Invasions

Nordic Society Oikos

InvasionsAuthor(s): Mark WilliamsonSource: Ecography, Vol. 22, No. 1 (Feb., 1999), pp. 5-12Published by: Wiley on behalf of Nordic Society OikosStable URL: http://www.jstor.org/stable/3683202 .

Accessed: 03/07/2014 14:22

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Wiley and Nordic Society Oikos are collaborating with JSTOR to digitize, preserve and extend access toEcography.

http://www.jstor.org

This content downloaded from 81.243.89.192 on Thu, 3 Jul 2014 14:22:19 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=black

http://www.jstor.org/action/showPublisher?publisherCode=oikos

http://www.jstor.org/stable/3683202?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


ECOGRAPHY 22: 5-12. Copenhagen 1999

Invasions

Mark Williamson

Ja

F

o s .

Williamson, M. 1999. Invasions. - Ecography 22: 5-12.

Because of the damage some do, including threatening biodiversity, invasive species are a current focus of interest, particularly outside Europe. Yet invasion biology is still struggling to become analytic. Progress has been made in quantifying the proportion of invaders that succeed in various ways, and in measuring the degree of that success. Several examples, where the course of invasion has been explained, show the variation in biological processes involved. A detailed analysis, often with experiments and modelling, is necessary for a satisfactory explanation. Prediction is harder than explanation. Attempts at predicting invasions have generally been unsatisfac- tory. Ten reasons that may contribute to this are discussed. Progress will require defining more precisely what is to be predicted and measuring more quantitatively the ecological properties of species. Even so, predicting the ecological behaviour of a species in a new environment may be effectively impossible.

M. Williamson, Dept of Biology, Univ. of York, York, U.K. YOIO 5DD ([email protected]).

Where are we and how did we get here?

Invasions are the subject of much interest right now, largely because of the damage done by some invasive species. Although there are such invaders in Europe, the impact of invaders is more conspicuous elsewhere. Examples such as the rabbit Oryctolagus cuniculus (particularly in Australia), the brown tree snake Boiga irregularis on Guam, Zebra mussel Dreissena polymor- pha in North America, prickly pear cacti Opuntia spp. in Africa and Australia are all very familiar (Williamson 1996). The South American tree Miconia calvescens, which has swamped much of Tahiti's native vegetation (Meyer and Florence 1996) and is established and a serious threat in Hawaii, is an example of the many invaders of high impact that are less well known. Mack (1996) has attempted to map those parts of the world where invasive plant species now dominate the landscape. The largest areas are in North and South America and Australia, with other important areas in

Africa, India and on various islands. Invaders are often claimed to be the second largest threat (globally), after habitat destruction, to biodiversity (Glowka et al.

1994). So it is not surprising that the Convention on Biodi-

versity (CBD) exhorts the Contracting Parties to "pre- vent the introduction, control or eradicate those alien

species which threaten ecosystems, habitats or species" (Glowka et al. 1994). To help this, SCOPE, the Scien- tific Committee on Problems of the Environment, has launched GISP (Dean 1998), the Global Invasive Spe- cies Programme, with support from the United Nations (GEF, the Global Environmental Facility, and UNEP, the United Nations Environment Programme), IUCN (the World Conservation Union) and others.

There is a problem with the CBD exhortation. Many of the species introduced with agriculture and forestry are beneficial and desirable, while many species introduced accidentally have negligible effects. Identifying

This is an invited Minireview on the occasion of the 50th anniversary of the Nordic Ecological Society Oikos.

Copyright ? ECOGRAPHY 1999 ISSN 0906-7590 Printed in Ireland - all rights reserved

ECOGRAPHY 22:1 (1999) 5



the species already present that should be controlled or eradicated is relatively straightforward; identifying those that should not be introduced is much harder. This is the problem of prediction, a problem that shows up the limitations of our ecological understanding, a problem I discuss in detail below.

It is often said that invasions are less of a problem in Europe (and China) because agriculture started there, and many, but far from all, invasive problems follow the introduction of European agriculture. This is particularly true for plants; animals pests, particularly mammals such as rats and mice, while often associated with

agriculture do not generally start from there. What is undoubtedly true is that invasions are seen as much more of a threat in the US than in Europe. Invaders are indeed the second threat to imperilled species in the US (Wilcove et al. 1998). In Europe pollution, agricultural practices, urbanisation and other threats are probably at least as important. While Edwards (1998) urges proper planning before the management of any invasive species in Europe, Holt (1998) states bluntly that "Alien species invasion is the engine driving Hawaii's extinction crisis". The President of the United States has established a high-level council to coordinate efforts against non-native plants and animals in his nation

