Complexity and the Biosphere

Te Pūnaha Matatini is applying network analysis, complexity theory, and dynamical systems methodologies to understand the biosphere; developing models that couple the interactions between biodiversity, the economy, and human decision-making.

The diversity of life on Earth is the planet’s most striking feature; recent estimates are that fewer than a million of approximately eight million animal species have been described. Biodiversity exists at a large range of physical scales: multicellular eukaryotes have linear dimensions that range in size from tens of microns to tens of metres, and metazoans encompass 17 orders of magnitude by volume. The ability of next generation sequencing technologies to efficiently and simultaneously analyse massive numbers of DNA molecules has allowed the diversity and ecology of microbial communities to be examined in previously unfeasible depth and detail. This vast new resource for understanding the hidden majority of species that contribute to New Zealand’s terrestrial ecosystems and ecosystem services will require new tools for its analysis and visualization. The research in this theme will inform government policy and decision-making, and will assist the New Zealand public in better understanding their relationship with our unique flora and fauna.


2018-2020 Projects:

Mapping antimicrobial use and its impacts on human and environmental health

Antibiotics are a cornerstone of modern medicine, used to treat infectious diseases and prevent infection in vulnerable humans and animals. Most of the antibiotics we and our animals ingest are not broken down, but pass through the body and into our faeces, ending up in wastewater treatment plants and the environment. Here they provide a rich breeding ground for antibiotic resistant bacteria. A map of these potential breeding grounds would help in the development of policies to tackle antibiotic resistance. Such policies are crucial as a 2014 report by the World Health Organization (WHO) concluded that within a decade, antibiotic resistance will make routine surgery, organ transplantation and cancer treatment life-threatening. In this project, we will build a picture of antibiotic use in New Zealand which includes both human and agricultural use. Data for human use is available from Pharmac and is available as prescriptions by district health board. Use by our agricultural sector has only ever been available as summaries of sales date. Antibiotic use in people is high in New Zealand. By comparison, our use of antibiotics in food production is low. This is mainly because the majority of our large animals are raised outside on grass. Antibiotics are used in poultry farming and also by the dairy industry to prevent mastitis. Antibiotics were also widely used to control the bacterial PSA disease that struck our kiwi fruit industry. In this project, we will also develop methods to identify whether resistance genes are present in the potential breeding grounds.

Three-year outcome: Development of a map of antibiotic use in New Zealand for use by the cross-agency group charged with establishing a national strategic plan to minimise the incidence of antimicrobial resistance across New Zealand.


Improving biosecurity outcomes via general surveillance

New Zealand spends millions of dollars each year to protect its borders from new incursions of potentially damaging exotic species that threaten both productive and natural ecosystems. High-risk species and high-risk pathways have been identified, in part, by combining climatic and biotic variables such as available hosts with abiotic variables like trade pathways. This information is used to inform a range of official surveillance programmes, such as the High Risk Site Surveillance (HRSS) scheme that is managed by the Ministry for Primary Industries. Official surveys are limited in scope, and they either target certain high risk species, e.g., fruit flies or gypsy moth, or high risk sites. However, it is apparent that there are many other known and unknown threats. Therefore, broader general surveillance activities, including observations by members of the public, are an important component of protecting New Zealand from new exotic species. Current public participation in surveillance is largely ad-hoc with a small number of species-specific campaigns, e.g., for the brown marmorated stink bug. One way to improve public participation is to provide ‘intelligence’-based alerts that can be conveyed to the public using the mobile technologies developed within the BioHeritage National Science Challenge. A suite of available data collected by key agencies, particularly MPI, can be used to identify current key threats. Such alerts will target public effort to a broader spectrum of threats thereby increasing the likelihood that new incursions will be found early, thereby increasing the probability of successful eradication. This project, a collaboration with the BioHeritage National Science Challenge, will evaluate biosecurity data from relevant central and regional government agencies, e.g. MPI, and select data suitable for our modelling approach.

