UC Santa Cruz UC Santa Cruz Electronic Theses and Dissertations

Title Strategic planning for biodiversity and ecosystem services: Assessing targets and actions in seabird and water conservation

Permalink https://escholarship.org/uc/item/5gk057z6

Author Ruiz, Diana Madrigal

Publication Date 2020

Peer reviewed|Thesis/dissertation

eScholarship.org Powered by the California Digital Library University of California UNIVERSITY OF CALIFORNIA

SANTA CRUZ

STRATEGIC PLANNING FOR BIODIVERSITY AND ECOSYSTEM SERVICES: ASSESSING TARGETS AND ACTIONS IN SEABIRD AND WATER CONSERVATION

A dissertation submitted in partial satisfaction of the requirements for the degree of

DOCTOR OF PHILOSOPHY in

ECOLOGY AND EVOLUTIONARY BIOLOGY

by

Diana Madrigal Ruiz

December 2020

The Dissertation of Diana Madrigal Ruiz is approved:

______Professor Donald A. Croll, chair

______Professor Bernie R. Tershy

______Professor M. Tim Tinker

______Professor James A. Estes

______Quentin Williams Acting Vice Provost and Dean of Graduate Studies

Table of Contents

List of Tables ...... iv List of Figures ...... v Abstract ...... vi Acknowledgements ...... viii CHAPTER 1 - Introdiction ...... 1

CHAPTER 2 - Turning off the tap: Common domestic water conservation actions insufficient to alleviate drought in the United States of America …………………..5 Abstract ...... 5 Introduction ...... 5 Materials and methods ...... 7 Results ...... 13 Discussion ...... 21

CHAPTER 3 - Improving threatened species assessments by scaling up population viability analysis for threatened seabirds …………………………………….23 Abstract ...... 23 Introduction ...... 24 Methods...... 29 Results ...... 32 Discussion ...... 43 Supplemental Materials ...... 48

CHAPTER 4 - Using meta-population models to guide conservation action ………56 Abstract ...... 56 Introduction ...... 57 Methods...... 60 Results ...... 67 Discussion ...... 79 Supplemental Materials ...... 84 CHAPTER 5 - Conclusion ...... 136 Bibliography ...... 140

iii List of Tables

Table 2.1. Standardized potential monthly water savings per household from commonly promoted domestic water conservation actions ...... 12 Table 3.1, Summary of IUCN Red List criteria for threatened categorizations and present study criteria E adaptation (Table 2.1 IUCN, 2019)...... 27 Table 3.2. Number and percent of species within current Red List threat statuses and seabird mPVA projected status (based on mean projected quasi-extinction risk and criterion E thresholds)...... 39 Table 3.A.1. BirdLife International 2020 threatened seabird species list and rationale for exclusion of 15 species from seabird mPVA database...... 48 Table 3.A.2 . Confusion matrix for ordinal logistic regression test data ...... 55 Table 4.1. Species threat status, mean projected quasi-extinction risk, final mean abundance, and associated intervention scenario...... 70 Table 4.2. Mean credible decrease in projected quasi-extinction risk and proportional increase in mean final abundances under all scenarios relative to baseline...... 77 Table 4.A.1 Threatened Species List and Overlap between New Zealand Department of Conservation Action Plan and seabird mPVA...... 84 Table 4.A.2 Confirmed and Probable Breeding Site List by Species and Country ... 89 Table 4.A.3. Scenario treatment by species and priority ...... 93 Table 4.A.4. Database and literature review of attempted seabird colony movements ...... 99 Table 4.A.5. Instances of viability measures and hierarchical intervention scenario priority congruence...... 127 Table 4.A.6. Log transformed increase in abundance under All scenarios relative to Baseline scenario ...... 131 Table 4.A.7. Prescribed intervention actions by species and action and associated viability gains ...... 132

iv List of Figures

Figure 2.1. Water stress status of U.S. counties in 2010...... 15 Figure 2.2 Counties with potential water stress relief from promoted domestic actions...... 17 Figure 2.3. 2010 County level dominant water withdrawal sectors...... 20 Figure 3.1. Mean projected quasi-extinction risk for 99 threatened seabird species under baseline conditions...... 35 Figure 3.2. Mean of individual species mean projected quasi-extinction risk categorized by current Red List status ...... 37 Figure 3.3a +b. Current vs. our modeled threatened seabird Red List status by family ...... 41 Figure 3.A.1. Predicted probability of Red List status designation over scale quasi- extinction risk...... 55 Figure 4.1. Baseline mean projected quasi-extinction risks and IUCN Red List status for 27 threatened New Zealand seabird species...... 68 Figure 4.2. Decrease in mean projected quasi-extinction risk by intervention scenario priority...... 76

v ABSTRACT

Strategic planning for biodiversity and ecosystem services:

Assessing targets and actions in seabird and water conservation

by

Diana Madrigal Ruiz

Biodiversity and ecosystem function conservation face unprecedented challenges. Underfunding, incomplete policies, and backlogged assessments hinder conservation planning (Myers et al., 2000; James et al., 1999). In this dissertation, I seek to develop data-driven methods to inform strategic conservation of water and seabirds. Fresh water shortages threaten ecosystem function and human health

(Georgakakos et al., 2014; Jiménez Cisneros et al., 2015). In Chapter 2 , I addressed effective water conservation by consumers across the contiguous United States. I determined county-level water stress by relating annual water withdrawal to availability. I adjusted water stress by the potential savings from full adoption of common domestic water conservation actions, such as installing low-flow showerheads. I projected that, during a drought year, the majority of counties would remain water-stressed despite savings. Further, I identified the agriculture sector as the most common dominant water user, suggesting that meaningful consumer water savings should reduce dietary water use. Biodiversity loss is threatened with an average population decline of 68% for assessed species since 1970 (Almond et al.,

2020). A foundational step in conservation is determining species’ relative extinction risk. In Chapter 3 I applied a metapopulation viability analysis (mPVA), engineered

vi in partnership with the Conservation Action Lab and Tim Tinker, to 99 threatened seabird species. I indexed species by relative extinction risk and identified extremely threatened species. I compared the mPVA projections with IUCN Red List threat statuses to identify species that should be prioritized for reassessment and to inform resource allocations. After species have been targeted and prior to implementation, interventions should be evaluated for potential benefit. In Chapter 4 I used the mPVA to analyze the prescribed actions in New Zealand’s national recovery plan for

27 threatened seabird species. I simulated common conservation actions, such as translocation and invasive species removals, at the breeding population level and compared this to status quo scenarios. I found that the majority of species were not predicted to significantly benefit from prescribed interventions. This dissertation guides effective conservation planning by demonstrating means to assess current risk and project the benefits of conservation actions.

vii ACKNOWLEDGEMENTS

The text of this dissertation includes reprints of the following previously published open-access material: Chapter 2: Ruiz, DM, Tallis, H, Tershy, BR, Croll,

DA. 2020. Turning off the tap: Common domestic water conservation actions insufficient to alleviate drought in the United States of America. PLOS ONE . 15(3), e0229798.

First, I would also like to thank the generous funding sources that made my graduate education and research possible, including the U.S. Department of

Education Graduate Assistance in Areas of National Need Fellowship, the Eugene

Cota-Robles Fellowship, the Graduate Division of the University of California Santa

Cruz, the David and Lucile Packard Foundation, and the National Fish and Wildlife

Foundation.

I sincerely thank my academic advisors, Don (Diego) Croll and Bernie Tershy for their guidance and confidence in me. I will continue to emulate Don’s big-picture perspective and Bernie’s ability to make any idea exciting and accessible. The

Conservation Action Lab was my first choice as it emphasizes training leaders in applied conservation research. The coaching I received exceeded my expectations in technical, professional, and personal development.

Next, I thank my remaining dissertation mentors. Tim Tinker advised me as a committee member and coauthor, always ensuring I understood the principle underlying his notes and methods. Jim Estes served on both my comprehensive exam and reading committees and brought a wide ecological lens to my research. As a

viii coauthor on my first scientific publication, Heather Tallis offered me genuine kindness and support. She is a true champion of early career scientists. Kelly Zilliacus helped me in many roles: coauthor, lab manager, lab mate, and friend. Kelly was impossibly quick to assist my progress from teaching me sound data management, generating figures, to providing feedback on manuscripts.

I could not be more grateful for my lab mates and friends who helped me navigate the program, shared their experiences, and saw me through the inevitable challenges. ¡Gracias Luz de Wit, Luli Martinez, Abe Borker, Susy Honig, Erin

McCreless, Angela Quiros, Dena Spatz, Asha de Vos, Joe Cutler, Melissa Cronin, and

Beth Howard!

I thank the many EEB administrators and faculty who helped me throughout the years. Pete Raimondi carefully guided me through statistical analyses. Rita Mehta managed disbursement of my fellowships. Judy Straub, Kris West, Sarah Amador and

Veronica Larkin gave me reliable administrative support.

I would not have attempted a Ph.D. without those who championed me prior. I thank Jim and Shirley Modini who set the north star of my conservation ethic. Sherry

Adams gave me my first field experience and honest advice. Always with kindness,

Judy Johnston held high expectations of me and sought opportunities for me to achieve them. During my time with Audubon Canyon Ranch, John Petersen, Jesse

Grantham, and John Kelly fostered my growth as a leader and treated me like a colleague despite my certain inexperience. Sarah Swope, John Harris, and Jenn

Smith, my professors at Mills College, made ecological principles enthralling,

ix showed me patience, and were instrumental in guiding my graduate school applications. Susie Bennett entrusted me with a variety of field studies at the National

Park Service when I was a novice volunteer.

I also thank my personal support system. In addition to those above, I could not have graduated without my lifelong friends: Aja Heisler, Mia Winkler, Paulo

Quadri, Chris Schwind, and Alex Smith. My east coast, Fryxell-Johnson-Smith-

Uremovich family got me over the finish line, nourishing me with thoughtfulness, laughter and hot Cheetos. For cheerful encouragement, which brightened my time, I thank my brother-in-law Josh (Broshua) Banerje, and the world’s best pups, Pepe and

Rio. I owe every achievement to my Ruiz family. My parents, Tomas and Elidia, inspire me everyday by proving that dedication and hard work can yield incredible success, like their own. I aspire to live up to the sacrifices they have paid for me. My sisters, Grisel and Erika, will always be my role models, providing a template to dream big, thrive outside my comfort zone, and follow my values. Finally, I am forever grateful to have met and married another EEB alum, Dave Fryxell. Through companionship and example, Dave keeps me smiling and eager to take on the next chapter.

x CHAPTER 1

Introduction

An estimated one million plant and animal species are threatened with extinction (IPBES, 2019). Biodiversity loss impedes ecosystem services, including those affecting human well-being (Hooper et al., 2012; Dobson et al., 2006). Yet current extinction rates are estimated to be 1,000 times higher than natural background rates with future rates expected to be 10,000 times higher (de Vos et al.,

2015). At the same time, conservation is underfunded (James et al., 1999; Myers et al., 2000). Strategic analyses are needed to better inform conservation planning for biodiversity protection and sustained ecosystem services. In this dissertation, I develop analyses to address challenges for effective strategies in two critical, but distinct, topics: fresh water conservation and threatened seabird conservation.