(Anon. 1998). The US Congress produced an excellent report on terrestrial and freshwater invasions (Anon. 1993). The report resulted from the sudden and major impact of zebra mussel from just before 1990; the loss in the US from that invasion by 1991 was estimated at over three billion dollars. This is indeed different from Europe, where the impact of this invader is negligible (Williamson 1996). Land managers in the US are nota- bly conscious of the need to control invasive species (Randall et al. 1996, Hiebert 1997, Randall unpubl.). In

England, in contrast, managers tackle without distinction native species, e.g. bracken Pteridium aquilinum, ancient invaders e.g. ground elder Aegopodium poda- graria and modern (post-mediaeval) ones e.g. Rhodo- dendron ponticum (Williamson 1998a). But whatever the local perception, invasions are, almost everywhere, an increasing problem.

Although there had been many studies of invaders earlier, invasion biology emerged as an important and recognisable part of ecology as a result of Elton's (1958) book. Elton was always surprised at the academic impact because the book arose from a set of

semi-popular broadcast lectures. It shows the time was ripe. Since then there has been the 1980's SCOPE programme on the Ecology of Biological Invasions, which produced 15 sets of papers (listed in Williamson 1996, synthesis in Drake et al. 1989), the US Congress report (Anon. 1993), a series of international meetings on plant invasions (the latest is reported in Starfinger et al. 1998), various other symposia (e.g. Kareiva 1996, Carey et al. 1996, Sandlund et al. 1996), collective works concentrating on management (Luken and

Thieret 1997, Simberloff et al. 1997), two complemen- tary advanced texts (Williamson 1996, Shigesada and Kawasaki 1997), and two journals have been launched (Richardson 1998, Carlton 1999). Py'ek (1995) shows that ca 100 papers a year are being published in this field. Invasion biology is an active and growing discipline.

Where are we going? From that quick overview, two points emerge. The first is that invasions are important to many managers, and that the science of managing known problem invaders is advancing satisfactorily. One of the aims of GISP is to systematise advice on dealing with invasive species. The second is that the academic study of invasions is still struggling to become quantitative, analytic and systematic (Parker and Reichard 1998). As I said (Williamson 1996) much published work is "case histories, surveys of case histories and statistically weak generalisations". To examine how the situation could be improved, I will divide the study of invasions into five approaches (Table 1).

By quantification, I mean here the statistical study of the frequency and success of invasions. The tens rule (Williamson and Fitter 1996a) divides invasions into arbitrary stages of imported, casual, established and pest and says that roughly 10% (between 5 and 20%) progress from one stage to the next. It is based on empirical studies of British plants (Williamson 1993), but emphasises the need for clear definitions and realis- tic confidence limits. Variations and exceptions show the need to allow for different types of organisms in different places and for propagule pressure. Rejminek (1996a) and Lonsdale (in press) make more wide rang- ing comparative studies. Despite the possibility of inconsistent standards of description, they are able to show, contrary to standard dogma, that species richness, generally found in tropical areas, is associated with more invasion not less. In contrast, fitting dogma, islands are more invasible, but not because of impover-

Table 1. Different approaches.

Approach Methodology Information relevant

Quantification overall success locality/habitat/niche rates

Measurement success in differ- range/abundance/ef- ent ways fect per unit

Management perception and cost effectiveness experience

Explanation e.g. ANOVA % variance explained Prediction risk analysis see below

Risk analysis requires:

1. hazard identification 2. probability estimation 3. accounting for perception

6 ECOGRAPHY 22:1 (1999)



ishment. It is good to find that nature reserves are less invaded. In this work it is important to allow for area:

species-area effects pervade ecology. Lonsdale suggests that it is quite likely that most of the variation in the rate of species establishment relates to species properties, ecosystem properties and propagule pressure but the information is not there, as yet, to formulate that mathematically. Unlike the tens rule, Rejminek's and Lonsdale's work is based only on the invaders that have succeeded.

Quantification explains nothing. It orders the empirical data to allow explanation, to challenge prediction.

Measurement is the term I use for a different sort of

quantification, the success in various ways of different invaders (Starfinger 1998). It is, if you like, the next level of description. Going beyond asking what proportion of invaders establish, measurement quantifies, for instance, their variation in abundance or in distribution. Measurement of different impacts (Parker et al. unpubl.) highlights the crudity of the pest part of the tens rule; different aspects of being a pest can be expected to occur with different frequencies. Measure- ment of the extent of establishment in different types of habitats, particularly habitats with different degrees of naturalness, extends a different part of the tens rule. Measurement of the different impacts of plants and mammals in Britain (Williamson 1998a, b) shows that each has a different spectrum, that there is generally a positive correlation of impacts, and that they can be ordinated in a series from purely agricultural to purely ecological. The impacts studied were agricultural weediness, control costs, perceptions of pestiness (Perrins et al. 1992a), abundance and distribution.

