Progressing our understanding of Rhinonicteris over the past 18 years - - PowerPoint PPT Presentation

Progressing our understanding of Rhinonicteris over the past 18 years

Dr Kyle Armstrong

How far we have come

Postgraduate Research at UWA

• Field surveys – distribution, habitat use and spatial pop structure
• Phylogeographic and taxonomic comparisons

Rise of acoustic surveys with AnaBat

• Project specific surveys
• Monitoring studies

Reinventing acoustic surveys with ‘full spectrum’

• Balancing expediency and accuracy in analysis
• Automated and semi-automated approaches
• New standards needed

Mixing research with mining

• Population genetic studies

Contributing to conservation and management

• BAP, MAP, SPRAT, Commonwealth survey guidelines

Postdoctoral work

• Resolving taxonomic status  ABRS

Field surveys – distribution, habitat use and spatial pop structure

(Armstrong 2000, 2001, 2003; Armstrong and Anstee 2000)

Postgraduate Research

Data from Armstrong 2003 PhD thesis; KN Armstrong and Gaia Resources unpub ms

• Pilbara high freq phonic type

(Armstrong and Coles 2007, Armstrong and Kerry 2011)

• Some subtle morphological differences

(Armstrong 2002, 2005)

• Previous genetic work (mtDNA) showed

some genetic divergence (Armstrong 2006) Taxonomy

• Bat Action Plan, EPBC Listing

Postgraduate Research

• First comprehensive acoustic surveys were undertaken for

several clients between 2001 – 2003 by Biota Environmental Sciences

• AnaBat-based surveys became the standard in the

following years

• Monitoring studies with BHP Billiton, Atlas Iron, Rio Tinto

Iron Ore based on comprehensive standardised approach

• Monitoring during drilling programmes

(Armstrong 2010)

Rise of acoustic surveys with AnaBat

Inferring usage from activity

Possible roosting – peaks after dusk, much activity before dawn. High

• importance. Permanent diurnal roost,

transient diurnal roost, maternity roost. Drinking? – visiting pool before foraging, before roosting. Important for maintaining presence, useful for monitoring. Not necessarily a diurnal roost. Night roost – low level visitation during the night, refuge for resting? Lowest activity. Less critical habitat, but of value. Night roost.

Advantages

• Good mic zone of reception, esp. for ultra-high frequency bats
• Greater sensitivity, all things considered
• Better resolution of calls for Rhinonicteris and Macroderma, harmonics for
• thers, some still difficult
• Several brands and models available
• Lots of functionality with accompanying software

Challenges

• Many gigabytes of data to analyse in short timeframes, similar budget
• expectation. Balancing expediency and accuracy in analysis
• Careful consideration of the effects of
• sensitivity settings,

 thresholding and parsing

• filter use,

 file scrubbers,

• format conversion, and
• ther automated processes including for species-level identification.
• New standards needed  ABS
• Principles of good survey design

Reinventing acoustic surveys with ‘full spectrum’

Auto-manual hybrid process

• Download and convert WAC to WAV
• SoundID – non-trivial
• Visual Basic script in Excel for data

management

• Manual checking of each marked WAV for

target

• Compile and graph

Automated analysis in SoundID

Example monitoring project

(Atlas Iron)

• About 750 GB (3 weeks from 6 units)
• 290,532 WAVs (5 sec or less)
• 519,803 hits (above 100 kHz)
• 520 matches
• 244 WAVs with matches
• 100 WAVs with target
• CPU time 179 hours = 22 office days
• More for a person

Small benchmark test

• 16,000 WAV files
• Manual searching for Rhinonicteris
• Gave 76 WAV, 0% false neg/pos
• Automated SoundID analysis 39% detection
• Convert to ZCA 29% detection
• Improved with a new SoundID version, improved
• ptimisation, 70% detection
• But auto detection rate back to 30% (13,000 file test set) in
• pen habitat rather than cave entrance
• ZCA from Kaleidoscope had greater detection than SoundID.

But not appropriate for cave entrances.

• Similar patterns between WAV and ZCA, both better than

recording in ZCA with AnaBat equipment.

• Connectedness of colonies, population level genetic diversity
• Good genetic diversity, elevated inbreeding only in peripatric West

Pilbara group

• MtDNA suggests high female philoptary, with two subpops
• Microsatellites suggest males move throughout the Pilbara, responsible

gene flow

• Demonstrated opportunity for avoidance of human disturbances (low

level chronic, non-catastrophic) BUT females less likely to move

Mixing research with mining

What we still face

• How do we deal with uncertainty?
• What detection methods are appropriate?
• How can we maintain rigour in acoustic surveys?
• How to overcome practical constraints to info gathering?
• How can we confirm roosts?
• How do we place a value on roosts?
• How do we protect colonies with large aggregations?
• How do we ensure there are no concomitant impacts?
• How important is foraging habitat?
• How far can they move? And is this useful info?
• How close can mining come to a roost? Intractable?
• What should monitoring achieve? Triggers? Contingencies?
• What should we do when a roost coincides with an ore body?

Key threatening processes

Mammal Action Plan Vulnerable A3c; D2 Area of occupancy <10 km2, declining Unknown number of mature individuals 6-7 subpopulations  actually just one population, two subpops … Key threats include:

• Deterioration of old mines (includes flooding)
• Human disturbance of roosts
• Habitat loss (caves, mines)

Plus

• Roadkill near roosts
• Concomitant effects

Conservation objectives (draft Mammal Action Plan)

• 1. Maintain viability of colonies throughout the population
• 2. Review and collate unpublished information collected by

mining companies

• 3. Develop a regional management plan that prevents extensive

destruction of roost sites

• 4. Acquire/manage land for conservation purposes where

significant roosts occur

• 5. Encourage non-invasive or low impact activities that confirm

the diurnal use of potential roost sites following acoustic detection at cave sites

• 6. Ensure any exploratory surveys, monitoring programmes and

research projects do not introduce disturbance

Plugging the gaps – highest priority needs

From BAP, SPRAT, MAP, Comm Cons Advice, my more recent thoughts

Research

• Estimating abundance, studies of social behaviour, diet – PhD
• Complete taxonomy – being undertaken currently

Proactive and focussed management

• Roost protection, esp. old disused mines
• Develop a set of standard guidelines for mining proponents to

follow, flowchart, regional management plan (incl ghost bat), collated information, update Commonwealth documents

• Long term monitoring of known roosts, and engagement with

private programmes Rigorous consultancy work

• Non-invasive surveys with appropriate gear and visitation
• Roost confirmation as a standard survey approach
• Artificial roost experiments
• Long term monitoring where appropriate

• summary of what is known based on published science and
• bservations with good evidence in unpublished reports.
• flowchart for approach, provide a standardised path for

developers to follow

• management options, guidance for offsets
• buffer zone size guidance
• guidance for monitoring projects - design constraints, triggers

and appropriate contingency responses

• artificial habitat guidance note - when and how it would be

appropriate, and a strategy outline

• monitoring programme at key sites to assess concurrent impacts,

to establish baseline and maintain vigilance

Regional management plan

Effect of ‘non-invasive’ monitoring

Let’s be careful when we think about research topics. Is the knowledge likely to have a net positive effect?