W. M. Lonsdale

GLOBAL PATTERNS OF PLANT INVASIONS AND THE CONCEPT OF INVASIBILITY

With a simple model, I show that comparisons of invasibility between regions are impossible to make unless one can control for all of the variables besides invasibility that influence exotic richness, including the rates of immigration of species and the characteristics of the invading species themselves. Using data from the literature for 184 sites around the world, I found that nature reserves had one-half of the exotic fraction of sites outside reserves, and island sites had nearly three times the exotic fraction of mainland sites. However, the exotic fraction and the number of exotics were also dependent on site area, and this had to be taken into account to make valid comparisons between sites. The number of native species was used as a surrogate for site area and habitat diversity. Nearly 70% of the variation in the number of exotic species was accounted for by a multiple regression containing the following predictors: the number of native species, whether the site was an island or on the mainland, and whether or not it was a nature reserve. After controlling for scale, there were significant differences among biomes, but not continents, in their level of invasion. Multiple biome regions and temperate agricultural or urban sites were among the most invaded biomes, and deserts and savannas were among the least. However, there was considerable within-group variation in the mean degree of invasion. Scale-controlled analysis also showed that the New World is significantly more invaded than the Old World, but only when site native richness (probably a surrogate for habitat diversity) is factored out. Contrary to expectation, communities richer in native species had more, not fewer, exotics. For mainland sites, the degree of invasion increased with latitude, but there was no such relationship for islands. Although islands are more invaded than mainland sites, this is apparently not because of low native species richness, as the islands in this data set were no less rich in native species than were mainland sites of similar area. The number of exotic species in nature reserves increases with the number of visitors. However, it is difficult to draw conclusions about relative invasibility, invasion potential, or the roles of dispersal and disturbance from any of these results. Most of the observed patterns here and in the literature could potentially be explained by differences between regions in species properties, ecosystem properties, or propagule pressure.

Alien vegetation and native biota in tropical Australia: the impact of Mimosa pigra

Abstract In the coastal areas of northern Australia, the introduced shrub Mimosa pigra has been spreading rapidly in native ecosystems during the past decade. An almost monospecific tall shrubland replaces sedgeland initially, then riparian, aquatic, paperback and monsoon forest communities. The flora and fauna of two areas, one infested for about five years and the other for three, were compared with uninfested areas nearby. Although small mammals were more abundant at least in the short term, and frogs seemed little affected, many birds and lizards were lower in abundance. Furthermore, it is likely that the advantage of the new habitat for small mammals is in the form of shelter from avian predation and that this would disappear as the shrubland spreads and removes the foods resources of nearby native habitats. A massive loss of animal and plant species may well occur if the spread of this aggressive weed is not halted.

The self-thinning rule: dead or alive?

The self-thinning rule predicts that for a crowded, even-aged plant popu- lation, a log-log plot of total plant mass against plant density will give a straight line of slope - x. It has been described as one of the more general principles of plant population biology, but the evidence supporting it has recently come under close critical scrutiny. Recent reevaluations have concluded that the slope is much more variable than previous authors have claimed, that straight lines are the exception rather than the rule, and that the slope varies with aspects of the biology of the plant. Using a range of statistical tech- niques, I examined the published evidence to test the strength of these conclusions. Much of the recently reported variability in slope has resulted from the inclusion of inappropriate data sets: not all populations for which biomass and density data are available are undergoing self-thinning. I also found that there was no evidence for relationships between shade tolerance or taxonomic groups and the slope, and that the reported rela- tionships between the slope and various allometric growth constants, though real, were weak. The combined data for all populations are not consistent with an interspecific rela- tionship of slope - x, the slope being somewhat shallower, at -0.379, possibly because only the stem mass of trees is generally measured. I conclude that there is no evidence at present for a - % power rule of self-thinning, but that final rejection of the idea that there is an ideal slope (which may or may not be - x), awaits experiments in which resource levels are carefully controlled.

