Not all of the many hundreds of species designated ‘invasive’ have detrimental impacts on their adopted environments, but a lot of them do. Now the unintended consequences of globalisation are exacerbating the problem. Can information and communications technology help track the challenges?
With its heart-shaped leaves and fragrant white flowers, the Mikania micrantha vine that’s so widespread across Hong Kong’s rolling country parks appears at first sight to be an inoffensive – even attractive – garden creeper.
But the South American native has become one of the city’s most noxious invasive plants since it was first introduced in the late 19th Century by gardeners who admired its pretty blooms. Known as ‘mile-a-minute weed’ for its prodigious growth of up to 8cm a day, the vine smothers and chokes other plants and trees, as well as releasing chemicals into the soil and air that slow the growth of neighbouring vegetation.
Mikania has been a thorn in the side of the city’s agricultural and conservation authority for several decades and it has conceded that, at least for now, it can only control its spread rather than eliminate it entirely. Mikania micrantha is just one example of the thousands of ‘invasive’ or ‘alien’ species across the world posing a huge threat to the local ecosystems which they adopt, by pushing out native species, acting as predators or pests, and spreading disease. Worse, in an early attempt to tackle the weed, Hong Kong scientists introduced rhesus macaques in the 1930s to prey on the vine; but the monkeys have since colonised swathes of countryside, becoming an invasive species themselves.
Many readers will recall media coverage of the UK’s burgeoning ring-necked parakeet roosts, now blamed for driving away indigenous birds with the volume of its massed squawking, or maybe the American red swamp crayfish reportedly “terrorising” swimmers in Hampstead Heath’s ponds.
The problem of invasive species is one that is growing – and giving environmental bodies and government agencies a headache. From mussels to knotweed, beetles to fish, thousands of alien species are flourishing in foreign climes at an alarming rate, causing devastating environmental, economic and social damage. Now authorities are increasingly turning to researchers to enable them to better understand and tackle the problem, and those researchers are in turn finding innovative ways to deploy information and communications technology (ICT) to meet the challenges.
It is estimated that there are some 12,000 invasive species in Europe, including microorganisms, animals and plants, that cost the economy an estimated €12bn per year. The European Environment Agency says that financial impacts are due to lost production, adverse effect on tourism and damage caused. European states desperately expend even more resources in fighting them through control and eradication programmes. “Only a proportion of alien species cause problems,” says Dr Helen Roy at the UK Natural Environment Research Council (Nerc) Centre for Ecology & Hydrology, “but those that do can cause far-reaching and costly problems.” Due to the lack of natural predators available to control the populations of new alien species, they can have a disastrous impact on crops, livestock and food production.
The USA suffers an estimated $120bn loss (at least) annually in lost crops, property damage, and environmental degradation (Cornell University). Across Europe, policy-makers, academics and research scientists have looked at solving the problem through raising awareness of the issue, introducing regulatory framework, developing technology for early detection and warning systems and improving knowledge transfer, among other objectives.
On a global scale, the Invasive Species Specialist Group (ISSG), an international network of science and policy experts, has set up the Global Invasive Species Database (GISD), an online resource that aims to act as an early detection and rapid response tool.
Invasives also affect human health, with some species carrying and transmitting potentially dangerous diseases. For example, the tiger mosquito, originally from south-east Asia but now a global presence after invading several countries, is known to carry the dangerous West Nile virus; while feral cats in Australia – recently estimated by the Commonwealth Scientific and Industrial Research Organisation to kill a massive 75 million native animals every night – can transmit rabies and hookworm.
Globalisation’s species spread
Even as governing bodies take steps to manage known invasive species, new invaders are constantly and unpredictably popping up – as in the case of the toxic Asian black-spined toad that’s currently invading Madagascar. Threatening snakes, lemurs and birds, as well as being able to transmit parasites to humans, the invasive amphibians are thought to have came ashore via shipping containers.
There are many examples of how globalisation has exacerbated the invasives problem as the boom in international trade, air transport, travel and tourism introduces increasing numbers of alien species into ecosystems on the other side of the planet.
“Global bio-geographical barriers to the dispersal of organisms, such as oceans and big mountain ranges, have been blurred by human activities,” explains Dr Belinda Gallardo at the Doñana Biological Station, part of the Spanish Council for Scientific Research in Seville. The problem is predicted to get worse, “mainly because the introduction of alien species is correlated with the amount of goods transported, which increases over time”, agrees Dr Hanno Sandvik of the Norwegian University of Science and Technology.
Climate change factors, such as warming temperatures, increased sea levels and changes in CO2 concentrations, also have an impact as they facilitate the growth and migration of invasive species. Researchers at Queen’s University Belfast (QUB), at the forefront of invasive species response projects in Northern Ireland, Ireland and Scotland, have predicted that some invasive waterweeds, previously killed-off by low winter temperatures, will thrive over the next 70 years if temperatures rise. Dr Ruth Kelly, post-doctoral research student at the School of Biological Sciences at QUB, has led a study on invasive species in waterways using statistical analysis, environmental niche models and climate niche models: “Combining global climate and regional landscape models to improve prediction of invasion risk,” as she decribes it.
The main software used was MaxEnt (Maximum Entropy Modelling), a free, open-source programme for modelling species distributions that runs on common operating systems like Microsoft Windows 8.
