1 The nature of the plant community: a reductionist view 2 J. Bastow Wilson 3 Botany Department, University of Otago, P.O. Box 56, Dunedin, New Zealand. 4 Andrew D.Q. Agnew 5 Institute of Biological Sciences, University of Wales Aberystwyth, SY23 3DA, U.K. 6 Chapter 1: Plants are strange and wondrous things 7 1 From plants to communities ................................................................................................... 2 8 1.1 Features of all land plants that predetermine their natural history .................................. 5 9 1.2 What is a plant community? .......................................................................................... 11 10 2 The accession of species into mixtures ................................................................................. 13 11 2.1 Step A, Speciation: What is a species? .......................................................................... 13 12 2.2 Step B, Biogeography: The species pool ....................................................................... 15 13 2.3 Step C, Dispersal ........................................................................................................... 16 14 2.4 Step D, Environmental filtering / ecesis ........................................................................ 18 15 2.5 Step E, Interference filtering (mainly competition) ....................................................... 21 16 2.6 Step F: Assembly rules .................................................................................................. 24 17 3 Geographical boundaries ...................................................................................................... 25 18 4 Concepts of the space occupied by one species .................................................................... 26 19 4.1 The niche ....................................................................................................................... 26 20 4.2 Guilds ............................................................................................................................. 30 21 4.3 Stratification .................................................................................................................. 32 22 5 Conclusion ............................................................................................................................ 34 23 24 25 A problem I had repeatedly as I read the chapter was that I could not figure out who the authors 26 expected to be in the target audience. Many cases exist where jargon enters into the text without 27 prior definition and importance ideas from the field are assumed without explanation. This 28 suggests that the book must be intended for no one earlier in their career than roughly a third-year 29 graduate student (which will limit sales). At other times simple ideas are discussed in a way that 30 suggests advanced undergraduates. Either level, or one in between, would work, but parts of the Wilson & Agnew, chapter 1, Plants, page 2 of 34 31 book need to be adjusted to so that the target isat least consistent. The Preface states that “the 32 book is deliberately unbalanced,” but this is not a facet I think should adhere to this underlying 33 philosophy. 34 35 In general, I like the approach. It seems the authors are attempting to demonstrate the high 36 amount of uncertainty in the field of plant community ecology. However, I am hopeful that after 37 the deconstruction, a framework remains from which we can build upon. In general I would 38 recommend providing some definition for all major terms. For example, the authors did not 39 provide definitions for individual or species pool although there was a great deal of discussion 40 regarding different interpretations of these terms. 41 42 General comment: In general, I am enjoying this new take on community ecology. Since many of 43 the ideas are slightly different from traditional textbooks and do not quite fit in with intro ecology 44 concepts (e.g. throwing out the concept of “individual”, saying that all previous attempts have 45 “failed”, etc.), I see this book as more of a companion book to these more traditional texts. 46 47 48 1 From plants to communities 49 Our aim in this book is to explore the workings of plant communities and especially the 50 forces that limit the coexistence of some species and promote the coexistence of others. We are 51 searching for generalisations that can be applied to plant assemblages, working from the bottom 52 up. We shall rarely discuss animals: this book is about plants. First we explain our view of 53 vegetation and of the plants that comprise it. 54 The landforms of the earth result from an underlying geological diversity, moulded by 55 geomorphological forces and mostly clothed with vegetation. Even in arid climates, any 56 scattering of plants intrudes and holds the human eye. Like the architectural heritage of the built 57 environment, landscape has the power to be emotionally and spiritually uplifting, or depressing. 58 Our reaction depends on our cultural history, our background experience and often current 59 fashion. We, the authors, have been able to study vegetation during our full working lives, and it 60 has been enormously rewarding and emotionally satisfying. Such studies are in some way a 61 homage to nature and to God.[personal taste perhaps, but I think God has no place in a book 62 about science; nature is, in contrast, the process we are trying to study.] However, we also enjoy Wilson & Agnew, chapter 1, Plants, page 3 of 34 63 the application of science to the natural world, behoving us to seek the processes behind the 64 vegetation that we see, to search for general patterns, and to attempt the formulation of 65 community-level theories. 66 From the beginnings of plant ecology, some scientists have concentrated on describing 67 the myriad of combinations in which species occur (e.g. Lawesson 2004). Others have used a 68 reductionist approach, examining a process by which species A affects species B, but have sought 69 no deeper generalisations. Yet others have developed theories into which they hope the world 70 will fit (see Bio 2000 –not accessible; use different reference). Such is the complexity of plant 71 communities that, whether the theories have been primarily deductive (e.g. MacArthur 1969) or 72 empirical (e.g. Grime 1979), all have basically failed This is a big statement here with no backing 73 provided? Are the authors suggesting that we have learned nothing and can make no generalities 74 about plant ecology?. This book is an attempt to move reality and theories closer. The authors 75 state that deductive and empirical attempts to describe plant communities have basically failed. 76 However, there is no further explanation or qualification of that statement. Perhaps they failed at 77 proving that communities exist without exception? Maybe they failed at developing a “law” of 78 community ecology? 79 There are plenty of theories to test, some more trivial than others, but it seems none have 80 reliable truth. Suppose we take a group of students into the field, tell them that there is a ‘theory’ 81 that species richness is higher in ecotones (boundaries) and have them sample. Will they find 82 that? Probably not. Suppose we tell them of the opposite ‘theory’ that can be found in the 83 literature – species richness is lower in ecotones – will they take community ecology seriously as 84 a science? Suppose we draw out of the hat a theory on where species evenness will be high, or 85 where the relative abundance distribution will be a particular shape; will the students find it? 86 Probably not.[This book needs to address why not] The only reason that students put up with this 87 ‘science’ is that they, like us, find being in the field more pleasant than being in the lab. 88 Nevertheless, it is our duty as scientists to start solving these problems. Is a theory disproved if 89 it can’t be demonstrated within a class period? 90 We shall emphasise terrestrial vascular plants, because more is known about them, and 91 most of the processes to be found are found in them. However, it is likely that many of the same 92 principles apply to lower plants, down to macro-algae and plankton (Tilman 1981; Wilson et al. 93 1995 %689; Steel et al. 2004), and we shall take examples from any group of plants when we 94 fancy. Very rarely do we see a plant species persisting on its own even when we try to make it do Wilson & Agnew, chapter 1, Plants, page 4 of 34 95 so in a garden or farm, so this book is about plant communities. However, in keeping with our 96 reductionist approach we start by examining the importance and nature of plants. 97 The importance of plants 98 Plants, as the dominant carbon fixers in the biosphere, control all ecosystems. The 99 terrestrial part of the biosphere is overwhelmingly vascular plant cover. Plant communities have a 100 global entropic effect. Visible light from the sun is intercepted by our planet, and is dissipated at 101 longer wavelengths into space (D.H. Miller 1981). This represents a gain in entropy, i.e. a trend 102 towards homogenization of the universe. Yet life captures energy and by its maintenance and 103 replication creates order/decreases entropy. The accumulation of biomass and organic matter 104 suggest that at the level of planet entropy is decreased. The plant covering of the Earth increases, 105 be it almost immeasurably, this entropy gain. It does this by fixing a tiny part of solar energy into 106 organic matter and, through evolutionary processes, maximising the efficiency of its utilisation 107 (Ulanowicz and Hannon 1987) so that even more energy is re-radiated as long wave radiation. 108 [Chemical energy storage is small compared to absorption and reradiation, which accomplishes 109 the same thing, and that storage might well be construed to be the opposite of increased entropy, 110 bought at the price of some of the much larger entropy increase associated with absorption and 111 reradiation.] This is quite temporary for an individual living plant, but forests hold a long-term 112 store of energy as reduced carbon and terrestrial plant products can remain for longer in soil, peat 113 and eventually in subfossil and fossil deposits. The result is the maintenance of the oxygenated 114 atmospheric state, of no small importance to us all [and itself a stronger example of neg-entropy 115 storage ]. In fact, the vegetation cover has multifarious feedbacks on the climate (Hayden 1998). 116 Plant communities affect the rocks and soil too, exercising major geomorphological 117 controls on the earth’s land surface and landforms. They intercept precipitation and wind, damp 118 down environmental fluctuations, reduce erosional rates, affect soil formation and dominate 119 geochemical cycles (Trudgill 1977). Local and regional hydrology are profoundly affected by 120 vegetation through evapotranspiration, which reduces the amount of water available in soil and 121 the catchment outflow. Plant cover may accrue wind- and water-borne deposits and thus build 122 landscapes. Every plant affects the local environment in ways that are again multifarious (Eviner 123 and Chapin 2003). This is the ‘reaction’ of Clements (1904; 1916) and Gleason (1927). 