Fresh water shortages increasingly threaten ecosystem services and human health (Jiménez Cisneros et al., 2015). Nearly 80% of the global population is under high water scarcity threat and 65% of river discharge and associated species are under at least moderate water scarcity threat (Vörösmarty et al., 2010). In Chapter 2 , I address the problem of effective fresh water conservation by consumers. Specifically,

I test whether the potential benefits from full adoption of domestic water conservation action are sufficient to alleviate water stress across the continental U.S. To do this, I determined annual water stress at the county-level by calculating a water withdrawal to water availability ratio. I then adjusted water withdrawal by estimated household savings from full adoption of 13 commonly promoted water conservations actions,

1 such as turning off faucets when washing dishes and installing low-flow toilets. I found that the majority of counties would not be alleviated of water stress despite savings from conservation actions. Further, I determined county-level dominant water use sectors finding that the agriculture sector withdrew the majority of water in more counties than other sectors. These findings present the first national evaluation of water conservation campaign efficacy while suggesting that substantive consumer- driven water savings should target reducing dietary water use.

Biodiversity loss can disrupt ecosystem services, trophic interactions, and ecosystem resilience (Estes and Duggins, 1995; Tilman, 1996; Worm et al., 2006).

Still, populations of mammals, birds, reptiles, and other vertebrates have declined on average by 68% since 1970 (Almond et al., 2020). Extinction risk is not only influenced by population size, but also correlates with life history traits such as large body size, high trophic position, migratory status, and slow sexual maturation (Pimm et al., 1988; Purvis et al., 2000). Unsurprisingly, seabirds, which generally track these traits, are highly threatened, accounting for over 25% of historical marine extinctions with 31% of species currently listed as threatened by the IUCN Red List of

Threatened Species (Critically Endangered, Endangered, Vulnerable) (Dulvy et al.,

2003; IUCN, 2020). If extinction trends continue, the consequential roles seabirds play in ecosystem processes may be lost such as, fertilizing near shore communities, creating wide-ranging nutrient connectivity, spurring ecotourism, and generating tons of harvestable, nutrient-rich guano (Sanson, 1994; Croll et al., 2005; Honig and

Mahoney, 2016).

2 Effective conservation planning relies on accurate extinction risk assessments to inform investment. In Chapter 3 I evaluate the relative vulnerability of 99 threatened seabird species over the next 100 years using a metapopulation viability analysis (mPVA) developed in partnership with Dr. Tim Tinker, Dr. Don Croll, Dr.

Bernie Tershy, and Kelly Zilliacus. The mPVA uses a stage-structured matrix to project metapopulation trajectories while accounting for parameter uncertainty and environmental and demographic stochasticity. Estimated final abundance of 10,000 simulations and the proportion of simulations to fall below a quasi-extinction threshold define viability measures. I identified species at the highest relative extinction risk and additional species that require further research to reliably gauge vulnerability. These findings were contrasted with IUCN Red List assessments to identify species projected at higher risk than their current listing, and understand discrepancies in different Red List assessment criteria. These results inform research needs and resource allocation.

Once species have been targeted, management gains should be tested prior to implementation. In Chapter 4 I develop a framework to estimate intervention benefits and assess the accuracy of perceived effect based on prescription priority.

Using the mPVA, I examined the predicted benefits from prescribed actions in the

New Zealand national plan of action for 27 threatened seabird species. Prescribed actions included common management interventions, such as translocation and invasive species removals. I found that under comprehensive conservation scenarios, extinction risk was not mitigated for the majority of species. I also found mismatches

3 in the frequency of prescribed actions and projected effect. This analysis provides a means to prioritize threatened species and conservation actions to efficiently meet conservation targets.

4 CHAPTER 2

Turning off the tap: Common domestic water conservation actions insufficient to alleviate drought in the United States of America

Abstract

Climate change is exacerbating drought and water stress in several global regions, including some parts of the United States. During times of drought in the

U.S., municipal governments, private water suppliers and non-profits commonly deploy advocacy campaigns and incentive programs targeting reductions in residential water use through actions including: repairing leaks, shutting off taps, and installing new water-saving appliances. We asked whether these campaigns have the potential to alleviate water stress during drought at the county scale by estimating the potential impact of full adoption of such actions. In 2010, we show that the maximum potential use reductions from these residential actions may only alleviate water stress in 6% (174) of U.S. counties. The potential impact of domestic programs is limited by the relative dominance of agriculture water withdrawal, the primary water user in

50% of U.S. counties. While residential actions do achieve some water demand savings, they are not sufficient to alter water stress in the majority of the continental

U.S. We recommend redirecting advocacy efforts and incentives to individual behaviors that can influence agricultural water use.

Introduction

Drought intensity and frequency are increasing in some regions of the United

States (Georgakakos et al., 2014) resulting in increasing public attention to water

5 conservation. For example, regionally, 30-40% of the western U.S. has experienced sustained drought in recent decades (Pulwarty et al., 2005a) and the concurrence of drought and heat waves has increased in duration and frequency across the US from

1990-2010 (Mazdiyasni and AghaKouchak, 2015). Over the same timeframe, human demand has increased in these areas leading to predictions that consumptive needs may not be met, such as along the Colorado River (Bates et al., 2008; Pulwarty et al.,

2005b; Pulwarty and Melis, 2001; Tilman, 1999). These stresses have prompted greater public interest in the contributing to water conservation, with the United

States ranking as the country with the highest proportional of internet search for the terms “water savings “in 2018 (GoogleTrends, 2018). Adults surveyed nationwide in

2013 indicated that 87% were willing to conserve water to combat drought and over

35% of respondents believed the public should be the first to conserve water ahead of industry, cities, and agriculture (Stoutenborough and Vedlitz, 2014).

Common public advocacy programs for water conservation promote individual-based strategies directed at reducing domestic water use (San Francisco

Public Utilities, 2016; Arizona Department, 2011). However, hydrologic and agronomic studies indicate that the agricultural sector, not the domestic sector, typically dominates water withdrawals (Postel, 2000; Postel et al., 1996). Within the agricultural sector, different food products and farming systems have widely varying water footprints and diet and product choices have the potential to exert a large influence on agricultural water demands (Tilman et al., 2002; Hoekstra and

Mekonnen, 2012; Mekonnen and Hoekstra, 2012). Despite this knowledge, advocacy

6 and incentive programs continue to focus on individual domestic behaviors and water conservation options on the apparent assumption that these activities have the potential to significantly contribute to water stress alleviation during drought. To date, no studies have evaluated whether these measures can reduce total water withdrawals sufficiently to alleviate water stress across the U.S. If currently promoted actions do not contribute sufficiently to alleviate water stress, it may be more effective and efficient to socialize and incentivize other individual actions that have greater impact. The most commonly promoted public water conservation actions include: repairing household leaks, taking shorter showers, closing faucets whenever possible, running dishwashers and washing machines only when full, and installing low-flow appliances (San Francisco Public Utilities, 2016; Environmental Protection

Agency, 2015; EPA, 2016b, 2013c; Department of Natural Resources, 2014; EPA,

2016a, 2013b; Alliance for Water Efficiency, 2016; EPA, 2013a; EnergyStar, 2003;

South Staff Water, 2009). We examined whether full adoption of these actions in

2010 (a drought year) could lead to annual water savings sufficient to relieve water stress at the county scale across the continental U.S. Our intent is not to prescribe specific actions within a county or characterize detailed hydrologic processes and water stress conditions. Rather, we ask at a high level whether strongly-supported domestic efficiency practices are likely to alleviate water stress in the U.S.

Materials and methods

We applied a simplified approach to estimate order of magnitude impacts of domestic efficiency methods on county scale water stress in the lower 48 states of the

7 U.S. We determined which counties were water stressed in 2010 and calculated adjusted county level water use by assuming total household adoption of promoted domestic water conservation actions. Adjusted withdrawal calculations were used to determine which counties would change status from water stressed to non-water stressed under full adoption of common domestic efficiency practices in 2010.

Water use sector

Sector-specific water use data were compiled from United State Geological

Survey (USGS) records for 2010 (Maupin et al., 2014), the most recent drought year with complete water monitoring data. Drought in the contiguous U.S. was more extensive between 2001-2010than at any other time since the 1950s (Peterson et al.,

2013). Total annual withdrawal rates (saline +fresh) for all US counties were binned into three sectors: industry, agriculture, and domestic. The industrial sector included water withdrawal used for industry, mining, and thermoelectric power. The agricultural sector captured water used for irrigation, livestock, and aquaculture. The domestic sector included water used for domestic and public supply. We calculated percent water withdrawal by sector and designated dominant water use sector

(accounting for 50% or more of total county withdrawals) for each county.

Water stress

We defined water stress based on the ratio of water withdrawal to water availability for each county (Roy et al., 2005; Averyt et al., 2011; Sun et al., 2008a,

2008b). We considered any county with a water stress ratio at or above 0.4 to be water stressed, adopting the commonly used thresholds in the water drought and

8 scarcity literature (Mazdiyasni and AghaKouchak, 2015; Bates et al., 2008; Pulwarty et al., 2005b; Pulwarty and Melis, 2001). Ideally, water stress values would be calculated at a high temporal resolution to capture variations in important hydrologic drivers (rainfall, baseflow, groundwater recharge rates, evapotranspiration rates, etc.) and demand drivers (temperature, growing seasons, competing demands, water prices, etc.). However, monthly or daily water withdrawal (demand) data are not publicly available at the county level in the U.S. Given the paucity of temporally variable demand data, we opted for a simple water stress calculation rather than complex modeling exercises. As our intent was to understand whether domestic efficiency changes are of the order of magnitude to alleviate water stress, we consider this a sufficient order of magnitude approximation approach.