When to Ignore Advice: Invasion Predictions and Decision Theory

Organisms generally become pests at a low rate. As a consequence of this low ‘base-rate probability’, the large majority of organisms rejected in any random sample of potential introductions would probably be harmless, despite the fairly high accuracy of some recently proposed risk assessment systems for exotic introductions. Here we distinguish between a systems accuracy (the proportion of a group of known pest species that would be correctly identified as pests) and reliability (the rate of false positives and false negatives produced once the base-rate is taken into account). We next adapt a decision theory analysis of earthquake prediction to explore when we would be best advised to ignore the recommendations of a screening system for exotic introductions. In one scenario, we show that a pest risk assessment system with an accuracy of 85% would be better ignored, unless the damage caused by introducing a pest is eight times or more that caused by not introducing a harmless organism that is potentially useful. Furthermore, because of the base-rate effect, in certain situations it may be more efficient to focus on identifying potential invaders from amongst already naturalized species than from amongst species at the importation stage.

Tourist vehicles as vectors of weed seeds in Kakadu National Park, Northern Australia

Abstract A total of 1960 seeds was collected from 304 tourist vehicles entering Kakadu National Park at monthly intervals between May 1989 and May 1990. Individual cars carried up to 789 seeds and 15 species, but most cars carried one or no seeds. The proportion of cars carrying seeds, and the total number of seeds entering per month, did not vary strongly with seasons, despite the fall in numbers of cars entering during the wet season. The numbers of seeds, and number of occurrences, of different weed species on the cars were unrelated to the abundance of the weeds in the Park found previously. However, those weed species that were found on tourist cars occurred at three times as many sites in the Park as those that were not, suggesting that movement of seeds by tourist cars may be partly responsible for weed infestations. Ten species of known tropical weeds were found amongst the samples, including Pennisetum polystachion, Sida acuta and Hyptis suaveolens , as well as 14 species not known in the Park. Propagules of the major tropical weeds, Mimosa pigra and Salvinia molesta , were not found amongst the samples. In view of the low density of weed seeds on tourist cars, it is recommended that resources are best spent on detecting and eradicating existing weed infestations, rather than on attempting to prevent this form of seed movement.

PLANT GEOMETRY AND SELF-THINNING

SUMMARY (1) Monocultures of Agrostemma githago, Cichorium endivium and Festuca pratensis were grown at high densities to investigate the effects of plant geometry on self-thinning. Whilst populations of all three species followed the -3/2 power law when weight was plotted against density on logarithmic scales the intercepts were different and fell in the order Festuca > Agrostemma > Cichorium. These differences are related to plant geometry. (2) Published reports show that coniferous trees generally have higher intercepts than deciduous trees and that grasses have higher intercepts than dicotyledons. (3) The mean volume of canopy available to each plant was also found to be related to density by the -3/2 power law. However, the intercepts for a number of species on a volume-density plot on logarithmic scales all fell within one order of magnitude in comparison with two orders of magnitude on a weight-density plot. (4) The biomass which could be packed into a given volume of space generally increased with canopy height, but the range was nevertheless relatively narrow (1 .5-5.2 kg m-3) for the three species. Only part of the variation in intercept between species on a weight-density plot could be related to differences in biomass packing. (5) Variation in the biomass per unit volume through time means that the gradient of the thinning line on a weight-density plot may be different to that on a volume plot. It is suggested that the -3/2 power law might be better stated in terms of canopy volume rather than weight.

A neighbourhood approach to self-thinning

SummaryThe fate of individual plants of Helianthus annuus was monitored in populations sown at a density of 650 seeds m-2. Ninety six percent of the seeds germinated but only fifty two percent of the seedlings survived to the end of the experiment due to self-thinning. Whilst the seedlings emerged over only four days the cotyledons then took up to eight days to emerge from the seed case and open fully. The survival of individuals depended primarily on cotyledon-opening time but plants which occupied very small areas (<900 mm2) at germination were also less likely to survice. The increased likelihood of mortality amongst the late germinating individuals was probably as a result of shade from older, larger neighbours. It is suggested that one-sided interference for light during self-thinning will increase the importance of early establishment relative to neighbour effects in determining the survival of plants except at very high levels of crowding. The complex of environmental and genetic factors acting to maximise the growth rate of the very young seedling relative to its neighbours will clearly play an important part in determining the position of a plant in the size hierarchy of a population.