124 [Reference will not be clear to the reader ] 125 126 Plants are also almost the sole basis for the food chain. Reichle et al. (1975) itemise the four essential parts of ecosystem function as: (1) energy input in photosynthesis (‘energy base’), Wilson & Agnew, chapter 1, Plants, page 5 of 34 127 (2) the capital of energy [not ecosystem ‘function’, though perhaps an ecosystem ‘service’, 128 whatever that is]in photosynthetic biomass (‘reservoir of energy’) – I would rework #2 as energy 129 flow through the trophic structure, (3) cycling, especially of elements, and (4) the control of the 130 rates of these and other processes by factors such as temperature and the availability of 131 heterotrophs (‘rate regulation’) – #4 also seems not a function, but a background condition. On 132 land, green vascular plants comprise almost the whole of the energy base and the resevoir of 133 energy, and they make major contributions to cycling and rate regulation. 134 1.1 Features of all land plants that predetermine their natural history 135 Terrestrial green plants are so familiar to us that we often lose our sense of wonder at 136 them, even as their features become more extraordinary as our knowledge of biology deepens. 137 We argue that: 138 1. Land plants root in the soil to obtain mineral nutrients, water and anchorage. Therefore, 139 they are sedentary, so defence from herbivores can be only by structure and chemistry, 140 not by escape. – Two separate things going on here, and it confuses things to fully merge 141 them. One is obtaining nutrients and water from the environment and the other is 142 anchorage. They need not be linked. Plants anchor in other places, like as epiphytes. 143 This provides, among other things, a stable place from which to array aboveground parts 144 for light foraging. Only when we invoke a direct linkage like mycorrhize does fixed 145 location become important for resource extraction 146 2. This puts a selective premium on cell walls that are low in food value to herbivores, 147 basically cellulose, which is also strong enough to support cell turgor. However, cellulose 148 cannot be efficiently recycled. Therefore plants almost always have to grow by the 149 replacement of modules, such as leaves. Discarded modules are a necessary byproduct, 150 comprising litter. -- Would be productive to contrast the modular growth of plants with 151 the generally non-modular growth of most higher animals? 152 3. Because of modular growth, the number of cell divisions between generations (i.e. 153 gamete-to-gamete) is indeterminate and large [perhaps note nondeterminate animals like 154 fish and turtles in contrast with determinate animals like mammals, birds, and most 155 dinosaurs]. In the process of module production somatic mutations can occur, so all 156 ‘individuals’ are potentially genetic mosaics. Genetic mosaic is not a readily-available 157 term in the average student’s mind. It would be good to briefly define this term here, or 158 use another, more obvious, group of words . The germ cells are defined only just before Wilson & Agnew, chapter 1, Plants, page 6 of 34 159 the meiotic process, so they include these somatic mutations. This contrasts with animals, 160 where the germ cells are defined at an early stage and migrate to the gonads (Gilbert 161 1997) with few cell divisions between one generation and the next and hence little 162 opportunity for somatic mutations to accumulate and be passed on. [I hope you will return 163 to why these three things are important as we move through the book ] 164 Another result of modular growth is movement. Motile animals move around but, having grown, 165 usually stay within approximately the same adult body, replacing organs cell-by-cell or molecule- 166 by-molecule until death. Plants are sedentary [or certainly do not ‘actively’ move] , but their 167 organs and elements of their living transport system have a limited length of useful life and must 168 be replaced by new ones (Larcher 1980).[this varies – some plants – especially annuals – are 169 determinant in growth – nonetheless, they do grow ] The photosynthetic rate of a leaf is maximal 170 early in its life and declines thereafter, so leaves and their supporting organs are generally 171 replaced several times during the lifespan of a plant.[in perennial woody plants at least] These 172 replacement leaves are formed distally on the stem, or on side branches. This means that plants 173 can never persist in an unchanged physical space; they must grow and in the process explore and 174 expand into new space. [again, a subset of plants] Even cacti must increase in size during their 175 life (de Kroon and van Groenendael 1997). This remorseless renewal of all modules of growth, 176 the discard of old dead plants as litter and exploration of new space results in disturbance to 177 neighbours. In other words: plants move, animals don’t.[sounds cute and is thought provoking, 178 but does not hold up to careful scrutiny] [Discussion of symmetric vs asymmetric competition 179 seems to be relevant here.] I think this should be changed to something like “plants move in a 180 different way than animals do”. The authors have a pretty good argument that plants do move, 181 but the converse that animals do not move is not really true, or at least the text provides no 182 background for this statement. 183 This seems to be a point Wilson has been dying to make, and is therefore blinded by his 184 own logic. There are many ways to dissect the functions, assemblages, and relationships of living 185 organisms, and I feel that making this statement leaves the reader not only confused, but also 186 short of the whole truth. Anyone can manipulate the definition of movement to make a point 187 188 Some colonial, sedentaryi animals are similar to plants in that they must grow to stay 189 alive: some Urochordata (tunicates), corals and Porifera (sponges). As a result they have several 190 similarities to plants. They have similar genetic characteristics. They filter water for carbon just Wilson & Agnew, chapter 1, Plants, page 7 of 34 191 as plants can be said to be filtering water and air. The sedentary tunicates and corals have 192 exoskeletons somewhat resistant to decay and predation (tunicin and calcium carbonate 193 respectively), comparable to the epidermis of plants. However, there are differences. The 194 modules causing the mandatory growth of corals are not discarded in the way leaves are, though 195 the xylem in the heartwood of trees is retained too.[but the filtering parts are replaced! And don’t 196 forget the symbiotic algae in the corals are at least as important as the filtering of stuff from the 197 water. ] Although animals that accumulate calcium carbonate have profound effects on marine 198 geomorphology and the biosphere, no animals on land have byproducts similar to the litter of 199 dead waste parts produced by living plants (chapt. 2, sect. 2 below). In the arthropods there is a 200 periodically-shed exoskeleton that includes cellulose-like material, but there is sufficient protein 201 in the exoskeleton to make it readily decomposable and anyway the biomass of herbivores can 202 never approach that of the primary producers and therefore cannot modify the environment of 203 entire systems as can plants. Is it worth thinking about aquatic and marine systems where the 204 biomass of photosynthetic organisms can be lower than that of the next trophic level and perhaps 205 that higher trphic level and modify the environment more because of the amount nutrient that gets 206 stored/removed by its actions? Lots of stuff ends up on the sea floor, and it is not decomposed 207 and recycled, but for very different reasons than on land.. I feel that this minor section which 208 speaks of the biomass left behind by plants is also a little misleading. It may be true that 209 historical plant biomass surpasses animal biomass. However, there is no recognition of the mass 210 limestone across the earth, which was contributed by animals in the past. How these amounts 211 compare isn’t the point for me—it’s an issue of Wilson giving inaccurate examples to make his 212 point 213 The problem of the individual 214 The problem of recognising individuals in plant populations is longstanding. It is reflected 215 in discussions of the terms biotype, genet and ramet [here we again have a jargon issue that needs 216 to be rsolved in terms of the expected background of the target audience. I don’t think we can 217 expect all the readers to know these terms (Harper 1977) as well as more philosophical 218 discussions of the nature of the plant individual (Firn 2004).[A rather oblique reference to a 219 rather little know but interesting paper – the appropriateness will vary with the target audience. 220 You seem to be aiming at professionals and advanced graduate students, whereas I think you 221 should aim at new graduate students as they are the ones who can be shaped and who have the 222 time to read books.] In an annual with no vegetative reproduction it is clear what an individual is. Wilson & Agnew, chapter 1, Plants, page 8 of 34 223 In vegetatively-reproducing plants with ramets gradually becoming independent (Marshall 1996), 224 perhaps with the clone then splitting into several discrete patches (Harberd 1962), ‘individual’ 225 has no demographic meaning. The same issue arises with the apomictic offspring in genera such 226 as Crepis (hawksbeard), Poa and Taraxacum (dandelion) that are potentially identical in 227 genotype.[Do identical twins in humans function as individuals? Usually! I don’t think we want 228 to link the definition of individual too tightly to genetic identity.] Another problem with applying 229 the animal ‘individual’ concept to plants is that whilst most animals [only because most animals 230 are insects and we conveniently ignore their larval stages because they are harder to find and less 231 pretty.] are relatively constant in size at any particular age, individuals of one plant genotype can 232 differ in biomass by several orders of magnitude (Harper 1977). There is some evidence that the 233 root system of an individual genet or even ramet can differentiate between roots from its own 234 parent and other individuals of its own or other species. The experimental evidence of Gersani et 235 al. (2001) using Glycine max (soybean) plants, of Gruntman and Novoplansky (2004) in Buchloe 236 dactyloides (buffalo grass) and the neurotransmission speculations of Baluska et al. (2004) are 237 fascinating in this respect and need confirmation. [Even in an advanced textbook these probably 238 need to be defined rather than assuming the papers have been read by the reader. ] 239 The section on genetic change seems only tangentially related here. I would have preferred a 240 summary paragraph on the definitions of the individual and the consequences of using these 241 definitions for interpretation in plant community ecology. 