Dealing with Trade-Offs in destructive sampling designs for occupancy surveys

Stefano Canessa, Geoffrey W. Heard, Peter Robertson, Ian R.K. Sluiter

Research output: Contribution to journalReview article

3 Downloads (Pure)

Abstract

Occupancy surveys should be designed to minimise false absences. This is commonly achieved by increasing replication or increasing the efficiency of surveys. In the case of destructive sampling designs, in which searches of individual microhabitats represent the repeat surveys, minimising false absences leads to an inherent trade-off. Surveyors can sample more low quality microhabitats, bearing the resultant financial costs and producing wider-spread impacts, or they can target high quality microhabitats were the focal species is more likely to be found and risk more severe impacts on local habitat quality. We show how this trade-off can be solved with a decision-theoretic approach, using the Millewa Skink Hemiergis millewae from southern Australia as a case study. Hemiergis millewae is an endangered reptile that is best detected using destructive sampling of grass hummocks. Within sites that were known to be occupied by H. millewae, logistic regression modelling revealed that lizards were more frequently detected in large hummocks. If this model is an accurate representation of the detection process, searching large hummocks is more efficient and requires less replication, but this strategy also entails destruction of the best microhabitats for the species. We developed an optimisation tool to calculate the minimum combination of the number and size of hummocks to search to achieve a given cumulative probability of detecting the species at a site, incorporating weights to reflect the sensitivity of the results to a surveyor's priorities. The optimisation showed that placing high weight on minimising volume necessitates impractical replication, whereas placing high weight on minimising replication requires searching very large hummocks which are less common and may be vital for H. millewae. While destructive sampling methods are sometimes necessary, surveyors must be conscious of the ecological impacts of these methods. This study provides a simple tool for identifying sampling strategies that minimise those impacts.

Original languageEnglish
Article numbere0120340
Pages (from-to)1-11
Number of pages11
JournalPLoS One
Volume10
Issue number3
DOIs
Publication statusPublished - 11 Mar 2015

Fingerprint

microhabitats
Sampling
Weights and Measures
Reptiles
Bearings (structural)
Lizards
Poaceae
sampling
Ecosystem
Logistic Models
Scincidae
Logistics
Costs and Cost Analysis
reptiles
lizards
Surveys and Questionnaires
case studies
grasses
Costs
habitats

Cite this

Canessa, Stefano ; Heard, Geoffrey W. ; Robertson, Peter ; Sluiter, Ian R.K. / Dealing with Trade-Offs in destructive sampling designs for occupancy surveys. In: PLoS One. 2015 ; Vol. 10, No. 3. pp. 1-11.
@article{eaacd22072b84cd4b2b398e1b7b98d4e,
title = "Dealing with Trade-Offs in destructive sampling designs for occupancy surveys",
abstract = "Occupancy surveys should be designed to minimise false absences. This is commonly achieved by increasing replication or increasing the efficiency of surveys. In the case of destructive sampling designs, in which searches of individual microhabitats represent the repeat surveys, minimising false absences leads to an inherent trade-off. Surveyors can sample more low quality microhabitats, bearing the resultant financial costs and producing wider-spread impacts, or they can target high quality microhabitats were the focal species is more likely to be found and risk more severe impacts on local habitat quality. We show how this trade-off can be solved with a decision-theoretic approach, using the Millewa Skink Hemiergis millewae from southern Australia as a case study. Hemiergis millewae is an endangered reptile that is best detected using destructive sampling of grass hummocks. Within sites that were known to be occupied by H. millewae, logistic regression modelling revealed that lizards were more frequently detected in large hummocks. If this model is an accurate representation of the detection process, searching large hummocks is more efficient and requires less replication, but this strategy also entails destruction of the best microhabitats for the species. We developed an optimisation tool to calculate the minimum combination of the number and size of hummocks to search to achieve a given cumulative probability of detecting the species at a site, incorporating weights to reflect the sensitivity of the results to a surveyor's priorities. The optimisation showed that placing high weight on minimising volume necessitates impractical replication, whereas placing high weight on minimising replication requires searching very large hummocks which are less common and may be vital for H. millewae. While destructive sampling methods are sometimes necessary, surveyors must be conscious of the ecological impacts of these methods. This study provides a simple tool for identifying sampling strategies that minimise those impacts.",
author = "Stefano Canessa and Heard, {Geoffrey W.} and Peter Robertson and Sluiter, {Ian R.K.}",
year = "2015",
month = "3",
day = "11",
doi = "10.1371/journal.pone.0120340",
language = "English",
volume = "10",
pages = "1--11",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science (PLoS)",
number = "3",

}

Dealing with Trade-Offs in destructive sampling designs for occupancy surveys. / Canessa, Stefano; Heard, Geoffrey W.; Robertson, Peter; Sluiter, Ian R.K.