Water withdrawal rates were defined as total annual saline and fresh water withdrawals (Maupin et al., 2014). Water availability was calculated monthly from

USGS data (Department of Natural Resources, 2014) as the difference between total monthly evapotranspiration (L) and total monthly precipitation (L). Monthly county water availability values were then summed across the year to generate an annual water availability metric for each county. Reported annual water withdrawal values were then related to these annual availability values to estimate a water stress ratio per county as described above.

This simplified approach does not represent important temporal relationships between rainfall, extraction, return flows, or groundwater, and only considers locally available surface water supplies, overlooking major inter-basin water transfers. In

9 addition, analyses were bounded by political governance boundaries (counties) rather than watersheds. Such simplified calculations have been used in other contexts to explore relative water stress conditions (Oki and Kanae, 2006; Vörösmarty et al.,

2000; Alcamo et al., 2003), and were deemed sufficient here to approximate the significance of domestic efficiency approaches and to provide an estimate in the absence of more temporally resolved withdrawal data.

Conservation-adjusted withdrawals

We calculated the potential water savings per household from full adoption of twelve commonly recommended and reliably measured water conservation actions

(Table 1). This criteria led to the exclusion of actions not found to be regularly recommended by water governing or advocacy groups or for which consistent savings could be calculated, for example, limiting toilet flushing, washing cars on lawns, or changing household water pressure. While our study does not incorporate all possible domestic water conservation actions, it represents those most commonly recommended by government and activist groups. Assuming full adoption of these actions provides an upper bound estimate of the largest impact that common water efficiency campaigns could have on domestic water use. We gathered estimates of potential savings per U.S. household for each action and applied the most conservative estimates for all calculations. When data permitted, projected savings from new installations were adjusted to exclude households already fitted with water efficient appliances (S1 Table). Likewise, actions influenced by age or sex were also adjusted to reflect the county-level demographics of the population assumed to adopt

10 efficiency actions. Annual single household savings was based on U.S. average household occupancy in 2010 (Lofquist et al., 2012). Annual single household savings estimates were multiplied by the number of households in each county to estimate county-level potential annual water savings. We then determined the impact of these domestic actions on water stress by subtracting our estimated county-level potential annual water savings from reported total annual withdrawal rates and recalculating water stress for all counties as above. We could not account for differences in savings that might result from variance in household water pressure because we could not identify a data source capturing this variance and installation of pressure regulating devices at the county level.

11 Table 2.1. Standardized potential monthly water savings per household from commonly promoted domestic water conservation actions Water Conservation Action and Reference Water Savings (L household -1

-1 Repair household leaks month ) (Environmental Protection Agency, 2015; Perlman, 2015) 3155 Shorten shower time from 8 minutes to 5 minutes (EPA, 2015) 2930 Close faucet when brushing teeth (EPA, 2016b) 2344 Install WaterSense labeled low flow toilets (EPA, 2013c) 1322 Install WaterSense showerheads (EPA, 2016a) 726 Install ENERGYSTAR clothes washer (EnergyStar, 2003) 431 Only run full loads in dishwasher (Department of Natural Resources, 2014) 399 Close faucet when shaving (Department of Natural Resources, 2014) 333 Install WaterSense irrigation controller (EPA, 2013b) 324 Install ENERGYSTAR dishwasher (Alliance for Water Efficiency, 2016) 316 Install WaterSense bathroom faucet and aerators (EPA, 2013a) 158 Close faucet while hand washing or rinsing dishes (South Staff Water, 2009) 46 Total Maximum Potential Water Savings 14,213

Water savings for interventions estimates were primarily drawn from water governing agencies reports such as the US Ecological Protection Agency and local governments. Further research is needed to characterize county specific current water use practices and potential gains. Rebate programs provide important information on

12 water efficient replacements; however miss households that do not participate in the programs. Here estimates on potential water savings assume that national water use practices averages can be applied across counties. For the nationwide view of this study, these assumptions are permissible.

Results

In 2010, the majority (92%) of the 3,109 US counties in the lower 48 states were water stressed based on our basic water stress indicator (Figure 2.1). As 2010 was a drought year, this high rate of county-level water stress was expected. Estimated reductions in domestic water withdrawals from full household adoption of commonly supported water saving measures would release 174 counties (6% of all continental counties) from water stress (Figure 2.2). Domestic water use dominated in these mostly rural, eastern and mid western counties.

13 Figure 2.1. Water stress status of U.S. counties in 2010, a drought year. The majority of counties (2,853) in the continental United States were water stressed (red), while only 256 counties were free from water stress (blue).

14

15 Figure 2.2 Counties with potential water stress relief from promoted domestic actions. Only 174 (6%) of all counties have the potential to be relieved from water stress (yellow) by full adoption of conventionally supported household water conservation actions.

16

17

Including the potential impact of domestic water saving actions, 86% of counties remained water stressed. This shortfall is likely explained by the dominance of the agricultural sector withdrawals in most counties (Figure 2.3). Of the 2,242 counties with a dominant sector, agriculture was dominant in approximately twice as many counties (1,103 counties) as either the domestic (570 counties) or industrial

(569 counties) sectors.

18 Figure 2.3. 2010 County level dominant water withdrawal sectors. Blue indicates >50% water withdrawal attributed to the domestic sector. Red indicates >50% of water withdrawal attributed to the industrial sector. Green indicates >50% of water withdrawal attributed to the agricultural sector.

19

20

Discussion

Given these findings, the potential impact of domestic savings may be inflated relative to potential savings in other sectors. The national dominance of agricultural withdrawals suggests that water stress will not be alleviated in most counties unless agricultural water efficiencies are achieved. Further emphasizing the need to address agricultural water withdrawals, this sector’s water use dominates in regions projected to experience the greatest future increases in drought (Georgakakos et al., 2014; Chen et al., 2013; Wehner et al., 2011).

There is a substantial literature describing measures for individuals to influence agricultural sector water withdrawals. For example, individuals may tailor their diet to reduce their water footprint. Additionally, consumers could pressure suppliers to reveal information about how well matched crops are to growing region conditions.

For example, cotton, paddy rice, alfalfa and other high water use crops grown in wetter regions are less likely to stress local water supplies than the same crops grown in drier regions. Several major companies such as Wal-Mart and Kellogg’s have started to improve their supply chain tracking and could begin to provide this kind of information if buyers demanded it. Consumers could also reduce their portion of food waste, and encourage farmers to reduce waste that happens in harvesting and processing, minimizing the amount of water ‘lost’ through the production and disposal of uneaten food (Water Saving Tips- Food Choices, 2016, Water

Conservation Facts And Tips, 2016; Lundqvist et al., 2008).

21 However, these potential water conservation measures are seldom recommended in water-saving campaigns, while the domestic actions evaluated here dominate. Our analyses suggest that domestic water savings advocacy and incentive programs will fail the majority of the time in the United States because domestic water use is not the dominant driver of water stress, and available household savings measures are not sufficient to transition the majority of counties out of water stress during a drought.

Promoting individual water conservation actions targeted at reducing agricultural withdrawal have a much higher potential to significantly improve water savings and promote longer term water security.

22 CHAPTER 3

Improving threatened species assessments by scaling up population viability analysis for threatened seabirds

Abstract

A key step to prioritize threatened species for conservation is understanding their relative extinction risk. The IUCN Red List of Threatened Species is a standardized list of species and relative extinction risk. It uses five criteria to assess extinction risk, but only the rarely used criterion E, involves quantitative analysis. We applied a meta-population viability analysis to 99 IUCN Red List threatened seabirds using risk thresholds consistent with criterion E. We found 33 seabirds had >50% mean projected extinction risks over 100 years, and 25 species had mean projected extinction risks above thresholds for their current threat statuses. For these species, our analysis captures viability threats missed by shorter timeframes and criteria based on symptoms of extinction rather than risk. Ordinal logistic regression showed that overall, our measure of extinction risk was positively related to Red List status, although only 19% of categorizations aligned between current Red List status and our mPVA categorization. Red List and mPVA assessment mismatches indicate differences in conservatism between demographically-explicit projections of our mPVA model and the non-demographic criteria of the Red List. The families

Phalacrocoracidae, Spheniscidae, and Alcidae included the greatest proportion of species with projected extinction risks greater than 50%. This application of the seabird mPVA provides a systematic and complimentary analysis to Red List

23 assessments, and is well suited to evaluate outcomes of conservation interventions.

Thus, we propose a process for bolstering threatened species assessments and conservation planning through the scaling up and generalization of population viability analysis for threatened seabird conservation.

Introduction

An estimated one million plant and animal species are threatened with extinction (IPBES, 2019). Extinction threatens a growing number of species, and this biodiversity loss impacts important ecosystem processes and reduces human well- being (de Vos et al., 2015; Dobson et al., 2006; Worm et al., 2006). At the same time, inadequate resources are available to meet conservation need (Myers et al., 2000).

While it is estimated that 177.5 billion USD annually is needed to meet current conservation targets, only a fraction of this total is spent (~24.3 billion USD annually)

(Waldron et al., 2020; Tauli-Corpuz et al., 2018). One way to strategically apportion conservation funds is to focus on the species most threatened with extinction. The

IUCN Red List of Threatened Species (hereafter Red List) provides a globally agreed upon standard for assessing species extinction risk and currently categorizes ~32,000 species as threatened (Critically Endangered, Endangered, or Vulnerable) (IUCN,

2020).

Seabirds are a particularly threatened group of marine species. They account for over 25% of historical marine extinctions and are the most endangered marine group with 114 species (31%) currently listed as threatened on the Red List (Dulvy et al., 2003; IUCN, 2020). Aspects of their life history make them particularly

24 vulnerable to human impacts including habitat destruction, introduction of non-native predators on breeding islands, and fisheries bycatch. At the same time, seabirds play consequential roles in ecosystem services and functions including fertilizing near shore communities, creating wide-ranging nutrient connectivity, spurring ecotourism, and generating tons of harvestable, nutrient-rich guano (Sanson, 1994; Croll et al.,

2005; Honig and Mahoney, 2016). Thus, understanding the factors that predict relative extinction risk of seabird species is a logical step in prioritizing conservation efforts and mitigating population declines.

The Red List uses a set of five criteria to standardize extinction risk assessments across species from different taxa, life histories, and available population data (Table 2.1, IUCN, 2019). Criteria A – D represent indirect, qualitative indices of risk, largely based on life history correlates that are assumed or have been demonstrated to be predictive of extinction risk. Only criterion E is a direct quantitative analysis of extinction risk using some type of population modeling, such as population viability analysis (PVA). A recent study found that across taxa, no threatened species had been categorized based on criterion E (Collen et al., 2016).

