its broad, corrugated, hairy leaves and rose-purple daisy flowers on multiple stems. It is possible that while some species e.g. Anisotome, may be able to adapt to new conditions, Pleurophyllum is narrowly adapted to subantarctic conditions with little potential to change. This study investigates these possible limitations. Aim The aim of this study is to explore the phenomenon that is the subantarctic megaherb with a view to contribute towards the understanding of their evolution and development. It is hoped that the initial ecophysiological and biological investigations that follow may provide a background for further work in this little understood area. 6 References Aubert, S ., Assard, N., Boutin, J.-P., Frenot, Y. and Dorne, A.-J. 1999. Carbon metabolism in the subantarctic Kerguelen cabbage Pring/ea antiscorbutica R.Br.: environmental controls over carbohydrates and proline contents and relation to phenology. Plant, Cell and Environment 22: 243-254. Barrett, S.C.H. 1996. The reproductive biology and genetics of island plants. Philosophical Transactions of the Royal Society London B. 35 I: 725-733. Barrett, S.C.H. and Husband, B.C. 1990. The genetics of plant migration and colonisation. Pp. 254-277. In Plant population genetics, breeding, and genetic resources. Eds. A.H.D. Brown, M.T. Clegg, A.L. Kahler and B.S. Weir. Sunderland Massachusetts: Sinauer Associates, Inc. Billings, W.D. 1974. Arctic and alpine vegetation: plant adaptations to cold summer climates. Pp. 403-443. In Arctic and Alpine Environments. Eds. J.D. Ives and R.G. Barry. Methuen, London. Bjorkman, 0 . 198 la. Responses to different quantum flux densities. Pp. 55-107. In Encyclopaedia of Plant Physiology NS Volume 12A. Physiological Plant Ecology 1. Eds O.I.G. Lange, P.S. Nobel, C.B. Osmond and H. Ziegler. Springer, Berlin. Bjorkman, 0 . 1981b. The response of photosynthesis to temperature. Pp. 273-301. In Plants and their atmospheric environment. Eds J. Grace, P.G. Jarvis. 21st symposium British Ecological Society, Edinburgh 1979. Blackwell Scientific Publications Oxford. Bliss, L.C. 1979. Vascular plant vegetation of the southern circumpolar region in relation to antarctic, alpine and arctic vegetation. Canadian Journal of Botany 57: 2167- 2178. Buchanan, J. 1883. Campbell Island and its Flora. Transactions and Proceedings of the New Zealand Institutel6: 398-400. Carlquist, S. 1966. The biota of long-distance dispersal. IV. Genetic systems in the floras of oceanic islands. Evolution 20: 433-455. Cockayne, L. 1903. A botanical excursion during midwinter to the Southern Islands of New Zealand. Transactions and Proceedings of the New Zealand Institute 36: 225-333. 7 De Lisle, J.F. 1965. The climate of the Auckland Islands, Campbell Island and Macquarie Island. Proceedings of the New Zealand Ecological Society 12: 37-44. Godley, E.J. 1975. Flora and vegetation. Pp. 177-229. In Biogeography and Ecology in New Zealand. Ed. G. Kuschel. Monographs in Biology 27. Junk, The Hague. Godley, E.J. 1979. Flower biology in New Zealand. New Zealand Journal of Botany 17: 441-66. Godley, E.J. 1982. Breeding systems in New Zealand plants 6. Gentiana antarctica and G. antipoda. New Zealand Journal of Botany 20: 405-420. Goldstein, G. , Meinzer, F. and Monasterio, M. 1984. The role of capacitance in the water balance of Andean giant rosette species. Plant, Cell and Environment 7: 179-186. Goldstein, G., Rada, F. , Canales, M.O., and Zabala, 0 . 1989. Leaf gas exchange of two giant caulescent rosette species. Acta Oecologica/Oecologia Plantarum JO: 359- 370. Gressitt, J.L, Rennell, K.P and Wise, K.A.J . 1964. Insects of Campbell Island. Ecology. Pacific Insects Monograph 7: 515-530. Hedberg, 0 . 1964. Features of Afroalpine plant ecology. Acta Phytogeographica Suecica 49: 1-144. Hooker, J .D. 1844. The botany of the Antarctic voyage of H.M. Discovery Ships Erebus, and Terror in the years 1839-1843, under the command of Captain Sir James Clark Ross; Volume I: Flora Antarctica; Pt. l: Botany of Lord Auckland's Group and Campbell's Island. Reeve, London, U.K. Korner, Ch. and Diemer, M. 1987. In situ photosynthesis responses to light, temperature and carbon dioxide in herbaceous plants from low and high altitude. Functional Ecology I: 179-194. Leppick, E.E. 1977. The evolution of capitulum types of the Compositae in the light of insect-flower interaction. Pp. 61-89. In The Biology and Chemistry of the Compositae . Volume I. Eds. V.H. Heywood, J.B. Harborne and B.L.Tumer. Academic Press, London. Lloyd, D.G. 1982. Variation and evolution of plant species on the outlying islands of New Zealand, with particular reference to Cotula featherstonii. Taxon 31: 478- 487. Lloyd, D.G. 1985. Progress in understanding the natural history of New Zealand plants. New Zealand Journal of Botany 23: 707-722. 8 Mabberley, D.J. 1979. Pachycaul plants and islands. Pp. 259-291. In Plants and Islands. Ed. D. Bramwell. Academic Press, London. Mark, A.F., Dickinson, K., Hofstede, R. and Halloy, S. 2000. Patterns in New Zealand alpine vegetation, plant distribution, life forms and environments: Temperate oceanic and tropical high mountain affinities. P. 48. Abstract from III Southern Connection Congress. Wickliffe Press Ltd., New Zealand. Meinzer, F. and Goldstein, G. 1985. Some consequences of leaf pubescence m the Andean giant rosette plant Espeletia timotensis. Ecology 66: 512-520. Melcher, P.J., Goldstein, G., Meinzer, G.C. , Minyard B., Giambelluca, T.W. and Loope, L.L. 1994. Determinants of thermal balance in the Hawaiian giant rosette plant, Argyroxiphium sandwicense. Oecologia 98: 412-418. Meurk, C.D. 1982. Regeneration of subantarctic plants on Campbell Island following exclusion of sheep. New Zealand Journal of Ecology 5: 51-58. Meurk, C.D. and Given, D.R. 1990. Vegetation map of Campbell Island 1:25 OOO. DSIR Land Resources, Christchurch, New Zealand. 