242 It is unclear what the relevance of this discussion about the individual is. The concept of 243 individual seems to have been thrown out because it may not always mean the same thing. The 244 fact that you would define an individual differently for a clonal plant than you would for an 245 animal doesn’t mean that the concept isn’t valid or that different uses of the concept might not be 246 important to address particular questions, it just means that you need to be clear as to what you 247 mean by it in a given use. 248 249 Somatic mutations complicate the issue further. There can be mutations as Taraxacum 250 plants reproduce (King and Schaal 1990) so the apomictic offspring need not be genetically 251 identical. Somatic mutations can occur in vegetatively-reproducing plants and during growth 252 (Gill et al. 1995). Using the plant cell sizes in the classic Strasburger's textbook of botany 253 (Harder et al. 1965) with a conservative estimate of mean cambial cell length of 0.1 mm it is clear 254 that there could be of the order of 220 cell divisions between separate sectors of growth in a tree, Wilson & Agnew, chapter 1, Plants, page 9 of 34 255 with a consequent probability of mitotic errors. Therefore, even an apparently ‘individual’ plant 256 cannot reliably be taken as a single genotype, and has to be regarded as a colony of apical 257 meristems, even a colony of apical meristem segments (Fig. 1.1). Could you not say the same 258 thing of every organism? Are we all not “colonies” of cells with slightly different genetic codes? 259 Every apex and therefore each flower can be genetically unique, or perhaps every sector within 260 an apex (Newbury et al. 2000). [In this day of molecular studies it is not sufficient to simply 261 assert this diversity, but rather papers documenting the degree of variation should be cited/]The 262 modules of a physiological individual such as a tree also differ in their environment (e.g. light 263 intensity) and often the cause of that variation (shade, in our example) can be the individual itself 264 (self-shading). [For this to be ecologically interesting we need some evidence of selection within 265 the plant, rather than simple genetic diversity – that is, show us that it matters for something.] 266 However, physiological interdependence between the modules overcomes this to some extent. Litter Fig. 1.1: A stylised dicotyledonous plant as a colony of active and inactive apices 267 268 We conclude that because of vegetative reproduction and apomixis, variation in size, somatic 269 mutation and plasticity, the animal concept of ‘individual’ is not appropriate or useful in plants. 270 [more helpful to clearly articulate the differences and then indicate what we mean by an 271 individual. ] This doesn’t mean that there isn’t a plant version that would be. The statement that 272 the concept of ‘individual’ is not appropriate for plants is taking things a step too far. Certainly 273 this concept is useful to ecologists, but the definition should depend on the question being asked. 274 If a genetic analysis is being performed, then an individual may be any genetically-unique shoot. 275 For a survey of regeneration following disturbance, perhaps each shoot should be counted as an 276 individual. Wilson & Agnew, chapter 1, Plants, page 10 of 34 277 278 The features we have been discussing make genetic change difficult.[explain why they 279 make genetic change difficult – seems like you are referring to all that within-plant genetic 280 diversity, but perhaps referring in a confusing way back earlier to plastic and indeterminant 281 growth – unclear at any rate] We must ask why plants need genetic change when they can change 282 plastically. One answer to this paradox has been the controversial theory of ‘genetic assimilation’ 283 (Pigliucci and Murren 2003): that plastic changes can become incorporated into the genotype. 284 Bradshaw’s (1973) answer was that plants are genetically ‘sown into their winter underwear’, 285 because their plastic response to an adverse environmental shock would be too slow. [a bit too 286 obscure to follow without more explanation; wit is nice but clarity of expression is perhaps better 287 ]A third answer is that they do not actually become adapted genetically: Rapson and Wilson 288 (1988; 1992) found that though significant genetic differences had developed in Agrostis 289 capillaris (bent) in southern New Zealand since it was introduced in 1853, there was no sign that 290 populations were differentially adapted to the habitat they were growing in [can this be tested in a 291 two-suided way, or can we only find or fail to find specific evidence of adaptation – easy to not 292 test the right thing when there is somuch to test.]. Perhaps genetic conservatism is a result of 293 duplication of alleles on chromosomes and of duplication of genomes (polyploidy). Of course, 294 populations and eventually species do change sometimes, giving in some cases dramatic ecotypic 295 adaptation and eventually leading to the 400,000 flowering plant species that we see today and 296 the millions that rest in peace. 297 Interaction with other trophic levels 298 Plants are mostly autotrophic, but they interact with all other trophic levels. Since our 299 thrust is plant communities, we shall generally discuss this only so far as it mediates plant-plant 300 interactions. Wardle (2002) has discussed interactions with decomposers. [seems insufficient to 301 simply allude to work without any mention of what was found] Plants meet and usually withstand 302 challenges from herbivores and diseases, usually in two totally differing environments: the 303 relatively humid soil below ground and the comparative aridity of sunlight above ground (chapts. 304 2 and 4). Many plants rely on animals for pollination and dispersal (chapts. 2 and 4).Top 305 carnivores will have indirect effects. Mycorrhizae are crucial for many species, and will be 306 discussed especially in chapter 2. In addition to their rôle in nutrition and water acquisition, it 307 seems that vesicular-arbuscular mycorrhizae (VAM) can restrict the development of pathogen 308 loads in their host (Larsen and Bødker 2001). Endophytic fungi and bacteria are also widespread Wilson & Agnew, chapter 1, Plants, page 11 of 34 309 and have a multiplicity of effects on plant growth. There is usually an extensive microflora in the 310 phyllosphere and in the rhizosphere. Some plants form special relationships with ants, to which 311 we shall refer. In some plants mites inhabit small pits in leaves (domatia), and apparently protect 312 the plant against other herbivores or against pathogenic fungi (Grostal and O’Dowd 1994). This 313 brief list of interactions is surely far from exhaustive. [seems to tease the reader –we perhaps 314 need more at the level of impact on the community] 315 1.2 What is a plant community? 316 To test community theories we need communities. Unfortunately it is not possible to 317 provide a definition of ‘community’ that includes areal extent, uniformity of environment, 318 closeness to equilibrium, etc. All sorts of species mixtures exist, in all sorts of environments, and 319 there are no discontinuities in the hierarchy of this variation. Furthermore, species mixtures are 320 constantly changing. We believe the plethora of terms that have been applied to species mixtures 321 (phytocoenose, association, nodum, etc.) represents attempts to persuade vegetation ecologists 322 that the study of this aspect of the natural world can yield general statements and predictive rules, 323 but it cannot. [They do provide contingent generalizations in the sense of May. The smaller the 324 distances in time and space, the greater our ability to predict. It is at the level global level that 325 predication . I agree that there are several seemingly ambiguous or unnecessary terms for 326 communities in ecology, and that it can sometimes be difficult to discern the boundaries between 327 community types on a landscape. However, I do not think these terms necessarily exist to 328 persuade us that community ecology can “yield general statements and predictive rules”. I think 329 that most of these concepts exist so that we can talk about plant community ecology concepts 330 more easily. 331 332 How close can we get to defining ‘plant community’? A degree of repeatability between 333 samples (i.e. quadrats) would be a useful restriction, but this again is difficult to prescribe (chapt. 334 6, sect. 2 below). We need to specify scale at some point in our argument; Gleason (1936) 335 suggested it should be one plant of a largest species but we do not feel able to insist on this.. 336 Mueller-Dombois and Ellenberg (1974) give a historical summary and agree that no rigid 337 definition is possible. However, they distinguish between conceptual communities which are the 338 abstract units of plant community classification and ‘concrete’ communities that are the actual 339 plant species mixtures encountered in the field. We hope that all our discussions can be related to 340 real, actual examples of plant communities, the concrete ones, for we are not persuaded of the Wilson & Agnew, chapter 1, Plants, page 12 of 34 341 relevance of conceptual communities.[But, are you not seeking generality????] We could use the 342 splendidly neutral and practical statement of Tansley and Chip (1926) that “A plant community 343 may be defined as any naturally growing collection of plants which, for the purposes of the study 344 of vegetation, can be usefully treated as an entity.” 345 To include environmental relations, stability and change in the community, and spatial contiguity 346 we here see the plant community as: Naturally generated plant stands where the environment of 347 the individuals of one species potentially, predictably and persistently includes individuals of its 348 own and usually a restricted number of other species.[ wow – he used the bad word ] 349 The definition of plant community is confusing. The wording “potentially, predictably 350 and persistently” is the most confusing part of the definition, as it is difficult to understand how 351 and when something can “potentially” occur, as well as “predictably and persistently” occur. 352 353 354 355 [It seems rather odd to make a definition of community using the term “individual” after just spending an entire section throwing that term out as inappropriate for plants!] I like the Tansley and Chip definition much better than the Wilson and Agnew one. The Wilson and Agnew one seems to lead the reader even more confused with vagueness. 356 357 This excludes mixtures deliberately planted, such as a mixed shrubbery, but planted 358 gardens and agricultural fields can contain a rich weed flora and are valid objects of study. Of 359 course, indirect human intervention such as fertilisation and the release of grazers is quite 360 acceptable: they often mimic perturbations in natural communities, and in any case it is 361 fascinating to see how a mixture of species responds (e.g. Fuhlendorf and Smeins 1997; 362 Silvertown 2006). I didn’t have the impression that the authors agreed with their definition of 363 plant community. It is unclear to me still what the authors take to be a “community”. They say 364 earlier that they “are not persuaded of the relevance of conceptual communities”. but this seems 365 like a very conceptual def. to me. 366 We are trying to make sense of nature, starting with a vision about plants and plant 367 communities, and looking for underlying predictability and repeatability so we can claim 368 community ecology as a science. As the great Robert MacArthur (1972) said: “To do science is to 369 search for repeated patterns”. [But in that same book MacArthur makes the point that the intrinsic 370 complexity of ecosystems forces ecologists to search for generalizations that are contingent on 371 numerous and often quite specific initial conditions.] The major difficulty for us is that we do not 372 know what sort of pattern to look for (chapt. 