Effective Ecological Monitoring

David Lindenmayer,Gene Likens

• Publisher: CSIRO PUBLISHING
• Year: 2018
• ISBN:
  9781486308941
• Edition: 2
• Topic: Biological Sciences
• Subtopic: Environmental Conservation & Protection
• Index: Biological Sciences

Long-term monitoring programs are fundamental to understanding the natural environment and managing major environmental problems. Yet they are often done very poorly and ineffectively. This second edition of the highly acclaimed Effective Ecological Monitoring describes what makes monitoring programs successful and how to ensure that long-term monitoring studies persist.

The book has been fully revised and updated but remains concise, illustrating key aspects of effective monitoring with case studies and examples. It includes new sections comparing surveillance-based and question-based monitoring, analysing environmental observation networks, and provides examples of adaptive monitoring.

Based on the authors' 80 years of collective experience in running long-term research and monitoring programs, Effective Ecological Monitoring is a valuable resource for the natural resource management, ecological and environmental science and policy communities.

1

Introduction

Organisms, including humans, depend upon the integrity of ecosystems for their wellbeing and survival. High-quality ecological information collected over long periods provides valuable insights into changes in ecosystem structure, ecological processes, and the services ecosystems provide. Without this information, we would have no knowledge about the changing status of the life support system of the planet (Muir 2010; Lindenmayer et al. 2015b). Nor will we know how effective management has been for actions like weed control, restoration programs, recovery efforts for endangered species, or abatement of nutrient enrichment. For example, a lack of effective monitoring means that Australia is unable to determine the effectiveness of the millions of dollars it invests in biodiversity management (State of Environment 2017). Similarly, there is no credible monitoring program for bees anywhere in the world, despite the serious global declines in this group and the critical role they play in pollination (Inoye et al. 2017). Indeed, the public has every right to ask what scientists and resource managers have been doing in spending of public money over many decades! For the purposes of this book, we consider such information, collected conscientiously and continuously for at least 10 years, and then analysed, to be long-term monitoring (see also Box 1.1).

We see the broad area of monitoring falling into three types:

• Curiosity-driven or passive monitoring. This type is monitoring, sometimes also called surveillance monitoring (e.g. see Wintle et al. 2010) is typically devoid of specified questions, with little or no purpose other than curiosity and no underlying experimental design. It may be done out of inquisitiveness, but often has limited usefulness in addressing environmental problems or in discovering how the world works (but see Olsen et al. 1993). Passive monitoring has these limitations because: (1) it is not hypothesis-driven, and (2) it lacks management interventions or different experimental treatments that facilitate scientific understanding about ecosystem responses to natural or human disturbance. Curiosity-driven, passive or surveillance monitoring could be purely mindless or it could be based on other motivations (see Chapter 4).

Box 1.1. What is not long-term monitoring?

Although one can easily do runs from a simulation model and make projections for thousands of years into the future, we do not consider simulation modelling per se to be long-term work. Studies that substitute space for time (e.g. snapshot investigations (Diamond 1986) or retrospective investigations (Cambridge et al. 2007)) can provide long-term perspectives but we do not consider them to be long-term work per se. We also do not consider haphazard revisits to a site after a prolonged absence to be programs of long-term research or monitoring (e.g. Currie and Parry 1999; Smith et al. 2007). Contrary to widely held beliefs in some organisations, we also do not consider that simply measuring something in the environment constitutes monitoring. Rather, as we outline in the following chapters, monitoring needs to be driven by questions, an experimental design, a conceptual framework, and data integrity through repeatable application of appropriate field protocols (usually the same ones employed through time) (Lindenmayer and Likens 2010).

We acknowledge there are many nuances to discussions about what constitutes ‘long term’ in ecological research and monitoring. The review by Stankey et al. (2003) is instructive. They note that some workers consider long-term studies to be those that continue beyond the generation time of dominant organisms in an ecosystem or sufficiently long to quantify the key processes that structure the ecosystem under investigation. This definition would mean that studies of bacterial assemblages with very rapid generation times would be long-term investigations if they were to persist for a year, a month or even a week. Conversely, a 300-year study of stands of giant sequoia (Sequoiadendron giganteum) in which the dominant trees may live for over 1000 years would not qualify as a long-term study. These considerations are important because they emphasise the variable lifespan of different organisms, but they are not feasible to use for many ecosystem analyses. For the purposes of this book, we use a practical, operational definition and consider long-term as monitoring efforts that continue beyond 10 years.