Management, too, is concerned with impact. Two schemes bring out a further range of impacts that should be considered. Hiebert's (1997) scheme, soon to be available on the world wide web, is designed for the US National Park system; Randall et al.'s (1996), fully in Randall unpubl.) for the non-governmental US Na- ture Conservancy. Both are based on observation of the present status of invasive plants, and so are in no sense predictive. They seek to prioritise targets for control, seeking to maximise cost-effectiveness rather than min- imising total cost. So, beyond simple abundance and distribution, they are concerned with effects on ecosystem processes, impact on community composition and the naturalness and conservation significance of the vegetation invaded and the present feasibility of control. The schemes are new, and measurements of these impacts have not been published.

Explanation

Many people have tried to explain why some species are successful invaders while others are not. The 1980's

Table 2. Correlates of invasion success.

Historical Previous success (quite good) Propagule pressure Populational Abundance (moderate) Range

Intrinsic rate of increase Absence of enemies

Individual Climatic matching (often poor) Taxonomic isolation

Anthropogenic Size and other measures Wide niche Empty niche Genetic characteristics

SCOPE programme asked the questions "What factors determine whether a species will become an invader or not?" and "What site properties determine whether an

ecological system will be prone to, or resistant to, invasions?" (Williamson 1996). These are questions asking for explanations rather than predictions, but even so the programme did not produce any agreed answers. There is, of course, a strong interaction between the two questions. For instance, those studying invasions

by land plants often stress disturbance as a factor. But

plants of disturbed ground, unless they have the potential to spread to more natural habitats, are of little concern to natural area managers. For agriculture, usually a process of disturbance, such plants constitute most weeds. The importance of disturbance and disturbed habitats varies with the type of invasive plant considered.

Table 2 gives a speculative classification of some factors that have been associated with invasion success. Some justification of the classification will be considered below, when considering prediction, but there is a need for better data, for quantitative assessments of the usefulness of different factors. All these factors are discussed in Williamson (1996). What impresses me is that the most useful factors, the historical ones, are not strictly biological at all. Although the distinction is fuzzy, it seems that populational characters are more easily associated with establishment than the characters of individuals. Possible reasons for this emerge from a handful of detailed quantitative studies of the population dynamics of successful invaders.

A clear demonstration of the importance of several interacting factors determining the density of a single species is the seven year study of Ohgushi and Sawada (1998) on the population dynamics of an introduced ladybird beetle Epilancha niponica feeding on a thistle Cirsium nipponicum. The introduced population, in a botanic garden, was a mere 10 km beyond the native range, which is in forest, where other populations, feeding on C. kagamontanum, had been studied. Less arthropod predation on the introduced population leads to higher densities, defoliation by and starvation of the beetles, and density-dependent reductions of

ECOGRAPHY 22:1 (1999) 7



fecundity and survival of the beetles. The principal predators and parasitoids of the beetle were left behind in the forest. The reasons for the beetle not having reached the botanic garden unaided are apparently that it is a poor disperser (as measured experimentally) and that the 10 km gap is relatively free of thistles.

Another seven year study (Schoener and Spiller 1996) showed that lizards, Anolis sagrei, have a devastating effect on the number and diversity of their web-spider prey when introduced on small Bahamian islands. The vegetated areas of the islands were only ca 100 m'2. Here invasion success by the lizard could have been predicted for any island with sufficient prey density. The pattern across small islands seems to result from stochastic extinction of lizard populations, probably mostly in storms. Both these studies would imply that successful invasion merely requires a sufficient food supply, but neither involved competitors.

Six studies show the range of biological complication that can go with the competitive displacement of a native by an invader. Competition can be down the food chain, for shared resources, horizontally, by direct behavioural interaction between the species, or upwards, from shared predators or parasites; or by any combination of these three directions, giving seven modes in all. Upwards competition is usually known nowadays as apparent competition. The six studies give evidence for at least five of these modes.

With squirrels, Sciurus carolinensis and S. vulgaris, in Britain the critical factor seems to be the ability of the invading American S. carolinensis to digest acorns, which S. vulgaris can not do (Kenward and Holm 1993). That could be called, rather loosely, competition for food. Okubo et al. (1989) showed, in a modelling study, that the important demographic differences were in fecundity rather than survival and in the equilibrium population density. This invasion is also notable for the

disagreement, over seven decades or more, between

experts as to whether there was indeed competitive displacement (Williamson 1996). If you do not know what the phenomenon is, it is impossible to explain it, let alone model it.

