2016 Hooded Plover Biennial Count Report

Report on the 2018 Biennial Hooded Plover Count

Amy Adams, Grainne Maguire and Joris Driessen

CONTENTS

Summary ...... 3 Introduction ...... 5 Methods ...... 6 Coordination ...... 6 Survey design ...... 6 New online data portal ...... 7 Use of defined routes ...... 8 Threat assessment ...... 9 Data entry and analysis ...... 11 Data vetting ...... 11 Mapping ...... 12 Results ...... 13 Survey timing and effort ...... 13 Overall beach-nesting shorebird numbers ...... 13 Banded and flagged birds ...... 21 Hooded Plover coverage and densities ...... 21 Evidence of breeding ...... 28 Threat assessments ...... 31 Invasive weeds ...... 34 Volunteers ...... 35 Discussion and Recommendations ...... 37 Assessing population trajectory change over time ...... 37 Learnings from the launch of the new online data portal ...... 38 The value of threat data ...... 38 Coverage of remote coastline ...... 39 Importance of flagged and banded bird data ...... 39 Targeted surveys and further data analysis ...... 40 Acknowledgements ...... 41 References ...... 42 Appendices ...... 44

SUMMARY

The Hooded Plover (Thinornis cucullatus cucullatus) Biennial Count, occurring since 1980, rallies hundreds of volunteers across eastern mainland Australia to survey suitable ocean beach habitat for Hooded Plovers (eastern subspecies) over several weeks in November. During this count, all other species of beach-nesting shorebirds are also recorded, enabling an assessment of the use of ocean beach habitats by these species. Fixed routes, first established in 2010, are surveyed during the biennial count so that direct comparisons of species abundance can be made across years. During the 2018 count: • 2,709 kilometres of suitable coastline (i.e. Hooded Plover habitat) was identified in New South Wales, Victoria and South Australia, of which 2,333 km (86%) was surveyed. • A total of 1,515 Hooded Plovers (1,489 adults and 26 juveniles) were counted, comprising approximately 51% of the estimated world population of 3,000 birds (eastern subspecies) and 84% of the estimated mainland breeding population of 1,800 birds. • Regionally, Hooded Plover numbers were distributed as follows: o In Victoria (76% of habitat surveyed): 739 Hooded Plovers (737 adults and 2 juveniles). o In South Australia (91% of habitat surveyed): 716 Hooded Plovers (695 adults and 21 juveniles). o In New South Wales (86% of habitat surveyed): 60 Hooded Plovers (57 adults and 3 juveniles).

A comparison with the 2016 count (1,563 Hooded Plovers; 2,291 km surveyed) revealed there were 48 fewer Hooded Plovers during the 2018 count (2,333 km surveyed). This apparent decrease could be attributed in part to 14% of suitable habitat not being surveyed in the 2018 count but also in part to stabilisation of the population after a boom breeding season in 2015/16, where juvenile survival in the first few years is known to be lower than adult survival. Within regions, there appeared to be a significantly higher number of Hooded Plovers in the regions of Discovery Bay, Yambuk to Swan Lake,

NSW Border to Point Hicks, Wilsons Prom to Waratah Bay, Mueller River to Lake Tyers, Seaspray to Corner Inlet in Victoria, and New South Wales South. A slight increase in Hooded Plovers was observed in the Fleurieu Peninsula and Eyre Peninsula regions in South Australia. The Queenscliff to Lorne region remained stable, while all other regions experienced decreases in Hooded Plover numbers, some even despite higher survey coverage compared with the 2016 count.

The highest densities of Hooded Plovers were recorded in the regions of Warrnambool to Yambuk (2.49 birds/km) in far west Victoria, followed by San Remo to Inverloch (2.00 birds/km) on the Bass Coast, Mornington Peninsula (1.88 birds/km), Wilsons Prom to Waratah Bay (1.42 birds/km), Kangaroo Island (1.28 birds/km), Princetown to Warrnambool (1.24 birds/km), Yambuk to Swan Lake (1.21 birds/km), Seaspray to Corner Inlet (1.17 birds/km) and Phillip Island (1.12 birds/km). Most of these regions have been flagged as having the highest densities in the two previous count years, even if the order of significance has altered slightly, with the exception of Wilsons Prom to Waratah Bay, Yambuk to Swan Lake, and Seaspray to Corner Inlet regions.

INTRODUCTION

The biennial Hooded Plover counts began in 1980 after initiation by the Australasian Wader Studies Group and from the 1990s, they were coordinated by Birds Australia, and then BirdLife Australia. The biennial count is a census-style count occurring over a single weekend in mid-November aimed at providing an accurate estimate of population size and distribution of the mainland population. Timing of the count coincides with when most Hooded Plovers are firmly established on their breeding territories, minimising the possibility of inaccurate counting due to bird movements. The count has always included Victorian and South Australian coastlines and has expanded in recent years to include most of the southern New South Wales coastline extending as far north as Jervis Bay. The count does not include the Tasmanian coastline which is divided into regions and surveyed by BirdLife Tasmania over varying timeframes due to lack of volunteers.

This report details the results of the 2018 Hooded Plover biennial count, held on 17th–18th November 2018. The count was successful with many regions experiencing ideal weather conditions. A total of 367 participants undertook 539 surveys, totalling approximately 1,310 hours of surveying (not including travel to and from their designated survey routes). A total of 2,333 kilometres (86%) of suitable coastline habitat was surveyed, extending from 250 kilometres west of Ceduna in South Australia to just south of Jervis Bay in New South Wales.

Currently, the population of the eastern subspecies of the Hooded Plover (Thinornis cucullatus cucullatus) which occurs in South Australia, Victoria, Tasmania and New South Wales, is estimated at 3,000 birds, and listed as Vulnerable [Category C1 + 2a (ii)] by The Action Plan for Australian Birds 2010 (Garnett et al. 2011). In late 2014, the eastern subspecies was listed as Vulnerable on the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act), recognising substantial population declines, and the need for more extensive management measures to improve its conservation status.

The Hooded Plover population in Victoria has previously been estimated at between 450 and 550 birds, based on counts between 2006 and 2010. In New South Wales, the population was estimated at 50 birds, on the basis of regular surveys during the breeding season (New South Wales National Parks and Wildlife Service), and in South Australia, estimates of between 600 and 800 birds in the population have been made based on data collected from biennial counts and extrapolations made for areas that have not been surveyed.

METHODS

Coordination

A part-time (1 day per fortnight) count coordinator based at BirdLife Australia coordinated the biennial count and was primarily responsible for liaising with regional coordinators, volunteers and land managers to ensure that the count occurred within the specified timeframe across all three states. For the first time, this was a paid role through funding from the Australian Government’s National Landcare Program. A major component of the count coordinator’s role is to ensure that all of the data is collected using standardised data sheets, survey instructions and following strict survey routes determined in previous counts. Furthermore, the coordinator ensures that all collected count data is sent in and collated, entered, vetted, analysed and mapped. The count coordinator then assembles this information into a report. For the 2018 count, Amy Adams was the count coordinator. Joris Driessen assisted with mapping the data.

Survey design

No changes were made to the well-established survey methods in 2018.

The timing of the count and the time span within which surveys are conducted, are two critical aspects of the biennial count. Bird movements are likely to be minimal since the count is timed to coincide with the peak of the breeding season when territory occupancy is highest (mid to late November), and most pairs are at least on their first nesting attempt for the season. However, if the survey window were to be particularly wide, there is an increased risk of double counting due to possible bird movements (e.g. as a result of birds not yet being settled on a territory or experiencing failed breeding). Therefore, it is important that the bulk of the census is undertaken in as short a time span as possible.

In summary, volunteers were asked to survey a predefined section of coastline in mid-November 2018 in suitable weather and tide conditions, recording all observations of beach-nesting birds, including terns. To avoid double-counting, survey routes were counted in one direction only. Information on evidence of nesting and the presence of threats and invasive weeds was also collected for each observation.

The data sheets used by survey participants in 2018 were similar to those used in the 2016 count (Appendix 1). A few changes to the data sheets were made prior to the 2018 count with the aim of

reducing inconsistencies between participants. In addition, data collected about vegetation in the vicinity of bird sightings was altered between the 2016 and 2018 counts. In 2016, we asked participants to record the density of vegetation on the nearby dune (if present) and the dominant species present. In 2018, we realised it would be more important to record the presence or absence of weed species to be able to map the distribution of weeds across the spatial range of Hooded Plovers. We therefore asked participants to individually note the presence of the main weeds: Sea Spurge (Euphorbia paralias), Marram Grass (Ammophila arenaria), Sea Wheat-grass (Thinopyrum junceiforme), Pyp grass (Ehrharta villosa), and Beach Daisy or Cape Beach Daisy (Arctotheca populifolia), with the option to add additional weed species observed where required.