• Mandated monitoring. This is monitoring for which environmental data must be gathered as a stipulated requirement of government legislation or a political directive, such as monitoring of weather, air quality or river flow funded by governments and conducted by government agencies. For example, populations of many threatened species listed under the US Endangered Species Act must be subject to monitoring, making such species unusual among those rare and/or threatened globally that are typically not subject to rigorous monitoring. The challenges associated with monitoring of rare and endangered species are discussed in Chapter 5. Similarly, Mexico has developed a mandated monitoring and reporting framework for assessing land condition and degradation in that country (Garcia-Alaniz et al. 2017). Rigid quality assurance/quality control protocols are usually strictly stipulated in this type of monitoring. Mandated monitoring is driven by the potential to answer broad and often practical questions. These questions almost always are posed post-hoc in regard to some environmental problem and are not derived from a conceptual model of a particular system or process. Mandated monitoring does not attempt to identify or understand the mechanism influencing a change in an ecosystem or an entity. Rather, the focus is usually to identify trends in a given entity (e.g. whether environmental conditions are getting ‘better or worse’).

• Question-driven monitoring. This type of monitoring is guided by a conceptual model of an ecosystem or some other entity (e.g. a population of organisms) and guided by a rigorous experimental design. The use of a conceptual model will typically result in a priori predictions that are then tested as part of the monitoring program. This kind of monitoring might be undertaken by a single investigator or it might fall under a program like the US Long-Term Research in Environmental Biology (LTREB) program (https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=13544). In question-driven monitoring, mechanisms can be discovered whereby prospective scenarios of trends can be calculated and modelled (see Chapters 3 and 4). Often such learning is informed by strongly contrasting management interventions (Carpenter et al. 1995) and in statistical parlance such studies might be best termed ‘longitudinal studies with interventions’ (Cunningham and Lindenmayer 2017). This approach can lead to robust predictive capacity and enable an investigator to pose new questions – an advantageous part of the adaptive monitoring paradigm (sensu Lindenmayer and Likens 2009) that we outline in Chapter 3. This predictive capacity can be of immense value for ecologists, resource managers and decision makers (Lindenmayer et al. 2015b). In contrast, in curiosity-driven or passive monitoring, and in mandated monitoring to a lesser extent, the predictive and prospective capability of monitoring can be limited such as simply extending trend lines without supporting data.

Obviously, there can be overlap between broad categories of monitoring. For example, a rigorous statistical framework can characterise both mandated and question-driven monitoring (Cunningham and Lindenmayer 2017), although many long-term studies were never initially designed to be long-term studies (see Box 1.2). There can be advantages and disadvantages of mandated monitoring programs and question-driven research and monitoring. As we discuss later in this chapter, mandated monitoring is often coarse scale, leading to assessments of resource condition, but providing limited understanding of ecological mechanisms that have given rise to that condition. Question-driven monitoring is often the converse. It is finer scaled and often process based (see examples across Australian ecosystems in Lindenmayer et al. 2014c), but it is very difficult to make valid spatial extrapolations to larger scales (e.g. at the state, province or national level). Thus, in Chapter 5, we discuss the challenges posed by the integration of data, insights and management recommendations from these broad kinds of monitoring programs.

Some of the ecological values and uses of long-term datasets

Countless scientific articles, books, management plans and other documents have been written about the need to do long-term monitoring (Strayer et al. 1986; Likens 1989, 1992; Goldsmith 1991; Spellerberg 1994; Thompson et al. 1998; Lovett et al. 2007; Krebs 2008; Muller et al. 2011; Lindenmayer et al. 2012; Rowland and Vojta 2013; Hamilton 2015; Kupschus et al. 2016; Reynolds et al. 2016). As part of writing this book, a search of the ecological literature published between 1985 and early 2017 produced more than 131 300 articles with the term ‘monitoring’ in the title or abstract. There are even entire journals directly or indirectly focused on long-term research and monitoring programs. Environmental Monitoring and Assessment and Journal of Environmental Monitoring are just two of many examples.

In this vast literature, many ecologists and managers of natural resources have readily acknowledged the importance of long-term research, which normally incorporates monitoring, for improved understanding and management of complex systems (Likens 1989; Muller et al. 2011). Indeed, long-term research and monitoring contribute disproportionately to ecological knowledge and policy relative to other kinds of studies (Hughes et al. 2017). Long-term data are valuable for many reasons. For example, they are fundamentally important for (after Lindenmayer et al. 2012):

• documenting and providing baselines against which change or extremes can be evaluated (e.g. climate change)

• evaluating ecological responses to disturbance, such as from experimental manipulations

• detecting and evaluating changes in ecosystem structure and function, such as change in forest biomass accumulation or change in nutrient limitation of aquatic productivity

• generating new and important questions about population, community and ecosystem dynamics

• providing empirical data for testing ecological theory and models

Box 1.2. A key caveat: many long-term studies were never initially intended to become long-term studies

Much of this book discusses the design, implementation and maintenance of long-term research...