By Jane Catford (This article was first published in the July 2013 issue of Decision Point, The Monthly Magazine of the Environmental Decisions Group)
In a rapidly changing world full of strange new assemblages of species, the goal posts for conservation aren’t as clear as they once were. What is it we’re trying to conserve, and against what threat? And where do we look for guidance on how to do it when the rules appear to be changing faster than ever? Dealing with novel ecosystems in a time of accelerating change is an enormous challenge. However, two recent collaborations I’ve been involved in have come up with some helpful guidelines on how to proceed. One examines how to predict characteristics of novel ecosystems (Catford et al. 2013, discussed in this story). The other focusses on measuring levels of biological invasion as one way of gauging the extent of ecosystem novelty (and is discussed on pages 10-11 of Decision Point issue 71).
New playing fields
The character of ecosystems and their biological communities is changing as humans modify the environment, move species around the world and cause some species to go extinct. In these novel (or emerging) ecosystems, species occur in combinations and abundances that have not occurred previously and this can prompt changes in ecosystem structure and function (Hobbs et al 2009).
Consider that vacant paddock down the road. More than likely it contains a range of plants that are both native and non-native that never existed as an assemblage fifty years ago. Or, if you want to scale up, consider a country like New Zealand. It has around 7,000 native plant species and some 25,000 exotic species (of which over 2,500 are deemed to be naturalised). With such a high percentage of alien flora, most ecosystems in New Zealand can be considered novel.
Traditionally, most conservation activities have focused on conserving particular suites of species in particular locations (Hobbs et al 2009). Such a goal, however, may become unrealistic as ecosystems and their species assemblages change. To use a sporting analogy, changing the game will shift the goal posts—and human actions are certainly changing the game.
To be able to identify realistic goal posts (i.e., realistic management objectives) it’s necessary to know a little bit about the game that will be played. As such, predictions about the characteristics of future ecosystems are crucial.
Box 1 Anticipating ecological can help management
Humans have been building cities for a long time. As a result, many of the ecological impacts of urbanization are well known and can be predicted ahead of a new urban development. Fortunately, this knowledge can be used to guide urban planning so as to reduce some of the ecological impacts of urbanization.
Facilitated by environmental policy and legislation, Water Sensitive Urban Design is increasingly being used to minimize impacts on aquatic ecosystems and cat curfews in urban areas that border native bushland are helping to reduce predation of native wildlife.
Predicting likely impacts caused by other forms of environmental change can similarly be used to guide environmental policy and the management of novel ecosystems.
Envisioning future ecosystems
In 2011, I attended a workshop hosted by the National Climate Change Adaptation Research Facility (NCCARF) that focused on riparian ecosystems under climate change. Among the various discussions at the workshop, some colleagues and I started discussing how riparian ecosystems might be affected by climate change and ways in which their abiotic and biotic characteristics are likely to change. It soon became clear that envisioning future ecosystems is no easy task, so we set about trying to come up with an approach by which to do so.
Our approach is set out in a paper that was recently published in the journal Ecosystems (Catford et al. 2013). It’s based around four recommendations (see Box 2), which we present in the first part of the paper. We then considered four case studies from contrasting environments to illustrate the approach and to determine:
- whether certain characteristics make some ecosystems more susceptible to climate-induced shifts in community structure than others; and
- which aspect of climate change seems to have the greatest effect on community structure and therefore should be a research priority.
Focusing on changes in community structure, we used qualitative process models to predict likely abiotic and biotic changes in four case study systems: tropical coastal floodplains, temperate streams, high mountain streams and urban riparian zones. We concentrated on functional groups rather than individual species and consider dispersal constraints and the capacity for genetic adaptation.
Box 2 How to predict characteristics of novel ecosystems
Use process models to predict characteristics of future ecosystems, but integrate their predictions with those from other models.
Because they are based on mechanistic understanding, predictions from process (or mechanistic or causal) models are more general than those based purely on correlative relationships and are more likely to remain valid under novel environmental conditions.
Use functional groups, as well as species, to predict the types of communities that will develop in the future.
Predictions based on functional groups (collections of taxa grouped according to shared traits, life history characteristics or ways they interact with the environment) should be used to complement predictions about individual species and vegetation types. Based on traits rather than the identity of individual taxa, functional groups enable findings to be generalised and compared across ecosystems and biomes. Depending on the type of functional classification used, they enable implications of environmental change for ecosystem structure and function to be more apparent.
Use of analogue systems to help envisage the sorts of ecosystems that might develop in the future.
Although novel ecosystems are by definition ‘new’, it may be possible to use ecosystems with similar attributes (or those that have experienced similar conditions at some point in time) to help inform predictions. Flow regulation and diversion of water from rivers may result in similar effects as the reduction in precipitation associated with climate change. In this sense, information about ecological changes in regulated rivers could be used to inform predictions about climate change effects in unregulated rivers. As well as modified systems, ecosystems in different biogeographic regions may serve as suitable comparisons. For example, many lowland rivers in Australia’s temperate south-east may become more like dryland rivers in arid and semi-arid Australia.
Of course, where an environmental change has occurred previously, it makes sense to learn directly from that experience.
Incorporate information about taxon migration rates, dispersal patterns and genetic adaptation.
Modelling approaches need to include information about migration, dispersal pathways (including the role of humans as dispersal vectors), proximity of source populations and the ability of a species or population to adapt to altered conditions. It is likely that species with higher phenotypic or genotypic plasticity will tolerate changes in climate better than species with low genetic diversity or plasticity.
The scenarios we generated as we applied our approach to these four contrasting ecosystems began turning up a few common factors. Our scenarios suggest that climatic changes will reduce the diversity of native species, facilitate non-native invasion (especially warm-season grasses), increase fragmentation and result in simplified and less distinctive riparian ecosystems.
“Climatic changes will reduce the diversity of indigenous species, facilitate non-indigenous invasion, increase fragmentation and result in simplified”and less distinctive riparian ecosystems.”
Using our approach to better understand what the future might hold for different ecosystems also allowed us to test the assumptions we made and, in large measure, confirmed our choices. Compared to models based on correlations between biota and environmental conditions, the process models we used were built on a mechanistic understanding. Process-based models (like Bayesian belief networks) are more likely to remain valid under novel climatic conditions. And predictions based on species’ functional traits will facilitate regional comparisons and can highlight effects of climate change on ecosystem structure and function. Ecosystems that have experienced similar modification to that expected under climate change (e.g., altered flow regimes of regulated rivers) can also be used to help inform and evaluate predictions.
Conservation in the past has focused on conserving particular suites of species in particular locations. Looking ahead, such a goal may be unrealistic as ecosystems and their species assemblages change. Instead, it may be more pragmatic to focus on maximizing genetic, species and functional diversity, maintaining the biogeochemical configuration of a system or increasing habitat complexity.
The ability to accurately predict the characteristics of future ecosystems is essential for setting realistic management objectives. It also represents an exciting area of research.