New online data portal

In 2018, a new online data portal form was developed within BirdLife Australia’s Birdata for the Hooded Plover biennial count. This was the first time that the biennial count survey has had an online data entry capacity since its inception, enabling greater efficiencies in data transfer to the National Office, reducing the likelihood of data entry errors and reducing time lags in entering individual data sheets with the data entry effort spread across many. It also enabled the count coordinator to quickly identify data that had not yet been submitted, so that missing data could be chased up closer to the survey participation dates before it was potentially lost or misplaced.

Participants were provided with detailed instructions on how to obtain a login to Birdata (note: this is a different system and data collection methodology to the My Beach Bird Portal) and how to enter their biennial count data via the ‘Beach-nesting Birds’ program form in Birdata (see Appendix 1). Support was provided to ensure new users were able to enter their data, and following the data submission period, user feedback was collated and shared with the developer, to adapt the portal for improved capability. In 2019, the portal was made available in the Birdata app for iPhones and Androids, ready for the 2020 biennial count.

Development of the Beach-nesting Birds portal within Birdata was supported by BirdLife Australia, through funding from the Australian Government’s National Landcare program, Adelaide and Mount Lofty Ranges Natural Resources Management Board, the Victorian Government’s Biodiversity On- Ground Action Icon Species Grants program, The NSW Environmental Trust and the generosity of BirdLife Australia donors.

Use of defined routes

Although Hooded Plover biennial counts have been undertaken since 1980, the capacity to compare each count over time has been limited (Glover 2008), particularly because in earlier years there was no determination of the lengths and proportions of suitable habitat surveyed on each count. The 2010 biennial count was the first to report the length of suitable coastline surveyed and to relate this to the density of birds observed. From continuing these methods, it is now possible to make further meaningful interpretations of what bird numbers might mean between years. As described in Ewers et al. (2011), it is essential to be able to quantify the length of the coastline surveyed in order to assist with interpretation of bird numbers.

For the purposes of organising such an extensive census, the entire coastline of the south-eastern mainland of Australia was divided into 24 regions based across three states. These regions are based on historical count regions, and often land management or Natural Resources Management (NRM) regional boundaries. They are not equal in size or availability of suitable habitat.

Each region was assigned one to two regional coordinators (local land managers or volunteers) who organised count participants to survey the routes in their designated region. In many cases, regional coordinators have fulfilled the same role for several biennial counts, providing much needed local knowledge and consistency across years. Regional coordinators were asked to assign people to survey as many of the survey routes in their designated region as possible, and to inform the count coordinator if any routes were not going to be covered prior to the count weekend so alternative arrangements could be investigated.

Using the 2016 set of survey route start and finish points, each regional coordinator was provided with survey maps for the routes in their region prior to the count weekend. Each map covered what appeared to be suitable Hooded Plover habitat based on historical range, expert knowledge and assessment of habitat features (typically high energy/surf beaches backed by dunes). In 2010, extensive feedback from regional coordinators was collated to modify survey routes to exclude non- suitable habitat. Further feedback was sought from regional coordinators prior to the 2018 count, particularly because time was being invested in creating and uploading defined routes in to Birdata as polygons, to ensure no significant amendments were required. Adjustments were made to 49 routes: with 9 routes being renamed for ease of recognition, 22 routes split, 4 routes having coordinates modified, and 13 new routes being added (Appendix 2).

The benefits of following this process were: 1) Simplified coordination - Having clear maps and a corresponding set of start- and finish- coordinates for each survey route would make it easier for coordinators to match-up participants with survey routes. It also made it clearer and simpler for volunteers to survey their routes. In 2014, route start and end points were superimposed on screenshots of Google Maps aerial images to create basic survey maps for distribution and use on the survey day. 2) Routes not surveyed twice or overlooked - This potential problem was particularly relevant at the boundaries of neighbouring regions during earlier counts but is no longer considered an issue. 3) Standardised surveys - It is easier to get meaningful data from survey results when using the same routes for each count. If a route is not surveyed in a particular year, it will be taken into consideration when determining temporal changes and explaining discrepancies in numbers.

This was the fifth count where fixed routes have been used with success. In some areas there is probably a case to be made for removing certain beaches from the standard list or to reduce survey effort in areas with low habitat quality, and/or very low bird densities (see Discussion).

Threat assessment

Of as much value as understanding the abundance and distribution of Hooded Plovers and other beach-nesting species on ocean beaches, is understanding the threat levels to which each site where these birds occur (observation location) is exposed. Effective conservation management is built around mitigating threats at breeding sites, so it is critical to know what these threats are and how threatened sites are relative to one another.

While a proportion of the mainland Hooded Plover population is monitored intensively during the breeding season and threats are recorded during each visit using the MyBeachBird online portal, for other sites that are rarely visited, the biennial count provides an opportunity to assess the range and relative severity of threats that the birds may be encountering.

Whenever a beach-nesting bird was observed, participants were asked to note all of the key threats present on the beach within 100 m distance of the observation. This data is used to devise a crude

scoring system for threats at sites and to devise heat maps to signal how threatened the birds are at each site.

The threat score was calculated on the basis of the presence and type of threat:

5= Vehicles/ 4= Dogs off leash/ 3 = Dogs on leash/ Horses/Stock Dune use Evidence of people/Dog prints/Cats/Foxes/Deer

Threat types are scored individually and summed to provide an overall score for a particular observation. For vehicles, for which three subtypes exist (4WD, trail bike, quad bike/ATV), the presence of each of these is scored separately. Similarly, ‘evidence of people’ consists of two subtypes: presence of human prints above the high tide line and presence of people on the beach.

Threats given a score of higher than 3 are rated as having a greater impact because they: a. have multiple impacts on the birds, their eggs and chicks as well as their physical habitat; b. are generally present across a greater cross-section of the birds’ habitat (i.e. water’s edge, beach and dunes); c. are more difficult to mitigate (e.g. roaming stock, unregulated horse or vehicle access); and, d. are known to inhibit successful breeding.

Five score categories were used, in line with the previous two biennial counts (Driessen and Maguire 2015), ranging from very low threat levels (green) to extreme threat levels (purple, generally only encountered in a suburban beach context or at a recreational hotspot): • Green, score of 3; • Yellow, score of 4-8; • Orange, score of 9-13; • Red, score of 14-23; and, • Purple, score of 24 or more.

Through grouping the threat scores into fixed categories for each biennial count, approximate trends in threat levels across regions and years emerge. This is useful both as a high-level indicator – i.e. trends in threat levels across different states – as well as a local conservation management aid.

Data entry and analysis

BirdLife Australia’s Birdata web portal was used to record and manage the biennial count data using the custom-built Beach-nesting Birds program. Count data were entered directly into Birdata by either the participant, regional coordinator or BirdLife Australia National Office volunteers (with 5 dedicated volunteers regularly entering data with assistance from the Beach-nesting Birds team). The data were then exported into Microsoft Excel for analysis and to produce maps.

Data vetting

There was still a strong need for data vetting, especially in relation to Global Positioning System (GPS) coordinates. There are different formats of coordinates that describe a position on a map, and they are available as options on hand-held GPS units. The biennial count instructions and data sheets contained examples of the format of coordinates that should be used for the count. The format “decimal degrees” was preferred (DDD.DDDDD°; e.g. 38.540903°S 145.438145°E) as much of our data is collected using this format to allow for consistency across years. A proportion of the data still came back in the formats of degrees minutes seconds (DDD° MM' SS.SS”; e.g. 38° 32' 27.25"S 145° 26' 17.32"E), degrees and decimal minutes (DDD° MM.MMM; e.g. 38° 32.454'S 145° 26.289'E) and Universal Transverse Mercator (UTM; e.g. 55 H 363882.80m E 5733011.78m S). All of these different formats represent the same spot on a map, but when multiple formats are used, conversion calculators are required to convert locations into decimal degrees where an element of accuracy is lost during conversion as well as significantly slowing down data processing.

Another frequently occurring error with data recording was that the proper name of the defined route was not used or was incomplete on the data sheet, so that volunteers entering the data had to plot start and finish points to discover the correct route surveyed or had to contact the participant where start and finish points had also not been provided. Occasionally, a single data sheet would be used to record multiple surveys, particularly where no bird observations were made, which also meant contacting the observer to clarify the correct data (e.g. dates, times, method of surveying) for each route.