1 sheet. Meurk, C.D., Foggo, M.N. and Wilson, J.B. 1994. The vegetation of Subantarctic Campbell Island. New Zealand Journal of Ecology 18: 123-168. Monasterio, M. and Sarmiento, L. 1991. Adaptive radiation of Espeletia in the cold Andean tropics. Tree 6: 387-391. New Zealand Meteorological Service, 1980. Summaries of Climatologial Observations to 1980. Ministry of Transport, New Zealand. Schulze, E.-D., Beck, E., Scheibe, R. and Zeigler, P. 1985. Carbon dioxide assimilation and stomata! response of afroalpine giant rosette plants. Oecologia (Berlin) 65: 207-213. Troll, C. 1960. The relationship between the climates, ecology and plant geography of the southern cold Temperate Zone and of the tropical high mountains. Transactions of the Royal Society London Series B. 152: 529-532. Vitousek, P. 1988. Diversity and biological invasions of oceanic islands. Pp.181 - 189. In Biodiversity. Eds. E.O. Wilson and F.M. Peter. National Academy of Sciences, Washington DC. Wallace, L.L. and Harrison, A.T. 1978. Carbohydrate mobilization and movement in alpine plants. American Journal of Botany 65: 1035-1040. 9 Walton, D.W.H. 1984. The terrestrial environment. Pp. 1-60. In Antarctic Ecology: Volume I. Ed. R.M. Laws. Academic Press, London. Wardle, P. 1974. Alpine timberlines. Pp. 371-402. In Arctic and Alpine Environments. Eds. J.D. Ives and R.G. Barry. Methuen, London. Wardle, P. 1978. Origin of the New Zealand mountain flora, with special reference to trans-Tasman relationships. New Zealand Journal of Botany 16: 535-550. 10 CHAPTER TWO Y.J . NICHOLLS and G .L. RAPSON Eco logy, Institute of Natural Resources. Massey University, Palmerston North , New Zealand. Email: V.J .Nicholls@ mas sey .ac .nz SHORT COMMUNICATION BIOMASS ALLOCATION IN SUBANTARCTIC ISLAND MEGAHERBS, PLEUROPHYLLUM SPECIOSUM (ASTERACEAE) AND ANISOTOME LAT/FOL/A (APIACEAE) Summary: We analysed bio mass allocation of Pleurophyllum speciosum (Asteraceae) and Anisotome latifolia (Apiaceae) to explore the 'megaherb ' phenomeno n, the apparent importance of large- leaved , colourful forbs on southern oceanic offsho re islands. The two species had similar shoot dry weights , with high leaf:stem ratios. Even w ithin the megaherb form the re are differences in shoot allocations, with Pleurophy/lum investi ng more biomass in rhi zome than fol iage, compared wi th Anisotome. The megaherb form might be attributable to responses to the physical environment, involving the pre-emption of resources such as li ght, nutri ents, water, or space; alternative ly it may be re lated to the pauc ity of woody species at this latitude. Keywords: Megaherb; subantarcti c; Southern Ocean; island ; allocation; phenology: strategy; leaf:stem ratio. Introduction Situated in the Southern Ocean south of New Zealand between 47°40' and 52°38' S latitude are a small and scanered group of volcanic and sedimentary islands, the c urrent representatives of subantarctic landmasses present at these latitudes possibly since the end of the Miocene (Marshall and Browne, 1909) . Their flora con tains several so-called ' megaherb ' species, members of the genera Pleurophyllum (Asteraceae), Anisotome (Apiaceae). Bulbinella (Liliaceae) and Stilboca rpa (Arali aceae). These herbaceous perennial forbs have large growth forms (often more than I metre high or wide). with large leaves and very co lou rful floral displays (Hooker. 1844). Their striking growth fo rm (Fig. I ) appears extraordinary compared with other herbaceous perennials, and may be an adaptation to their southern oceanic island e nvironment. Six randomly chosen plants each of Pleurophyllum speciosum and Anisotome latif olia from Campbell Island were destructively harvested and resource al locations studied . These spec ies were then compared with other herbaceous perennials to identify biomass allocation patterns that might be characteristic of megaherbs . Methods Campbell Island li es 700 km south of Bluff, New Zealand at 52°33'S, 169°09 ' E. It is an 11 OOO ha windswept island with moderate rainfall (mean of 1361 mm y( 1 di stributed throughout the year) and low annual sunshine hours (659). Its oceanic cl imate results in a mean monthly maximum temperature for the warmest month of 15.9 C and minimum for the co ldest month of -2.7 C (NZ Meteorological Service, 1983). Pleurophyllum speciosum (Hook. f.) is a rosette herb (up to 50 cm high and 100 cm wide) with large (up to 75 cm or more wide), co rrugated leaves. A sing le plant may produce up to 16 scapes containing 10 o r more capitula with pink florets, which are pollinated by insects and possibly also wi nd (pers. abs.). While considered to be evergreen, the plant reduces in size over winter, w ith the outer leaves dying off, and spring regrowth is possibly suppo rted by the large rhi zome. Anisotome latifolia (Hook. f. ) is a taller (70 cm), evergreen plant with long petio les supporting the pinnate laminae; one : crown produces usually one dioec ious scape. The habitat of these megaherbs varies from high altitude turf-meadow to maritime megaherb-tussock grassland (Meurk, Foggo and Wilson , I 994b). The sampled habitat of Anisotome and Pleurophy llum was an open subalpine e nvironment comprising tussocks (Chionochloa and Paa) and occasional shrubs of Dracophyllum, Coprosma, Myrs ine and Polystichum (bordering 'Poa litorosa meadow and Chionochloa ', and 'Tundra mosaic ' as mapped by Meurk and Given ( 1990)). The site was at approximately 52° 33 ' S and I 69° 09 ' E at approximately 140 m above sea level. Six randomly chosen plants (with consideration to extraction logistics) of both species were excavated and destructi vely harvested in early February 1997. New Zealand Journal of" Ecolo!iy ( 1999) 23( 1 ): 87-93 ©New Zealand Ecological Society 11 EW ZEALAND JOURNAL OF ECOLOGY, VOL 23. NO. I, 1999 aJ b) Figure I: (u) A Kroup of'Pleurophyllum speciosum; (b ) a lone Anisotome latifolia. Roots proved difficult to harvest a nd were omitted from further ana lysis. Harvest of the megaherbs was afte r peak flowering time: the flowers had begun to dehisce in Pleurophyllum and the male scapes of Anisotome were beginning to rot; consequently some scape material could not be collected. The plants were sorted into foliage. reproductive matter (scapes and capitu la) . rhizome, and stem, and weighed fresh. They were then laid out to a ir-dry prior to transportation back to New Zealand. The lamina:petiole ratio depicts allocations within the leaf to dedicated supporting vs. light harvesting tissue. While the lamina-petiole boundary is clear in Anisotome latifolia, the petiole in Pleurophyllym speciosum was arbitrarily defined as the basal part of the leaf below the expanded blade, which is considerably paler and more hairy than the lamina. In New Zealand each leaf of each plant was subdivided into lamina and petiole ; reproductive material was divided into scapes, pedicels and capitula. All material was then dried for 72 hours at 65 Cina vacuum oven (- I 5atm.) and weighed. The quantity of ti ssue missing due to other sampling was visually assessed at time of weighing and values for the weighed portions adjusted. Statistical analysis of ti ssue weights and allocations was by Analysis of variance (d. f. = I 0) using SYSTAT (SYSTAT, 1992). Differences in a llocations between large (assumed to be older) and small (assumed to be youn ger) leaves were examined by ranking leaves by size , with each plant standardised over the range O - 1. For mean lamina:petiole ratio, leaves of each plant were first grouped into quartiles by biomass. Results Pleurophy llum and Anisotome had contrasting fresh:dry weight ratios (Pleurophyllum: mean± S.E. = 10. 13 ± 0.43, Anisorome: 5.75 ± 0.66), indicating that Pleurophyllum is more succulent. 12 NICHOLLS and RAPSON: BIOMASS IN MEGAHERBS Table I: Mean 1issue al/oca1ions (% of shoo/ dry mass), and lea{"s1em and lwnilw.pe1iole mean ra1ios ± s1andard errors. Roof rrwlerial is omiued because of'harves/ difficullies, bu1 sub1errunean rhizomes were excuva1ed. Reproduc1ive /issues include sc:apes and pedicels. Proportion of shoot dry mass(%) Dry ma s ratio Species Stem Rhizome Leaf Reproductive Leaf:stem Lamina:petiole Pleurophyllum specio.wm Anisotome /a1ifolia I 2 20 8 The two species of megaherb did not differ significantly in their total shoot dry weights (rhizome, stem, reproductive and foliage tissue) , being 220 g per plant for P/europhyllum and 180 g for Anisotome (P = 0.1 8. Error M.S. = 2287.7). There is very little variation in dry weight between the individual plants measured (Pleurophy/lum: S.E. = 16.69; Anisotome: S.E. = 21.99). Reproductive tissue (including scape) in Pleurophyllum makes up three times the proportion of shoot compared with Anisotome (Table I). Anisowme is dioecious, and as collection occured late in the summer when male flowers were dying off not all reproductive material was sampled. The single female plant has a notably higher allocation to reproductive biomass (20%) compared with an average of 6% for the male plants. Allocations to the non-reproductive shoot are comparable between the two species (mean± S.E. for Pleurophyllum = 162g ± 15.96, and 163g ± 7.93 for Anisorome; P = 0.9, Error M.S. = 1728.9), though Pleurophyllum has a higher allocation to rhizome (P = 0.00 I. Error M.S. = 17 .45), and Anisotome invests a high percentage of resources in foliage (Table I; P = 0.000, Error M.S. = 43.85). P/europhyllum has similar leaf numbers to Anisotome (mean± S.E. = 17 .0 ± 1.5 and 20.5 ± 2.7 respectively, P = 0.2, Error M.S. = 28.65). There is a higher proportion of leaves in the lowest biomass c lass in P/europhyllum compared with Anisotome. though the rest of the range of leaf sizes is similar (Fig. 2). A difference in leaf size patterns between the two species is suggested (Kolmogorov-Smimov, using the combined data for all plants: P = 0.07) . Both species have similar leaf:stem biomass ratios (Table I : P = 0.98, Error M.S. = 585.19). Pleurophyllum has a mean lamina:petiole ratio of 5.45 , three times the ratio for Anisotome. In Anisotome the laminae of the smallest quartile of leaves contribute a large proportion of the leaf (Fig. J; P < 0.001, Error d.f. = 38, Error M.S . = 1.58). This trend changes during foliage development with the largest leaves placing proportionately more resources into petioles. In Pleurophyllum these ratios have a reverse trend, with the laminae dominating the allocations of the largest leaves. 