In: PLoS One, Vol. 10, No. 3, e0120340, 11.03.2015, p. 1-11.

Research output: Contribution to journalReview article

TY - JOUR

T1 - Dealing with Trade-Offs in destructive sampling designs for occupancy surveys

AU - Canessa, Stefano

AU - Heard, Geoffrey W.

AU - Robertson, Peter

AU - Sluiter, Ian R.K.

PY - 2015/3/11

Y1 - 2015/3/11

N2 - Occupancy surveys should be designed to minimise false absences. This is commonly achieved by increasing replication or increasing the efficiency of surveys. In the case of destructive sampling designs, in which searches of individual microhabitats represent the repeat surveys, minimising false absences leads to an inherent trade-off. Surveyors can sample more low quality microhabitats, bearing the resultant financial costs and producing wider-spread impacts, or they can target high quality microhabitats were the focal species is more likely to be found and risk more severe impacts on local habitat quality. We show how this trade-off can be solved with a decision-theoretic approach, using the Millewa Skink Hemiergis millewae from southern Australia as a case study. Hemiergis millewae is an endangered reptile that is best detected using destructive sampling of grass hummocks. Within sites that were known to be occupied by H. millewae, logistic regression modelling revealed that lizards were more frequently detected in large hummocks. If this model is an accurate representation of the detection process, searching large hummocks is more efficient and requires less replication, but this strategy also entails destruction of the best microhabitats for the species. We developed an optimisation tool to calculate the minimum combination of the number and size of hummocks to search to achieve a given cumulative probability of detecting the species at a site, incorporating weights to reflect the sensitivity of the results to a surveyor's priorities. The optimisation showed that placing high weight on minimising volume necessitates impractical replication, whereas placing high weight on minimising replication requires searching very large hummocks which are less common and may be vital for H. millewae. While destructive sampling methods are sometimes necessary, surveyors must be conscious of the ecological impacts of these methods. This study provides a simple tool for identifying sampling strategies that minimise those impacts.

AB - Occupancy surveys should be designed to minimise false absences. This is commonly achieved by increasing replication or increasing the efficiency of surveys. In the case of destructive sampling designs, in which searches of individual microhabitats represent the repeat surveys, minimising false absences leads to an inherent trade-off. Surveyors can sample more low quality microhabitats, bearing the resultant financial costs and producing wider-spread impacts, or they can target high quality microhabitats were the focal species is more likely to be found and risk more severe impacts on local habitat quality. We show how this trade-off can be solved with a decision-theoretic approach, using the Millewa Skink Hemiergis millewae from southern Australia as a case study. Hemiergis millewae is an endangered reptile that is best detected using destructive sampling of grass hummocks. Within sites that were known to be occupied by H. millewae, logistic regression modelling revealed that lizards were more frequently detected in large hummocks. If this model is an accurate representation of the detection process, searching large hummocks is more efficient and requires less replication, but this strategy also entails destruction of the best microhabitats for the species. We developed an optimisation tool to calculate the minimum combination of the number and size of hummocks to search to achieve a given cumulative probability of detecting the species at a site, incorporating weights to reflect the sensitivity of the results to a surveyor's priorities. The optimisation showed that placing high weight on minimising volume necessitates impractical replication, whereas placing high weight on minimising replication requires searching very large hummocks which are less common and may be vital for H. millewae. While destructive sampling methods are sometimes necessary, surveyors must be conscious of the ecological impacts of these methods. This study provides a simple tool for identifying sampling strategies that minimise those impacts.

UR - http://www.scopus.com/inward/record.url?scp=84924347525&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84924347525&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0120340

DO - 10.1371/journal.pone.0120340

M3 - Review article

C2 - 25760868

AN - SCOPUS:84924347525

VL - 10

SP - 1

EP - 11

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 3

M1 - e0120340

ER -