Birdata uses predefined polygons to encompass each survey route which participants plot their observations into. This was the first time that the defined routes were converted to polygons (with a 100 m buffer) and used for data entry. We learnt that the polygon boundaries did not always encompass the entire surveyed region (for example due to extensive sand flats occurring outside of

the standardised buffer due to variation from satellite imagery over time) and participants often forced their ‘out-of-bounds’ observations into the predefined boundaries and recorded the actual GPS co-ordinates in the notes section. Manual vetting was required to identify these instances followed by modification of the polygon boundaries and the correction of the record so that the true GPS coordinates of the observations were recorded. Where there was an error in transcribing the GPS coordinates onto the data sheet and there was no marked map attached, the survey participants were contacted directly. Participants were contacted during the data entry period (December 2018 to April 2019), by which time people could not always recall exact locations of sightings. Where insufficient information was available, observations were plotted along the mid-point of the survey route. All survey observations were vetted for anomalies and queried where required, ensuring a high level of data quality.

Mapping

All existing survey routes were digitised in a Geographic Information System (GIS) environment, using ArcGIS 10.7 software. Existing spatial coastline data was used to provide an accurate basis for the complexity of the coastline along each survey route – i.e. each route was digitised in accordance with the layout of the landscape, not ‘as the crow flies’. Subsequently all surveyed and non-surveyed routes (or sections thereof) were similarly digitised in GIS, providing an overview of regional coverage. Count data (observations) were imported into ArcGIS to allow for the creation of maps as well as spatial querying of the dataset.

RESULTS

Survey timing and effort

The majority of surveys for the 2018 biennial count were undertaken in the third and fourth weeks of November, coinciding with the count target weekend (Figure 1). In fact, 35% of all data was collected during the count weekend. Within four days (the day before to the day after the count weekend), 53% of all data was collected, while 80% of all count data was collected within 10 days of the count weekend. This outstanding effort from participants across the three states resulted in 96% of all data being collected during November.

350

300

250

200

150 Number of surveys of Number 100

0 W1 W2 W3 W4 W5 W1 W2 W3 W4 Nov Dec

Week number of month

Figure 1. Distribution of the 2018 biennial count surveys in time.

Overall beach-nesting shorebird numbers

A total of 1,515 Hooded Plovers were counted (1,489 adults and 26 juveniles) across 86% of suitable coastline of south-eastern mainland Australia (Table 1; Appendices 3-5). The 2018 biennial count recorded 84% of the estimated number of birds in the mainland population and 51% of the world population (eastern subspecies). In a global context, counts for Hooded Plover, Red-capped Plover (Charadrius ruficapillus), Sooty Oystercatcher (Haematopus fuliginosus) and Pied Oystercatcher (Haematopus longirostris) represent internationally important numbers (Table 1). While the count is

not intended to census the populations of the latter three species, it does reveal that ocean beach habitats on the eastern mainland provide important habitat for them and for Pied Oystercatchers in particular, surveys a high proportion of their preferred habitat.

Table 1. Overview of the 2018 biennial count results in an international context.

Red-capped Pied Sooty Hooded Plover Plover Oystercatcher Oystercatcher Victoria 739 453 637 184 South Australia 716 2,598 943 393 NSW 60 68 93 49 Total 1,515 3,119 1,673 626 Global population* 3,000 95,000 11,000 11,500 Importance 51% 3% 15% 5%

*estimates from IUCN Red List of Threatened Species (BirdLife International, 2016) and Wetlands International (2019).

A regional and state breakdown of total counts reveals that for Hooded Plovers, similar to previous counts, the regions of Warrnambool to Yambuk (VIC), Kangaroo Island, Yorke Peninsula and Eyre Peninsula (all SA) all support significant numbers of the species (Table 2, Figure 2a). It is important to note that the size of these regions relative to their Hooded Plover populations varies significantly and thus Figure 2b of the density of birds relative to area provides a meaningful depiction of high value sections of coastline.

Juvenile Hooded Plovers made up approximately 1.7% of the total count. Given the timing of the count within the breeding season, only August to mid-September nesters would have flying juveniles by mid- November. These would be the earliest nesting attempts and so this proportion of juveniles is not unexpected. Interestingly, New South Wales South (5.9% of count made up of juveniles), Eyre Peninsula (4.8% of count made up of juveniles) and Kangaroo Island (4.6% of count made up of juveniles) had a higher proportion of juveniles than other regions. This may indicate several things, either more pairs begin nesting earlier in the season in these regions, or survival of chicks is higher in these regions, or in some cases, these regions represent areas where juveniles prefer to congregate.

As with previous counts, the number of Red-capped Plovers on ocean beaches in South Australia is considerably higher compared with Victorian and southern New South Wales beaches (Table 2, Figures 3a and b). In South Australia, Red-capped Plovers were recorded in large numbers in the regions of

Yorke Peninsula, Eyre Peninsula and Ceduna and West, and of the total recorded, 5.2% were juveniles. Red-capped Plovers are known to start breeding as early as late July in some parts of Australia which may explain the high proportion of juveniles recorded in these regions. In Victoria, only the beaches of the Warrnambool to Yambuk region supported relatively significant numbers of Red-capped Plovers, whereas they were scattered in relatively low numbers elsewhere. This species occupies a broad range of habitats that were not surveyed, including low-energy beaches and wetlands. These habitats may instead be the preferred habitat for this species in Victoria.

Relatively large numbers of Pied Oystercatchers were recorded in South Australia, far-western Victoria and Corner Inlet (Table 2, Figures 4a and b). Corner Inlet, The Coorong, Kangaroo Island, Yorke Peninsula and Eyre Peninsula each support internationally important numbers of Pied Oystercatchers (exceeding 1% of the global population). Pied Oystercatchers were also the most abundant beach- nesting bird species recorded in the two New South Wales regions, collectively. Given its Endangered conservation status in New South Wales (fewer than 200 breeding pairs in the state; Office of Environment & Heritage, NSW 2019), this highlights the value of the southern beaches for the species.

Sooty Oystercatchers were found in relatively large numbers in the regions of Yorke Peninsula, Eyre Peninsula and Ceduna and West (all in South Australia) however, densities were quite low (Table 2, Figures 5a and b). This high abundance relative to other parts of the coast may be attributed to the rocky coastline and the presence of offshore islands in these regions, as this is the preferred habitat of the species. The highest numbers for Victoria were recorded in the region of Wilsons Prom to Waratah Bay and in the northern transect of the southern region in NSW. By no means does this biennial count give a reliable indication of the population of Sooty Oystercatchers in the southern mainland of Australia, as it does not target their prime habitat.

The Beach Stone-curlew (Esacus magnirostris) which was first recorded during the 2014 biennial count in East Gippsland out of its normal breeding range and recorded again during the 2016 biennial count (in the region of Mueller River to Lake Tyers) was not recorded in the 2018 biennial count. Since 2014, there have been several reports of Beach Stone-curlew sightings in Victoria and South Australia which could be due to either the permanent dispersal of a few individual birds or seasonal local migration, south of their normal range.

Table 2. Results of the 2018 Hooded Plover biennial count (by state and region).

Hooded Plover Red-capped Plover Pied Oystercatcher Sooty Oystercatcher REGION Adult Juv. Total Adult Juv. Total Adult Juv. Total Adult Juv. Total

Victoria 1. NSW Border to Point Hicks 29 0 29 32 6 38 55 0 55 10 0 10 2. Mueller River to Lake Tyers 51 0 51 67 0 67 84 0 84 5 0 5 3. Lake Tyers to Seaspray 12 0 12 10 0 10 31 0 31 1 0 1 4. Seaspray to Corner Inlet 53 0 53 70 2 72 230 0 230 30 0 30 5. Wilsons Prom to Waratah Bay 75 0 75 17 0 17 15 0 15 45 0 45 6. Venus Bay 18 0 18 0 0 0 3 0 3 9 0 9 7. San Remo to Inverloch 71 0 71 1 0 1 4 0 4 5 0 5 8. Phillip Island 44 0 44 0 0 0 17 0 17 35 1 36 9. Mornington Peninsula 77 1 78 17 0 17 2 0 2 33 0 33 10. Queenscliff to Lorne 47 0 47 23 0 23 1 0 1 0 0 0 11. Lorne to Princetown 32 0 32 0 0 0 0 0 0 0 0 0 12. Princetown to Warrnambool 22 0 22 14 0 14 0 0 0 0 0 0 13. Warrnambool to Yambuk 113 1 114 164 2 166 43 0 43 10 0 10 14. Yambuk to Swan Lake 75 0 75 22 0 22 50 0 50 0 0 0 15. Discovery Bay 18 0 18 6 0 6 102 0 102 0 0 0 VIC Total 737 2 739 443 10 453 637 0 637 183 1 184 South Australia 16. Southeast South Australia 51 0 51 194 0 194 66 0 66 17 0 17 17. Coorong 28 1 29 171 1 172 159 2 161 21 0 21 18. Fleurieu Peninsula 53 1 54 13 1 14 3 0 3 20 0 20 19. Kangaroo Island 166 8 174 63 2 65 172 6 178 36 0 36 20. Yorke Peninsula 248 4 252 976 37 1,013 117 2 119 102 2 104 21. Eyre Peninsula 140 7 147 591 56 647 344 7 351 122 3 125 22. Ceduna and West 9 0 9 475 18 493 64 1 65 69 1 70 SA Total 695 21 716 2,483 115 2,598 925 18 943 387 6 393 New South Wales 23. New South Wales South 32 2 34 39 4 43 59 2 61 12 0 12 24. New South Wales North 25 1 26 24 1 25 28 4 32 36 1 37 NSW Total 57 3 60 63 5 68 87 6 93 48 1 49 Grand Total 1,489 26 1,515 2,989 130 3,119 1,649 24 1,673 618 8 626