There are several explanations for this deficiency. Analyses such as PVA used in quantitative criterion E usually require extensive data in order to reliably project population dynamics and account for the demographic impacts of multiple threats. In addition to these data requirements, quantitative modeling expertise is required.

Additionally, because criteria A-D rely upon incomplete indicators of species condition (primarily population size and population structure), thresholds in these

25 criteria are generally designed to be more precautionary when compared with the population modeling approach used in criterion E (Gärdenfors, 2000). Lastly, Red

List guidelines direct species to be assigned the highest qualifying threat status (i.e. lowest listing threshold) by any criteria. As a result, most species are assessed based upon proxies of extinction risk (criteria A-D), and seldom assessed using PVA style models (criterion E).

26 Table 3.1, Summary of IUCN Red List criteria for threatened species categorizations and present study criteria E adaptation (Table 2.1 IUCN, 2019). EoO= Extent of occupancy, AoO= Area of occupancy, yrs.=years Criterion Critically Endangered Endangered Vulnerable A, Population ≥ 90% (and ceased) or ≥ 80% ≥ 70% (and ceased) or ≥ 50% ≥ 50% (and ceased) or ≥ 30% decline observed, estimated, inferred, observed, estimated, inferred, observed, estimated, inferred, suspected or projected suspected or projected suspected or projected B, Geographic EoO 100 km 2 or AoO < 10 km 2 EoO < 5,000 km 2 or AoO < 500 EoO< 20,000 km 2 or AoO< 2,000 range and any two of: severe km 2 and any two of: severe km 2 and any two of: severe fragmentation or few locations, fragmentation or few locations, fragmentation or few locations, continued decline in habitat, continued decline in habitat, continued decline in habitat, extreme fluctuations in extreme fluctuations in occurrence extreme fluctuations in occurrence occurrence or occupancy or occupancy or occupancy

27 C, Small < 250 mature individuals and < 2,500 mature individuals and < 10,000 mature individuals and population and any of: an observed, estimated or any one of: an observed, estimated any one of: an observed, estimated decline projected decline of 25% in the or projected decline of 20% in the or projected decline of 10% in the longer of 3 yrs. or 1 generation, longer of 5 yrs. or 2 generations, longer of 10 yrs. or 3 generations, an observed, estimated or an observed, estimated or an observed, estimated or projected decline and low or projected decline and low or projected decline and low or concentrated mature individuals concentrated mature individuals in concentrated mature individuals in in subpopulations, or extreme subpopulations, or extreme subpopulations, or extreme fluctuations fluctuations fluctuations D, Very small < 50 mature individuals < 250 mature individuals < 1,00 mature individuals or AoO or restricted < 20 km 2 or number of locations ≤ population 5 E, Quantitative Probability of extinction in the Probability of extinction in the Probability of extinction in the analysis wild ≥ 50% the longer of 10 yrs. wild ≥ 20% in the longer of 20 yrs. wild ≥ 10% in 100 yrs. or 3 generations (100 yrs. max.) or 5 generations (100 yrs. max.) Seabird mPVA Quasi-extinction risk ≥ 50% in Quasi-extinction risk ≥ 20% in 100 Quasi-extinction risk ≥ 10% in 100 analysis 100 yrs. yrs. yrs.

We developed a generalized meta-population viability analysis (hereafter seabird mPVA) to assess relative viability of 99 threatened species of seabirds. This tool is freely available ( https://nhydra.shinyapps.io/mPVA1/ ), and compliments Red

List assessments, which largely rely upon symptoms of endangerment (criteria A-D).

A demographic, age-structured PVA model is well suited to predict seabird viability given trends in population structure and life history traits, and the fact that threats often impact specific life history stages. Seabirds are an ideal group to scale-up the use of PVAs due to their broadly similar and relatively well-known life history patterns, well-identified threats, and high concentration of threatened species in well- defined geographic locations (oceanic islands). Specifically, seabirds are generally long-lived, slow to reproduce, have high breeding site fidelity, and discrete life stages

(Schreiber and Burger, 2001). The major threats to seabirds: habitat loss, invasive species predation, human disturbance, overexploitation, pollution, and incidental bycatch from fisheries (Croxall et al., 2012b), are similarly amenable to PVA modeling.

Our seabird mPVA is designed to evaluate the relative extinction risk of each species under current status as well as under various conservation scenarios. Here we focus on determining the baseline viability of the 99 threatened species under current conditions and compare these viabilities to IUCN Red List threat assessments. We compare our direct risk assessments based upon mPVA models (criterion E analog) with those based upon proxies for extinction risk (criteria A-D). This approach facilitates examining discrepancies in threat status based upon criteria and identifies

28 the most vulnerable of threatened seabird species relative to current priority standards. The seabird mPVA is the first application of PVA across nearly all threatened species within a highly endangered taxon to set conservation priorities.

Methods

Model basics & baseline results

We conducted analyses of 99 seabird species, all of which are insular breeders and are classified as Vulnerable, Endangered, or Critically Endangered by the Red

List (version 2020.2). These species represent 87% of currently threatened seabird species (Supplemental Materials, Table 3.A.1). Our seabird mPVA is demographically structured and spatially explicit, and thus requires data on both seabird population characteristics and as well as information on breeding locations.

Our threatened seabird database (UCSC, 2020) contains extensive information on seabird population status, trends, life history traits and threats, and is continually updated as new data becomes available, with the most recent version completed in

July 2020. Data for breeding locations and island-based characteristics were extracted from the Threatened Island Biodiversity Database ( http://tib.islandconservation.org )

(TIB Partners, 2018).

The species included in this study span 10 taxonomic families and 1,046 known or probable subpopulations. We used Bayesian methods to estimate parameters required for the mPVA, including fecundity, stage-specific survival, annual breeding probability, and dispersal at the species-wide or species-island level.

Our Bayesian approach accounted for the paucity of data for many species by

29 utilizing available data from related species as priors for data poor species. We modeled species-specific threat impacts to survival using a proportional hazards approach, which readily allows incorporation of multiple competing threats (e.g., the cumulative effects of suites of invasive species) to calculate stage-specific survival rates. Survival, growth and fecundity rates were combined to parameterize a stage- structured projection matrix used to track annual transitions among three independent life history stages: 1) sub-adults, 2) breeding adults, and 3) non-breeding adults. Note that because the model had an annual time step, we integrated the within-year transitions of immature life stages (eggs, chicks and fledglings) into the reproduction term. Transition matrices for multiple breeding sites were then combined into a meta- population matrix that described demographic transitions within breeding colonies as well as spatial transitions between breeding colonies (Caswell, 2001). A detailed description of seabird mPVA structure and mechanics is provided online at https://nhydra.shinyapps.io/mPVA1/.

To parameterize the model we gathered data from published studies, grey literature, and expert review. These data were used to develop distributions of estimates for each parameter. Prior distributions were then customized for each species by using a weighting approach for taxonomic relatedness (i.e. data collected from the focal species had greatest weight, while data from less related species had lower weight, and data from other taxonomic families were excluded altogether). This approach naturally resulted in well-informed priors for species that were data-rich and more diffuse priors for data-poor species that were reliant on life history data from

30 other taxa. We then fit the model for each species to the observed data on abundance and trends (based on the IUCN Red list) to obtain posterior distributions of mPVA parameters that were consistent with published information (Supplemental Materials,

Appendix A).

Projected quasi-extinction risk and Red List status

Our seabird mPVA was used to projected meta-populations forward for 100 years and replicated 10,000 times across two hierarchical loops to account for environmental and demographic stochasticity (inner loop) and parameter uncertainty

(outer loop). The proportion of inner loop simulations that fell below a quasi- extinction threshold for each iteration of the outer loop provided a point estimate of projected quasi-extinction risk; the distribution of these point estimates across outer loop iterations (each of which drew a different combination of parameter values from the species-specific joint posterior distributions) were used to compute 95% confidence intervals around the quasi-extinction risk estimates. We standardized quasi-extinction thresholds based on literature review, but we note that guidelines for quasi-extinction thresholds are variable for seabirds. Therefore, we chose thresholds that are considered conservative for large and rare species. We set quasi-extinction thresholds as 50 adult females (100 individuals) for species with an initial population above 200 pairs or 10 adult females (20 individuals) for species with an initial population ≤ 200 pairs. We related species mean projected quasi-extinction risks to

Red List thresholds described in criterion E (Table 3.1).

31 Under criterion E, the Red List prescribes three thresholds for categories of extinction risk: Critically Endangered (≥ 50% probability of extinction in 10 years or

3 generations, whichever is longer, with a 100 years maximum), Endangered (≥ 20% probability of extinction in 20 years or 5 generations, whichever is longer, with a 100 year maximum), and Vulnerable (≥ 10% probability of extinction in 100 years). The

Red List defines extinction as population size reaching zero. The average generation length for seabirds in this study is 16.4 years. We identified 23 instances for these species where the 100-year maximum would be met. We considered 100 years to be a practical timeframe for conservation planning and biologically relevant. Therefore, we did not directly apply Red List thresholds to seabird mPVA outcomes in favor of a standardized timeframe (100 years) and a conservation-minded quasi-extinction threshold.

We conducted an ordinal logistic regression to determine if mean projected quasi-extinction risk is positively related to Red List status (R package MASS ver.

7.3-51.6) and a Brant test to test the parallel regression assumption (R package brant ver. 0.3-0).

Results

Seabird mPVA outputs and Red List status

Of the 99 threatened seabird species in this study, 33 had a mean projected quasi-extinction risk ≥ 50% (Critically Endangered risk threshold) over 100 years under current (baseline) conditions. Of these, 16 species had a projected quasi- extinction risk of ≥75% with 95% probability (Figure 3.1). Mean projected quasi-

32 extinction risk represents the relative likelihood of model-projected abundances dropping below quasi-extinction threshold, and should not be treated as an absolute prediction of risk. The 95% CI provide an indication of the relative degree of uncertainty around these mean projected values, an important consideration for any management decisions. Of the 66 threatened seabird species with a mean projected quasi-extinction risk below 50%, seven had a 95% confidence interval that exceeded

50% extinction risk, indicating that we cannot confidently assume that the actual mean risk is indeed less than 50%. Mean projected quasi-extinction risk met thresholds associated with Red List categories for Vulnerable (10% - 19%) and

Endangered (20% - 49%) in six and seven species, respectively. Mean projected quasi-extinction risk was below 10% for 53 species (Figure 3.1).