5 below). One issue in dealing with samples of plant Wilson & Agnew, chapter 1, Plants, page 13 of 34 373 mixtures is the concept of phantom species. These are species present in the general area (“in the 374 community”), potentially available in samples but not actually recorded.[the species pool?] This 375 may be a valid concern for animal communities where species at low density can be around and 376 sometimes walk/swim/fly though the sample area/volume, but happened not to be there at the 377 recording time. This is less relevant for plant communities and we follow Pielou’s (1990) 378 suggestion: “a biological collection … should be treated as a universe in its own right”, rejecting 379 the concept of phantom species as a figment of the theoretical ecologist’s imagination. 380 2 The accession of species into mixtures 381 382 We discuss the processes that initiate plant communities in six steps, in some developmental order: 383 A. Speciation: Life has originated, and the species must have evolved. 384 B. Biogeography: The species must be in the regional species pool. 385 C. Dispersal: The species in the regional species pool must reach the particular site. 386 D. Environmental filtering / ecesis: The species must be able to germinate/develop from its 387 propagule and then grow to reproduction under the physical environmental conditions 388 prevailing. 389 E. Productivity and biotic filtering: The species must be able to ecise and reproduce under 390 the general interference pressure from the other species present: competition etc. (chapt. 2 391 below). 392 393 394 F. Assembly rules: The species must withstand restrictions from the particular species or types of species present (chapt. 5 below). 2.1 Step A, Speciation: What is a species? Wilson & Agnew, chapter 1, Plants, page 14 of 34 395 We shall deal only peripherally with sub-specific evolution and not at all with the 396 evolution of species, but they are the first required taxonomic category above the plant. The 397 recognition, description and diagnosis of species allow us to predict much of a plant’s It would be Species pool (metacommunity) Dispersal Challenge Niche constructed Niche available Niche unavailable No entry! Population Establishment Fig. 1.2: Pathways from the species pool to community entry. 398 This figure is completely reliant on the niche theory. There’s no attempt to “oust” the neutral 399 theory, which still gets a fair amount of research attention. Am I out of the loop here? It would 400 be nice it the figure and the steps A-F followed each other more directly. morphology and 401 behaviour after the identification of a scrap (the use of ‘morphospecies’ does not allow this). This 402 predictability was the basis for the development of the science of Botany in the eighteenth 403 century, and our ability to describe the vegetation around us. Unlike with animals, plant 404 taxonomists (Stace 1989) are happy to allow species with only incomplete restrictions to gene 405 exchange. However, each species is required to have a distinct phenotype. It must therefore have 406 a unique environmental tolerance and a unique reaction on the environment, even if the 407 difference from other species is sometimes small. Wilson & Agnew, chapter 1, Plants, page 15 of 34 408 409 2.2 Step B, Biogeography: The species pool The plant community can, in the short term, comprise only species present in the region, 410 which is the species pool (Fig. 1.2). The pool is difficult to define and quite as difficult to 411 determine, because we never know the distances over which species have the ability to disperse 412 or the frequency of dispersal events. However, different processes do occur on different spatial 413 scales. The regional species distribution for many species comprises a metapopulation: a series of 414 populations that are partly independent but connected by occasional migration events. In practice 415 the metapopulations of many species will show similar distributions due to similar habitat 416 requirements, giving a metacommunity (Holyoak et al. 2005). It is a nice distinction as to 417 whether a disseminule arrives via long-distance dispersal or from the metacommunity hinterland, 418 and in any case the resulting processes of establishment must be similar. 419 Questions about the species pool are dependent on the time frame: how long are we 420 prepared to wait for the species to arrive? Were time the only limitation to dispersal, disseminules 421 from far and wide would arrive anywhere, 400,000 species, and clearly this does not happen. 422 Continents have very different floras. Many European tree species have failed to occupy their 423 potential ranges in spite of several thousand years in which to spread across the continent 424 (Svenning and Skov 2004). The school of panbiogeography sees many present-day restrictions in 425 distribution between and within land areas as a reflection of the geography millions of years ago 426 (Fig. 1.3), and its analyses of species distributions that have repeatedly been borne out by 427 subsequent geological discoveries (Heads 2005). Clements and Shelford (1939) agree that 428 whereas migration of propagules is common, establishment of them is “altogether exceptional”. 429 [some discussion of waifs and mass effect seems to fit here] There are also restrictions on the 430 scale of hundreds of years: Matlack (2005) modelled the distribution of species in eastern USA 431 and concluded that the frequencies of species in the modern landscape was controlled by the time 432 available for spread in the last 300 years, with vertebrate-dispersed species occupying 433 considerably more of their potential geographical range than other species. It is often unclear on 434 which timescale the distribution limitation has occurred; for example a gap in the distribution of 435 Nothofagus spp. in the South Island of New Zealand (Fig. 1.3) has variously been correlated with 436 geological movements (Heads 1989), the last glaciation (Wardle 1980) and the current 437 environment (Haase 1990). Therefore, the closest we can come to definition is to say that over 438 realistic time spans most members of the area’s species pool could arrive, and we have to explain 439 the restricted subset of species found in each plant community. Wilson & Agnew, chapter 1, Plants, page 16 of 34 440 The concept of the species pool has sometimes included only species suited to the 441 environment of the habitat in question. In Europe, species have sometimes been excluded from 442 the pool using their Ellenberg ecological-tolerance rating (Ellenberg 1974). These values, 443 originally crude, have been progressively refined. Outside Europe, little information exists on 444 species tolerances for whole floras. A confounding question is whether the species pool is 445 defined before or after interference. If the species pool comprises all those species 446 physiologically able to tolerate the physical conditions at the site, it would include many never 447 found there, because of interference (Steps E and F). Many reports concerning filtering and Fig. 1.3a: Disjunct distribution (●) of the subshrub Kelleria laxa in South Island, New Zealand, interpreted as an originally contiguous distribution torn apart by tectonic movement ( ) along the Alpine Fault 2-10 million years ago and the ‘beech gap’ ( ). From Heads (1989). 448 interference assume that the species pool includes species that can tolerate the environment, but 449 cannot stand interference. However, the species lists are often taken from post-interference 450 communities and so the argument is circular. An example of this syndrome is when climate 451 change models of future vegetation are based on physiological parameters derived from 452 distributions (i.e. from the realised niches; sect. 4.1 below), and used in models as physiological 453 parameters (e.g. Sykes and Prentice 1996). [approaches to species pools seems to assume prior 454 knowledge] What are the authors going to use as their definition of the species pool? 455 2.3 Step C, Dispersal 456 Propagules 457 458 Propagule types are various. Within the angiosperms, seeds can be produced sexually (after meiosis and fertilisation), apomictically (with no meiosis and no involvement by pollen) or Wilson & Agnew, chapter 1, Plants, page 17 of 34 459 by pseudogamy (pollen is needed for seed development, and fertilises the endosperm, but the 460 embryo itself is produced apomictically). Vegetative reproduction can occur via bulbils, stolons, 461 rhizomes, layering of branches (e.g. Salix cinerea, willow tree), root suckers, etc. There is no 462 basic distinction between the apomictic seeds of Taraxacum spp. (dandelion), ‘vegetative 463 reproduction’ such as the production of Kalanchoe daigremontiana plantlets from the leaf 464 margin, the growth of an Elytrigia repens (couch grass) clone by rhizomes, the growth of a 465 Populus tremuloides (aspen) clone by root suckers and the growth of an axillary bud on a tree 466 branch to give new leaf modules. All replicate an original genotype but after many mitotic 467 divisions which can accumulate errors. [common theme without relevance noted] 468 The immediate fate of these propagules is various. Bulbils and viviparous seeds both 469 develop as plantlets on the parent. Ramets produced by stolon or rhizome are initially dependent 470 on the parent, then for a period are physiologically independent unless a change occurs, such as 471 defoliation or shading, when ramets subsidise each other (Marshall 1996), and then become fully 472 independent as the connecting stolons/rhizomes wither. Seeds are usually dispersed by wind, 473 water or animals, although a few plants produce hypogeal seeds (i.e. belowground). Tree and 474 herb sectors behave similarly to clonal tillers with limited integration, except that there is a 475 greater tendency for branches to overtop one another competitively (Novoplansky 1996). 476 Migration 477 Dispersal is the means by which species move around the landscape. The two critical 478 considerations are the distance and frequency with which disseminules move outside their source 479 habitat, which is negatively related to disseminule size, and their potential for establishment as a 480 seedling in a new site amongst existing plants which is positively related to disseminule size 481 (Salisbury 1942). 482 483 484 Plant dispersal usually has a long tail (Fig. 1.4), i.e. it is leptokurtic, and is often best fitted by a negative exponential function.[tail even longer than that??] Indeed, there is work by James Clark (1999, 1998 among them) that explores a diversity 485 of functions and found that other distributions often fit better. That is, most dispersal of 486 disseminules of every type is surprisingly short-distance, with rare long-distance events, as Carey 487 and Watkinson (1993) found for mechanical scatter of the seeds of an annual festucoid grass and 488 Matlack (2005) for dispersal of ingested seeds. The reason is probably that most species are 489 dispersed by two or more mechanisms: for example Agnew and Flux (1970) found in the Rift 490 Valley, Kenya, that though many grass disseminules had a large wing apparently adapted for Wilson & Agnew, chapter 1, Plants, page 18 of 34 491 wind dispersal, the longer distance dispersal seemed to occur when the fruit became entangled in 492 the coats of Lepus capensis (hares). Occasionally, the direction can be towards suitable habitat, 493 for example ants dispersing seeds along their runways (Huxley and Cutler 1991). In general, the 494 number of disseminules arriving (the ‘propagule pressure’) will not matter: a smaller number will 495 delay an invasion but will not prevent it.