300

250

200

150

100

Hooded Plover count Plover Hooded 50

PRI

DIS

FLE

PHI

EYR LAK

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW

States and regions (a)

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Hooded Plover density (birds/km) density Plover Hooded

PRI

DIS

FLE

PHI

EYR LAK

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW

States and regions (b)

Figure 2. Hooded Plover counts (a) and densities (b) of adult birds by region from the 2018 biennial count. Densities are presented with ± standard errors and regions are arranged from west to east for ease of interpretation.

1000 900 800 700 600 500 400

300 capped Plover count Plover capped - 200

Red 100

PRI

DIS

FLE

PHI

EYR LAK

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW

States and regions

(a)

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5

1.0 cappeddensity Plover (birds/km) - 0.5

Red 0.0

PRI

DIS

FLE

PHI

EYR LAK

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW States and regions

(b)

Figure 3. Red-capped Plover counts (a) and densities (b) of adult birds by region from the 2018 biennial count. Densities are presented with ± standard errors and regions are arranged from west to east for ease of interpretation.

350

300

250

200

150

100

Pied Oystercatcher count Oystercatcher Pied 50

PRI

DIS

FLE

PHI

EYR LAK

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW

States and regions (a)

7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5

Pied Oystercatcher density (birds/km) density Oystercatcher Pied 0.0

PRI

DIS

FLE

PHI

LAK

EYR

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW States and regions

(b)

Figure 4. Pied Oystercatcher counts (a) and densities (b) of adult birds by region from the 2018 biennial count. Densities are presented with ± standard errors and regions are arranged from west to east for ease of interpretation.

140

120

100

20 Sooty Oystercatcher count Oystercatcher Sooty

PRI

DIS

FLE

PHI

EYR LAK

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW

States and regions (a)

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5

0.0

Sooty Oystercatcher density (birds/km) density Oystercatcher Sooty

PRI

DIS

FLE

PHI

EYR LAK

SEA

WIL

STH

LOR

CED

SAN

YOR

VEN

KAN

BOR

NTH

QUE

COO

YAM

MUE

WAR MOR SA VIC NSW

States and regions

(b)

Figure 5. Sooty Oystercatcher counts (a) and densities (b) of adult birds by region from the 2018 biennial count. Densities are presented with ± standard errors and regions are arranged from west to east for ease of interpretation.

Banded and flagged birds

Banding and flagging of beach-nesting bird species occurs throughout southern Australia and is undertaken by a number of different study groups including BirdLife Australia’s Beach-nesting Birds team, New South Wales National Parks and Wildlife Service (NSW NPWS) staff, Phillip Island Nature Parks (PINP) staff and Friends of Shorebirds South East/Victorian Wader Study Group volunteers. Participants are encouraged to read engraved flags and note down different coloured flag combinations on birds during the biennial count. During the 2018 count, 289 observations of banded or flagged beach-nesting birds were reported. Of these observations, 220 were of marked Hooded Plovers, 1 of Red-capped Plovers, 66 of Pied Oystercatchers and 2 of Sooty Oystercatchers. Reading flag codes especially on engraved flags can be difficult for a number of reasons (e.g. bird is too distant, bird flies off, insufficient power of binoculars, etc.). Nearly 72% of total observations of flagged birds comprised fully read flag codes or combinations and this information is invaluable and helps in gaining a better understanding of bird survival and movements.

Hooded Plover coverage and densities

As evident in Figures 2-5, when comparing regions, bird numbers are less informative than the density values that can be derived from the bird numbers and the proportion of suitable habitat surveyed. It is also essential that only the adult bird numbers are used in calculating densities because juveniles may disperse from their natal territories hence inflating density values if included in the calculation. Table 3 represents the length of habitat surveyed and the density of Hooded Plovers in each region. Approximately 86% of suitable Hooded Plover habitat was surveyed which is only slightly lower than 88% coverage achieved in the 2016 count.

The Hooded Plover hotspots on the south-eastern mainland that have had densities exceeding 1 bird per kilometre and 2 birds per kilometre in some cases, in previous biennial counts produced similar results in the 2018 count. The coastline between Warrnambool to Yambuk had the highest density (2.49 birds/km), followed by San Remo to Inverloch (2.00 birds/km) on the Bass Coast, Mornington Peninsula (1.88 birds/km), Wilsons Prom to Waratah Bay (1.42 birds/km), Kangaroo Island in South Australia (1.28 birds/km), Princetown to Warrnambool on the Shipwreck Coast (1.24 birds/km), Yambuk to Swan Lake (1.21 birds/km), Seaspray to Corner Inlet (1.17 birds/km), and Phillip Island (1.12 birds/km). Significant variation in densities across the coast indicates that habitat for Hooded Plovers is not uniform in quality, and that regions with high densities are likely to reflect high quality habitat. Research into habitat quality and preference reveal that Hooded Plovers are selective of

particular habitat features, both terrestrial (amount of dune and foredune habitat) and offshore (amount of intertidal and submerged rocks), as well as food availability which is undoubtedly linked to these features (Cuttriss et al. 2015; Ehmke et al. 2016).

Table 3. Regional habitat coverage and adult Hooded Plover densities for the 2018 biennial count.

Region Habitat Habitat Habitat Density length (km) surveyed (km) surveyed (%) (birds/km) 1. NSW Border to Point Hicks 53.9 52.4 97.3 0.55 2. Mueller River to Lake Tyers 121.9 118.0 96.8 0.43 3. Lake Tyers to Seaspray 104.5 15.5 14.8 0.78 4. Seaspray to Corner Inlet 91.2 45.3 49.7 1.17 5. Wilsons Prom to Waratah Bay 62.8 52.7 84.0 1.42 6. Venus Bay 40.5 23.8 58.7 0.76 7. San Remo to Inverloch 36.5 35.4 97.0 2.00 8. Phillip Island 39.2 39.2 100.0 1.12 9. Mornington Peninsula 40.9 40.9 100.0 1.88 10. Queenscliff to Lorne 89.8 75.6 84.2 0.62 11. Lorne to Princetown 47.7 30.2 62.3 1.06 12. Princetown to Warrnambool 19.4 17.8 91.6 1.24 13. Warrnambool to Yambuk 51.5 45.4 88.0 2.49 14. Yambuk to Swan Lake 67.3 61.9 92.0 1.21 15. Discovery Bay 37.0 37.0 100.0 0.49 16. Southeast South Australia 212.2 212.2 100.0 0.24 17. Coorong 173.2 173.2 100.0 0.16 18. Fleurieu Peninsula 81.5 74.4 91.2 0.71 19. Kangaroo Island 135.8 129.3 95.2 1.28 20. Yorke Peninsula 346.6 309.7 89.4 0.80 21. Eyre Peninsula 496.6 406.6 81.9 0.34 22. Ceduna and West 195.5 195.5 100.0 0.05 23. New South Wales South 105.2 94.1 89.5 0.34 24. New South Wales North 58.6 46.8 80.0 0.53 Victoria 904.1 691.1 76.4 1.07 South Australia 1641.4 1500.9 91.4 0.46 New South Wales 163.8 140.9 86.0 0.40 TOTAL 2709.3 2332.9 86.1 0.64

Comparison with previous years

From 2010 onwards, we have adjusted routes to more accurately reflect potential habitat, and so there have been changes in densities recorded (Table 4). Minor variation in density occurred between 2010 and 2012, despite an additional 200 km being surveyed. In 2014, an additional 469 km (a 25% increase) were surveyed, largely related to new routes identified in remoter regions such as Eyre Peninsula. This inclusion resulted in 135 more Hooded Plovers being recorded however the substantial increase in coverage resulted in a lower overall density. The 2016 count saw a further 116 km added to the fixed routes but a 2% decrease in coverage compared with the 2014 count. Remarkably, 178 more Hooded Plovers were recorded causing the overall density to increase markedly. This reflects a genuine increase in population numbers related to a boom breeding season in 2015/16. A further 99 km was added to the fixed routes in the 2018 count which saw a 2% increase in coverage compared with the 2016 count, however the number of Hooded Plovers recorded decreased by 31, with the overall density also decreasing (Table 4).