Another case which now seems to be competition for food is the displacement, around houses but not in forests, of the asexual gecko Lepidodactylus lugubris by the sexual Hemidactylus frenatus (Petren and Case

1996) on oceanic islands in the Pacific. There is a behavioural component to this, an agonistic dominance by L. frenatus, that was earlier thought sufficient to explain the competition (Petren et al. 1993, Case et al. 1994). Experiments were needed to show that this was not so. Very few invasions have been studied experimentally.

The European speckled wood butterfly Parage aegeria has been on the island of Madeira for only two decades or so. It is displacing the native endemic P. xiphia, which is probably the evolved descendant of a

much earlier invasion by P. aegeria. The mechanism seems to be largely, if not entirely, by territorial fights between males (Jones et al. 1998). It is not clear if P. xiphia will survive in the indigenous laurel forest, but it seems fairly certain to die out outside it.

There are two studies of invading crayfish, where the important interaction is via predation, but modulated by direct competition. In Sweden, the American

Pacifastacus leniusculus is displacing the native Astacus astacus. Sdderback (1994) showed experimentally that P. leniusculus is the superior competitor for refuges from predation by Perca fluviatilis European perch, a fish. The anti-predator responses are similar, but A. astacus was taken more often. In Britain, this displacement has been ascribed to a plague caused by the fungus Aphanomyces astaci, and it is likely that that effect is more important in Sweden too than competition to avoid predation (Williamson 1996). In Wiscon- sin, the larger native Oronectes virilis and the smaller earlier invader 0. propinquus are being replaced by a new medium sized invader 0. rusticus (Garvey et al. 1994). The main fish predator is Micropterus salmoides, largemouth bass, which prefers small individuals but nevertheless preys most on 0. virilis. Again, it seems to be a question of direct competition for refuges.

The last study of competition involves only competition to avoid an enemy. One invading leafhopper is displacing a native on an introduced crop plant. On vines in California, Settle and Wilson (1990) showed experimentally that Erythroneura variabilis out-com- petes E. elegantula via the shared myrmarid egg para- site Anagrus epos. These six examples show competition for resources, for resources and directly, directly, via enemies and directly, by enemies alone; five modes in all.

These cases show that population analysis and, usually, experiments are necessary to explain the success of an invader. Curiously, only one of them is based on

modelling. The gecko case shows that explanations may be modified by more work. Unfortunately it is not possible from these papers to examine the factors in Table 2; there is mostly no information on previous history or abundance and range in the native habitat for instance. What all these competitive studies show is that considerable detail is often needed to explain a successful invasion. The devil is in the detail.

Prediction

Explanation is not prediction. Lawton (1996) noted this too. While explanation is usually based on statistical analysis, prediction is more akin to a risk analysis (Table 1). It is well known that social scientists and engineers differ in how to do such an analysis (Anon. 1992), but the stages listed in Table 1 combine the two

8 ECOGRAPHY 22:1 (1999)



approaches. In predicting invasion potential, it is desirable to be clear about whether establishment or impact is to be predicted and, in either case, what type. That is

equivalent to identifying hazards. The quantitative prediction corresponds to the risk probability. The third

stage varies with the problem. Engineers want to know if a bridge will stand certain conditions, social scientists are much concerned with the perception of risks, such as that from BSE (mad cow disease). In invasions, assessing predicted abundance and distribution are

largely independent of value judgements, but scoring the conservation status of habitats and the pestiness of

species are often heavily affected by perceptions. Before discussing the generalities of why prediction is

difficult in detail, consider some situations where prediction has been attempted or would have been desirable.

A major set of deliberate introductions are those for

biological control. The intention is to have an impact, normally the reduction of the population of another invader. Yet it is well known that, despite much skilled work, the rate of success, both in establishing biocontrol species and achieving an impact, is low (Sheppard 1992, Anon. 1995). Various quantifications are possible; mine (Williamson 1996) is that in comparison with the tens rule, biocontrol produces a threes rule; about a third of species establish and about a third of these have some effect. Modulations of this are needed for the apparency of the target species and for propagule pressure, but the rate of success for such expensive, scientific, introductions is an indication of how poor prediction is.

An example of failure to predict, particularly relevant to the release of genetically engineered organisms, is Southern Corn Leaf Blight (Williamson 1996, Crow 1998). This is a disease of corn or maize Zea mays caused by a fungus Bipolaris maydis, which is normally a negligible pathogen. In 1971 it caused a major epi- demic, devastating in some areas. This was because most of the corn contained a mitochondrial gene T- urfl3 for male sterility, a desirable trait for breeders. The gene produces a protein in the mitochondrial wall that is both the cause of male sterility and the target for the toxin produced by B. maydis. The genetic change, a recombination, is more restricted than those normally used in genetic engineering. The effect, a simple ines-

capable pleiotropic effect of a gene in what was an invasion of sorts, was unpredictable with our present knowledge.