Three-year outcome: To predict spatial and temporal biosecurity risks throughout New Zealand to provide intelligence alerts that inform public participation in biosecurity surveillance.


Mai i ngā maunga ki te tai – From the mountains to the sea – enhancing conservation using mātauranga

New Zealand’s natural environment is a global treasure yet our unique ecosystems are facing increasing threats and pressures. Invasive species, a changing climate, and intensification of agriculture are examples of processes that are causing ongoing environmental degradation. We need to find new ways to conserve threatened species and endangered ecosystems by using the best aspects of conventional science and mātauranga. Through our existing relationships we have found that despite the previous dedicated work and effort that has been put into creating engagement models, not all Māori communities who aspire to realise their conservation goals are having their needs adequately met. We anticipate this work will contribute to addressing that, specifically in the undertaking and completion of a conservation project, and more broadly in the creation of an effective engagement model. There are few documented examples of effective collaborations between Māori and western-trained scientists but Forster (2011) and Henwood et al. (2016) provide two successful exemplars of wetland restoration. Scientists often feel motivated to reach out to tangata whenua and Māori custodians may have needs that scientists can address but building effective working relationships is challenging because of cultural and social barriers. This project will involve approaching an iwi to define local environmental issues. We will then work together to integrate traditional approaches and mātauranga with the latest scientific equipment and data so scientists and tangata whenua can learn from each other to address a pressing environmental challenge.

Three-year outcome: development of a test-case and framework for strong working relationships between scientists and tangata whenua with a common conservation goal; communication of our findings to research organisations that do not yet have accessible engagement frameworks in place.



How can you improve something that you can’t measure? In the battle to improve the conservation status of New Zealand birds, many of which are endangered or worse, there is a critical problem: we don’t actually know how many of them are present in any given area. This presents several problems, not least that the success of interventions is virtually impossible to judge. Since many of our birds are cryptic (hard to see) it seems better to use the sounds they make – in calls and song – than their appearance in order to infer the size of populations. Acoustic recorders suitable for monitoring birdsong are readily and cheaply available, such as those produced by the Department of Conservation (DOC). However, the recordings are noisy, and currently require extremely tedious manual processing. The need for automatic analysis of these recordings is clear.

While there has been a lot of work recognising small numbers of bird calls from clean, manually processed recordings, we appear to be the only group trying to solve the real-world problem. From recognising the calls, the next problem is to infer the possible number of birds in an area. This requires knowledge of the terrain, the bird species, and statistical inference, together with longitudinal data over time. This is a classic inverse problem, with conservation biology providing the constraints.

Three-year outcome: development of a usable tool for DOC to successfully monitor bird populations at scale.


Citizen science for Predator-free New Zealand

Citizen science is recognised internationally as a valuable source of information and a means of building community participation in environmental management. It creates opportunities for collecting data with unprecedented frequency at much larger scales than has been possible with conventional and costly data-collection techniques. New Zealanders’ strong cultural identification with the environment gives this country a unique advantage in harnessing non-specialist expertise and provides the opportunity for New Zealand to become a world leader in applying citizen science to decision making for primary industries and natural resource management.

However Citizen Science data has many drawbacks; they are often highly biased both spatially and in terms of the type of data collected. With this in mind this project will use existing datasets, for example from the NatureWatch database, to assess what scientific questions can be answered through Citizen Science. Typically, biologists use information provided by non-specialists to estimate the current distribution and abundance of species, particularly those that are easily recognizable, e.g. birds like tui or invasive species like the brown marmorated stink bug. However, for wildlife management it is often necessary to know the likelihood a monitoring technique can detect changes in abundance or distribution that trigger intervention, either to protect a valued species or to control a pest species. The challenge is to match the characteristics of monitoring data, e.g. citizen science data, to the requirements of the manager.

Three-year outcome: development of a best-practice approach to the utility and uptake of citizen science data, integrated into the iNaturalist platform.


