10.5061/DRYAD.C2FQZ617C
Prowse, Thomas
0000-0002-4093-767X
University of Adelaide
O'Connor, Patrick
0000-0002-8966-9370
University of Adelaide
Collard, Stuart
University of Adelaide
Peters, Kristian
Natural Resources Adelaide and Mount Lofty Ranges
Possingham, Hugh
The Nature Conservancy
Data from: Optimising monitoring for trend detection after 16 years of
woodland-bird surveys
Dryad
dataset
2021
Ecological monitoring
Natural Resources Adelaide and Mount Lofty Ranges*
Natural Resources Adelaide and Mount Lofty Ranges
2021-03-04T00:00:00Z
2021-03-04T00:00:00Z
en
1035226 bytes
2
CC0 1.0 Universal (CC0 1.0) Public Domain Dedication
1. Long-term biodiversity monitoring programs provide important
information about species’ trajectories and broader environmental change.
Often constrained by funding and organisational capability and commitment,
monitoring programs need to be optimised to maximise ecological and
economic efficiencies, as part of sound adaptive management. 2. The
monitoring design requirements for detecting biodiversity trends, across
assemblages of species with different traits, can be informed by
historical datasets. Using data from a landscape-scale (c. 2,500 km2) bird
monitoring program encompassing 151 sites visited 3 times annually over 16
years, we used resampling to simulate different monitoring designs. We
quantified the capacity of modified monitoring regimes to detect
population trends for 65 bird species with different densities,
detectabilities and specialisations. 3. The majority (58%) of species
exhibited a significant decline in relative abundance, with the ability to
detect trends proportional to the length of the time series used for
analysis. The percentage of trends detected decreased as survey sites or
sessions were dropped from the monitoring dataset. Statistically
significant trends remained undetected for an additional 2.5 species for
every 10% of sites excluded randomly from the program. As monitoring
effort was reduced, the precision of trend estimates for rare species was
particularly compromised. 4. Conducting bird surveys every second year
would produce better results than an equivalent reduction in effort
achieved by surveying only half the sites each year, but could compromise
the sustainability of the program. If the number of survey sites were
reduced, trend detection would be optimised by retaining the spatial
extent of the surveys (i.e., by dropping sites from well-surveyed regions
rather than excluding outlying, isolated sites), but the cost savings of
this approach would be small. 5. Synthesis and applications. Reduced
monitoring effort will compromise trend detection for rare species or
species that are difficult to observe, and declining species that will
soon become rare. Consequently, monitoring effort that is considered
‘surplus’ today could provide critical data for detecting species-level
trends and prioritising conservation interventions in the future. Further,
sampling efficiencies are not all-important; we must also consider the
impacts of survey design modification on the social and political
sustainability of existing monitoring programs.
Citizen-science woodland bird monitoring dataset conducted annually for
151 sites in the Mount Lofty Ranges, South Australia, from 2001 to 2016.