Table 4. Comparison of adult Hooded Plover count totals and densities between 2010 and 2018.

Fixed route Actual habitat Density Total HP length (km) surveyed (km) (birds/km) 2010 1,164 2,134 1,755 (82%) 0.66 2012 1,207 2,334 1,871 (80%) 0.65 2014 1,342 2,494 2,340 (94%) 0.57 2016 1,520 2,610 2,291 (88%) 0.66 2018 1,489 2,709 2,333 (86%) 0.64 Difference 2010- +43 +200 +116 (+7%) -0.01 2012 count Difference 2012- +135 +160 +469 (+25%) -0.08 2014 count Difference 2014- +178 +116 -49 (-2%) +0.09 2016 count Difference 2016- -31 +99 +42 (+2%) -0.02 2018 count

Breaking down the above total density values further to the regional level reveals some major differences between the average densities (2010-2016) compared with the 2018 densities (Table 5). However, since density is a combination of number of birds sighted and length of coastline surveyed, a significant difference in density may be a consequence of change in either of these values, or both. Therefore, separating the density values into their components, and then comparing those directly provides a better insight into why some densities appear to have changed dramatically within ten years (between five counts) while others have remained stable (Table 6). Theoretically, for each count, if all habitat surveyed was equally suitable, the difference in proportions between the coverage and the number of Hooded Plovers should be negligible; in other words, surveying an extra 20% of ‘suitable’ coastline should yield approximately 20% extra Hooded Plovers counted, leading to a minor discrepancy.

Table 5. Adult Hooded Plover densities from 2010 to 2018, arranged in order of decreasing percentage of difference between the average densities (2010-2016) and 2018 densities.

Zone Region Density (birds/km) Percentage difference 2010 2012 2014 2016 2018 (2018 to avg.) 3 Lake Tyers to Seaspray 0.15 0.16 0.13 0.22 0.78 374 4 Seaspray to Corner Inlet 0.29 0.30 0.25 0.51 1.17 246 5 Wilsons Prom to Waratah 0.98 1.02 0.91 1.00 1.42 45 Bay 14 Yambuk to Swan Lake 1.13 0.82 0.67 0.72 1.21 45 11 Lorne to Princetown 0.68 0.77 0.54 1.08 1.06 38 10 Queenscliff to Lorne 0.39 0.40 0.51 0.52 0.62 36 1 NSW Border to Point Hicks 0.35 0.40 0.54 0.42 0.55 28 15 Discovery Bay 0.36 0.92 0.24 0.15 0.49 17 18 Fleurieu Peninsula 0.69 0.50 0.57 0.68 0.71 16 17 Coorong 0.17 0.13 0.12 0.14 0.16 15 23 New South Wales South * 0.34 0.23 0.32 0.34 15 7 San Remo to Inverloch 1.88 1.58 1.63 2.19 2.00 10 2 Mueller River to Lake Tyers 0.30 0.59 0.33 0.47 0.43 2 24 New South Wales North 0.59 0.42 0.46 0.62 0.53 1 13 Warrnambool to Yambuk 2.85 2.26 2.44 2.82 2.49 -4 8 Phillip Island 1.10 1.18 1.19 1.21 1.12 -4 21 Eyre Peninsula 0.34 0.37 0.42 0.40 0.34 -11 12 Princetown to Warrnambool 1.03 1.25 1.14 2.17 1.24 -11 16 Southeast South Australia 0.27 0.34 0.21 0.27 0.24 -12 20 Yorke Peninsula 0.98 1.05 0.75 0.92 0.80 -14 6 Venus Bay 0.91 0.89 1.04 0.82 0.76 -17 9 Mornington Peninsula 2.60 2.41 2.08 2.09 1.88 -18 19 Kangaroo Island 1.57 1.67 1.40 1.64 1.28 -18 22 Ceduna and West 0.26 0.14 0.26 0.12 0.05 -74 * Insufficient information is available for analysis of densities.

Table 6. Comparison between coverage and the numbers of Hooded Plovers (adults and juveniles) in 2016 and 2018, by region. The ‘Difference in coverage’ column shows how much more or less coastline was surveyed in 2018 than in 2016, e.g. a negative value indicates less coverage in 2018. Similarly, the ‘Difference in HP total’ column compares 2018 and 2016 totals of Hooded Plovers, e.g. a negative value indicates fewer birds in 2018. The ‘Discrepancy’ column shows the difference between these two values which should, theoretically, be very small if the routes surveyed in that region are similar in quality. The regions are arranged in order of decreasing ‘Discrepancy’.

Zone Region State Difference in Difference in Discrepancy coverage (%) HP total (%) (%) 15 Discovery Bay VIC 169 800 631 14 Yambuk to Swan Lake VIC 16 97 81 4 Seaspray to Corner Inlet VIC -49 15 65 1 NSW Border to Point Hicks VIC 21 61 40 5 Wilsons Prom to Waratah Bay VIC -6 34 40 3 Lake Tyers to Seaspray VIC -85 -48 37 10 Queenscliff to Lorne VIC -16 0 16 23 New South Wales South NSW 4 17 13 18 Fleurieu Peninsula SA -4 2 6 17 Coorong SA 0 -3 -3 11 Lorne to Princetown VIC 2 -3 -5 8 Phillip Island VIC 1 -6 -7 6 Venus Bay VIC -41 -49 -8 24 New South Wales North NSW -6 -16 -10 7 San Remo to Inverloch VIC 5 -5 -10 13 Warrnambool to Yambuk VIC -4 -15 -11 2 Mueller River to Lake Tyers VIC 36 24 -12 16 Southeast South Australia SA 0 -12 -12 9 Mornington Peninsula VIC 7 -6 -13 20 Yorke Peninsula SA 4 -10 -15 21 Eyre Peninsula SA 22 2 -19 19 Kangaroo Island SA 5 -17 -22 12 Princetown to Warrnambool VIC 49 -24 -73 22 Ceduna and West SA 25 -55 -80

AVERAGE +6 +33 +26

Nine of the 24 regions had greater than a 20% difference in coverage between the 2016 and 2018 counts – an increase compared to the 2016 biennial count when five regions fell into this category. Six of these nine regions had 21-169% more coverage due to funding support through the Australian Government’s National Landcare Program, enabling better access and support to monitor remote regions such as Discovery Bay, Eyre Peninsula and Ceduna and West. The other three regions had decreased coverage (41-85% less) compared with 2016, largely due to logistical issues and scarcity of volunteers (Venus Bay and central East Gippsland). Similarly, nine of the 24 regions had greater than a 20% difference in Hooded Plover numbers between the 2016 and 2018 counts which was slightly fewer than the 12 regions in 2016. Five of these regions recorded increases while the remaining four related to a decrease in observed numbers.

Once the difference in coverage and bird numbers has been taken into account, large discrepancies (Table 6) can be interpreted as either a real change in the local numbers of Hooded Plovers or unsuitable habitat has been surveyed in regions where habitat coverage increased. Large negative discrepancies trigger potential concern and a need to explore our local knowledge and data from these areas. Results from a couple of regions suggest an actual decline in bird numbers. Namely, the South East South Australia region had the same percentage of habitat surveyed between the 2018 and 2016 counts but there was a significant decrease in Hooded Plovers (12% fewer). While the Ceduna and West region had 25% more habitat surveyed, there were 55% fewer Hooded Plovers, but this may instead be due to large expanses of unsuitable habitat being included plus recent on-ground intensive surveys have revealed that the birds can move large distances from day to day and be particularly cryptic along this coastline. Fewer birds, despite increased coverage, were noted in Princetown to Warrnambool region (49% more habitat surveyed but 24% fewer Hooded Plovers), Mornington Peninsula (7% more habitat surveyed but 6% fewer Hooded Plovers), Kangaroo Island (5% more habitat surveyed but 17% fewer Hooded Plovers), San Remo to Inverloch (5% more habitat surveyed but 5% fewer Hooded Plovers) and Yorke Peninsula (4% more habitat surveyed but 10% fewer Hooded Plovers). In some of these regions we have intensive breeding season monitoring occurring and have noted loss of some pairs from sites and/or reductions in non-breeding flock sizes (e.g. flock sizes at Portsea and in to Point Nepean have reduced over the last couple of seasons).