Reichard and Hamilton (1997) assay how to predict the establishment of woody plants (other than conifers and those in Florida) in North America. Rejminek and colleagues (Rejminek 1995, 1996b, Rejmanek and Richardson 1996) attempt a more general prediction for all seed plants, though concentrating on pines. A simple instance of the contradictions found in predictions is that Rejminek emphasises small seed size, Reichard

and Hamilton find no effect from it. What they do find is that previous introduction (positively) and native

range, if any, in North America (negatively) are important predictive characters of establishment success, which fits Table 2. They are far more successful in

predicting invasive species, i.e. successful establishment, (97.1 and 93.8% in different analyses) than non-invasive ones (70.8 and 56.5%). Note that they use the terms naturalised and introduced for what the tens rule calls

imported, and invasive for naturalised or established, so

missing the introduced or casual stage, and that they could not identify many imported species which were listed only under their common names. The identification of failed invaders is almost always difficult.

Reichard and Hamilton (1997) report no predictors of impact and apparently found none even though 37% of their invaders are pests. An attempt to find predictors of weediness in British annual plants (Perrins et al.

1992b) was also unsuccessful. Again using a step-wise discriminant procedure and rather more systematic test-

ing than Reichard and Hamilton, they found 71, 75 and 78% success in different analyses. These percentages are too low for practical use, but more seriously there was no consistency in the characters found useful in the different analyses, though maximum seed width, seed bank type, maximum relative growth rate and the month flowering starts were picked by all three. These are all called individual characters in Table 2, and such characters typically behave inconsistently from one

analysis to another. An example of that with populational characters is Bergelson's (1994) finding that fe-

cundity was irrelevant to invasive success in an

experiment; high fecundity is often thought and some- times found (Roy 1990) to be a good predictor.

More indications of the inconsistency of individual characters come from comparing three studies contrast-

ing successful British invaders with British natives.

Crawley et al. (1996) are the only one of the three to use a phylogenetic correction. Thompson et al. (1995) find invaders (which are spreading) to be somewhat like those natives which are spreading. Williamson and Fitter (1996b) use a somewhat more comprehensive data base. But the three studies are mostly inconsistent. For instance, Crawley et al. find large seeds aid invasions, both Thompson et al. and Williamson and Fitter (like Reichard and Hamilton), that seed size is irrelevant; Rejmanek's group find small seeds invasive. There is some consistency. All three agree that trees (phanero- phytes, tall plants, erect stems) are more invasive than other plants.

Why is it so hard to predict invasions? Maybe this is part of the problem of finding general laws in ecology (Lawton 1999). So far, the only consistent predictor is success in previous invasions. Yet we know that that can fail badly. Zebra mussel is an innocuous invader in Europe, a devastating one in America. That is consis-

ECOGRAPHY 22:1 (1999) 9



tent with Reichard and Hamilton's only being able to

predict establishment. There are many reasons why prediction is harder than explanation, set out in Table 3, and with invasions even explanation has been poor and patchy. Given the variety of ways in which a species can establish (Randall et al. 1996, Hiebert 1997, Randall unpubl.) and can have an impact (Parker et al.

unpubl., Williamson 1998a) the General reason in Table 3 is important.

The Statistical reasons in Table 3 are mostly well known, if ignored. I will, briefly, discuss them in turn. A small but significant r2 is one way of finding a useful

explanation that is a poor predictor. Williamson (1993) showed that that applied to Baker characters for weeds, which is why Baker characters should not be cited in risk assessments of genetically engineered crops. Ex-

trapolation is particularly dangerous in multidimen- sional analyses, including the use of discriminant functions, and with non-linear or transformed functions (Altman and Bland 1998). Statistical shrinkage is the

phenomenon that "the fit of a regression predictor is nearly always worse than its fit to the original data" (Copas 1983). Techniques for dealing with it are still being developed (Mielke et al. 1997, Schumacher et al. 1997). Base rate effects are the production of a dispro- portionate number of false positives when the expected frequency is low (Matthews 1996, 1997, Smith et al.

unpubl.). No prediction that fails to include all the variables in the case to be predicted can hope to succeed, yet each new invader introduces its own novel set of characters.