33 Figure 3.1. Mean projected quasi-extinction risk and 95% confidence interval at year 100 for 99 threatened seabird species under baseline (current) conditions. Colors indicate current Red List threat status (blue = Vulnerable, yellow = Endangered, red = Critically Endangered). Horizontal lines indicate extinction risk minimum thresholds for Red List statuses associated with criterion E (0.10 = Vulnerable, 0.20 = Endangered, 0.50 = Critically Endangered).

34

35 Of the 33 species with a mean projected quasi-extinction risk ≥ 50%, 13

(39%) are categorized as Vulnerable on the Red List, 11 (33%) as Endangered, and nine (27%) as Critically Endangered. Within current threat statuses, Critically

Endangered had the highest proportion (53) of species with a mean projected quasi- extinction risk ≥ 50%, followed by Endangered (31) and Vulnerable (28). Of the seven species with a mean projected quasi-extinction risk ranging 20% - 49%, five are currently listed as Endangered, the status associated with this range. Similarly, five species with a mean projected quasi-extinction risk ranging 10%-19%, defined as

Vulnerable, are currently listed Vulnerable. Of the species with a mean projected quasi-extinction risk below 10% (hereafter unthreatened) the majority (n=28, 53%) are currently listed as Vulnerable, 19 (36%) as Endangered, and six (11%) as

Critically Endangered.

The Brant test demonstrated that the parallel regression assumption was upheld ( P value = 0.27) for the ordinal logistic regression. The coefficients of the logistic regression model were converted into an odds ratio for ease of interpretation.

An increase in mean projected quasi-extinction risk was associated with an increase in the odds of higher threat Red List status, with an odds ratio of 3.11 (95% CI 1.14,

8.61 P value =0.03). The predicted probability for each Red List status based on the ordinal logistic model shows an intuitive trend, but poor predictability of Critically

Endangered status (Supplemental Materials, Figure 3.A.1, Table 3.A.2). It is additionally apparent that projected quasi-extinction risk tracks the hierarchy of Red

36 List statuses when examining the mean of quasi-extinction risk scores within each status (Figure 3.2).

Figure 3.2. Mean of individual species mean ( ±±± SE) projected quasi-extinction risk for 99 threatened seabird species at year 100 categorized by current Red List status.

Implications for seabird conservation

Our results support a general correlation between Red List status and mean projected quasi-extinction risk. However at the individual species level, mean projected quasi-extinction risk does not align with ranges accompanying current threat status in 90 (81%) species. Of these, we projected 25 species to be more threatened than their current status and we estimated 26 species to be at least two

37 positions below their current status (Table 3.2). As a conservation tool, aimed to reduce extinctions, the Red List is intentionally precautionary. Thus, as expected, the statuses we estimated using thresholds for Red List quantitative analysis (criterion E) tended to be less precautionary than assessments based on criteria A-D. Fifty-four percent of species we estimated fell below a 10% mean projected quasi-extinction risk (Figure 3.1). The most extreme mismatches in current Red List statuses and modeled viability were represented by the 13 species currently listed as Vulnerable but we projected to have a mean quasi-extinction risk ≥ 50% (i.e. more threatened than current IUCN status) and the six species (Christmas Frigatebird, Galapagos

Petrel, Magenta Petrel, Newell’s Shearwater, Waved Albatross, Whenua Hou Diving- petrel) currently listed as Critically Endangered with mean projected quasi-extinction risk < 10% (i.e. less threatened than current IUCN status).

38

Table 3.2. Number and percent of species within current Red List threat statuses and seabird mPVA projected status (based on mean projected quasi-extinction risk and criterion E thresholds). Colors indicate degree of congruence: aligned results (green), species projected at one status position higher than current status (light red), species projected at two status positions higher than current status (dark red), species projected at one status position lower than current status (light grey), and species projected at least two status positions lower than current status (dark grey).

Vulnerable Red Endangered Critically List Red List Endangered (n=47) (n=35) Red List (n=17) Unthreatened 28 (60%) 19 (54%) 6 (35%) mPVA

Vulnerable 5 (11%) 0 (0%) 1 (6%) mPVA

Endangered 1 (2%) 5 (14%) 1 (6%) mPVA

Critically 13 (28%) 11 (31%) 9 (53%) Endangered mPVA

Phalacrocoracidae (cormorants & shags) had the greatest number (10) and percentage (33%) of species (threatened and non threatened) with a > 50% mean projected quasi-extinction risk, followed by 17% of Spheniscidae (penguins) and 16% of Alcidae (auks) (Figure 3.3b). Procellariidae (petrels, shearwaters) had the second greatest number of species (7) with a > 50% mean projected quasi-extinction risk and the greatest number of species (31) with a mean projected quasi-extinction risk below

10%.

39 Figure 3.3a +b. Current vs. our modeled threatened seabird Red List status by family. Figure 3.3a (top): Red List bar indicates number of species currently within threat statuses per family assessed by criteria A-D. mPVA bar indicates number of species projected within threat statuses by seabird mPVA at year 100 using extinction risk thresholds in criterion E. Figure 3.3b (bottom): Red List bar indicates percent of extant species (threatened and nonthreatened) per current Red List status. mPVA bar indicates projected percent of extant species following method above. (critically endangered=red, ≥ 50%; endangered=yellow, ≥ 20-49%; vulnerable=blue, ≥ 10%, unthreatened=grey, <10%) *Focal species comprise 87% of currently threatened seabird species. See Supplemental Materials, Table 3.A.1 for cause of exclusion for remaining 15 species. *Extant species per family derived from BirdLife Seabird Species List 2020.

40

41 42 Discussion

The IUCN Red List assessment process includes five major criteria to determine threat status. Four of these (criteria A-D) are based upon proxies of extinction risk (e.g. population decline, size, and structure) while one (criterion E) is based upon quantitative modeling of extinction risk. In practice, no Red List threatened species has been listed based upon quantitative modeling of extinction risk

(Collen et al., 2016). This is understandable as for many species there is insufficient information to conduct a rigorous quantitative model while population trends, size, and structure are often more readily available. Criteria A-D often produce a more precautionary approach, which could be argued is the correct way to consider irreversible extinctions. However, it is also important to understand how species’ status assessments may change if quantitative modeling is used within the assessment framework. This was feasible in the case of threatened seabirds because population statuses and demographic data are relatively well known and coherent across species, particularly within families. We found that inclusion of quantitative modeling can lead to some important differences from assessments using criteria A-D and novel insights relevant to conservation priorities.

While our seabird mPVA results were significantly correlated with current

Red List statuses, some key differences emerged. Overall, we found perfect coherence between a status assignment based on mPVA model projections and Red

List status in only 19 of the 99 (19%) species we examined. Our model results indicated that 25 species (25%) should be classified at higher Red List status and 55

43 (56%) should be classified at lower status than their current Red List classification.

For example, we found agreement between our model and the Red List for nine of 17

(53%) species listed on the Red List as Critically Endangered. Perhaps more importantly, we found coherence in only five of 47 (11%) species listed`` as

Vulnerable, with our model estimating that 14 (30%) species should be classified as

Endangered (1; 2%) or Critically Endangered (13; 28%). This suggests that in some cases Red List assessments may be less precautionary than quantitatively modeled estimates.

These mismatches underscore how the differences in the parameters used and timeframes of assessment across listing criteria may affect status. For example, seven of the 13 species our model projected as Critically Endangered (mean projected quasi-extinction risk ≥ 50%) are currently listed as Vulnerable on the Red List based on criterion D: “restricted area of occupancy or the number of locations with a plausible future threat that could drive the taxon to CR or EX in a very short time”

(IUCN, 2020). It is plausible that these species may have relatively low immediate extinction risk, but within longer timeframes (i.e. 100 years) extinction risk will become considerably greater unless some change to the current set of environmental and demographic conditions occurs. Extending criteria timeframes, such as the century presented here captures the broader extinction risk these populations are projected to experience. The 25 species projected to be at higher risk than their current Red List status should be prioritized for threat status reassessment.

44 Lower classifications projected by our model compared to current Red List status for many species likely emerge from the precautionary emphasis of Red List assessments based on restricted data (criteria A-D) versus our quantitative assessment based upon the approach described for criterion E. Current Red List assessments are based on criteria A-D, which describe symptoms of decline in shorter timeframes, and our seabird mPVA results assessed extinction risk out 100 years. It is important to note that the present analysis relies on mean estimates of extinction risk and excludes the uncertainty in the mean estimation, which ultimately should be investigated further and included in decision-making. Still, resource allocation decisions should consider the predicted lower relative risk in the 26 species projected to be at least two positions below their current threat status.

Adding quantitative modeling to listing assessments could improve understanding of extinction risk and the relative importance of different threats.

Combining quantitative models with the other criteria can serve to determine aggregate extinction risk. Assessments made using our quantitative modeling approach over 100 years doubled the potential list of Critically Endangered species, but at the same time reduced the number species potentially listed as Endangered and

Vulnerable species by 84%.

Strategic seabird conservation should focus on the species most threatened with extinction. We identified 33 Red List threatened seabirds with a mean projected quasi-extinction risk ≥ 50% within the next 100 years. The seven species with broad confidence intervals that exceeded the 50% quasi-extinction risk threshold should

45 also be considered at high relative risk of extinction. These species should also be the focus of additional research to increase the accuracy and precision of model parameters, including improving information on distribution, exposure to threats, movement and dispersal behavior, and vital rate estimates. Recovery plans should be developed for the 17 threatened species with mean projected quasi-extinction risk ≥

50% and no reported plans in-place (IUCN, 2020).

Across seabird families, our model found that the Procellariidae, Alcidae,

Spheniscidae, and Phalacrocoracidae contained the largest number of species at high quasi-extinction risk. For Procellarids, Alcids, and Spheniscids, this may be attributed to low fecundity, restricted metapopulation structure, and for Procellarids and Alcids, high vulnerability to introduced species. In contrast, Phalacrocoracids are relatively fecund and widespread breeders. However, this is a speciose group and therefore perhaps more likely to include threatened species. It may be particularly important to focus conservation efforts on Procellarids and Alcids as our model also predicted those families to contain a relative large proportion of Critically Endangered species

(15 total; 10 and 5, respectively).

