Elsevier

Ecological Indicators

Volume 94, Part 1, November 2018, Pages 274-282
Ecological Indicators

Original Articles
A framework for deriving measures of chronic anthropogenic disturbance: Surrogate, direct, single and multi-metric indices in Brazilian Caatinga

https://doi.org/10.1016/j.ecolind.2018.07.001Get rights and content

Highlights

  • We present a conceptual framework for deriving multi-metric measures of disturbance.

  • It considers multi-scale integration of disturbance data.

  • We apply this framework to Catimbau National Park in the Brazilian Caatinga.

  • Multi-metric disturbance indices were validated with ant data.

  • Our approach facilitates the analysis of resource use impacts on biodiversity.

Abstract

The development of multi-metric indices of chronic anthropogenic disturbance (CAD) from disparate disturbance indicators represents a major challenge for understanding the impacts of CAD on biodiversity, especially in tropical dry areas where livelihoods of local populations are highly dependent on natural resources. We present a conceptual framework for deriving variably integrated, multi-metric measures of CAD from disparate disturbance indicators. Our framework has three steps: (1) identifying the main sources of CAD in the target region, and quantifying them using data of varying levels of spatial and intensity precision; (2) classifying the sources of disturbance into general disturbance pressures, and deriving an index for each; and (3) combining the individual disturbance pressure indices into a fully integrated index that characterizes the overall level of CAD. We apply this framework to Catimbau National Park in the Brazilian Caatinga, using 12 primary data sources to derive disturbance pressure indices relating to livestock, wood extraction and people pressure. The meaningfulness of pressure and overall CAD indices were validated by reference to variation in ant communities. Our analysis revealed notable findings. First, indirect measures from the geographic and socio-ecological context were poorly correlated with direct, field-based measurements, and were therefore of questionable reliability. Second, the three main disturbance pressures were largely independent of each other, which points to complex patterns of resource use by local communities. Third, different weightings of component disturbance pressure indices had little influence on the Global index, making our Global CAD index somewhat insensitive to assessments of the relative importance of different disturbance pressures. Finally, our results caution against a reliance on multivariate ordination to derive integrated indices of disturbance from disparate data sources. Our multi-scale integration of disturbance data can facilitate the analysis of the resource use effects on biodiversity, contributing to effective conservation management and sustainable livelihood development.

Introduction

Disturbance is a key factor influencing the structure of ecological assemblages and evolution of species within ecosystems (Dornelas, 2010, Ponge, 2013). Over recent decades, increasing levels of anthropogenic disturbance have been a major driver of biodiversity loss at local, regional and global scales (Sala et al., 2000, Fahrig, 2003, Fischer and Lindenmayer, 2007, Chazal and Rounsevell, 2009). In turn, biodiversity loss is jeopardizing the sustainability of ecological processes and the provision of ecosystem goods and services (Cardinale et al., 2012, Mitchell et al., 2015). There is thus an urgent need to quantify and predict the ecological effects of anthropogenic disturbance to guide conservation efforts and the management of ecological resources.

Chronic anthropogenic disturbance (CAD) involving activities such as grazing by livestock, firewood collection and exploitation of non-timber forest products is the most widespread form of environmental change in developing countries (Singh, 1998, Gunderson, 2000, Ribeiro et al., 2015, Ribeiro et al., 2016, Ribeiro-Neto et al., 2016). It is especially prevalent in dry areas of the tropical world because ecosystems typically support dense and low-income rural populations that depend on forest resources for their livelihoods (Singh, 1998, Davidar et al., 2010, Specht et al., 2015, Rito et al., 2017). Areas with a long history of past and present human occupancy usually result in a complex mosaic of differently disturbed patches, such that measurements of CAD often need to integrate very different and uncorrelated sources of disturbances (Martorell and Peters, 2005, Ribeiro et al., 2015, Rito et al., 2017). The development of multi-metric CAD indices from disparate disturbance indicators represents a major challenge for understanding the impacts of CAD.