52 8 1 27 9 61.55 ± 9.29 61.19± 10.42 5.45 ± 0.50 1.80 ± 0.18 a) 50 40 '- Q) 30 .c E :::J C iii 20 Q) ....J 10 0 0.0 0.5 1.0 Leaf biomass classes b) 50 40 '- Q) 30 .c E :::J C iii 20 Q) ....J 10 0 0.0 0.5 1.0 Leaf biomass classes Figure 2: Frequency dis1ribu1ion of leaf biomass in six classes for (a) Pleurophyllum and (b) Anisotome, slandardised per plan/from O - I , and tn/al/ed over six plan/.~ of each species. 13 NEW ZEALAND JOURNAL OF ECOLOGY , VOL. 23 , NO. I , 1999 9 • 0 0 • 0 0 • • Q-4----~---~----.------, 0 2 3 4 Quartiles of leaf biomass Figure 3: Mean lwnina:periole rarios for each quarrile of leaves when leaves are ranked by biomass.from smallesr ( I ) lo lar>:esr (4)/i>r Pleurophyllum (0) und Anisotome (e ). Discussion Herbaceous species with large leaves are a conspicuous feature of the vegetation o n southern oceanic islands and ha ve long been comme nted upo n by botanists and natura li sts (e .g .. Hooker. I 844. Cockay ne. I 903, Meurk et al .. I 994b) . Do these spec ies rea ll y have unique features that set th e m apart from othe r herbaceous species'l Strategies of Anisotome and Pleurophyllum The two megaherb spec ies studied ha ve very s imilar standing crops at the time of harves t (early February , i.e .. late summer). They show a hi gher lea f:stem ratio (6 1.2 for Anisotome and 6 1.5 for Pleurophvllum) re lative to other herbaceous plants (2.5 ± 1.65. n = 10: Hickman. I 975: Hickman and Pite lka, I 975: Bostoc k and Benton, I 979; Abrahamson and Caswell. I 982: Gross. 1983; Joll s, 1984). Allocations to laminae within the leaf show considerable variation between the two species, with the petiole being small in the rosette of Pleurophvl/um. and an important component o f bio mass in Anisorome. Rhizo me storage of carbohydrates is a key factor in maintenance and possible support of vegetative ex pansion. at least in alpines (Mooney and Billings. i 960: c.f. Hadley and Rosen, 1974). The shoot allocation (excluding reproductive bi o mass) to rhizome in Anisorome is considerably lower than in Pleurophyllum. Anisotome overwinters at a size comparable with the summer form (with " retained vegetative tissues" - Pugliese and Kozlowski , 1990), so that this species 'exerts much influence on the winter physiognomy of the meadow' (Cockayne, 1903 ). Pleurophyllum sheds its outer leaves and stays in " perpetual somatic youth" (Pugliese and Kozlowski , 1990), reducing from around 75 cm over summer to a winter rosette approximately 25 cm in diameter (Cockayne. 1903) . Pleurophyllum may then require a rhizome-stored energy supply to support its initial regrowth in the spring. Thus the strategies of the two megaherbs do vary substantively . A characteristic and competitively successful strategy of tall herbs in open grasslands is rapid vegetati ve growth during the growing season. often resulting in a large above-ground mass (Al-Mufti er al., 1977); many of the leaves will be formed during that growing season (Yoshie. 1995). Pleurophyllum and particularly Aniso1ome may have such a competitive strategy with leaf size distribution. and thus possibly time o f initiation, apparently beinv continuous throu gho ut the growing season. This suggests that any winter cessation of leaf productio n may be controlled by temperature, rather than intrinsically. with growth of new foliage possib,y being responsive at any time of the year to warmer than usual temperatures. In compari son. phenology is more rigidly controlled by the seasons in alpine areas (Mark. 1970), which are more constrained climatically, with shorter growing seasons and more ex treme temperatures. The herbaceous form typical of mainland alpine regions is still adaptive in the less extreme southern oceanic island environments. No autumn fl ower bud initiation was observed here. Thi s is unlike some New Zealand alpine plants where fl oral initial s are set during the prev ious growing season and sometimes even at the beginning of the season (Mark . 1970). Evolution of the megaherb form Large-leaved forbs are not rare . The mainland of New Zealand has a large orbicular-leaved buttercup, Ranunculus lyallii (Hook. f. ), leaf diameter up to 30 cm, and an apiad, Aniso/ome lyalli (Hook. f. ) with leaf length up to 60 cm (Allan, 1961 ). Myoso1idium horlensia (DecneJ, a large forget-me-not. is endemic to the Chatham Islands (off New Zealand ). The Chilean Gunnera (lamina diameter 100 cm with a petiole often over 150 cm long ), Hawaiian silversword, Argyroxiphiwn (Goldstein and Meinzer, l 983 ), and Kenyan Dendrosenecio and Lobelia (Schulze et al. , 1985: Fetene et al., i998) are other examples. Yet the syndrome of the southern islands appears different with corrugations. stereom tissue. 14 NICHOLLS and RAPSON: BIOMASS IN MEGAHERBS hairy and occasionally coriaceous laminae, a rosette form, fleshy root system, and colourful flowers . Pleuroph_vllum is endemic to these southern ocean is lands, while one species of Stilbocarpa extends to southern South Is land , New Zealand. These genera contain only macrophyllous forbs . The other putative megaherb species have congeners throughout New Zealand; yet the megaphylls appear to be outside the no rmal size range of their genera . While it is possible that the megaherb phenomena might be a chance evolutionary occurrence, megaphylls may convey several possible se lective advantages in these subantarcti c environments. Nutrient availability cou ld be limiting as it vari es with moisture ho lding capacity and acidity of peat. the main growing medium on the is land (Meurk and Foggo, 1988). Large leaves may intercept nutrients from marine aeroso ls (Meurk et al. , 1994a) , channell ing resources directly to the stem base and onto roots (Enright, I 987: Agnew et a l .. 