Context

The Australasian bittern is a Nationally Critical wetland species that occurs throughout New Zealand. To assess the rate and patterns of its decline, and to guide its future management, a database of historical sightings has been collated and monitoring methods developed for this cryptic species.

Key findings

In order to assess the rate of decline of Australasian bitterns across New Zealand, a database of historical sightings has recently been collated along, with development of methods for monitoring this cryptic species. These tools can now be used to assess the rate and patterns of decline and guide future management of this critically endangered wetland species.

• Historical records show that there has been an overall decline in bitterns across New Zealand in the last century, with this species being present in 1054 grid cells on maps from pre-1900 but only 560 cells on maps post-1990 (Figure 1). There has also been a shift in distribution towards the lowland and coastal wetlands of the North Island.
• Approximately 90% of inland palustrine wetlands have been drained in New Zealand since the mid-1800s, leading to the loss and degradation of wetland habitat, which has been a significant contributor to this c. 50% reduction in bittern distribution across the country (Ausseil et al., 2008, 2011).
• Bitterns are being actively monitored in the Whangamarino and Awarua wetland sites, which are managed by the Arawai Kakariki (AK) restoration project, as well as within several of the Department of Conservation’s (DOC’s) Ecosystem Management Units (EMUs) across New Zealand (Figure 2).
• Native threatened wetland bird populations, including bittern, are impacted by the common suite of invasive predatory mammals (O’Donnell et al., 2015).
• Trapping of mammalian predators in the AK wetlands is currently being trialled to assess the effectiveness of predator management on improving bittern populations.

Figure 1: Presence of bittern pre 1900s and post 1990 in 25 km2 grid cells.

Figure 2: Bittern call count rates at the Whangamarino and Awarua AK sites. Note the low call rate at Awarua in 2015 is likely affected by bad weather during the survey period.

Definition and methodology

Database

The bittern database was collated from a wide range of sources including publications, museum and local authority records, personal communication and other databases resulting in a total of 4179 records. Each record was given a rating of 1 (high accuracy) to 4 (low accuracy) and assigned to a time period. Records rated 4 for accuracy were excluded on the grounds they were non specific in location. To avoid sampling bias through time, records were assigned to both the time period they were recorded in and all time periods prior on the assumption that habitat was being lost rather than created (i.e. where bittern are found today they would also have been found historically). Spatial analysis was applied using a 5x5 km grid of New Zealand.

Bittern monitoring techniques

Protocols for four most cost-effective techniques have been developed (O’Donnell, 2011; O’Donnell et al., 2015, 2013; O’Donnell and Williams, 2015). These include close approach and acoustic triangulation methods for small wetlands and automatic call recorders for large wetlands. Research has determined optimum time of year, time of day and environmental conditions in which sampling should occur (Ausseil et al., 2008; O’Donnell et al., 2013).

Monitoring in wetlands

Australasian bittern were monitored at both the Whangamarino wetland (2010-2016) and Awarua/Waituna wetlands (2013-2016) using the methods outlined in (O’Donnell et al., 2013). In addition, monitoring has commenced at a number of DOC EMUs (Te Waihora/Lake Ellesmere, Cascade Wetland, Lake Hatuma, Mangarakau Wetland, Waikato Peat Lakes) with additions planned in the future.

Data quality

This measure is classified as a case study. Further years of monitoring are required to make conclusions about trends in bittern numbers.

Relevance

This measure relates to indicator 1.4.2 Security of threatened and at risk taxa

Accuracy

These estimates are considered accurate enough to measure changes in distribution at a national scale, based on the large sample size (>4000 reports), a high level of accuracy (>50% of reports from exact locations) and large numbers of observers engaging in the survey (>600). However, it should be noted that the index measures only provide information on the adult males in the population, because it is only the males that call and thus can be counted, with no data on the numbers or survival of females and juveniles. Therefore, there is an implicit assumption that trends in the number of adult males at a specific wetland reflect what is happening across the population.

This measure complies with the data quality guidelines used in New Zealand’s Environmental Reporting framework.

Outcomes Monitoring Framework

The Department of Conservation’s (DOC’s) Outcomes Monitoring Framework provides a platform on which DOC and others can assess outcomes in a clear, structured and transparent way (Lee et al., 2005). It has been developed as a logical hierarchy that is based on broad, overarching Outcomes, beneath which are nested Outcome Objectives, Indicators, Measures and Data Elements to provide ever increasing levels of detail. The framework is scalable, as the indicators and measures remain compatible and consistent whether applied locally, regionally or nationally. The recently updated framework provides a roadmap for gathering information to meet the specific objectives of DOC and other agencies (McGlone and Dalley, 2015). The provision of a national framework with agreed outcomes, indicators and measures supports collaboration with land management and regulatory agencies, allowing for more integrated environmental policy and ‘State of the Environment’ reporting. DOC has partially implemented a national monitoring and reporting system, whereby priority indicators and measures are routinely used to report on progress against the objectives and outcomes. This factsheet reports on a measure for the 2016/2017 year.

Glossary of terms

Palustrine wetlands are inland wetlands that lack flowing water, contain ocean-derived salts at concentrations of < 0.5 parts per thousand and are non-tidal.

DOC Ecosystem Management Units (EMUs) are places that have been identified as being important for management due to the types and condition of the ecosystems and species present. They may be on land/water of any tenure (not only PCL) and often include several connected ecosystems and communities of threatened species.

With the close approach method, an observer quietly paddles a kayak towards each booming male bittern to map its location whilst avoiding disturbance. To increase confidence in the bird’s location, it can be approached from more than one direction or even circled.

With the acoustic triangulation method, at least two observers listen for booming male bitterns from two different stations that are 500 m apart on the shoreline and repeatedly take compass bearings of the directions of calling birds over a 1-hr period at dusk. The location of each different bittern that was heard is then plotted on a map and the estimated locations are compared to decide whether observers were hearing the same or different birds.