[propagule pressure and disturbance can together cause 496 considerable influx] The exception is when an Allee effect is operating. 70 Number of seeds 60 50 40 30 20 10 0 -80 497 -60 -40 -20 0 20 40 60 80 100 120 Distance (m) Fig. 1.3. Fig. 1.4. The leptokurtic curve of dispersal: Juncus effusus seed rain around a single plant. 498 The population of dormant seeds in the soil or in aerial fruits – the ‘seed pool’ or ‘seed 499 bank’ – is a buffer against elimination of a species. This is important for an annual species in an 500 adverse year, and for species surviving through disasters such as fire. The seed pool is not only 501 local, it can have a metapopulation structure comparable to that of adults when seeds are moved 502 around by floods or large herbivores such as elephants. This can be seen as spatial mass effect 503 [not defined](chapt. 4, sect. 12 below). 504 2.4 Step D, Environmental filtering / ecesis 505 Propagule germination and establishment 506 The germination of a propagule starts phase of an invasion when it is the challenged by 507 the conditions in a new site (Fig. 1.2). Awkward sentence Sometimes seeds germinate only under 508 conditions more mesic than those tolerated by their parents, and these events may be rare in 509 stressed habitats. For example, many halophytes need unusually low salinity on a saltmarsh 510 before they germinate, when their more glycophytic seedlings can become established 511 (Alexander and Dunton 2002). Arid land species show very precise adaptations for effective Wilson & Agnew, chapter 1, Plants, page 19 of 34 512 dispersal and germination in the highly variable rainfall patterns found in these habitats 513 (Gutterman 2002). For example, Pake and Venable (1996) found that different species of winter 514 annual in the Sonoran Desert tended to germinate in different years, and species tended to 515 germinate more in years that turned out to give them higher reproductive success. 516 Plants of all species need to pass through a juvenile phase before becoming reproductive. 517 This part of the challenge in occupying a new site is called ecesis (Clements 1904). As Clements 518 (1916) wrote: “Ecesis is the adjustment of the plant to a new home. It consists of three essential 519 processes, germination, growth, and reproduction. ... Ecesis comprises all the processes exhibited 520 by an invading germule from the time it enters a new area until it is thoroughly established there. 521 Hence it really includes competition and other types of interference, except in the case of 522 pioneers in bare areas.”. This definition is of course far too broad, because it seems to include the 523 whole life cycle. We need to separate out the first stages for our reductionist view, so we restrict 524 ‘ecesis’ to post-germination survival, growth and establishment: the part of the species filtering 525 process that determines which species survive the initial dispersal and germination phases of 526 plant community establishment. [reference to Grubb’s regeneration niche?] 527 Move to start of STEP D and combine with the environmental filtering section. 528 529 530 Invasion patterns Move invasion patterns prior to Step D Invasion can be seen on all scales, from movement to and fro across 2 m within a decade 531 in links (Olff et al. 2000) to movement over thousands of years. The patterns of invasion seen are 532 much the same at any scale, and fall into three types. Phalanx invasion is dense and over a broad, 533 solid front. Guerrilla invasion comprises invasion by isolated individuals, which gradually fills in 534 the space available (Hutchings 1986). It is difficult to measure invasion processes because 535 a-priori assessment of habitat suitability is problematic. However, an example of an invasion 536 where at least part of the flora used guerrilla invasion is the advance of the herb flora of ancient 537 forest into adjacent secondary growth in Sweden, where Brunet et al. (2000) concluded that 538 distance (equivalent to time for dispersal) and the soil environment almost equally controlled 539 species composition. However, by far the most common seems to be a third type of invasion – a 540 combination of the former two – infiltration invasion in which there is occasional long-distance 541 dispersal (as in guerrilla invasion), followed by local short-range dispersal from these foci (as in 542 phalanx) (Fig. 1.5; Wilson and Lee 1989). This matches the leptokurtic / two-mechanism 543 dispersal typical of plant disseminules.[nucleation well documented in primary succession 544 studies, such as Mt St. Helens] It can be seen at a variety of scales, e.g. over kilometres (Lee et al. Wilson & Agnew, chapter 1, Plants, page 20 of 34 545 1991) to centimetres. A small-scale example occurs when most of a tussock grass’ tillers are 546 produced within the leaf sheath, but a few are pushed greater distances by animal hooves 547 (Harberd 1962), an example of the leptokurtic / two-mechanism dispersal to which we referred in 548 section 2.3 above. Egler (1977) describes all these patterns, with details, though using different 549 terms. Expanding foci Guerrilla individuals g clumps 550 551 552 553 Fig. 1.5: Infiltration invasion by Olearia lyallii in the Auckland Islands. Environmental filtering No habitat holds all the available species from its hinterland pool. There are physical 554 environmental conditions, such as soil type, hydrology, climatic regimes and altitude, that 555 prevent the immigrants’ growing (Honnay et al. 2001). This has been called environmental 556 filtering or abiotic filtering (Weiher and Keddy 1995). It occurs largely during ecesis. The 557 existence of this filter is obvious. Sophisticated methods can be used to record the response 558 surface (e.g. Bio et al. 1998), but it remains what Warming (1909) described as “this easy task”. 559 In this book, we generally take the physical restrictions as given, and concentrate on those 560 community processes that control species composition. However, we need to return to this topic 561 in considering the niche. 562 Reaction 563 The physical environment is not strictly abiotic, for the receptor community can alter the 564 environment to create or close invasible sites, a process for which Clements (1904) coined the 565 term ‘reaction’: “By the term reaction is understood the effect which a plant or a community 566 exerts upon its habitat. … Direct reactions of importance are confined almost wholly to physical 567 factors” (Clements 1916) This term was first used much earlier in the text but not defined there, I Wilson & Agnew, chapter 1, Plants, page 21 of 34 568 suggest moving this def to the first use of the work. Any organism must cause reaction, though 569 the environmental modification varies from slight to major, and the causal species' autoresponse 570 from negative (facilitation: Clements 1916) to positive (switch: Wilson & Agnew 1992). 571 Reaction is the basis of almost all plant-plant interactions (Clements 1904). Since Clements, 572 other terms have been used for the same effect, such as ‘ecosystem engineer’ (Jones et al. 1994), 573 and ‘niche construction’ (Laland et al. 1996). They seem to be later synonyms, but we shall 574 sometimes use the latter when discussing the niche. 575 Sometimes, the favourable microsites for establishment are gaps, but Ryser (1993) found 576 in a temperate calcareous grassland that the favourable microsites were those where the reaction 577 of established plants in the community provided shelter from frost; unvegetated gaps were not 578 colonised. Litter production is a common mode of reaction, forming the seedbed of an invader, 579 and enabling or inhibiting its germination (chapt. 2 below). needs rewording to emphasize that 580 these are examples of reaction. 581 2.5 Step E, Productivity and biotic filtering 582 Communities develop a regime of carbon cycling, of which the autotrophic production is 583 our concern. We devote space to productivity here because it is easy to think rather loosely about 584 it. Productivity is “The potential rate of incorporation or generation of energy or organic 585 matter … per unit area …” (Lincoln et al. 1982), but there is a lot of complication behind this 586 definition: 587 588 589 590 1. Carbon is fixed from CO2 by the C3, C4 or CAM mechanisms [define, discuss?]. We suppose this is gross productivity. 2. Immediately, some of the fixed C is lost by photorespiration, though in C4 plants it is retained in the leaf, and can be re-assimilated. 591 3. Later, at night, some of the fixed C is lost by dark respiration. 592 4. The remaining C is transported to sinks at sites of cell division (secondary cambium, root, 593 shoot apex, inflorescence, etc.), where it is incorporated into cell wall tissue, storage 594 carbohydrates or cytoplasm, with shared potential fates. In this process, it can be: 595 A. Lost by respiration whilst still incorporated in soluble C compounds, etc. 596 B. Lost to aphids on the way: Heizmann et al. (2001) wrote: “during midday depression 597 of photosynthesis, a high percentage of the total C delivery was provided to the 598 leaves by the transpiration stream (83 to 91%). Apparently, attack by phloem-feeding 599 aphids lowered the assimilate transport from roots to shoots; as a consequence the Wilson & Agnew, chapter 1, Plants, page 22 of 34 600 portion of C available to the leaves from xylem transport amounted to only 12 to 601 16%.” 602 C. Leached from leaves as soluble C compounds (Czech and Kappen 1997). 603 D. Lost by roots as exudate of soluble C compounds (Kuzyakov and Siniakina 2001) and 604 605 606 607 as mycorrhizal growth and respiration (Johnson et al. 2002) E. Converted into plant material mainly as cell wall. This material, with the cell contents, can be: i. Eaten by pests (vertebrates, invertebrates and pathogens). 608 ii. Removed by allogenic or autogenic damage or abscised. The amount lost varies 609 from parts of leaves to tree branches: cell walls plus modified cell contents. 610 iii. Lost at the death of tissues (wood, bark), organs (roots, leaves flowers and fruit), 611 612 or the whole plant. F. In living tissues, C in carbohydrate storage and cytoplasm can be lost by respiration as 613 the root becomes old or the leaf becomes shaded and/or old, or it can be translocated 614 with attendant respiratory costs. However, cell wall C cannot be lost this way, 615 because no autolysis of cellulose or lignin occurs within living plants. (This contrasts 616 with animal tissue, where all C is part of the labile pool except for some dermal 617 structures.) 