Ideally, CAD metrics would be based on direct measurements of land-use intensity in the field (Martorell and Peters, 2005), especially when we are interested in spatially-explicit impacts. However, this is often not feasible, and so a range of indirect metrics have been used as surrogates (Sagar et al., 2003, Martorell and Peters, 2005, Leal et al., 2014, Leal et al., 2015, Ribeiro et al., 2015, Ribeiro et al., 2016, Ribeiro-Neto et al., 2016, Schulz et al., 2016), usually focusing on single types of disturbances (Ribeiro et al., 2015, Rito et al., 2017). Such metrics can be based on locally-derived socio-economic and socio-ecological information (Medeiros et al., 2012a), or from distance-based geographic and population data with the assumptions that higher population densities and closer distances to towns and roads equate to higher intensity of land-use (Ahrends et al., 2010, Leal et al., 2014, Ribeiro et al., 2015). Socio-ecological data can provide a robust indication of disturbance intensity at the landscape level (Ostrom and Cox, 2010), but lack spatial precision. Geographic data are more spatially explicit, but provide a very imprecise measure of disturbance intensity (Rogan et al., 2007, Barlow et al., 2016).

CAD metrics can be used for different purposes that require different levels of data integration. Some studies aim to analyze the role of a particular disturbance (e.g. grazing by livestock or firewood collection), or the relative importance of different disturbances, as a contribution to a mechanistic understanding of the drivers of ecosystem dynamics (e.g., Specht et al., 2015, Eldridge et al., 2016, Schulz et al., 2016, Zhou et al., 2016). This requires metrics that are specific to particular disturbances. Other studies are more interested in the overall impact of human disturbance on ecosystems, and so require a fully-integrated index that provides a metric of overall CAD (Schoolmaster et al., 2012).

The serving of multiple objectives using both indirect and direct sources of information requires a hierarchical framework that uses multi-level integration of data of varying precision (Schoolmaster et al., 2012). We propose a conceptual framework for deriving multi-metric measures of CAD that serves this purpose (Fig. 1). Our framework uses a three-step process. The first step is to identify the main sources of chronic disturbance and classify them into general disturbance pressures (Pressure index x1 to Pressure index xn in Fig. 1). The second step is to use available sources of information to derive a metric for each disturbance pressure. We focus on metrics that are proxies of disturbance pressure intensity (‘universal metrics’, sensu Schoolmaster et al., 2012) rather than measures of disturbance impact (e.g., Stoddard et al., 2008, Miller et al., 2016). The available information follows a gradient of data precision, from less-precise but more traditionally-used indirect measures based on geographic and socio-ecological surrogates, to more-precise and spatially explicit field-based measurements of disturbance intensity. A single metric may be based on a single source of information, or integrate multiple sources. Finally, the individual disturbance pressure metrics are then combined to form an integrated metric that characterizes the overall level of CAD (Fig. 1).

Our study has three aims. First, we illustrate how our conceptual framework can be populated, using information on CAD in Catimbau National Park in the Caatinga domain of northeastern Brazil. Caatinga is a mix of dry forest and thorn scrub vegetation, and is the world’s most diverse semi-arid biome (Leal et al., 2005, Moro et al., 2016). It is one of the most endangered ecosystems of Brazil due to historical unsustainable exploitation of natural resources by an ever-growing human population (Leal et al., 2005, Albuquerque et al., 2017). Caatinga supports very dense (26 inhabitants per km2; Medeiros et al., 2012a) and low-income (Ab’Sáber, 1999) rural populations that are highly dependent on forest resources for their livelihoods (Davidar et al., 2010, Djoudi et al., 2015). Second, we use variation in ant communities to test of the validity of our indices in terms of biodiversity impacts. Ants are a globally dominant terrestrial faunal group, and are widely used as indicators of broader ecological change (Andersen and Majer, 2004). Finally, we analyze for our Catimbau case study how the different disturbance indices at different levels of data integration behave along the disturbance axis, and the extent to which they provide redundant or independent information.