The Biological reasons in Table 3 are also well known. The appropriate use of the phylogenetic correction is still disputed (Westoby et al. 1995) and some

techniques require the use of the nonexistent origin of a

log-log plot (Williamson 1997), but there is no dispute that some such correction should be tried when explain- ing or predicting invaders. Williamson and Brown

(1986), Daehler (1998) and Py'ek (1998) all showed that some families of plants are more invasive than others, so an analysis of species without allowing for this is

Table 3. Some reasons why prediction may fail (short explanations are in the text).

General: Target not precise enough

Statistical: significance statistical but not useful extrapolation statistical shrinkage base rate effect new variables

Biological: lack of phylogenetic correction time lags or delays non-linearity (& chaos) complexity situation specific detail

suspect. Time lags or delays (Kowarik 1995, Williamson 1996) have often been reported in invasions; some are over 100 yr. Using information from early stages will generally fail to predict later stages satisfactorily. Non-linearity is well known to produce unexpected consequences in population dynamics, even if chaos in population size is a rare phenomenon (Ellner and Turchin 1995). Perhaps the complexity (Bar-Yam 1997) of ecosystems, shown in the competition studies considered above, is more important. It may be that complex systems behave counter-intuitively more often than non-linear ones. Contrariwise, it could be said that those competition studies show more clearly the importance of species specific detail, of finding the devil in the detail. As Lawton (1999) says "if we want to predict in detail the population dynamics ... then there is no alter- native but to study the species in detail".

Conclusion It could be that invasions just are unpredictable in the way that earthquakes are. Much money has been spent on trying to predict earthquakes; their geophysics is better understood than the biology of invasions; yet it is now agreed that individual earthquakes are intrinsically unpredictable (Matthews 1997, Geller et al. 1997). Nev- ertheless, better explanation and understanding of invasions is certainly possible. There have been far too few studies of the population dynamics of invaders, too few estimates of their demographic parameters, too little experimentally manipulation of invasive situations. There is still a need for better and wider quantification and measurement of what invasive species do. In the context of preserving biodiversity, such studies are ur- gently needed. It might even be found that some invasions actually are predictable.

Acknowledgements - I am grateful to Alastair Fitter, Richard Law and Carey Smith for their helpful comments, to Jeb Byers and Elizabeth Chornesky for information and to Ingrid Parker and all participants of the NCEAS Invasion Biology meetings for stimulus.

References Altman, D. G. and Bland, J. M. 1998. Generalisation and

extrapolation. - BMJ 317: 409-410. Anon. 1992. Risk: analysis, perception and management. -

Royal Society, London. Anon. 1993. U.S. Congress, Office of Technology Assessment,

[US OTA]. Harmful non-indigenous species in the United States, OTA-F-565. - U.S. Gov. Printing Off.

Anon. 1995. U.S. Congress, Office of Technology Assessment, [US OTA]. Biologiaclly based technologies for pest control, OTA-ENV-363. - U.S. Gov. Printing Off.

Anon. 1998. War declared on aliens. - Science 281: 761. Bar-Yam, Y. 1997. Dynamics of complex systems. - Addison-

Wesley, Reading, Mass.

10 ECOGRAPHY 22:1 (1999)



Bergelson, J. 1994. Changes in fecundity do not predict invasiveness: a model study of transgenic plants. - Ecology 75: 249-252.

Carey, J. R. et al. (eds) 1996. Invasion biology. - Biol. Conserv. 78: 1-214.

Carlton, J. T. (ed.) 1999. Biological invasions [vol. 1]. - Kluwer, Dordrecht.

Case, T. J., Bolger, D. T. and Petren, K. 1994. Invasions and competitive displacement among house geckos in the tropical Pacific. - Ecology 75: 464-477.

Copas, J. B. 1983. Regression, prediction and shrinkage. - J. Roy. Stat. Soc. B 45: 311-354.

Crawley, M. J., Harvey, P. H. and Purvis, A. 1996. Compara- tive ecology of the native and alien floras of the British Isles. - Philos. Trans. R. Soc. Lond. B. 351: 1251-1259.

Crow, J. F. 1998. 90 years ago: the beginning of hybrid maize. - Genetics 148: 923-928.

Daehler, C. 1998. The taxonomic distribution of invasive angiosperm plants: ecological insights and comparison to agricultural weeds. - Biol. Conserv. 84: 167-180.

Dean, W. R. J. 1998. Space invaders: modelling the distribution, impacts and control of alien organisms. - Trends Ecol. Evol. 13: 256-258.

Drake, J. A. et al. 1989. Biological invasions: a global perspec- tive. SCOPE 37. - John Wiley.

Edwards, K. R. 1998. A critique of the general approach to invasive plant species. - In: Starfinger, U. et al. (eds), Plant invasions: ecological consequences and human responses. Backhuys, Leiden, pp. 85-94.