Modelling effects of individual heterogeneity on emergent population characteristics

There is a growing recognition that, like humans, animals show consistent variation in behaviour among individuals, often described as ‘personality’. This individual heterogeneity can lead to significant behavioural differences between members of the same species that can have important consequences for population-level processes and ecological interactions. However, it is not clear how these behavioural variations contribute to the emergent dynamics of a population. Factors affecting consistent inter-individual variation in behaviour could be environmental (e.g. habitat, predation pressure, food availability, social environment) or intrinsic (e.g. physiological differences). For instance, ‘shyness’ can result from individuals living in high-risk environments, while individuals at invasion fronts often exhibit increased aggression, activity and boldness. Animal personality has been studied in several species of interest in New Zealand, e.g. little blue penguins, hihi, brushtail possums, and rodents.

We will explore how consistent inter-individual heterogeneity in behaviour, i.e. different ‘personalities’, affects the emergent behaviour of a population. This problem will initially be addressed in an ecological setting, where personality could affect, for example, trapability, movement, disturbance, harvest and reintroduction. We will use stochastic, individual-based models that explicitly incorporate different sources of individual heterogeneity. These modelling frameworks will then be applied to real-world data for particular case studies, e.g. with kiwi social interaction and movement data. Camera footage of burrows, transmitter/activity data (over 10-13 years), and bird mating status/history data are already available for kiwi, however additional data from other sites could also be collected to compare behaviour in kiwi populations that are subject to different predation or environmental pressures.

Three-year outcome: develop a theoretical framework which overcomes the limitations of current models, by incorporating consistent inter-individual behavioural variation and exploring its consequences for wildlife population dynamics. Apply the models to investigate the implications of behavioural variation for conservation initiatives and pest management.


Very large scale conservation: Pan-regional control of invasive predators

Pests are everywhere, from urban gardens to national parks, but agreeing how to control them is not always straightforward. One-third of New Zealand’s land area is classed as protected, the highest proportion of OECD countries, but this still leaves a huge area unprotected. Further losses of remaining native species are inevitable unless the extent of pest control increases markedly, especially outside currently protected areas where the conservation effort has a relatively low priority. This unprotected land may be urban, peri-urban or production but it still makes a welcoming home for many predators and leaving it without pest control is causing problems for many species of reptiles, invertebrates and birds, including some of our most iconic species such as kiwi. Much of this unprotected area is in the hands of private individuals who have a range of views on pest control so land management decisions have to be carefully negotiated and agreed to by a range of different stakeholders from cat-lovers to rabbit-haters. Experience has shown there are minimum landholder participation thresholds that need to be met to maintain ecosystem health and provide ecosystem services at the levels expected by society. For example, a coordinated effort is required so that pest reinvasion from a few untreated properties does not compromise pest control achieved by others. Also, a landscape perspective is needed to mesh pest control with land use so that connectivity (‘safe passage’) can be established for native species dispersing between fragments of suitable habitat. Because this biological connectivity necessarily crosses landowner boundaries, large scale pest control is inherently a spatial issue with social, environmental and economic components.  Without a well thought-out strategy, native biodiversity will continue to decline at local, regional and national scales.

In New Zealand, intensive control of invasive mammals is mostly restricted to large uninhabited areas, such as national parks, forests and near-shore islands, and very small fenced or unfenced eco-sanctuaries. The first major exception is the current Cape-to-City project, ‘large scale’ predator control covering 26,000 ha of agricultural land in Hawkes Bay. This is just a start for a much more ambitious project proposed by four regional councils for the southern part of the North Island. In addition, initiatives like ‘Predator-free New Zealand’ have a national-scale vision for pest eradication. A major challenge is to make this work in production, peri-urban and urban areas where participation and buy-in from private individuals will be essential.

Three-year outcome: develop socio-ecological models to assess the feasibility of scaling-up management of invasive mammals outside traditional conservation areas.