1993). Alternatively , Wardle ( 199 1) has suggested that a "greenhouse space" is set up between the large overlapping leaves of Pleuroph_vllum, with leaves acting as so lar pane ls and focussing radiation towards the growing apex. Such temperature increases are as much as 25 C above ambie nt air in the Hawaiian montane s ilverswords, Argyroxiphium (Melcher et al. , I 994 ). Detrimental effects of cold are further reduced by decreases in wind speed and accompanying reductions in transpiration losses that are associated with the rosette growth form (Regehr and Bazzaz, 1976). Light is also often the limiting resource in open si tes with tall herbaceous vegetation (Yoshie, 1995). In Anisotome the high allocation to petiole projects the leaves above the surrounding herb canopy, conferri ng a competitive advantage. By contrast, rosettes. such as Pleurophyllum , may suppress other competing herbs with their large ri gid leaves. The megaherb form might also be a response to the paucity of woody species. Yet the environment can sustain these as the tree daisy 0/earia lyalli occurs on the more northerl y of the islands and is invasive on the Auckland Islands (50° 45'S) (Lee et al.. I 99 1 ). The shrubby Dracophyllum, found on most subantarctic islands, including Campbell Island. tends to form thickets but mainly on the coastal fringes, while the Auckland Islands have a species of Metrosideros, also growing coastally . Elsewhere, megaherbs are aggressive, actively reoccupying their former habitat since the removal o f exotic grazers (Meurk, 1982). So the megaherb growth form must provide some adaptational advantages via pre-emption of resources and therefore be a derived syndrome. Certainly floral colour (other than white) is a derived element in subantarctic Abrotanella (Swenson and Bremer, 1997). Other aspects of the megaherb syndrome might also have evolved in situ . Conclusions This work is the first biomass allocation study of subantarcti c island megaherbs. These species invest a large proportion (50-80%) of their shoot biomass in leaf; consequently, their proportionate rhi zome and stem allocation is low. Even though strategies differ within the growth form, e.g., varying allocation within the leaf to lamina and petiole, these results support the suggestion that there is a 'megaherb ' phenomenon, a growth strategy different to other herbaceous perennials, e nabling these species to be a dominant and characteristic feature of subantarctic vegetation. Acknowledgements Th;mks to Mike Joy for ass istance on the island , to Colin Meurk and Geoff W all s fo r company, and wisdom, on a previous trip . to Re nu Kumar fo r ass isting with harvest, and to Ian Henderson and Alastair Roberston for help with stati stical analysis. Department o f Conservation (especially Carol West and Jeremy Carroll) provided permits and logistic support for the trip and Southern Heritage Expeditions and the Kapitan Khlebnikov supplied boat transport at bargain rates. Thanks to Bill Lee and an anonymous referee for helpful comments on an earli er draft. Thanks a lso for financial assistance to Professor S. Rumball (Science Dean), Massey University Research and Graduate Funds. and Department of Ecology Development Fund. References Agnew, A.D.Q.; Rapson, G .L. ; Sykes, M.T. ; Wilson, J.B . 1993 . The functional eco logy of Empodisma minus (Hook. f.) Johnson & Cutler in New Zealand ombrotrophic mires. New Phytologist I 24: 703-710. Abrahamson, W.G.; Caswell, H. 1982. On the comparative allocation of biomass. energy, and nutrients in plants. Ecology 63: 982-991 . Allan, H.H . 1961 . Flora of New Zealand, Volume I . Botany Division, Department of Scientific and Industrial Research , Wellington, New Zealand. 1085 pp. Al-Mufti. M.M.: Sydes. C.L.; Furness, S.B. ; Grime. J .P.; Band. S .R. 1977 . A quantitative analysis of 15 NEW ZEALAND JOURNAL OF ECOLOGY. VOL. 23. NO. I. 1999 shoot phenology and dominance in herbaceous vegetation. Journal of Ecology 65: 759-791. Bostock. S.J .; Be nto n. R.A . 1979. The reproductive strategies of five perennial Compositae. Journal of Ecology 67: 9 1- I 07. Cockayne. L. 1903. A botani cal excursion during mid winter to the So uthern Islands of New Zealand . Transacrions and Proceedings of rhe New "Zealand lnsritule 36: 225-333. Enright , .J . I 987. StemOow as a nutrient source for nikau palm. Rhopalost_vl is sapida, in a New Zealand forest. Ausrralian Journal of Ecology 12: 17-:?.4. Fetene. M.; Gashaw, M. : Nauke. P.: Beck. E. 1998. Microclimate and ecophysio logical significance of the tree- like life- form of Lobelia rh_vnchopewlum in a tropical a lpine environment. Oecologia 11 3: 332-340. Goldstein. G.: Meinze r, F. I 983 . Influence of insulating dead leaves and low temperatures on water balance in an Andean giant rose tte plant. Plant, Cell and En vironment 6: 649-656. Gross. K.L. 1983. Patte rns of resou rce allocation among five herbaceous perennial s. Bulletin of the Torrey Botanical Club 110: 345-352. Had ley, E. B.: Rosen. R.B . 