618 In the face of this complexity there is no consensus as to what productivity is or how it should be 619 measured. Logic and simplicity would suggest that the real definition of productivity [ecological 620 efficiency]is the amount of C that reaches the next trophic level, herbivores or decomposers. This 621 top down definition would comprise only E1-3 in the above schema. ‘Gross productivity’ can be 622 measured in the field through gas analysis (though this omits photorespiration: Step 2), and ‘net 623 productivity’ as this less all the later losses. Productivity is most often estimated by sequential 624 sampling. Suppose a habitat holding mature stable vegetation in an approximately steady state. If 625 we sample at one time and then resample the same area 12 months later, in many areas there 626 would be no change in the absence of climate change and apart from sampling error – neither 627 accumulation nor loss of biomass – so we would arrive at an estimate of zero productivity. This is 628 either accurate or misleadingly trivial, depending on your definition of productivity. However, all 629 systems show some seasonal development and change, and most estimates of plant productivity 630 rest on the successive sampling of harvest biomass (standing crop) during the season of maximal 631 growth (Perkins et al. 1978). This working approximation is a wild under-estimate of actual Wilson & Agnew, chapter 1, Plants, page 23 of 34 632 productivity, yet has physical presence and ecological meaning. Actually, in much discussion of 633 productivity, e.g. in testing for a humped-back curve (Grime 1979) [not defined??], standing crop 634 is used as a substitute, and it is a very poor one. 635 The productivity potential of a site controls what plant community develops in three ways. 636 Firstly, the readiness of the soil surface to provide sites for invasion and thus augment the 637 community: often with greater productivity more litter will be available affecting the ecesis of 638 invaders (sect. 2.4 above; chapt. 2 below). It should definitely be briefly revised here HOW the 639 ecesis of invaders is affected, e.g. positively or negatively? 640 Secondly, disturbance of the community through herbivory and often fire: high 641 productivity attracts herbivory, while pronounced dry seasons between productive growing 642 seasons favour fire. However, in general these first two factors affect the rate of invasion more 643 than the eventual fate of an invasion.[really? ]A statement like that should have references to 644 back it up. The third aspect is the competitive status of the community: greater productivity 645 means more competition and more difficulty for additional species to establish-what effect do 646 you hypothesize that this will have on the outcome and rate of invasion?. For example, Cantero et 647 al. (1999) concluded that the diversity of short grasslands in Argentina was affected by 648 surrounding species pools, while that of tall grasslands with more competition was not. This is 649 the interference filter, in which a species is able to tolerate the physical environment of a site, but 650 cannot grow well enough there to withstand the general level of interference present there: 651 competition, allelopathy, etc. It is an effect not specially dependent on the identity of the 652 associates. This is the classic distinction between the fundamental and realised niche (sect. 4.1 653 below). It is clearly a major factor, as can be seen by the ready cultivation of many species in 654 botanic gardens outside their natural edaphic and/or climatic range. 655 Organisms of other trophic levels affect ecesis and reproduction (sect. 1.1 above, 656 “Interaction with other trophic levels”). A species might be able to maintain a positive population 657 growth rate without heterotrophs, but be pushed into negative growth when pathogens or 658 herbivores take their toll. This could be environmentally dependent: a potentially fatal herbivore 659 might be absent because the environment is beyond its tolerance. On the other hand, a plant may 660 be unable to reproduce because a normally subventing pollinator is beyond its environmental 661 range, and hence absent. This would be a type of assembly rule (Step F), though in this book we 662 consider only plant-plant assembly rules as such. Wilson & Agnew, chapter 1, Plants, page 24 of 34 663 2.6 Step F: Assembly rules and micro-evolution 664 Did not provide enough background information for me. It could be written more generally reserving the specific 665 examples for chapter 5. 666 Assembly rules are "restrictions on the observed patterns of species presence or 667 abundance that are based on the presence or abundance of one or other species or groups of 668 species …" (Wilson 1999 %chapter). We discuss them in chapter 5. It is clearly a simplification 669 to take species as fixed units and we do so only to limit the scope of this book. To glimpse into 670 the world of ecotypes and micro-evolution as they affect community structure we examine the 671 work of Turkington and Harper (1979), taking plants of Trifolium repens (white clover) from 672 patches of a field dominated by four different grass species. When they planted them into boxes 673 of a standardised soil sown with the four grass species, each T. repens genotype was the best 674 performer against the species from whose neighbourhood it had been taken in the field – an 675 amazingly neat result. They interpreted this as genetic coadaptation within the field. A 676 mechanism such as the quality of transmitted light is possible (Thompson and Harper 1988). 677 Awkward paragraph 678 Aarssen and Turkington (1985 %605) performed a similar experiment in a pasture in 679 British Columbia, Canada, but using different patches/genotypes of Lolium perenne (ryegrass). 680 They obtained similar results for T. repens. However, they also examined variation within the 681 associated L. perenne and the results for it were the opposite – three of the four L. perenne 682 genotypes had their lowest competitive ability against the T. repens genotype with which they 683 had been growing in the field. This would tend to keep the grass/clover competitive abilities 684 balanced. There remains a fear that the effects were due to carry-over, i.e. maternal effects in the 685 vegetative material. Aarssen (1988) found that collecting seed rather than using vegetative 686 material (ramets) gave quite different results, which he attributed to screening of the gene pool 687 between seed and adult populations. Such screening has certainly been seen in heavy-metal 688 ecotypes, though under conditions where gene flow was high and the selective differentials 689 extreme. However, Turkington and Harper (1979) had used preconditioning periods of only 3 690 months. Evans and Turkington (1988) in Canada, collecting plants in a similar way to Turkington 691 and Harper (1979) – from below four different grass species – found morphological differences 692 between T. repens of the four origins after 4 months growth in a common garden which 693 disappeared after 27 months growth. Chanway et al. (1989) suggested that the difference between 694 T. repens material might be in the specific Rhizobium strains carried with it, not in the T. repens 695 itself. This could still be a force in structuring communities. Wilson & Agnew, chapter 1, Plants, page 25 of 34 696 The Turkington and Harper (1979) result would have been small-scale character 697 displacement. It is almost impossible to prove that character displacement has occurred because 698 the evidence must involve comparisons between areas, and those areas might differ in other ways 699 (Strong 1983). However, some cases are suggestive. If found, character displacement would be 700 evidence that species interactions were a strong force in genetic selection, and therefore also in 701 ecological selection, implying deterministic community structure. 702 3 Geographical boundaries 703 The behaviour of a species at its distributional limit can be fascinating. Often the habitat 704 range of a species becomes more restricted towards its boundary. Pigott (1970) reported that near 705 the limit of Ilex aquifolium (holly) in Britain it becomes increasingly restricted to forest, and 706 Cirsium acaule (stemless thistle) at its northern limit becomes confined to southern (warm) 707 aspects. On the other hand, Diekmann and Lawesson (1999) found four potential examples where 708 species had wider ecological amplitudes towards their range margin in northern Sweden, and 709 suggested that there is such climatic stress in that region that a smaller flora is present and 710 important competitors are absent. Usually the plants are smaller and less fecund towards the 711 limit, leading to populations becoming smaller and absent from apparently suitable habitat 712 patches (Carey et al. 1995; Nantel and Gagnon 1999; Jump and Woodward 2003). Lower 713 fecundity also makes populations more sensitive to disturbance. For example, fire restricts 714 Canadian Abies balsamea (balsam fir; Sirois 1997) and Pinus resinosa (red pine; Flannigan and 715 Bergeron 1998) to islands and isolated populations at the northern edge of their ranges. In some 716 cases there is a sudden cut-off point in a species with no reduction in vigour near the limit 717 (Lactuca serriola, prickly lettuce; Carter and Prince 1985). This may not be exceptional. Griggs 718 (1914) made an early and careful observational study of Sugar Creek, in a “tension zone” in Ohio 719 where over 120 species have geographical boundaries, asking whether populations became sparse 720 or less fecund in this region. He found no consistency of behaviour, but most edge-of-range 721 species were abundant and flowered and fruited successfully up to the geographical limit, as in L. 722 serriola. Griggs could only hypothesise that competition sharpened boundaries to make them 723 abrupt. There can be many different reasons for a species’ failing to expand its distribution, 724 sometimes surprising ones. Pigott and Huntley (1981) found that the environmental filter for Tilia 725 cordata (linden) at the northern limit of its range in England was that the pollen tube could not Wilson & Agnew, chapter 1, Plants, page 26 of 34 726 grow fast enough in the low temperature to reach the ovule, leaving a relictual population now 727 unable to reproduce by seed. At present it seems that the behaviour of species at the margins of their ranges is complex 728 729 and unpredictable. For example, if the range limit were due to individuals being selectively 730 eliminated we might expect lower variances of morphological measurements at geographic 731 edges, but Wilson et al. (1991 %780) could find no such effect [I find it interesting Bastow did 732 not hit on the classic bell-curve of abundance of Whittaker – since some species tend to be either 733 abundant or absent. One sacred cow was forgotten in the pasture]. 