Overall the change in numbers compared to survey coverage is a positive result, indicating that the overall population trend is still towards an increase over time (Table 6). In the region of Discovery Bay, there was a significant increase in coverage (169%) which resulted in an extremely significant increase in Hooded Plover numbers (800%). A similar pattern is evident in the Yambuk to Swan Lake region,

NSW border to Point Hicks region and the Mueller River to Lake Tyers region but with a much smaller increase in coverage and subsequent Hooded Plover numbers (16%, 21% and 36% in coverage respectively; 97%, 61% and 24% in Hooded Plover numbers respectively). Conversely, the regions of Seaspray to Corner Inlet and Wilsons Prom to Waratah Bay experienced a decrease in coverage (-49% and 6% respectively) but an increase in Hooded Plover numbers (15% and 34% respectively). We believe the latter relate to the areas that were surveyed being the highest value beaches in those regions where there are often large flocks sighted (i.e. unsurveyed stretches along 90 mile beach and a number of southern and south-eastern beaches on the Prom typically hold very few birds). In addition, there was improved investment in surveying within Corner Inlet due to funding from the West Gippsland Catchment Management Authority, which resulted in larger and highly experienced teams surveying these expansive islands.

Encouraging signs from the 2018 count relate to a genuine increase in the New South Wales South region where a minimal difference in coverage gave rise to a positive difference in Hooded Plover numbers. Most of these regions where an increase in Hooded Plover numbers was recorded have likely benefited from significant investment in recovery activities to mitigate threats where at least 90 fledglings were recruited into the population in 2017/18 for the major regions monitored in Victoria and South Australia (this data excludes NSW Border to Point Hicks, Mueller River to Lake Tyers, Lake Tyers to Seaspray, Seaspray to Corner Inlet, and Discovery Bay regions of Victoria, and all regions with the exception of South East South Australia and Fleurieu Peninsula in South Australia).

Evidence of breeding

Participants were requested to record evidence of breeding when it was observed. Of the observed Hooded Plovers, 37% were recorded as showing evidence of breeding (an increase from the 22% in 2016). Several people noted nests with eggs, abandoned, washed-out or old nest scrapes, as well as adult behaviour that suggested breeding (e.g. leading and distraction displays). However, overall there were low numbers of chick sightings, which is in line with the timing of the count being earlier in the breeding season (Table 7).

Count data are seldom useful for accurately assessing breeding activity, as Hooded Plovers are adept at hiding their nests and chicks, and because observers must cover a lot of ground during the survey, there is little spare time for observers to spend watching the birds’ behaviour and searching for nests. Nest monitoring is not an essential component of the biennial count and is typically covered by

participants within various organisations (e.g. BirdLife Australia, PINP and NSW NPWS) who embark on regular monitoring of breeding pairs in Victoria, South Australia and New South Wales. As part of the established monitoring program, Hooded Plover pairs are visited regularly and over time volunteers become proficient at recognising breeding behaviour of Hooded Plovers. The data collected through this method of monitoring enables us to quantify breeding success, to devise more accurate threat profiles based upon multiple visits to breeding sites, and for us to make comparisons between areas about breeding output, guiding us in our conservation efforts for the species.

Table 7. Evidence of Hooded Plover breeding recorded in each region during the 2018 biennial count. Values represent the number of pairs recorded within each nesting stage. Scrape: a small depression in the sand which does not contain eggs; Suspect nest: based on adult behaviours (leading, false brooding), a nest with eggs is suspected but never sighted; Nest with eggs: scrape containing eggs; Suspect chicks: based on adult behaviours (distraction displays, aggression), presence of chicks is suspected but never sighted; Chicks: chicks between 1-35 days old sighted. Suspect Nests with Suspect Region Scrapes Chicks nest eggs chicks Victoria 1. NSW Border to Point Hicks - 1 2 - - 2. Mueller River to Lake Tyers 2 3 1 - - 3. Lake Tyers to Seaspray - 1 - - 1 4. Seaspray to Corner Inlet 3 - 2 - 1 5. Wilsons Prom to Waratah Bay - 2 - - - 6. Venus Bay 2 - 1 - - 7. San Remo to Inverloch 14 2 5 - 1 8. Phillip Island 3 - 6 - - 9. Mornington Peninsula 5 2 11 - - 10. Queenscliff to Lorne 1 4 10 - 1 11. Lorne to Princetown 3 2 4 - - 12. Princetown to Warrnambool 1 - 10 - 1 13. Warrnambool to Yambuk 6 11 14 1 1 14. Yambuk to Swan Lake 4 1 3 - - 15. Discovery Bay 2 - 1 - - Total 46 29 70 1 6

South Australia 16. Southeast South Australia 2 3 6 1 - 17. Coorong 1 1 1 - 1 18. Fleurieu Peninsula 3 - 9 - 4 19. Kangaroo Island 3 9 12 - 7 20. Yorke Peninsula 4 22 8 3 5 21. Eyre Peninsula 5 7 8 1 3 22. Ceduna and West - - - - 1 Total 18 42 44 5 21

New South Wales 23. New South Wales South 1 1 3 - 1 24. New South Wales North - - 6 2 - Total 1 1 9 2 1 Grand Total 65 72 123 8 28

Threat assessments

Of all observations of beach-nesting birds, 80.7% of sightings included threat assessments. Out of the 2,118 observations of beach-nesting birds where threat data was collected, no threat of any kind was observed at 29% of these (611 observations). Fifty-three percent of these were recorded in remote areas in South Australia (Kangaroo Island, Yorke Peninsula and Eyre Peninsula) and a further 21% were recorded in relatively inaccessible areas in Victoria (parts of Wilsons Promontory National Park and the Corner Inlet sand islands). A summary of the percentage of sites in the three states falling within the different threat score categories reveals that in Victoria, there were more sites with green threat scores than any other threat score category and for any other state. In South Australia, sites with yellow threat scores were the most prevalent and in New South Wales, sites with orange threat scores were the most dominant category, the latter most certainly reinforcing why this population is Critically Endangered (Table 8).

Table 8. The percentage of sites (observations) with beach-nesting birds falling within each threat score category in 2018, by state.

Green (1-3) Yellow (4-8) Orange (9-13) Red (14-23) Purple (24+)

VIC 36.3% 24.5% 27.1% 10.6% 1.4% SA 19.1% 37.0% 29.9% 12.6% 1.3% NSW 24.3% 21.5% 39.3% 15.0% 0.0% ALL SITES 25.1% 31.9% 29.7% 12.1% 1.3%

Maps depicting threat score categories have been made for each of the 24 surveyed regions (see Appendices 3, 4 and 5). Since the 2010 count, these maps have revealed a similar pattern of distribution of threat score categories where sites with orange, red, or purple (high level) threat scores being generally located in close proximity to population centres or areas of high recreational use, while sites with green (low level) threat scores being typically located away from these busy areas.

A comparison of the spread of sites with different threat score categories between the 2016 and 2018 counts indicates an increase in the number of sites with low level (green) threat scores and high level (orange) threat scores, and a decrease in the number of sites with high level (yellow and red) threat scores (Table 9). The number of sites with the highest threat level (purple) appears to be approximately stable. At a state level, the most significant change of threat score distribution has

occurred in Victoria, where there has been a decrease in sites with red threat scores. There has been significant investment in Victorian site protection, which aligns with this lowering of threat scores.

Table 9. The percentage difference between spread of threats of sites between 2016 and 2018, by state.

Green (1-3) Yellow (4-8) Orange (9-13) Red (14-23) Purple (24+)

VIC 7.5 3.0 3.4 -14.8 0.7 SA 3.8 -10.0 11.2 -3.6 -1.5 NSW 10.2 -11.1 -1.0 1.9 0.0 ALL SITES 6.0 -6.9 8.2 -6.5 -0.8

Based on the 2018 count data, around a quarter of beach-nesting bird species territories are exposed to the threat of dogs or vehicles during the breeding season (Figures 6a and b). Forty-three percent of all territories had evidence of people being present, and 21% showed evidence of introduced mammalian predator presence (fox and/or cat; Figures 6c and d).

When examining data of just Hooded Plover observations, it is clear that there is substantial pressure from threats on the species. Considering the size of the mainland population (eastern subspecies) – around 1,500 adult birds as detected in this survey – it is particularly concerning that just over a half of all observations recorded during the biennial count showed evidence of people or dogs within 100 metres distance, which is within the buffer of disturbance (Figures 6e and g). Furthermore, 16% of observations had evidence of vehicle presence and 24% had evidence of introduced mammalian predators (Figures 6f and h).