1974. Carbohydrate and lipid contents of Celmisia pl ants in alpine snowbank and herbficld communities on Rock and Pillar Range. New Zealand. The American Midland Na turalist 91: 37 1-382. Hick man. J.C. 1975. Environmental unpred ictability and pl asti c energy allocation strategies in the annual Polygonum cascadense (Polygonaceae). Journal of Ecology 63 : 689-70 1. Hickman . J .C.: Pitelka, L.F. 1975. Dry weight indicates e nergy a llocation in ecological strategy analysis of plants. Oecologia 21: I 17- 121. Hooker. J .D. 1844. The borany of rhe Anrarcric vovage of H.M. Discovery Ships Erebus. and Terror in rhe years 1839- 1843, under rhe command of Captain Sir James Clark Ross; Volume I : Flora An1arc1ica; Pr. I: Bo/any of Lord Auckland 's Group and Campbell's Island. Reeve. London , U. K. 208 pp. Jolls, C.L. I 984. Contrasting resource allocation patterns in Sedum lanceolarum Torr .: biomass versus energy estimates . Oecologia 63: 57-62 . Lee. W .G .: Wil son. J.B .; Meurk . C.D.; Kennedy , P.C. 1991 . Invasion of the subantarctic Auckland Islands, New Zealand, by the asterad tree Olearia lyallii and its interaction with a resident myrtaceous tree Merrosideros umbellala . Journal of Bio geography 18: 493-508. Marshall, P. : Browne, B. 1909. The geology of Campbell Island and The Snares. Jn : Chilton, C. (Editor), The Subantarctic Islands of New Zealand, Volume 2, pp. 680-704. Philosophical Institute of Canterbury, New Zealand. Mark, A.F. 1970. Floral initiation and deve lopment in New Zealand alpine plants. New Zea land Journal of Botany 8: 67-75. Me lcher, P.J .; Goldstein , G .: Meinzer. F.C.: Minyard, B.; Giambe lluca. T.W.: Loope. L.L. 1994. Determinants of thermal balance in the Hawaiian giant rosette plant, Argyroxiphium sandwicense. Oecologia 98: 412-418. Meurk, C.D. 1982. Regeneration of subantarcti c plants on Campbell Island following exc lusion of sheep. New Zealand Journal of Ecology 5: 51-58 . Meurk, C.D .; Foggo, M.N. 1988. Vegetation res ponses to nutrients, climate and anima ls in New Zealand's 'Subantarcti c' Is lands. and general management implications. In : During, H.J . (Editor), Diversiry and pauern in plant communities, pp. 47-57. Dr. W. Junk , The Hague, The Netherlands. Meurk, C. D.: Foggo, M.N.; Thompson. B.M.: Bathurst, E.T.J.; Crompton. M. B. 1994a. Io n­ ri ch precipitatio n and vegetation pattern on subantarctic Campbell Is land . Arctic and Alpine Research 26: 28 1-289. Meurk, C.D.; Foggo, M.N: Wilson. J.B. 1994b. The vegetati on of subantarctic Campbell Is land. New Zealand Journal of Ecology I 8: 123-168. Meurk, C.D.; Given , D.R. 1990. Vegetarian map of Campbell Island I : 25 OOO. DSIR Land Resources, Christchurch, New Zea land . I sheet. Mooney, H.A .: Billings, W .D. 1960. The annual carbohydrate cycle of alpine plants as re lated to growth . American Journal of Botany 47: 594-598. New Zealand Meterological Service. 1983. Summaries of climatologica l observations to 1980. Misce llaneous Publicati on 177, Government Printer, We llington. New Zealand. 172 pp. Pugli ese, A. ; Kozlowski. J . 1990. Optimal patterns o f growth and reproducti on for perennial plants with persi sting or not persisting vegetative parts. Evolutionary Ecology 4 : 75-89. Regehr, D .L.; Bazzaz, F.A . 1976. Low temperature photosynthesis in successional winter annuals . Ecology 57: 1297- I 303 Schulze, E.-D.; Beck, E. ; Scheibe, R. : Ziegler, P. 1985 . Carbon dioxide assimilation and stomata! response of afroalpine giant rosette plants. Oecologia 65: 207-213. Swenson, U. ; Bremer, K. 1997. Patterns of floral evolution of four Asteraceae genera (Senecioneae, Blennospermatinae) and the 16 NICHOLLS and RAPSON : BIOMASS IN MEGAHERBS origin of white flowers in New Zealand. Systematic Biology 46 : 407-425. SYSTAT 1992. SYSTATfor Windows , Version 5. SYSTAT. Inc .. Evanston, Illinois, USA. Wardle, P. 199 1. Vegewtion of New Zealand. Cambridge Uni versity Press. Cambridge. UK. 672 pp. Yoshie, F. 1995. lntcrhabitat variati on in growth characteris ti cs of temperate herbaceous perennial s. Canadian Jou m a I of Botany 73 : 735-745. 17 Pollination in the subantarctic: Pleurophyllum Abstract The large, brightly coloured flowers of the megaherbs of the southern islands of New Zealand appear unusual in an environment supposedly unfavourable to pollinators. Studies of breeding systems of Pleurophyllum involving the out-crossing and selfing of flowers, and the exclusion of pollinators were carried out on Campbell Island, and observations made to determine the frequency of likely pollinators. While the insect life on Campbell Island is not diverse it was found that none of the Pleurophyllum species studied appeared to be pollen limited. Pleurophyllum shows an assortment of breeding systems. P. speciosum, the showiest, ray-flowered species, is an obligate out-crosser, showing no evidence of being self-compatible at all. The discoid P. criniferum is autonomously self-compatible, being able to set as much seed enclosed in a bag as with hand-selfing. The hybrid, P. criniferum x speciosum, shows partial self-compatibility and autonomous self-pollination, although sets more seed when out-crossed. It is suggested that a variety of breeding systems in Pleurophyllum provides a capacity for recombination of genetic material and potential for adaptive radiation. 