734 4 Concepts of the space occupied by one species 735 Our purpose in this book is to examine the way species fit together in a mixture. Two powerful 736 conceptual tools, developed for this purpose, are the niche and the guild. A table or diagram 737 would be useful for comparing niche and guild and also showing relationships between 738 alpha/beta niche/guild More discussion of the niche makes one thirst for more on neutral and why only niche 739 740 works…perhaps this is another (much smaller) book? 741 4.1 742 743 744 The niche The discussion of niches is interesting, and the concepts of alpha and beta niche were ones that seem useful, and original. The end point of a species' pilgrimage from the pool into a community is occupancy of a 745 niche. The niche is an old concept. Grinnell (1904) and Elton (1927) introduced the term, and 746 both used it to describe an area available within habitat space, broadly defined by physical and 747 trophic parameters. Hutchinson (1944) formalised this to “a region in n-dimensional hyperspace” 748 where the dimensions are all the environmental, resource or behavioural (e.g. phenology, 749 foraging) parameters that permit an organism to live. 750 Since Hutchinson’s overarching statement it has been tempting to regard species presence 751 as the only definition of the niche. Thus, Levins and Lewontin (1985) advocated that “ecological 752 niches are defined only by the organisms in them”. Olding-Smee et al. (2003 %book) believed 753 that for Hutchinson “a niche cannot exist without an occupant”. We see no reason to understand 754 Hutchinson thus. The crunch comes with the empty niche. Under the “the species is the niche” 755 concept “the idea of an ecological niche without an organism filling it loses all meaning” (Levins 756 and Lewontin 1985)-Surely this is a strawman argument. Would someone make the same claim Wilson & Agnew, chapter 1, Plants, page 27 of 34 757 today?. However, the empty niche is a necessary concept in theory, especially in relation to 758 invasions. The absurdity of the “the species is the niche” is seen by observing innovative 759 invaders. Did no niche for a cactus exist in central Australia until Opuntia stricta (prickly pear 760 cactus) was introduced (Hosking et al. 1994)? Was there no niche for a cactus-eating insect 761 before the moth Cactoblastis cactorum was introduced for biological control of O. stricta? Was 762 there no niche below the saltmeadow in British estuaries until Spartina ×townsendii / anglica 763 (cord grass) created itself by hybridisation in 1887? Was there no niche for an emergent tree in 764 Bonin Island shrublands until Pinus lutchuensis (a pine from elsewhere in Japan) was introduced 765 (Shimizu and Tabata 1985)? It seems better to regard all these as empty niches that were later 766 filled. To be sure, the identification of empty niches is very hard. There must be areas of 767 hyperspace that it is impossible for plants to fill: floating in the air, growing on the ice at the 768 South Pole or growing at 100 °C in hydrothermal steam vents? 769 Tilman (1997 %81) claimed to find evidence of empty niches. In 1991 he sowed seeds of up to 770 54 species into native grassland at Cedar Creek. Many became established [but for how long?]. 771 However, this did not cause extinctions among the species originally present in 1991: the 772 proportion of those lost was not correlated with the number of species added (r = +0.16, R2 = 773 2.6 %, not significant). Even more interestingly, the total cover of those species present in 1991 774 did not decrease (r = 0.04, R2 = 0.16%, not significant). The R2 values are impressively low so 775 we could conclude, as Tilman did, that the added species occupied empty niches. There is a 776 problem that the species composition probably co-varied with the species richness. The use of 777 ‘total cover’ is odd. If two leaves of different species are vertically aligned, both count towards 778 cover, and if two leaves of the same species are horizontally aligned both count, but if two leaves 779 of the same species are vertically aligned only one counts ?? I’m not certain what the the authors 780 are trying to describe here. I agree that total cover is generally not a very standardized method 781 when collecting field data. However, I think there are more methods in vegetation sampling that 782 are probably equally as flawed. 783 ‘Total cover’ is not a sensible concept. In this case, the ‘cover’ of each species was, 784 unfortunately, guessed. Cardboard cutouts were used to guide the guessing but we do not take 785 aids to guessing as removing the fact that cover was guessed. Philip Grime’s group at Sheffield 786 always uses objective measurements: presence/absence, local frequency, point quadrats or sorted 787 biomass as appropriate. Why can’t everyone? [seems self evident] We mention this issue because 788 it will repeatedly mar results that we report. We shall not shirk from pointing out when the data Wilson & Agnew, chapter 1, Plants, page 28 of 34 789 of this type are used, nor shall we use euphemisms like “estimated by eye”, because we believe 790 this practice is a blot on our science (even if we have occasionally been guilty ourselves in evil 791 places far away and naughty times forever gone) [“cover guessing” measures can be valid and 792 helpful in many contexts. As with many methods (most statistical methods, for example), you 793 need to be aware to the possible problems, the accuracy, the biases, and so forth. I am worried 794 that the book tends to be repeatedly dogmatic and negative without any attempt to provide a 795 balanced perspective – it is living up to its goal in the Preface of being unbalanced.] 796 How does the use of presence/absence data allow us to differentiate between dominant 797 and trace species within a plot? Not to mention that this throws out indicator species analysis, 798 ordination, community classification(a given), and a whole host of other measures for how 799 vegetation in one spot is different from vegetation in another.. Very soap-box. Some valid points 800 but I still think that there can be valid/appropriate uses of “estimated” cover. While it may be fine 801 (and quite Feasible!) to actually use repeatable, objective measures in herb/grass communitys, it 802 is not always a reasonable option. 803 The second reason for rejecting the “the species is the niche” concept is that it takes 804 species presence in the field as the de facto description of its niche, but this is affected by 805 interaction with other biota (competition, herbivory etc), introducing an imponderable set of 806 variables over space and time. It is more useful to separate biotic variables as restricting the 807 occurrence of a species to its realised niche, whereas its environmental tolerance defines its 808 fundamental niche or physiological tolerance (Hutchinson 1957). This distinction was known to 809 Tansley (1917) and Gleason (1917), though dismissed by Clements (1907) as “merely 810 migration”. The niche width of a group of species is usually considerably narrower, and hence 811 their niche overlap is less, when they are grown in an experimental mixed community than when 812 they are grown alone in the same conditions. That is, their realised niche is smaller than their 813 fundamental niche. Silvertown et al. (1999 %61) found this re-analysing data of Ellenberg’s with 814 six grass species and a water table gradient: the mean niche overlap was considerably higher in 815 the 6-species mixture than comparing monocultures, and the niche modes spread out to a range of 816 5-100 cm depth-to-water-table in the mixture compared to a range of 20-35 cm among the 817 monocultures. However, definitions become difficult because of reaction: any organism must 818 alter its own environment and this may cause niche construction, potentially leading to realised 819 niches that are larger than the fundamental ones (Odling-Smee, et al. 2003)[real or definitional 820 problem??] Wilson & Agnew, chapter 1, Plants, page 29 of 34 821 The niche includes a species' developmental requirements (temperature etc.), its material 822 requirements (resources) and its relations with neighbouring species. A complete circumscription 823 of these is almost impossible, requiring knowledge of every aspect of the species' physiology and 824 life history, but two types of niche can be distinguished. The beta (β) niche is the range of 825 physical environmental conditions under which the fitness of a species is maintained (Alley 826 1985), e.g. its temperature tolerance, and therefore its potential geographical limits. It is related to 827 Chesson’s (in press) concept of ‘environment’ as a factor that does not form a feedback loop, i.e. 828 is not appreciably affected by the organisms themselves. The alpha (α) niche represents the 829 resources used within a community/site, the “‘profession’ or functional role” (Alley 1985), e.g. 830 different rooting depths. Many methods of analysis, e.g. the calculation of niche width and 831 overlap, can be used for both alpha and beta niches, and there are areas of character overlap. 832 However, when we use the niche concept we generally need either one or the other. Much 833 ecological discussion has been confused by failing to take the distinction into account. I have to 834 say that this (the alpha and beta niches and guilds) is the part of the book that I have so far 835 enjoyed the most. Here is a presentation of a new/different way of looking at things rather than a 836 tearing down of what has been 837 838 839 The axes of the alpha niche are controlled by the morphology and physiology of the plant, its growth and its chemistry: 1. Morphology and its plasticity influence resource foraging and capture (light, nutrients, 840 water source), persistence (storage organs, wood), autogenic disturbance (through litter 841 and physical environmental effects), heat budget (convective, transpirative, radiative), 842 physical defence against herbivores (glands, hairs, thorns), pollination and dispersal 843 biology. An example of these factors is synusiae in forest, such as epiphytes and lianas, 844 and indeed stratification in almost all communities. Another example is the parasitic 845 habit. 846 2. Growth phenology as the plant's response to environmental signals comprises the 847 seasonality of growth and reproduction (pollination and dispersal). Examples are the 848 progression of flowering in temperate vegetation and leaf flushes in tropical forests. 849 3. The chemical functioning of a plant ultimately controls everything, but we may list as 850 examples phototype (C3, C4 or CAM), light requirement, mycotrophy, P sources via root 851 phosphatase exudate, N source (N2, NH4 or NO3) and chemical defence against herbivore 852 and pathogen challenge. Wilson & Agnew, chapter 1, Plants, page 30 of 34 853 4. Any of the above niche axes can influence plant/plant interactions, through interference or 854 subvention by neighbours (chapt. 2 below), for example in the morphological pre-emption 855 of soil resources, though the overgrowth of competitors and through the toxic chemical 856 countering of competition known as allelopathy. 857 858 5. Additional resources are gathered by the community. This is reaction causing niche construction. 859 Beta niche axes are the environmental features of the locality and its biota, that is to say the 860 habitat. Complexities between factors are more apparent than with alpha axes. Aspects are: 861 1. Climate delivers solar insolation, water availability, CO2 (though the latter is rarely local), 862 nutrients (when ombrotrophic), some pollination vectors and rate regulators such as 863 temperature. These interact with the chemistry below. Climate also delivers exposure to 864 atmospheric humidity, wind, aeration and snow. This affects morphology, for example as 865 a partial determinant of the Raunkiaer life form (the life form can also differ between 866 alpha niches, for example in forest stratification). 867 2. Chemical features of soils (calcareous versus non-calcareous, pH, salinity, etc.) affect 868 system function (nutrient availability, cycling); these overlap with geomorphology below. 869 In the short term, mineral nutrients are often dominant. 870 3. Geomorphology delivers allogenic disturbance and soil substrate. 871 4. Biota deliver allogenic disturbance, generalised herbivore pressure and animal pollinators. 