26% 24%

Dogs or prints present Vehicles 74% No dogs 76% No vehicles

(a) % dog occurrence, all species (b) % vehicle occurrence, all species

18%

2%1% 43% Fox present 57% People present Cat present No people present Both present 79% No signs

16%

32%

Dogs or prints present 68% Vehicles No dogs 84% No vehicles

(e) % dog occurrence, Hooded Plover (f) % vehicle occurrence, Hooded Plover

21%

2% 49% 1% 51% Fox present Cat present People present Both present 76% No people present No signs

(g) % people occurrence, Hooded Plover (h) % mammalian predator occurrence, Hooded Plover

Figure 6. Percentage of key threats recorded within 100 m of beach-nesting species’ observations (a- d, all species and e-h, Hooded Plover).

Invasive weeds

Weeds, such as Marram Grass (Ammophila arenaria; deliberately introduced from Europe for dune stabilisation purposes), Sea Spurge (Euphorbia paralias; originating from Europe and presumably introduced in shipping ballast water), and Sea Wheat-grass (Thinopyrum junceiforme; native to Europe and deliberately introduced for dune stabilisation purposes), have been identified as key species that change the structure of beach and foredune habitats in Australia (Cousens et al. 2013). These structural changes alter the resources available (foraging, nesting etc) to Hooded Plovers, leading to either direct impacts (increased predation, mortality or abandonment of beaches) or to more indirect impacts, such as reduced breeding success due to sub-optimal habitat, to the birds.

The density of vegetation estimated during the 2018 biennial count revealed that the majority of Hooded Plovers were sighted within habitats with sparse vegetation or no vegetation (77%, n=673 observations where vegetation density was assessed), which is their preferred nesting habitat. The remainder were present in heavily vegetated areas (23%) and at these sites, weeds were the dominant vegetation at 81% of sites. This included the following number of sites within regions: 24 Warrnambool to Yambuk, 13 Wilsons Prom to Waratah Bay, 10 Lorne to Princetown, 10 Phillip Island, 10 Yorke Peninsula, 8 Seaspray to Corner Inlet, 7 Southeast South Australia, 6 Queenscliff to Lorne, 6 Venus Bay, 6 Kangaroo Island, 5 San Remo to Inverloch, 4 Yambuk to Swan Lake, 4 Mornington Peninsula, 4 Discovery Bay, 3 Fleurieu Peninsula and 1 in the Coorong. Some of the highest weed infestations occurring in western Victoria and south-eastern South Australia are in line with the findings of Cousens et al. (2013), who indicated that the most extensive Marram Grass infestations occur here, making the dune largely unavailable as nesting habitat.

Overall, three major invasive weeds (Marram Grass, Sea Spurge, Sea Wheat-grass) occurred in 50% of all Hooded Plover observations where vegetation was assessed. Marram Grass is rated as being of greater threat to the Hooded Plover than other weeds due to Hooded Plovers showing strong avoidance of Marram vegetated dunes, whilst at low to moderate densities of Sea Spurge and Sea Wheat-grass infestations, they still place their nests amongst these weeds in the foredune and dune. However, Sea Spurge and Sea Wheat-grass are still weeds of serious concern, as once established at high densities, these too prohibit successful use of the foredunes and dunes by breeding Hooded Plovers. Also, of concern, is the spread of Cape Beach Daisy across the border from South Australia into Western Victoria, where this weed appears to be rapidly increasing its distribution. Maps of the

occurrence of the five main weed species of concern for beach-nesting bird species are provided in Appendices 3, 4 and 5.

Volunteers

The 2018 biennial count attracted 367 participants, 33 more participants than in 2016. Table 10 presents the number of participants in each region for 2016 and 2018 and the differences between years. Note that some participants surveyed routes in multiple regions. Victoria had the main growth in volunteer participation, mainly on the Mornington Peninsula, Warrnambool area and Bellarine-Surf Coast. These are key areas of long-term volunteer investment for BirdLife Australia, and the latter two regions have had significant investment from the Australian Government’s National Landcare Program via a partnership with Glenelg Hopkins Catchment Management Authority and BirdLife Australia. Disappointingly, central East Gippsland had the lowest participation, resulting in less survey coverage in this area. This area is in need of future investment in volunteer recruitment, training and support.

Participants with limited experience were coupled with experienced mentors to develop a solid understanding of species identification and behaviour of breeding birds. It can be difficult to engage volunteers in remote locations due to the long distances involved in getting to survey locations and perceived commitment levels. Being time poor was one factor that many people indicated would limit the likelihood that they would volunteer (Cullen 2012). The biennial count is a key pathway through which we can connect volunteers to the Beach-nesting Birds program and increase participation, particularly as it represents a limited, once-off commitment in a count year.

Table 10. The number of volunteers who participated in the 2016 and 2018 biennial counts.

Region 2016 participants 2018 participants Difference Victoria 1. NSW Border to Point Hicks 13 6 -7 2. Mueller River to Lake Tyers 6 11 5 3. Lake Tyers to Seaspray 21 6 -15 4. Seaspray to Corner Inlet 4 8 4 5. Wilsons Prom to Waratah Bay 6 8 2 6. Venus Bay 3 7 4 7. San Remo to Inverloch 7 11 4 8. Phillip Island 22 16 -6 9. Mornington Peninsula 8 18 10 10. Queenscliff to Lorne 22 31 9 11. Lorne to Princetown 3 8 5 12. Princetown to Warrnambool 5 6 1 13. Warrnambool to Yambuk 11 22 11 14. Yambuk to Swan Lake 2 7 5 15. Discovery Bay 1 4 3 Total 134 169 35 South Australia 16. Southeast South Australia 11 11 0 17. Coorong 3 6 3 18. Fleurieu Peninsula 24 12 -12 19. Kangaroo Island 39 63 24 20. Yorke Peninsula 41 41 0 21. Eyre Peninsula 38 32 -6 22. Ceduna and West 3 2 -1 Total 159 167 8 New South Wales 23. New South Wales South 22 24 2 24. New South Wales North 25 22 -3 Total 47 46 -1

DISCUSSION AND RECOMMENDATIONS

The biennial count data help us further our understanding of the extent of suitable Hooded Plover habitat, distribution and size of the population, and in detecting changes in occupancy over time. It was biennial count data that provided strong evidence of overall declines in the eastern mainland numbers of Hooded Plovers, and evidence for loss of occupancy which led to the nomination of the eastern subspecies of Hooded Plover for listing (Vulnerable) under the EPBC Act in 2014.

The biennial count data is complemented by nest monitoring data collected by volunteers through targeted and regular monitoring which reveals rates of recruitment and severity of threats at sites, in turn allowing us to conduct adaptive management to mitigate threats and improve breeding success. The two monitoring regimes are complementary and both provide different data used to assess the health of the Hooded Plover population; one at the population scale to see whether breeding success data translates in to actual trajectory change, and one at the scale of regional recruitment rates and threat trajectories, as tools for evaluating success in conservation actions and as triggers for adaptive management.

The biennial count has historically been largely an unfunded exercise, with the exception of regular investment from the Adelaide and Mount Lofty Ranges Natural Resources Management Board for regional coverage and reporting from 2010. For the first time in 2018, there was significant investment via multiple regional partnerships through the Australian Government’s National Landcare Program. This significantly improved coverage where there was funding, and also ensured that a project coordinator could manage the project and produce this report. In comparison, the unfunded 2016 count report has only just been launched in 2019 due to lack of resourcing. Given the need to track the trajectory of this threatened species and to understand whether investment in on-ground and other threat management actions are effective, maintaining the count over time is critical to monitoring conservation recovery.

Assessing population trajectory change over time

Although the biennial count report provides an approximate indication of population variation over time, the scope of these reports tends to be on the immediately preceding count, and thus at a temporal scale that cannot effectively report on population change. Given the last large-scale analysis (Victoria only) included data from the 1980s to 2006 (Glover 2008), it is timely to now include data from that point onward, particularly given fixed routes were established in 2010 making direct

comparisons between counts easier. An analysis of state and regional population trends over time is currently underway in a collaborative partnership between BirdLife Australia and Deakin University.

Learnings from the launch of the new online data portal

This was the first count where participants could directly enter their own data after the count, instead of posting forms in to be entered solely by the BirdLife Australia National Office. We were pleased that 64% of the surveys conducted had their sightings entered by the survey participants, and while there were teething problems with the polygon boundaries (i.e. in some instances these boundaries did not extend to cover the entire surveyed area), we have made adjustments to these to ensure survey polygons are more accurate for the next count. We will also make sure there are multiple training opportunities for using Birdata prior to the 2020 count, including the new phone app for direct entry in the field. This should dramatically improve data quality and reduce time lags in vetting, analysis and reporting.