Our multispecies approach enables direct comparisons across nearly 100 taxonomically similar species. This first taxonomic application of mPVA provides relative extinction risks across a highly threatened group. These results can be used to improve Red List categorizations and guide allocation of species conservation investment. While our study focuses primarily on projected quasi-extinction risk, conservation goals may be broader, such as restoring declining populations for

46 ecosystem function. Quantitative analyses can be adapted to assess viability under a range of management scenarios. The seabird mPVA is designed to evaluate island- specific invasive species eradications, translocations, and bycatch mortality mitigation.

This taxon-wide mPVA analysis also has important implications beyond seabird conservation. Criterion E is currently underutilized in Red List assessments, leading some to call for more widespread use (Akçakaya et al., 2006). We have demonstrated that applying criterion E at a broad scale is feasible and can provide reliable insights towards the most threatened species, mismatches among current assessment techniques, and strategic planning at the family-level. Quantitative modeling provides a standardized and consistent method of direct extinction risk estimation, with clear articulation of assumptions and an explicit treatment of uncertainty. Such analyses compliment existing assessments based solely on proxy metrics of threat and are a vital component of accurate threatened species management.

The Red List has assessed over 120,000 threatened and nonthreatened species.

Our seabird mPVA only evaluates threatened seabird species, but the framework could be reproduced for other well-studied taxa. This approach could be especially useful if applied to charismatic taxa (e.g. salmonoid fishes, cetaceans, sea turtles) to better allocate limited species conservation investment. Effective conservation planning relies on accurate extinction risk assessments to inform investment.

47 Supplemental Materials

Table 3.A.1. BirdLife International 2020 threatened seabird species list and rationale for exclusion of 15 species from seabird mPVA database. Current Included IUCN in seabird English Red List mPVA Rationale for mPVA Scientific name name Status database exclusion Onychoprion Aleutian 1988-2016 Least Concern or aleuticus Tern VU No Lower Risk/Least Concern Fratercula Atlantic 1988-2016 Least Concern or arctica Puffin VU No Lower Risk/Least Concern Puffinus Bannerman's Population number bannermani Shearwater EN No unknown Rissa Black-legged 1988-2016 Least Concern or tridactyla Kittiwake VU No Lower Risk/Least Concern 1988-2012 Near Threatened, Lower Sternula Risk/Near Threatened or balaenarum Damara Tern VU No Threatened (1988) Hydrobates Guadalupe Possibly extinct last seen macrodactylus Storm-petrel CR(PE) No 1912 Podiceps Horned 1988-2012 Least Concern or auritus Grebe VU No Non insular breeder Pterodroma Jamaican Possibly extinct, last seen caribbaea Petrel CR(PE) No 1879 Hydrobates Leach's leucorhous Storm-petrel VU No First assessment in 2017 Clangula Long-tailed hyemalis Duck VU No Non insular breeder Peruvian Sternula lorata Tern EN No Non insular breeder Saundersilarus Saunders's saundersi Gull VU No Non insular breeder Polysticta Steller's stelleri Eider VU No Non insular breeder Melanitta fusca Velvet Scoter VU No Non insular breeder White- Tachyeres headed leucocephalus Steamerduck VU No Non insular breeder Papasula Abbott's abbotti Booby EN Yes Spheniscus African EN Yes

48 Current Included IUCN in seabird English Red List mPVA Rationale for mPVA Scientific name name Status database exclusion demersus Penguin Hydrobates Ainley's cheimomnestes Storm-petrel VU Yes Diomedea amsterdamensi Amsterdam s Albatross EN Yes Diomedea Antipodean antipodensis Albatross EN Yes Ascension Fregata aquila Frigatebird VU Yes Hydrobates Ashy Storm- homochroa petrel EN Yes Pterodroma Atlantic incerta Petrel EN Yes Atlantic Yellow- Thalassarche nosed chlororhynchos Albatross EN Yes Leucocarbo Auckland colensoi Shag VU Yes Puffinus Balearic mauretanicus Shearwater CR Yes Phalacrocorax Bank neglectus Cormorant EN Yes Pterodroma Barau's baraui Petrel EN Yes Pseudobulweri a becki Beck's Petrel CR Yes Pterodroma Bermuda cahow Petrel EN Yes Procellaria parkinsoni Black Petrel VU Yes Pterodroma Black- hasitata capped Petrel EN Yes Chlidonias Black- albostriatus fronted Tern EN Yes Leucocarbo ranfurlyi Bounty Shag VU Yes Puffinus bryani Bryan's CR Yes

49 Current Included IUCN in seabird English Red List mPVA Rationale for mPVA Scientific name name Status database exclusion Shearwater Ardenna Buller's bulleri Shearwater VU Yes Thalassarche Campbell impavida Albatross VU Yes Leucocarbo Campbell campbelli Shag VU Yes Phalacrocorax Cape capensis Cormorant EN Yes Morus capensis Cape Gannet EN Yes Thalassarche Chatham eremita Albatross VU Yes Pterodroma Chatham axillaris Petrel VU Yes Leucocarbo Chatham onslowi Shag CR Yes Thalasseus Chinese bernsteini Crested Tern CR Yes Fregata Christmas andrewsi Frigatebird CR Yes Pterodroma Collared brevipes Petrel VU Yes Pterodroma cookii Cook's Petrel VU Yes Synthliboramp Craveri's hus craveri Murrelet VU Yes Pterodroma Desertas deserta Petrel VU Yes Eudyptes Erect-crested sclateri Penguin EN Yes Sternula nereis Fairy Tern VU Yes Pseudobulweri a macgillivrayi Fiji Petrel CR Yes Eudyptes Fiordland pachyrhynchus Penguin VU Yes Nannopterum Flightless harrisi Cormorant VU Yes Spheniscus Galapagos EN Yes

50 Current Included IUCN in seabird English Red List mPVA Rationale for mPVA Scientific name name Status database exclusion mendiculus Penguin Pterodroma Galapagos phaeopygia Petrel CR Yes Thalassarche Grey-headed chrysostoma Albatross EN Yes Synthliboramp Guadalupe hus hypoleucus Murrelet EN Yes Pterodroma Hawaiian sandwichensis Petrel EN Yes Puffinus Heinroth's heinrothi Shearwater VU Yes Pterodroma Henderson atrata Petrel EN Yes Spheniscus Humboldt humboldti Penguin VU Yes Puffinus Hutton's huttoni Shearwater EN Yes Indian Yellow- Thalassarche nosed carteri Albatross EN Yes Synthliboramp hus Japanese wumizusume Murrelet VU Yes Juan Pterodroma Fernandez externa Petrel VU Yes Larus fuliginosus Lava Gull VU Yes Eudyptes Macaroni chrysolophus Penguin VU Yes Pachyptila MacGillivray macgillivrayi 's Prion EN Yes Pterodroma Magenta magentae Petrel CR Yes Brachyramphu Marbled s marmoratus Murrelet EN Yes Pterodroma Masatierra defilippiana Petrel VU Yes

51 Current Included IUCN in seabird English Red List mPVA Rationale for mPVA Scientific name name Status database exclusion Pseudobulweri Mascarene a aterrima Petrel CR Yes Hydrobates Matsudaira's matsudairae Storm-petrel VU Yes Hydrobates Monteiro's monteiroi Storm-petrel VU Yes Fregetta New Zealand maoriana Storm-petrel CR Yes Puffinus Newell's newelli Shearwater CR Yes Northern Eudyptes Rockhopper moseleyi Penguin EN Yes Northern Diomedea Royal sanfordi Albatross EN Yes Pelecanoides Peruvian garnotii Diving-petrel EN Yes Pterodroma Phoenix alba Petrel EN Yes Ardenna Pink-footed creatopus Shearwater VU Yes Phalacrocorax featherstoni Pitt Shag EN Yes Nesofregetta Polynesian fuliginosa Storm-petrel EN Yes Pterodroma Providence solandri Petrel VU Yes Pterodroma Pycroft's pycrofti Petrel VU Yes Puffinus Rapa myrtae Shearwater CR Yes Rissa Red-legged brevirostris Kittiwake VU Yes Leucocarbo Rough-faced carunculatus Shag VU Yes Thalassarche Salvin's salvini Albatross VU Yes Synthliboramp Scripps's VU Yes

52 Current Included IUCN in seabird English Red List mPVA Rationale for mPVA Scientific name name Status database exclusion hus scrippsi Murrelet Phoebastria Short-tailed albatrus Albatross VU Yes Eudyptes Snares robustus Penguin VU Yes Phalacrocorax Socotra nigrogularis Cormorant VU Yes Phoebetria Sooty fusca Albatross EN Yes Southern Eudyptes Rockhopper chrysocome Penguin VU Yes Southern Diomedea Royal epomophora Albatross VU Yes Procellaria Spectacled conspicillata Petrel VU Yes Pterodroma Stejneger's longirostris Petrel VU Yes Leucocarbo chalconotus Stewart Shag VU Yes Puffinus Townsend's auricularis Shearwater CR Yes Hydrobates Townsend's socorroensis Storm-petrel EN Yes Pterodroma Trindade arminjoniana Petrel VU Yes Diomedea Tristan dabbenena Albatross CR Yes Diomedea Wandering exulans Albatross VU Yes Phoebastria Waved irrorata Albatross CR Yes Procellaria Westland westlandica Petrel EN Yes Pelecanoides whenuahouensi Whenua Hou s Diving-petrel CR Yes Procellaria White- VU Yes

53 Current Included IUCN in seabird English Red List mPVA Rationale for mPVA Scientific name name Status database exclusion aequinoctialis chinned Petrel Pterodroma White- cervicalis necked Petrel VU Yes White- Pterodroma winged leucoptera Petrel VU Yes Puffinus Yelkouan yelkouan Shearwater VU Yes Megadyptes Yellow-eyed antipodes Penguin EN Yes Pterodroma madeira Zino's Petrel EN Yes

54

Figure 3.A.1. Predicted probability of Red List status designation over scale quasi-extinction risk based on ordinal logistic regression . Red denotes probability of species designated as Critically Endangered across quasi-extinction risk range, yellow denotes probability of Endangered status, and blue denotes probability of Vulnerable status.