Ellner, S. and Turchin, P. 1995. Chaos in a noisy world - new methods and evidence from time-series analysis. - Am. Nat. 145: 343-375.

Elton, C. S. 1958. The ecology of invasions by animals and plants. - Methuen, London.

Garvey, J. E., Stein, R. A. and Thomas, H. M. 1994. Assessing how fish predation and interspecific prey competition influ- ence a crayfish assemblage. - Ecology 75: 532-547.

Geller, R. J. et al. 1997. Earthquakes cannot be predicted. - Science 275: 1616-1617.

Glowka, L., Burhenne-Guilmin, F. and Synge, H. 1994. A guide to the convention on biological diversity. - IUCN, Gland.

Hiebert, R. D. 1997. Prioritizing invasive plants and planning for management. - In: Luken, J. 0. and Thieret, J. W. (eds), Assessment and management of plant invasions. Springer, pp. 195-212.

Holt, A. 1998. A conservation vision for the 21st century. - The Nature Conservancy of Hawaii Newsletter 17: 4-5.

Kareiva, P. (ed.) 1996. Developing a predictive ecology for non-indigenous species and ecological invasions. - Ecology 77: 1651-1697.

Kenward, R. E. and Holm, J. L. 1993. On the replacement of the red squirrel in Britain: a phytotoxic explanation. - Proc. R. Soc. Lond. B. 251: 187-194.

Kowarik, I. 1995. Time lags in biological invasions with regard to the success and failure of alien species. - In: Py'ek, P. et al. (eds), Plant invasions: general aspects and special problems. SPB Academic Publ., Amsterdam, pp. 15-38.

Jones, M. J. et al. 1998. Territorial behaviour in the speckled wood butterflies Pararge xiphia and P. aegeria of Madeira: a mechanism for interspecific competition. - Ecography 21: 297-305.

Lawton, J. 1996. Corncrake pie and prediction in ecology. - Oikos 76: 3-4.

Lawton, J. 1999. Are there general laws in ecology? - Oikos 84: 177-192.

Lonsdale, W. M. in press. Global patterns of plant invasions and the concept of invasibility. - Ecology.

Luken, J. O. and Thieret, J. W. 1997. Assessment and management of plant invasions. - Springer.

Mack, R. N. 1996. Plant invasions: early and continuing expressions of global change. - In: Huntley, B. et al. (eds), Past and future rapid environmental changes: the spatial

and evolutionary responses of terrestrial biota. Springer, pp. 205-216.

Matthews, R. A. J. 1996. Base-rate errors and rain forecasts. - Nature 382: 766.

Matthews, R. A. J. 1997. Decision-theoretic limits on earth- quake prediction. - Geophysics J. Int. 131: 526-529.

Meyer, J.-Y. and Florence, J. 1996. Tahitis's native flora endangered by the invasion of Miconia calvescens DC. (Melastomaceae). - J. Biogeogr. 23: 775-781.

Mielke, P. W. et al. 1997. A single-sample estimate of shrinkage in meteorological forecasting. - Weather Forecast. 12: 847-858.

Ohgushi, T. and Sawada, H. 1998. What changed the demog- raphy of an introduced population of an herbivorous lady beetle? - J. Anim. Ecol. 67: 679-688.

Okubo, A. et al. 1989. On the spatial spread of the grey squirrel in Britain. - Proc. R. Soc. Lond. B. 238: 113-125.

Parker, I. M. and Reichard, S. H. 1998. Critical issues in invasion biology for conservation science. - In: Fiedler, P. L. and Kareiva, P. M. (eds), Conservation biology for the coming decade, 2nd ed. Chapman and Hall, pp. 283-305.

Perrins, J., Williamson, M. and Fitter, A. 1992a. A survey of differing views of weed classification: implications for regu- lation of introductions. - Biol. Conserv. 60: 47-56.

Perrins, J., Williamson, M. and Fitter, A. 1992b. Do annual weeds have predictable characters? - Acta Oecol. 13: 517- 533.

Petren, K. and Case, T. J. 1996. An experimental demonstration of exploitation competition in an ongoing invasion. - Ecology 77: 118-132.

Petren, K., Bolger, D. T. and Case, T. J. 1993. Mechanisms in the competitive success of an invading sexual gecko over an asexual native. - Science 259: 354-357.

Py7ek, P. 1995. Recent trends in studies on plant invasions (1974-1993). - In: Pyiek, P. et al. (eds), Plant invasions: general aspects and special problems. SPB Academic Publ., Amsterdam, pp. 223-236.

Pyiek, P. 1998. Is there a taxonomic pattern to plant invasions? - Oikos 82: 282-294.