The value of threat data

The biennial count provides us with a rare opportunity to gain an understanding of threats at Hooded Plover sites (especially remote sites) that are not typically monitored by volunteers during nest monitoring in the breeding season. As the count involves visiting beaches in three states, covering thousands of kilometres, collecting information on threats is crucial especially in trying to assess the distribution of a particular threat e.g. a weed species. Although some volunteers may find filling out a great deal of information for one observation of a beach-nesting bird a bit excessive, it is imperative that we take the opportunity of visiting these sites (given the expense and effort to get there) and continue to collect this data to better inform our understanding of threat distribution. The field app within Birdata should make recording this data more efficient. A priority for the Beach-nesting Birds team is to make every effort to educate all parties involved about the value of such data.

This data genuinely reveals the major pressures the NSW Critically Endangered population of Hooded Plovers is under, as well as showing some positive signs in this count in terms of more sites moving into the lower threat categories due to our investment in threat mitigation actions. The data can be used to really focus our efforts and learn whether we are making a difference on reducing the threats to improve the birds’ outlook.

Coverage of remote coastline

In surveying remote coastlines, one of the toughest challenges is sourcing enough volunteers to assist. This is particularly relevant to areas such as far eastern Victoria, south-eastern South Australia, Yorke and Eyre Peninsulas. In some areas it may not be practical to engage volunteers due to access challenges, but instead we need to work with land managers/traditional owners or staff to survey these regions, e.g. far west South Australia. Some of the routes require hiking over multiple days and camping in the bush which may not suit some volunteers. Some routes require organising logistics (e.g. boat hire) and reasonable physical fitness to survey the entire route (e.g. Clonmel Island in the region of Seaspray to Corner Inlet). With sufficient notice, it may be possible to get a birding group or team of experienced observers to organise an ‘expedition’ to these remote areas to survey routes together.

Great coverage of survey routes was achieved in the 2018 count in most regions where remote coastlines were included, with the exception of central East Gippsland, which will become a priority region for coverage in the 2020 count. The other major concern lies in areas where a few stalwarts cover large areas over a number of years – e.g. Eyre Peninsula, Far West Victoria and South and East Gippsland. In such areas there is a clear risk of relying heavily on a few key people. Here, a long-term succession plan is required to ensure continuation of survey efforts and more engagement from within the local communities. This project will continue to be supported by BirdLife Australia, through funding from the Australian Government’s National Landcare Program for the 2020 and 2022 counts. This is the first time the biennial count has had significant funding investment to support coordination and coverage and is a great opportunity to accurately assess change in population trajectory over the next three counts.

Importance of flagged and banded bird data

Mead et al. (2013) emphasised the need to collect as much data on leg flags or bands as possible. During the 2018 count, 289 observations of flagged birds were made and 72% of these observations reported the full ID for leg flag engravings or colour flag combinations. This is an incredibly impressive effort and adds greatly to the value of the existing leg flagging projects, as this extensive survey offers a limited opportunity to survey the entire distribution of the species and find flagged birds that may have dispersed. This enables us to more confidently report on survival rates and dispersal movements.

Targeted surveys and further data analysis

This report highlights a few regions of concern where it appears there are declines in Hooded Plover numbers which cannot be attributed to lower coverage during the count. In several of these regions, intensive monitoring occurs during the breeding season and thus this additional data could be further analysed to explore changes in the number of breeding pairs and/or flock sizes, such as for the Mornington Peninsula and South East South Australia. Targeted and repeated surveys in the other regions which we know less about may be required to determine local population size and to identify any potential threats or changes that may be driving these changes. This includes the regions of Venus Bay, Princetown to Warrnambool, Kangaroo Island and Ceduna and West.

ACKNOWLEDGEMENTS

We are extremely grateful to all the participants who partook in the 2018 biennial count. It is an enormous feat to cover so many thousands of kilometres of coast within such a short time frame and getting this snapshot of Hooded Plover numbers is the best opportunity we have in understanding the status of the eastern mainland population. Few projects worldwide can claim that nearly half the world population of a threatened bird species is surveyed in a matter of weeks! The effort of our participants is something one cannot put a value on, and we cannot thank you enough for your outstanding efforts.

In particular, we would like to thank the regional coordinators of all 24 regions. They organised participants to survey routes and helped fill gaps when they could not find volunteers to fill them. Without the help of the regional coordinators, none of this would have been possible. Thanks also to Parks Victoria, and the Department of Environment, Land, Water and Planning (VIC), the Department of Environment, Water and Natural Resources (SA) and the Office of Environment & Heritage (NSW) who provided assistance in the form of quad bikes, boats and staff time. For some regions, we would be unable to access the areas or survey these challenging sites without the assistance of the land managers.

At BirdLife Australia, organisation of the count and the subsequent collating, entering and analysing the resultant data took a substantial amount of time. We are grateful to volunteers Bronwyn Baird, Maddy Taylor, Jonathon Breedon, Danielle de Wit and Catherine Scott for assisting with data entry.

This project is supported by BirdLife Australia in partnership with Glenelg Hopkins Catchment Management Authority, Natural Resources Adelaide and Mount Lofty Ranges, Natural Resources Northern and Yorke, and Natural Resources Alinytjara Wilurara, through funding from the Australian Government’s National Landcare Program.

REFERENCES

BirdLife International (2016). Thinornis cucullatus. The IUCN Red List of Threatened Species 2016: e.T22693883A93429190. http://dx.doi.org/10.2305/IUCN.UK.2016- 3.RLTS.T22693883A93429190.en. Downloaded on 07 March 2019.

Cousens, R., Kennedy, D., Maguire, G. and Williams, K. (2013). Just how bad are coastal weeds? Assessing the geo-eco-psycho-socio-economic impacts. Report to Rural Industries Research and Development Corporation. The University of Melbourne, Melbourne, Australia.

Cullen, M. (2012). Volunteer Satisfaction Survey 2012. BirdLife Australia, Carlton.

Cuttriss A., Maguire G.S., Ehmke G. and Weston M.A. (2015). Breeding habitat selection in an obligate beach bird: a test of the food resource hypothesis. Marine and Freshwater Research. 66; 841-846. https://doi.org/10.1071/MF14213.

Driessen, J. and Maguire, G. (2015). Report on the 2014 Biennial Hooded Plover Count. BirdLife Australia, Carlton.

Ehmke, G., Maguire, G.S., Bird, T., Ierodiaconou, D. and Weston, M.A. (2016). An obligate beach bird selects sub-, inter- and supra-tidal habitat elements. Estuarine, Coastal and Shelf Science. 181; 266- 276. https://doi.org/10.1016/j.ecss.2016.08.050.

Ewers, G., Esbert, N., Hardie, M., Ekanayake, K., Cullen, M. and Maguire, G. (2011). Report on the 2010 Biennial Hooded Plover Count. Birds Australia, Carlton.

Garnett, S., Szabo, J.K. and Dutson, G. (2011). The Action Plan for Australian Birds 2010. Birds Australia, CSIRO publishing, Collingwood.

Glover, H. (2008). Population trends of Hooded Plover Thinornis rubricollis along the Victorian coast, Australia. Honours Thesis. Deakin University, Melbourne.

Mead, R., Yarwood, Y., Cullen, M. and Maguire, G. (2013). Report on the 2012 Biennial Hooded Plover Count. BirdLife Australia, Carlton.

Office of Environment & Heritage, NSW Government (2019). Pied Oystercatcher – profile. https://www.environment.nsw.gov.au/threatenedSpeciesApp/profile.aspx?id=10386. Downloaded on 07 March 2019.

Wetlands International (2019). Waterbird Population Estimates. wpe.wetlands.org. Downloaded on 07 March 2019.

APPENDICES

Appendices 1 to 5 are provided separately due to file size and can be found on the BirdLife Australia website: http://birdlife.org.au/projects/beach-nesting-birds/biennial-count-reports

Appendix 1 Survey instructions, survey datasheet and Birdata data entry instructions used in the 2018 biennial count.

Appendix 2 Modifications to the survey routes for the 2018 biennial count.

Appendix 3 South Australian maps of routes surveyed, species sightings, threats assessed and weed occurrence maps.

Appendix 4 Victorian maps of routes surveyed, species sightings, threats assessed and weed occurrence maps.

Appendix 5 New South Wales maps of routes surveyed, species sightings, threats assessed and weed occurrence maps.