Table 3.A.2. Confusion matrix for ordinal logistic regression test data Predicted Critically Vulnerable Endangered Endangered Vulnerable 15 2 0 Endangered 5 2 0 Actual Critically Endangered 3 3 0

55 CHAPTER 4

Using meta-population models to guide conservation action

Abstract

Biodiversity conservation is limited by resources, data, and time for execution. To maximize efficacy, it is best if conservation plans are strategically evaluated for cost, feasibility, and likely impact prior to implementation. We present a framework to systematically examine the likely impact of proposed conservation plans for threatened taxa. As a case study of this framework we use the national

Action Plan for Seabird Conservation in New Zealand and 27 threatened seabirds identified for conservation interventions. To evaluate impact, we applied a recently developed seabird meta-population viability analysis model (seabird mPVA) that employs the most current population and adjustable demographic data to assess threatened seabird viability at a global scale under various management scenarios.

This publically available, web-based tool is intended to meet the needs of threatened seabird managers at the initial phase of conservation planning. We used the seabird mPVA to model population trends and potential seabird viability gains from conservation actions that include: bycatch mitigation, invasive species removal, and seabird translocation prescribed in the action plan for individual species. Our model’s ranking of New Zealand seabirds by current quasi-extinction vulnerability roughly correlated with the seabirds’ IUCN Red List status and New Zealand Threat

Classification System. We found modeled conservation impact of proposed actions and assigned priority to be generally positively correlated, but variable in magnitude.

56 If all prescribed conservations actions were implemented, our model predicted significant mitigation of quasi-extinction risk for nine species (Antipodean Albatross,

Auckland Island Shag, Black-fronted Tern, Fairy Tern, Rough-faced Shag, Northern

Royal Albatross, Pitt Island Shag, Stewart Island Shag, Yellow-eyed Penguin). This approach, and our model, can be adapted to other taxonomic groups to provide a consistent framing for the prioritization of species for conservation investment and predictions about the benefits of specific conservation actions for those species.

Introduction

Biodiversity loss imperils ecosystem function and services (Hooper et al.,

2012). Still, conservation goals, such as the international Aichi Biodiversity Targets are largely unmet and trends for biodiversity loss are steady or increasing (Krug et al.,

2014; El-Sheikh, 2018). Worldwide, conservation plans are underfunded and tied to national wealth rather than biodiversity and threats (Waldron et al., 2013, 2017).

Species assessments of extinction risk are a critical tool to help direct policy and legal protections, but they are generally poorly funded, often outdated, and grossly backlogged. For instance, in the United States, the Endangered Species Act provides legal protection for threatened species. However, over 500 species potentially warrant protection but await decisions on protections and decision times average 12 years (Greenwald et al., 2019; Puckett et al., 2016). Internationally, insufficient resources for the IUCN Red List of Threatened Species delay many species assessments, threatening their relevance to conservation planning (Rondinini et al., 2014). Moreover, many species assessed as threatened lack recovery plans,

57 including 46% of threatened (critically endangered, endangered, or vulnerable) seabird species (IUCN, 2020). Thus, managers are limited by underfunding, inadequate data on population trends and specific threats, and incomplete biological, legal, and policy frameworks to prioritize species and develop strategic conservation plans. Given this, successful conservation planning requires a data-driven approach to testing threatened species recovery plans.

One method to assess the current status of threatened species and the potential efficacy of recovery plans prior to implementation is through population viability analysis (PVA). PVAs have played an important role in informing future population trajectories, conservation policy, and potential actions (Morris and Doak, 2002;

Finkelstein et al., 2010; Lindenmayer et al., 1993). For example, through meta- population viability analysis (mPVA), researchers determined that the endangered piping plover depends on human-created habitats to rebound from flooding disturbances (Catlin et al., 2015). While PVAs have been widely adopted for single species, they have not been taken to scale across taxonomic groups or regions, perhaps because PVAs are often difficult and time consuming to develop and run.

We developed a comprehensive, generalizable, mPVA model (hereafter seabird mPVA) that predicts relative quasi-extinction risk in a consistent way across an entire group of species with similar life histories, and the predicted impacts of different conservation actions for those species. The freely accessible

(https://nhydra.shinyapps.io/mPVA1/), menu-driven online model draws from a comprehensive, though by no means complete, database of species life history traits,

58 and uses phylogenetically weighted data from closely related species to fill demographic data gaps. Here we apply our seabird mPVA to threatened New Zealand seabirds to establish baseline risk and test the efficacy of proposed management actions. We use the seabird mPVA to project the metapopulation trajectories of 27 threatened seabird species under a baseline (no-intervention) scenario and various proposed management scenarios. This allows us to determine relative extinction risk among species and estimate the relative efficacy of prescribed actions within species.

As the most threatened marine taxonomic group, seabirds provide a useful trial of our mPVA approach. While the seabird mPVA is equipped to model 99 of the

114 threatened seabird species, we limit this study to the 27 threatened seabirds that breed in New Zealand and have prescribed management interventions that can be analyzed with our model. New Zealand is an ideal location to trial our framework for several reasons. First, it is a hotspot for seabird diversity, hosting 96 of the world’s

346 seabird species, 33 of which are endemic to New Zealand (Croxall et al., 2012a).

Second, strategic conservation is urgently needed with 39% of New Zealand seabird species and 82% of its endemic seabird species classified as threatened (critically endangered, endangered, vulnerable) by the IUCN Red List (Robertson et al., 2017;

IUCN, 2020). Third, New Zealand’s offshore islands rank highest for global seabird conservation opportunities when considering the probabilities of extinction, relative endemism, and evolutionary distinctiveness of breeding seabirds species (Spatz et al.,

2017). Finally, New Zealand developed a national action plan for all threatened New

Zealand seabirds (Taylor, 2000). The Action plan for seabird conservation in New

59 Zealand (NZ Action Plan) describes the distribution, population size, conservation status, threats, past conservation actions, future research priorities, and proposes conservation actions for every threatened New Zealand seabird.

Seabirds are impacted by a wide range of threats, including human disturbance, harvest, invasive species, habitat loss, commercial fisheries bycatch and pollution (Croxall et al., 2012a). The NZ Action Plan prescribes specific interventions to offset these, including: establishing new colonies, mitigating bycatch mortality, removal of invasive predators, and measures to increase survival at certain life stages.

Our seabird mPVA was designed with the flexibility to model such conservation intervention scenarios. Users may adjust population distributions to simulate translocations, adjust at-sea mortality, remove specific invasive species’ effects, as well as adjust vital rates across life stages. Here, we use our seabird mPVA to examine the predicted benefits of each intervention proposed in the NZ Action as a means to model potential benefit before investing the time and resources required for implementation. While we limit this study to threatened New Zealand seabirds, our seabird mPVA can readily be applied to all threatened seabirds. Likewise, this framework for systematic review of conservation plans could be replicated within other threatened taxa.

Methods

Seabird mPVA

The seabird mPVA is a spatially explicit and demographically structured approach to evaluate seabird population viability in the context of current distribution,

60 abundance, trends and multiple threats. The seabird mPVA draws on published seabird vital rates data to model demographic transitions and project future trends for populations of threatened seabirds, while accounting for meta-population structure of geographically distributed (but demographically linked) breeding colonies and incorporating information on current trends and spatially-explicit threats. The seabird mPVA is based around a stage-structured projection matrix (Caswell, 2001). To ensure model generality across many different life history patterns (Desholm, 2009) we described demographic structure in terms of three broad life history stages: 1) sub- adults, 2) breeding adults, and 3) non-breeding adults. Spatial structure is incorporated by embedding demographic matrices for semi-discreet sub-populations

(generally islands or island groups) within a larger meta-matrix structure representing the dynamics of the entire species. The advantage of including spatial and demographic structure in our model is that the impacts of various threats (invasive species, fisheries bycatch) are both spatially explicit and stage-specific, and thus the conservation benefits of mitigation efforts (such as removal of invasive species, fishing regulations, etc.) can be best-evaluated by modeling their effects on the appropriate demographic stages and/or sub-populations, and then translating these into species-level impacts (Desholm, 2009).

The seabird mPVA was parameterized using publicly available data contained in the IUCN Red List of Threatened Species version 2020.2 (IUCN, 2020) and additional data contained in the Threatened Island Biodiversity Database (TIB

Partners, 2018), literature-reported values of seabird vital rates, and solicited expert

61 opinion. We adopted a Bayesian approach for estimating vital rates. Specifically, for each focal species we compiled published estimates of each demographic parameter required for the model, as well as uncertainty measures (standard errors) for these estimates. We weighted all published data by taxonomic relatedness between the focal species and the species for which the data originated. The combined distributions of existing parameter estimates were treated as prior distributions, which could be updated by comparing population projections generated from these priors with reported data on trends derived from the IUCN Red List (IUCN, 2020). For species in which there were only qualitative descriptions of current population trends, we allocated modal lambda values of 1.02, 1.00, 1.00, and 0.98 for Increasing,

Unknown, Stable, and Decreasing status, respectively (with uncertainty distributions around these modal values). To update priors, we used Markov chain Monte Carlo

(MCMC) methods to find the combination of values of demographic parameters most likely to produce the observed trends, resulting in posterior distributions for each parameter.

We expanded the base seabird mPVA model to incorporate additional information on specific threats, including the effects of invasive species and at-sea threats such as fisheries bycatch. To estimate the demographic effects of invasive species on baseline vital rates, we compared published time series of seabird abundance estimates at islands where invasive species occur vs. islands where they are absent, as well as before-after trends for islands where invasive species were removed (Jones et al., 2008). We employed a proportional hazards formulation to

62 combine the additional hazards associated with invasive species with baseline survival rates. Log hazards were calculated using a function that included covariates for invasive type, nesting type, body size, island size, and number of co-occurring invasive species (allowing for compensatory mortality at islands with >1 invasive species present). We used MCMC methods to find the hazard function parameters (in conjunction with baseline demographic parameters) most likely to produce the observed reductions in growth rates.

We used the parameterized seabird mPVA to project population dynamics of threatened and endangered seabirds, accounting for environmental stochasticity and parameter uncertainty, and with starting abundances initialized using the most recent

IUCN Red List status reports. Management actions were evaluated by simulating the effects on vital rates (e.g. in the case of invasive species removal actions, we simply removed the additional hazards associated with the invasive), and by comparing the distributions of model projections with and without the management action. Further description of the seabird mPVA mechanics can be at https://nhydra.shinyapps.io/mPVA1/.

Application of the seabird mPVA to the NZ Action Plan

There are currently 114 seabird species classified as vulnerable, endangered, critically endangered or critically endangered/possibly extinct by the IUCN Red List

(IUCN, 2020). Of these species, 99 are incorporated in the mPVA, with omissions due to extreme data deficiencies and species that breed on continents. The NZ Action

Plan describes 46 species and subspecies, 27 of which we included in this study. We

63 excluded species from this study that are not considered globally threatened, are not incorporated into the mPVA seabird database, or the NZ Action Plan only called for interventions outside of the scope of the seabird mPVA model scenarios

(Supplemental Materials Table 4.A.1). The 27 species included here consist of 337 confirmed and probable global breeding populations, with New Zealand territories containing 48% of these populations across 105 islands (Supplemental Materials

Table 4.A.2).