18 Introduction From as early as 1903 (Hutton, 1903; Cockayne, 1903) there has been considerable interest in the apparent disparity between the flowers of the mainland of New Zealand and those of the southern islands. The flora of the mainland is known for having generally 'unspecialised' floral characteristics (Lloyd, 1985). The flowers are usually small, pale and easily accessible and attract a wide range of relatively unspecialised groups of insects: Diptera, Lepidoptera and short-tongued bees. There are few butterflies and no long-tongued bees. The subantarctic flowers on the other hand, are known for their relatively large size, and large, colourful floral displays. These displays are normally associated with animal pollination (Godley, 1979), yet these species occur in a climate considered unfavourable to pollinators (Lloyd, 1985). On the Cape Expedition ( 1909) Hudson observed: "Campbell Island, in November at least, is not a good place for insects". Subjected to high winds and cool temperatures, the subantarctic islands are an environment in which one might expect more self-compatibility, autogamy and wind­ pollination (Leppick, 1977). The genus Pleurophyllum (Asteraceae) is endemic to the southern islands of New Zealand (the Auckland Islands, Campbell Islands) and Australia (Macquarie Island). It consists of three species - P. criniferum, P. speciosum, and P. hookeri and one hybrid P. criniferum x speciosum. All three species and the hybrid grow on Campbell and Auckland Islands (although the hybrid is unconfirmed on the latter) but only P. hookeri occurs on Macquarie. They are examples of the so-called "megaherb" flora of these islands, characterised by their large growth form (up to 1 m high), high foliage resource allocation (Nicholls and Rapson, 1999), brightly coloured flowers and stout rhizomes. These herbs are a distinctive feature of the subantarctic herbfields (Mitchell et al., 1999). Despite the apparent interest in flower colour, few pollinator observations have been carried out on New Zealand's subantarctic islands (Godley, 1982) and this is the first study carried out on pollination of Pleurophyllum. This study examines the breeding systems in Pleurophyllum, looking at self-compatibility, pollinator requirements and pollinator visitation rates . Two species of Pleurophyllum were studied - P. criniferum, P. 19 speciosum, and the hybrid P. criniferum x speciosum from December 1996 to February 1997. Study Organisms All species of Pleurophyllum reproduce sexually, but are also capable of limited asexual reproduction by increasing crown number, i.e. clumping, particularly in P.hookeri (this species was excluded from the study as it had already begun to set seed at the beginning of the study period). They all form rosettes, their leaves arising from the crown. P. criniferum is the tallest member of the genus, its leafy flowering scapes reaching 1 m high (Figure lA). Its floral display consists of a tall scape of rayless , dark brown/maroon capitula suspended on long pedicels. The capitula have an indumentum when in bud. P. speciosum is the largest species - its corrugated, hairy leaves are capable of reaching 75 cm or more across . The flowers, on shorter scapes than P. criniferum, are the showiest of the genus (Figure lB), with capitula 3-4 cm in diameter in varying shades of purple, and with the occasional white morph occurring (pers. obs.) . The hybrid of these two species, P. criniferum x speciosum (Figure 1 C), contains elements of both parents, the leaves being larger than P. criniferum, but not corrugated, and the capitula rayed, but not as prominently so as in P. speciosum and of a more intense maroon. The capitula also have indumentum similar to that of P. criniferum that is lacking in P. speciosum. In winter P. speciosum and P. hookeri reduce in size, retaining a small leaf rosette, while P. criniferum is completely deciduous. Numerous capitula of both P. speciosum and P. criniferum contained Lepidoptera larvae burrowing at the base of the florets (Figure 2). This occurred in a couple of bagged capitulae which were tied on before the florets opened (and which were excluded from the analysis), so are almost certainly not pollinators, but may be seed predators (pers. comm. A. Robertson). 20 A) B) C) Figure 1: Flowering innorescences of A) P. criniferum showing the coloured ray norcts ; B ) P. speciosum showing the discoid, pend ulous capi tula; and C) P. criniferwn x speciosum I\) showing charac teris tics of both . ........ Study Area The study took place on Campbell Island, one of the subantarctic islands 700km south of New Zealand in the Southern Ocean. It is 11 OOO ha in size with a vegetation consisting largely of tussock, herbfields and shrubland with a littoral forest of tall Dracophyllum (Meurk and Given, 1990). The atmosphere is damp with a moderate rainfall (mean 1361 mm y( 1 ) distributed throughout the year, and the annual sunshine hours are low (659y(1). The temperatures reflect its oceanic climate - the mean monthly maximum temperature for the warmest month is 15.9°C and minimum for the coldest month is -2.7°C (NZ Meteorological Service, 1983). The study populations were in two different sites. The P. criniferum population was coastal and associated with Carex appressa ('Paa litorosa maritime tall tussock grassland' as mapped by Meurk and Given [ 1990]). P. speciosum and P. speciosum x criniferum were studied at a higher altitude ( 120m above sea level) in a more open subalpine environment consisting of tussocks (Chionochloa and Poa) and occasional shrubs of Dracophyllum, Coprosma and Myrsine and the ferns Polystichum and Blechnum ("Poa litorosa meadow and Chionochloa" community as mapped by Meurk and Given [1990]) . 