Herbivory by insects in Australian tropical savannas: a review

Herbivorous insects are undoubtedly important in major grazing animals. Folivory and pre-dispersal seed pre- savanna ecosystems, but have been largely ignored in stud- dation by insects are extremely poorly known in Australian ies of herbivory in favour of native ungulates and domestic savannas, although the results of work on southern species cattle. In Australia, where the native mammalian herbivore of Eucalyptus and Acacia (the dominant genera of woody fauna is depauperate, attention has strongly focused on plants throughout Australia) are likely to be at least partly cattle production. In this review we consider three major relevant. Harvester ants are the most important post- classes of herbivorous insects, namely grazers, folivores dispersal seed predators: they consist primarily of omnivo- and seed predators, and synthesize informnation on (1) their rous species of Monomorium and Pheidole, but also include composition, diversity and abundance, (2) their ecological an endemic radiation of granivorous Meranoplus. The over- effects as herbivores, (3) their importance relative to that of all composition and abundance of herbivorous insects in herbivorous mammals, and (4) insect herbivory in Australia Australian savannas appears similar to that in other savan- compared with that in savannas elsewhere in the world. The nas. However, their ecological effects as herbivores are al- most important grazing insects are grasshoppers and har- most totally unknown, and ought to be a priority area for vester termites, although the latter are probably mostly det- future savanna research.

Quantifying uncertainty in predictions of invasiveness

Using the Australian weed risk assessment (WRA) model as an example, we applied a combination of bootstrapping and Bayesian techniques as a means for explicitly estimating the posterior probability of weediness as a function of an import risk assessment model screening score. Our approach provides estimates of uncertainty around model predictions, after correcting for verification bias arising from the original training dataset having a higher proportion of weed species than would be the norm, and incorporates uncertainty in current knowledge of the prior (base-rate) probability of weediness. The results confirm the high sensitivity of the posterior probability of weediness to the base-rate probability of weediness of plants proposed for importation, and demonstrate how uncertainty in this base-rate probability manifests itself in uncertainty surrounding predicted probabilities of weediness. This quantitative estimate of the weediness probability posed by taxa classified using the WRA model, including estimates of uncertainty around this probability for a given WRA score, would enable bio-economic modelling to contribute to the decision process, should this avenue be pursued. Regardless of whether or not this avenue is explored, the explicit estimates of uncertainty around weed classifications will enable managers to make better informed decisions regarding risk. When viewed in terms of likelihood of weed introduction, the current WRA model outcomes of ‘accept’, ‘further evaluate’, or ‘reject’, whilst not always accurate in terms of weed classification, appear consistent with a high expected cost of mistakenly introducing a weed. The methods presented have wider application to the quantitative prediction of invasive species for situations where the base-rate probability of invasiveness is subject to uncertainty, and the accuracy of the screening test imperfect

Worth the risk? Introduction of legumes can cause more harm than good: an Australian perspective

Weeds are serious threats to Australias primary production and biodiversity conservation. For example, a recent Australia Bureau of Statistics survey found that 47% of farmers across Australia have a significant weed problem. A literature review revealed that legumes represent a significant proportion of the national weed problem and most serious Australian legume weeds are exotic thicket-forming species that were deliberately introduced for their perceived beneficial properties, such as for shade and fodder, or even quite trivial reasons, such as garden ornamentals. The low economic value of the rangelands most of these species infest, compared with control costs, hinders chemical and mechanical control of these weeds, such that biological control, which takes time, is expensive to implement and has no guarantee of success, may represent the only economically viable alternative to abandoning vast tracts of land. We argue that, because the behaviour of an introduced species in a novel environment is so hard to forecast, better predictive techniques should be developed prior to further introductions of plant species into novel environments. We also discuss the potential of legumes currently being promoted in Australia to become weeds and suggest the recent trend of exporting Australian Acacia spp. to semiarid regions of Africa risks history repeating itself and the development of new weed problems that mirror those posed by Australian Acacia spp. in southern Africa.