The QUB team also used the R statistical programming language and ArcGIS mapping software, made by ESRI. ArcGIS is GIS (geographical information systems) software, which can be used to store geographic data, generate maps and analyse spatial data. The map data is organised into layers, so that users can choose which layers they want to view or query. The models themselves rely heavily on online data repositories such as the Global Biodiversity Information Facility (an international open data infrastructure where every occurrence of a species is recorded with source details), and Worldclim (developed at University of California, Berkeley), which offers global environmental layers commonly used to construct ecological niche models and potential geographic distributions.
“R is becoming popular in ecology and academic research as it is more user friendly than [programming environments like] Java and Python,” Dr Kelly explains. “R is set up for people who are able to access huge amounts of power. As our work is focused on Big Data, global and climate datasets, we need the most powerful ways of analysing this [data].”
ICT has increasingly played a key role in the management of invasive species through gathering, collating and analysing huge amounts of data, helping experts to develop strategies for prevention, control and ultimately eradication. Computer-based systems have been deployed to aid management of invasives since as far back as the 1960s, from software simulations used to forecast population spreads, to databases collating and storing information for statistical modelling. More recently, technologies such as GIS, Big Data analytics, distribution mapping, and so forth, have become more important in tackling the challenges of invasive species appearing on land, sea and air.
They’re out there, somewhere...
Generic ICT should be able to play a greater role going forward as sensor technology, for instance, can be used for species monitoring applications. Sensors can be used to detect environmental variances that are connected to the presence of invasive species. Even social media can have a role here, by providing a way for people to report sightings of species known to be potentially problematic, and using that data to inform trend maps across a given affected area.
Data modelling is another IT-enabled discipline that can help. “Data is absolutely essential to modelling the spread of invasive species,” says Dr Sandvik, who co-developed a semi-quantifiable methodology adopted by the Norwegian government, which provides experts with an objective classification of the potential impact of invasive species on environments.
Statistical models use species distribution data to predict the risks posed by each invasive species, including their likelihood of being established and spreading in an ecosystem, and the degree to which they will interact with native species or transform habitats. Aquatic ecologist Dr Gallardo spoke of the success of species distribution models (SDM) in predicting the dispersal of invasive species, both in “locating areas of likely first introduction, and the potential spread of species once established into new areas”.
SDMs suffer from “a number of limitations”, she says, and yet predictions derived from SDM “are surprisingly accurate… For instance, models based on the native range of the infestational and damaging zebra mussel (Dreissena polymorpha) were able to locate the first areas of introduction of this species into North America – the Great Lakes”. In the case of invasive amphipod crustacean Dikerogammarus villosus – the so-called ‘killer shrimp’ – models based on its European presence identified East Anglia and a strip along the south coast of Wales as where the species was actually spotted for the first time.
Most researchers tend to use off-the-shelf SDM software packages from developers such as MaxEnt, Biomod (a computer platform devised by Dr Wilfried Thuiller for ensemble forecasting of species distributions), MARS (multivariate adaptive regression spline), BIOCLIM (the first SDM package that was developed), support vector machines (SVMs), GARP (genetic algorithm for rule-set prediction), etc. These models are adapted (e.g., adding or modifying environment, habitat, climate layers) to suit the data.
MaxEnt, Desktop GARP and Biomod, are all examples of open source freeware, while others are licenced packages – MARS, for example, is trademarked and available to buy from data-mining and predictive analytics specialist Salford Systems; but the open-source implementation of MARS is called ‘Earth’. Some scientists working in invasive species research develop their own SDMs: for instance, Oregon State University built its own model to predict which non-native species might become invaders.
The Centre for Ecology & Hydrology’s Dr Helen Roy coordinates the UK Ladybird Survey, and uses datasets to explore the effects of the invasive Harlequin ladybird (Harmonia axyridis) on native ladybirds, whose population has declined since Harlequins were imported from east Asia in 2004 as a biological pest control. Scientists at the Centre for Ecology & Hydrology designed the app in partnership with Nature Locator at the University of Bristol, a team of scientists, developers, designers and usability experts who specialise in building custom-made geospatial smartphone apps for use in biological recording. They used open source software Indicia, a Google toolkit that simplifies the development of websites which allow data entry, mapping and reporting of wildlife records. This is linked into iRecord and the data is shared through the NBN Gateway.
Big Data analytics has become a powerful tool in monitoring invasive species, with ecologists saying that with the problem of invasive species being on such a large scale, it is essential to analyse lots of data through studying several ecosystems. Supercomputers can help conduct analysis of the terabytes of data available, allowing data to be processed much more quickly. The previously-mentioned GARP software programme, developed at the San Diego Supercomputer Centre, can perform modelling analysis using a genetic algorithm to predict the potential distribution of an invasive species, and these models can be visualised as distribution maps using GIS. The model has been deployed successfully in several studies, for example in 2007 when it predicted that large estuaries including Chesapeake Bay in the US were at risk of invasion of Chinese mitten crab (Eriocheir sinensis). Just one month later crabs were discovered in the estuary’s waters.
The public has also helped to crowdsource observation data of invasives through smartphone apps such as the iRecord Ladybird mentioned already. Via these apps, users capture geotagged data and photos of invasive species: that’s uploaded onto a central server for verification by scientists then projected onto distribution maps, helping to detect/track alien spread.
GISs are being used in helping to detect, survey and track invasive species. Data recorded using handheld GPS units can mark where an invasive species has been found or follow its path, then can be uploaded into software, which can be cross-referenced with environmental data to look for patterns in which invasions occur. This again allows experts to build models and visual maps to both predict future invasions and increase monitoring efforts in vulnerable areas to prevent invasion.