Randall, J. M. et al. 1996. Criteria for ranking invasive natural area weeds objectively. - Abstract, Natural Areas Assoc. meeting, St Charles, Illinois.

Reichard, S. H. and Hamilton, C. W. 1997. Predicting invasions of woody plants introduced into North America. - Conserv. Biol. 11: 193-203.

Rejmanek, M. 1995. What makes a species invasive? - In: Py'ek, P. et al. (eds), Plant invasions - general aspects and special problems. SPB Academic Publ., Amsterdam, pp. 3-13.

Rejmfinek, M. 1996a. Species richness and resistance to invasions. - In: Orians, G. H., Dirzo, R. and Cushman, J. H. (eds), Diversity and processes in tropical forest ecosystems., Springer, pp. 153-172.

Rejmainek, M. 1996b. A theory of seed plant invasiveness: the first sketch. - Biol. Conserv. 78: 171-181.

Rejmainek, M. and Richardson, D. M. 1996. What attributes make some plant species more invasive? - Ecology 77: 1655-1660.

Richardson, D. M. (ed.) 1998. Diversity and distributions [vol. 4, previous volumes called Biodiversity Letters]. - Blackwell.

Roy, J. 1990. In search of the characteristics of plant invaders. - In: di Castri, F., Hansen, A. J. and Debussche, M. (eds), Biological invasions in Europe and the Mediterranean basin. Kluwer, pp. 335-352.

Sandlund, O. T., Schei, P. J. and Viken, A. (eds) 1996. Proceedings of the Norway/UN conference on alien species. - Dir. Nat. Manage. Norwegian Inst. Nat. Res., Trondheim.

Schoener, T. W. and Spiller, D. A. 1996. Devastation of prey diversity by experimentally introduced predators in the field. - Nature 381: 691-694.

ECOGRAPHY 22:1 (1999) 11



Schumacher, M., Hollander, M. and Sauerbrei, W. 1997. Resampling and cross-validation techniques: a tool to re- duce bias caused by model building? - Stat. Med. 16: 2813-2827.

Settle, W. H. and Wilson, L. T. 1990. Invasion by the varie- gated leafhopper and biotic interactions: parasitism, competition, and apparent competition. - Ecology 71: 1461-1470.

Sheppard, A. W. 1992. Predicting biological weed control. - Trends Ecol. Evol. 7: 290-291.

Shigesada, N. and Kawasaki, K. 1997. Biological invasions: theory and practice. - Oxford Univ. Press.

Simberloff, D., Schmitz, D. C. and Brown, T. C. (eds) 1997. Strangers in paradise: impact and management of non-indigenous species in Florida. - Island Press, Washington. D.C.

Sdderbick, B. 1994. Interactions among juveniles of two freshwater crayfish species and a predatory fish. - Oecologia 100: 229-235.

Starfinger, U. 1998. The success of plant invasions. - In: Starfinger, U. et al. (eds), Plant invasions: ecological consequences and human responses. Backhuys, Leiden, pp. 33- 42.

Starfinger, U. et al. (eds) 1998. Plant invasions: ecological consequences and human responses. - Backhuys, Leiden.

Thompson, K., Hodgson, J. G. and Rich, T. C. G. 1995. Native and invasive plants: more of the same? - Ecogra- phy 18: 390-402.

Westoby, M., Leishman, M. and Lord, J. 1995. Issues of interpretation after relating comparative datasets to phy- logeny. - J. Ecol. 83: 892-893.

Wilcove, D. S. et al. 1998. Quantifying threats to imperiled species in the United States. - Bioscience 48: 607-615.

Williamson, M. 1993. Invaders, weeds and the risks from genetically modified organisms. - Experientia 49: 219-224.

Williamson, M. 1996. Biological invasions. - Chapman and Hall.

Williamson, M. 1997. The biology of rarity. - J. Ecol. 85: 552. Williamson, M. 1998a. Measuring the impact of plant invaders

in Britain. - In: Starfinger, U. et al. (eds), Plant invasions: ecological consequences and human responses. Backhuys, Leiden, pp. 57-68.

Williamson, M. 1998b. Which mammals are pests? - Imprint 25: 26-30.

Williamson, M. and Brown, K. 1986. The analysis and modelling of British invasions. - Philos. Trans. R. Soc. B. 314: 505-522.

Williamson, M. and Fitter, A. 1996a. The varying success of invaders. - Ecology 77: 1661-1666.

Williamson, M. and Fitter, A. 1996b. The characters of successful invaders. - Biol. Conserv. 78: 163-170.

12 ECOGRAPHY 22:1 (1999)



Date post:	27-Jan-2017
Category:	Documents
Upload:	mark-williamson
View:	212 times
Download:	0 times

Invasions

Documents