872 There can sometimes be overlap between the concepts of alpha-niche and beta-niche in terms of 873 characters: e.g. low growth is a feature of the ground alpha-niche in a forest but also of arctic 874 plants. However, the effects are opposite: species of the same beta niche will tend to co-occur 875 because they have the same environmental tolerances; species of the same alpha niche will have 876 no such tendency to co-occur, and if competitive exclusion is operating they will tend not to co- 877 occur (Wilson, submitted). 878 4.2 Guilds 879 The ecological term ‘guild’ was coined by Drude (1885) as the German 880 ‘Artengenossenschaften’ to refer to a group of species moving from one region to another, such 881 as exotic species. It was used thus by Clements (1904; 1905) and Wilson (1989 %223). Perhaps 882 independently, Schimper (1898; 1903) used the term ‘Artengenossenschaften’ / ‘guild’ to mean a 883 synusia (e.g. stratum) in a forest. Tansley (1920) used it in the same way, writing of “guilds of 884 the same dependent life-form, such for instance as lianes”. Root (1967) ignored these established Wilson & Agnew, chapter 1, Plants, page 31 of 34 885 usages and with animal assemblages in mind re-defined the guild as a “group of species using 886 similar resources in a similar way”. This is not directly useful for plants, since almost all use the 887 same resources (the sun’s energy, water, CO2, N, P, K and minor elements). The guild is a 888 category that is intended to be ecological rather than taxonomic, and Wilson (1999) defined it as: 889 “a group of species that are similar in some way that is ecologically relevant, or might be”. It is 890 unusual to find “Or might be” in a scientific definition; it is necessary here because we hardly 891 ever known at the beginning of an investigation whether the guilds we are using are the real ones, 892 and often not at the end (but see the discussion of intrinsic guilds: chapt. 5, sect. 7.6 below). The 893 phrase “or might be” in the definition of guild is troubling. I understand the point that some plant 894 species are put into guilds a priori so that the guild may or may not be actually relevant. 895 However, the idea is that the guild is relevant, and if it turns out not to be relevant, then we 896 should change the way we have grouped the species, not the definition of guild. In spite of the 897 lack of precedence of Root’s usage, and the impossibility of applying it strictly to plants, a 898 similar usage can have value: a guild as a group of species that occupy similar niches. 899 Wilson (1999) pointed out that there are two basic types of guilds, corresponding to the 900 distinction between alpha-niches and beta-niches. Again, the outcomes are opposite: species that 901 are in the same beta-guild and therefore have similar environmental tolerances will generally co- 902 occur; species that are in the same alpha-guild and therefore use similar resources will tend 903 exclude each other. The species within one alpha-guild are similar in their resource use. For 904 example, within northern European forests, species that are within the same alpha-guild might be 905 the trees Tilia cordata (linden), Quercus petraea (sessile oak) and Fagus sylvatica (beech). They 906 are using similar resources: the light at the top of the canopy during the summer half-year, as well 907 as nutrients and water from the full profile of the soil. Conversely, if species are present in one 908 community that are in different alpha-guilds, they might be able to partition resources within a 909 community, so a community might tend to comprise species from several alpha-guilds. For 910 example, Tilia cordata, the hemi-parasite Viscum album (mistletoe), the liana Hedera helix (ivy) 911 and the ground herb Mercurialis perennis (dog’s mercury) would be in different alpha-guilds 912 because they use different light/support/nutritional resources, and if we found them together we 913 might see it as alpha-niche differentiation. 914 The species in one beta-guild are similar in their ecophysiology and therefore their 915 tolerance (across space or time) of environmental conditions, such as the “guilds of edaphic and 916 topographical specialists” of Hubbell and Foster (1986 %314). After a species pool has passed Wilson & Agnew, chapter 1, Plants, page 32 of 34 917 through an environmental filter, the remaining species will be a beta guild; they have overlapping 918 beta-niches. For example, all sub-arctic saltmarsh species would be in the same beta-guild 919 because they occur in the same climatic and soil conditions. An example of species occurring in 920 different beta-guilds might be the temperate, mesic tree Tilia cordata, the subalpine Pinus 921 contorta (lodgepole pine), arid land trees/shrubs of Prosopis spp. (mesquite), the tropical 922 Cinchona officinalis (quinine) and a species of mangrove. They occur in different environmental 923 conditions (climate and/or soil), so they are necessarily found apart in space or time since we 924 cannot find different external environmental conditions simultaneously at one spot. Díaz et al. 925 (1998) recorded abundances of ‘plant functional types’ (PFTs) of 100 species along a climatic 926 gradient in Argentina and found that vegetative traits differed between climatic zones, 927 demonstrating that beta-guilds are filtered out from the available species pool. The species within 928 each zone will almost certainly belong to different alpha-guilds. 929 The concept of ‘functional type’ (used above by Diaz et al. 1988) and that of the ‘guild’ 930 can be essentially identical (Wilson 1999; Blondel 2003). The current use of PFTs as the 931 predicted variate in models assumes that we know the characters of the types are trying to 932 summarise. In spite of the term ‘functional’ which implies alpha-guilds, most workers have 933 apparently intended to create beta-guilds. However, the characters they have chosen have often 934 been alpha-niche ones. For example, Kleyer (2002) formed guilds (‘functional types’) “to relate 935 unique PFTs to landscape specific habitat factors and to generalize syndrome-environment 936 relations across landscapes” and used characters such as annual versus biennial versus perennial, 937 plant height, regeneration from detached shoots, having leptophyllous leaves, longevity of seed 938 pool that are as likely to occur within a community. A distinction between ‘response’ and ‘effect’ 939 guilds obscures the issue, because there is far more to the alpha-niche of a species than its 940 reaction (effect) on the environment. This situation has arisen from a failure to consider the 941 purpose of the guilds being formed, what type of guilds they will therefore be – alpha or beta – 942 and what characters are therefore appropriate. 943 4.3 Stratification 944 The most obvious alpha-guilds in plant communities are the guilds of Schimper (1898; 945 1903), synusiae (Fig. 1.6). Almost all plant communities are structured vertically. Aboveground, 946 the greater the vegetation cover, the more uniform and predictable is the vertical change in 947 microclimate. Highly structured forests have a stratum of separated, emergent trees, a more 948 continuous upper canopy, then sub-canopy trees, shrubs, tall herbs, creeping herbs and Wilson & Agnew, chapter 1, Plants, page 33 of 34 949 bryophytes, lianas and epiphytes (including lichens, bryophytes and higher plants). This 950 represents specialisation to the attenuation of light, water, CO2 and nutrient resources. All this is 951 accepted for forests, but there is also complex stratification in grasslands, for example in the wet 952 grasslands of Tierra del Fuego (Díaz Barradas et al. 2001) and even in lawns (Roxburgh et al. 953 1993 %699). Naturally all stratification by primary producers is echoed by stratification in 954 consumer communities. Fig. 1.5: Stratification: profile of a rain forest in British Guiana. From Richards (1964). 955 956 We might expect that similar patterning is happening below ground because litter is 957 deposited on the soil surface eventually adding to the water-holding capacity and mineral nutrient 958 status of the upper soil. Most water arrives at the soil surface and percolates down, acidified by 959 organic acids and CO2, hydrolising the mineral fragments in the soil and most importantly 960 releasing phosphate. Plant roots and respiration can affect this, for example releasing acids. 961 Water in deep soil, including artesian water, is available to deep roots and may rise up by 962 capillarity and hydraulic lift. This can lead to stratification of root systems. Succulents of New 963 and Old World deserts have surface roots adapted for the uptake in ephemeral rainstorms 964 (Whitford 2002). Dodd et al. (1984) surveyed 43 woody species from the open woodland in SW 965 Australia, and Timberlake and Calvert (1993) 96 shrubs and trees of Zimbabwe woodlands, both 966 finding that there were indeed species with consistently shallow systems and others with deep 967 taproots. Most species had both lateral superficial roots and descending taproots, but herbs can be 968 shallow-rooted and in herbaceous or mixed herbaceous/woody communities there can be 969 considerable stratification (Weaver and Clements 1929 %p213; Cody 1986 %381). Wilson & Agnew, chapter 1, Plants, page 34 of 34 970 5 Conclusion This opening chapter has described the basic material of plant communities: the plants 971 972 themselves. We argued that the characteristics of plants are that they are colonies of modules that 973 must constantly be replaced causing movement in space, they are potentially genetic mosaics and 974 they are plastic. As a consequence, the term ‘individual’ usually has no meaning for plants. [Is 975 this true? In many cases individuals are discrete and meaningful ] They also have features that 976 make their evolution different from that of animals. The huge majority of stands have more than 977 one plant species, and we have outlined the basic processes through which multi-species 978 communities establish and develop. The occupancy of a niche is axiomatic in a species’ presence 979 in a community, and the interactions between pool, dispersal and niche are all important in this 980 process (Fig. 1.2). There are recent demonstrations of this at a large scale in the Palm floras of 981 Ecuador and Peru (Vormisto et al. 2004), and at an intermediate scale in the Netherlands (Ozinga 982 et al. 2005). Our conclusion, which we hope the reader shares, is that there is enormous 983 complexity in the life of plants in spite of the simplicity implied in their common sedentary habit 984 and modular structure, and their almost universal trophic function. Our basic concern in this book 985 is to examine how the species fit together to form communities and basic concepts for this are the 986 niche and the guild. In the next chapter we examine the processes involved when one species 987 interacts with another, starting community development. 988 ILLUSTRATIONS 989 Fig. 1.1: The plant as a colony of active and inactive apices 990 Fig. 1.2: Pathways from the species pool to community entry 991 Fig. 1.3: Disjunct distribution of Kelleria laxa in South Island, New Zealand, interpreted as an 992 originally contiguous distribution torn apart by movement along the Alpine Fault 2-10 993 million years ago. From Heads (1989). 994 Fig. 1.4: Infiltration invasion by Olearia lyallii in the Auckland Islands. After Lee et al. (1991). 995 Fig. 1.5: The leptokurtic curve of dispersal: Juncus effusus seed rain around a single plant. 996 Fig. 1.6: Stratification: profile of a rain forest in Guiana: From Richards (1964). i ‘sessile’ in zoological terminology