22 Methods Breeding Systems The potential for self-compatibility and autogamy in each taxon was tested for experimentally. As the plants came into flower, 15 individuals of each species were chosen and four scapes on each plant were assigned to each of four treatments: hand-selfing ('bagged selfed'), hand-outcrossing ('bagged crossed'), pollinators excluded ('bagged') , and open pollination ('open'). For the hand-selfing and hand-outcrossing treatments, unopened capitula were enclosed in small mesh bags (Figure 3) and hand pollinated with a small paint brush with either the plant's own pollen or a mixture of pollen from other plants of the same species in the vicinity. To attempt to pollinate all florets open at the time, hand pollination was carried out at least twice on the same capitulum over the season as the capitula opened basipetally. Bags were made up of an open mesh of approximately 0.3 mm to allow the passage of wind but not insects. To determine the potential for autogamy, capitula were enclosed inside bags and left unpollinated. The open pollination capitula were left unbagged and untouched. Capitula were collected in early February after 6 weeks from the start of the study and stored in 95% alcohol until analysis . While most flowers had been pollinated by this time, the achenes were not necessarily well developed. In the lab, the pickled capitulae were cut in half and both faces inspected using a dissecting microscope. Ovaries were recorded either as set, empty, aborted or immature, based on apparent development of the seed. Proportion of seed set was estimated as the number of filled seed/(unfilled seed + aborted seed + filled seed). There was some difficulty in assessing the difference between immature and unfertilised ovules in the centre of each capitula so these inner florets were omitted from analysis. In addition, some heads were considered too young to determine whether ovaries were empty or just too immature. These were also not included in the analysis. 23 Pollinator observations Only P. speciosum was observed intensively as there was very little pollinator activity on P. criniferum or P. criniferum x speciosum. Groups of up to 6 capitula were observed for 25 periods of 10 minutes at a time and the number of visiting insects counted. Observations were made at varying times of the day between 9am and 4.30pm in varying overhead conditions (sunny or overcast) and wind conditions (light or moderate), and these conditions recorded. Visitation rates were corrected to visits per head per hour. A small sample of pollinators was collected, and later identified by Dr Ian Andrew, Massey University. Data analysis Seed set was analysed by means of a generalised linear model with a binomial error distribution and logit link function using S-PLUS version 4.5 (Mathsoft Inc. 1998). Data were treated as a randomised block design with the plants as blocks. A posteriori multiple comparisons of means using the Sidak ( 1967) method were used to compare the different treatments for each species. Visitation rate analysis was carried out on the log-transformed pollinator visits (log 10 visits+ 1) using ANOV A (SYSTAT, 1992). 24 Figure 2: Capitula of P. criniferum showi ng burrowing by Lepidoptera larvae. Figure 3: Pollination bags for hand-selfed , hand-outcrossed , and bagged treatments , attached to individual capitula of different scapes of P. criniferum x speciosum. 25 Results Breeding Systems The three taxa differed m their breeding systems and in pollination requirements. P. criniferum showed no significant difference between any of the treatments (Table 1; Figure 4A) suggesting full autonomous self-compatibility. In contrast, P. speciosum, shows a significant difference between the hand-selfed and the hand-crossed treatments (Figure 4B), indicating that it is strongly self-incompatible. This self-incompatibility is reinforced by its inability to self pollinate in the bagged treatment. The open treatment, where pollinators have free access to the heads, indicates that there is no pollen or pollinator limitation in this species. The hybrid, P. criniferum x speciosum, like P. speciosum, shows a significant difference between outcrossing and selfing (Figure 4C) but shows partial self-compatibility and is capable of autonomous self-pollination in bagged flowers. In none of the species does there appear to be any significant difference between the open-pollination treatment and hand-pollination suggesting there is no pollen limitation of seed production. Of interest is the apparent higher (although more variable) seed set in the hybrid compared with the others. P. criniferum x speciosum has 43% achenes filled overall, followed by P. criniferum with 31 % and P. speciosum has the lowest rate at 20%. Pollinator observations The visitation rate by all insects to P. speciosum was 6.2 visits per head per hour. Four types of insect were observed: a moth (unidentified), a hoverfly (Melangyna novaezealandiae, family Syrphidae), and 3 small midge-like flies (Australimyza anisotomae, family Australimyzidae or Camidae; Tetragoneura mzmma, family Mycetophilidae; and Chironomidae species). The latter tended to sit on the corolla for long lengths of time, and make occasional movements onto the florets. The efficacy of these pollinators was difficult to determine although the resident pollinators clearly were efficient pollinators as the pollination experiment gave no evidence of pollen limitation in these species. Blowflies were very numerous in the vicinity, but on only one occasion was one observed on a flower. 26 A a a a a "O a.> 1.0 - u::: VJ 0.8 -a.> C • a.> .c 0.6 - -r u -r 1.0 -= u::: VJ 0.8 -a.> C Q) T .c 0.6 -u