Section snippets

Study system

Catimbau National Park (8°24'00'' and 8°36'35'' S; 37°0′30″ and 37°1′40″ W; Appendix S1) experiences a hot semi-arid climate (Sociedade Nordestina de Ecologia, 2002). Annual rainfall ranges from 480 to1100 mm, with large inter-annual variability. The annual average temperature is approximately 23 °C. Most (70%) of the Park has sand quartzolic soils. The Park was established in 2002 and most of the residents at that time have remained, with ongoing dependence on the exploitation of natural

Sources of information

Most of the 12 original sources of information, and especially the seven that were selected for integration in multi-metric indices, showed large variability among plots (Appendix S3 and S4), and were poorly correlated with each other (Table 2). We did not find strong correlations between original variables from different types of information (i.e. geographic context, socio-ecological context, and direct field data). Significant correlations occurred only within each of the three types of

Discussion

We have developed a conceptual framework for deriving variably integrated, multi-metric measures of CAD from disparate disturbance indicators varying in their levels of data precision. This framework has three steps: (1) identifying the main sources of chronic disturbance in the target region, and quantifying them using data of varying levels of precision in relation to both space and intensity; (2) classifying the sources of disturbance into general disturbance pressures, and deriving an index

Conclusion

We have presented a framework for developing indices of chronic anthropogenic disturbance based on variably integrated metrics that use both indirect socio-ecological and geographic data sources and measurements taken directly in the field. The framework includes measures of original disturbance indicators, single disturbance pressures, as well as a measure of overall disturbance. We acknowledge that the our specific measurements cannot be generalized to other study systems. However, we believe

Acknowledgements

This study was developed during a discipline of the Programa de Pós-Graduação em Biologia Vegetal (PPGB-UFPE) and supported by the Brazilian agencies ‘Conselho Nacional de Desenvolvimento Científico e Tecnológico’ (CNPq, PELD 403770/2012-2, Universal 470480/2013-0, CNPq-DFG 490450/2013-0), ‘Fundação de Amparo a Ciência e Tecnologia do Estado de Pernambuco’ (FACEPE, APQ 0738-2.05/12, 0138-2.05/14 and 06012.05/15), and the British ‘Rufford Small Grants Foundation’ (RSG 17372-1). K.F.R. thanks the

References (50)

  • T. Urquiza-Haas et al.

    Regional scale variation in forest structure and biomass in the Yucatan Peninsula, Mexico: effects of forest disturbance

    For. Ecol. Manage.

    (2007)
  • A.N. Ab’Sáber

    Dossiê Nordeste seco

    Estudos avançados

    (1999)
  • A. Ahrends et al.

    Predictable waves of sequential forest degradation and biodiversity loss spreading from an African city

    Proc. Natl. Acad. Sci. U.S.A.

    (2010)
  • U.P. Albuquerque et al.

    People and natural resources in the Caatinga

  • A.N. Andersen et al.

    Ants show the way Down Under: invertebrates as bioindicators in land management

    Front. Ecol. Environ.

    (2004)
  • J. Barlow et al.

    Anthropogenic disturbance in tropical forests can double biodiversity loss from deforestation

    Nature

    (2016)
  • B.J. Cardinale et al.

    Biodiversity loss and its impact on humanity

    Nature

    (2012)
  • J. Chazal et al.

    Land-use and climate change within assessments of biodiversity change: a review

    Glob. Environ. Change

    (2009)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1988)
  • M. Dornelas

    Disturbance and change in biodiversity

    Philosop. Trans. R. Soc. B

    (2010)
  • H. Djoudi et al.

    Dry forests, livelihoods and poverty alleviation: understanding current trends

    Int. For. Rev.

    (2015)
  • D.J. Eldridge et al.

    Ecosystem structure, function, and composition in rangelands are negatively affected by livestock grazing

    Ecol. Appl.

    (2016)
  • L. Fahrig

    Effects of habitat fragmentation on biodiversity

    Ann. Rev. Ecol. Evol. System.

    (2003)
  • J. Fischer et al.

    Landscape modification and habitat fragmentation: a synthesis

    Global Ecol. Biogeograp.

    (2007)
  • P.H.S. Gonçalves et al.

    The most commonly available woody plant species are the most useful for human populations: a meta-analysis

    Ecol. Appl.

    (2016)
  • Cited by (66)

    View all citing articles on Scopus
    1

    Permanent address: Programa de Pós-Graduação em Biologia Vegetal, Universidade Federal de Pernambuco, Av. Prof. Moraes Rego s/no, Recife, PE 50670-901, Brazil.

    View full text