For each focal species, we ran a suite of model simulations for the baseline

(status quo) scenario, as well as alternative scenarios that corresponded to interventions dictated by the NZ Action Plan. We projected meta-population dynamics forward for 100 years, and replicated model runs 10,000 times to account for demographic and environmental stochasticity as well as parameter uncertainty (by drawing all seabird mPVA parameters from their respective posterior distributions).

We associated each species with a quasi-extinction threshold (an a priori threshold below which the species was considered extremely likely to go extinct) of 50 breeding pairs. We note that recommended quasi-extinction thresholds for seabirds vary within the literature, and the value we used is a conservative threshold below which extinction risk due to natural disaster or genetic diversity loss is likely. For each scenario, we determined the mean expected metapopulation abundance after 100 years along with 95% confidence intervals, and the mean proportion of simulations with population sizes that fell below the quasi extinction threshold at 100 years to yield mean projected quasi-extinction risk. It is important to note that the mean

64 projected quasi-extinction risk should not be interpreted as an absolute measure of extinction risk, but rather an index of relative vulnerability that can be used to compare conservation gains across alternative scenarios. The NZ Action classifies proposed interventions by priority rank: Essential Actions, High Actions, Medium

Actions, and Low Actions. Priority ranks set in the NZ Action Plan are intended to denote urgency. Each priority rank scenario was first simulated independently. Then we simulated all prescribed interventions concurrently, under the “All” scenario. In total, this consisted of 83 scenarios (Supplemental Table 4.A.3). Here we focus on

Baseline and All intervention scenarios.

The interventions proposed in the NZ Action Plan were often regionally specific, however the model generated the global trajectory of the seabird species.

Proposed actions included detailed measures, such as removing particular invasive predators from certain islands and translocations or reintroductions at specific sites. In these cases, we included as much detail described in the NZ Action Plan as possible in the scenario. In the case of translocations, we incorporated prescribed breeding site data (island size and location) whenever possible. When translocation was prescribed without a target site, we set the translocation site location as the midpoint between the two largest breeding colonies with a standardized size of 2 km 2. The number of individuals to be translocated was not specified in the NZ Action Plan. Therefore, we standardized the size of translocation populations to 400 juveniles to present an optimum scenario based on a review of published seabird translocation population sizes documenting 130 attempted seabird colony movements with an average of 373

65 individuals translocated (Jones and Kress, 2012; Miskelly et al., 2009; Lincoln Park

Zoo, 2010) (Supplemental Materials Table 4.A.4).

In the case of less detailed proposed actions, we made certain assumptions to apply standardized treatments. Thus, when the NZ Action Plan called for a reduction in bycatch, we adjusted at-sea mortality by the mean estimated total annual potential fatalities (APF) due to trawl, longline, and set-net fisheries within New Zealand territories (MPI, 2019). APF was reported as individuals without age class information. However, at-sea mortality adjustments are modeled to affect age classes in proportion to their abundance within the seabird mPVA. As the NZ Action Plan is a national plan, we deemed treatments based on regional fisheries impacts appropriate. At times, the NZ Action Plan called for measures to improve survival at specific life stages, such as construction of wind shields to increase hatching success.

In the absence of specific vital rate targets, we modeled “high” and “low” scenarios simulating 20% or 10% improvements to the vital rate of interest (Supplemental

Materials Table 4.A.3).

Statistical Analysis

To determine the credible difference of viability gains under management scenarios relative to baseline scenarios, we used a Bayesian model comparison approach. We examined the posterior predictive distribution of the difference of means projected under baseline and treatment scenarios and assessed whether the null value (0, indicating no difference) fell within credible parameter values (95% CI).

66 Therefore, rejection of the null value indicated the difference of means was credible and statistically significant.

Results

Baseline quasi-extinction risk and current threat statuses

Baseline (current) mean projected quasi-extinction risk and uncertainty varied considerably among the 27 species (Figure 4.1). Over a quarter (26%) of species scored a baseline mean projected quasi-extinction risk > 0.80 with reasonable certainty. Some species, notably the Pitt Island Shag and Black-fronted Tern, demonstrated high uncertainty in baseline mean projected quasi-extinction risk, indicating poor parameter confidence. Nearly half (48%) of species were projected to have a baseline mean quasi-extinction risk of 0 with high certainty.

67

Figure 4.1. Baseline mean projected quasi-extinction risks, 95% confidence intervals, and IUCN Red List status for 27 threatened New Zealand seabird species.

When species were ranked by baseline mean projected quasi-extinction risk and evaluated against current Red List status, we identified several inconsistencies, especially for species with the highest mean projected quasi-extinction risk (Table

4.1). Of the seven species with a baseline mean projected quasi-extinction risk > 0.80, four species (Fairy Tern, Stewart Island Shag, Rough-faced Shag, Auckland Island

68 Shag) are listed as vulnerable, the lowest category of threat included in this study.

Similarly, the Red List categorizes the Magenta Petrel as critically endangered, however we estimated a relatively low baseline mean quasi-extinction risk (0.01, 0-

0.05 95% CI). Three species (Grey-headed Albatross, Hutton’s Shearwater, Westland

Petrel) are categorized as endangered, while we predicted relatively low baseline mean projected quasi-extinction risk. The remaining 10 species with a baseline mean projected quasi-extinction risk of 0 are classified as vulnerable.

Of the seven species with the baseline mean projected quasi-extinction risk >

0.80, the New Zealand Threat Classification System (NZTCS) assessed six species as threatened (nationally critical n=3, nationally endangered n=2, or nationally vulnerable n=1). Meanwhile, the Stewart Island Shag, with a high baseline mean projected quasi-extinction risk (0.97, 0.91-1.00 95% CI), is considered to be at risk- recovering (NZTCS).

69 Table 4.1. Species threat status, mean projected quasi-extinction risk, final mean abundance, and associated intervention scenario. Mean Projected Mean Projected Threat Status Initial IUCN Quasi- Quasi- Mean Final Common IUCN / Modeled Population extinction extinction Risk Mean Final Abundance Name NZTCS Abundance Trajectory Scenario risk 95% CI Abundance 95% CI Chatham Critically Is. Shag Endangered 904 Decreasing Baseline 1 0.99 – 1 1 0 – 4 Threatened- Nationally Critical All 0.99 0.98 -1 2 0 – 10 Fairy Tern Vulnerable 8250 Decreasing Baseline 0.98 0.94 – 1 6 0 – 34 Threatened- Nationally 70 Critical* All 0.82 0.68 - 0.91 185 44 – 519 Stewart Is. Shag Vulnerable 5428 Decreasing Baseline 0.97 0.9 – 1 10 1 – 57 At Risk- Recovering All 0.9 0.79 - 0.96 56 11 – 177 Antipodean Endangered 87086 Decreasing Baseline 0.95 0.91 - 0.98 13 4 – 32 Albatross Threatened- Nationally Critical All 0.91 0.87 - 0.94 37 17 – 68 Rough- faced Vulnerable 802 Stable Baseline 0.93 0.81 - 0.98 30 6 – 91 Shag Threatened- Nationally Endangered All 0.85 0.7 - 0.94 67 21 – 160 Auckland Is. Vulnerable 3812 Stable Baseline 0.92 0.76 - 0.98 38 4 – 167

Mean Projected Mean Projected Threat Status Initial IUCN Quasi- Quasi- Mean Final Common IUCN / Modeled Population extinction extinction Risk Mean Final Abundance Name NZTCS Abundance Trajectory Scenario risk 95% CI Abundance 95% CI Shag Threatened- Nationally Vulnerable All 0.85 0.68 - 0.95 81 14 - 270 Yellow- eyed Endangered 6874 Decreasing Baseline 0.82 0.72 - 0.9 139 50 – 303 Penguin Threatened- Nationally Endangered All 0.78 0.69 - 0.86 225 97 – 439 Black- fronted Endangered 8162 Decreasing Baseline 0.63 0.07 - 0.98 51 3 – 327

71 Tern Threatened- Nationally Endangered All 0.01 0 - 0.05 6389 3792 – 9984 Pitt Is. Shag Endangered 1094 Decreasing Baseline 0.6 0.3 - 0.85 183 56 – 441 Threatened- Nationally Critical All 0.48 0.24 - 0.73 294 126 – 575 Northern Royal Endangered 30080 Decreasing Baseline 0.51 0.37 - 0.66 534 224 – 1062 Albatross At Risk- Naturally Uncommon All 0.3 0.19 - 0.43 1936 939 – 3492 Fiorland Crested Vulnerable 9520 Decreasing Baseline 0.18 0.04 - 0.44 2375 936 – 4910 Penguin Threatened- Nationally All 0.2 0.07 - 0.42 2957 1267 – 5793

Mean Projected Mean Projected Threat Status Initial IUCN Quasi- Quasi- Mean Final Common IUCN / Modeled Population extinction extinction Risk Mean Final Abundance Name NZTCS Abundance Trajectory Scenario risk 95% CI Abundance 95% CI Vulnerable Chatham Vulnerable 17652 Stable Baseline 0.17 0.08 - 0.31 1513 770- 2641 Albatross At Risk- Naturally Uncommon All 0.17 0.09 - 0.28 2058 1102 – 3464 Snares Crested Vulnerable 96912 Stable Baseline 0.01 0 - 0.08 18006 9020 – 31761 Penguin At Risk- Naturally Uncommon All 0.01 0 - 0.05 40073 23719 – 62758 72 Magenta Critically Petrel Endangered 136 Increasing Baseline 0.01 0 - 0.05 305 161 – 519 Threatened- Nationally Critical All 0.01 0 - 0.04 371 201 – 620 Black Petrel Vulnerable 8052 Stable Baseline 0 0 - 0.03 2163 1561 –2903 Threatened- Nationally Vulnerable All 0 0 - 0 5746 5003 – 6560 Buller's Vulnerable 2500000 Stable Baseline 0 0 - 0 544989 298842 – 902228 Shearwater At Risk- Naturally Uncommon All 0 0 - 0 611759 341066 – 1000775 Campbell Vulnerable 68434 Increasing Baseline 0 0 - 0 81858 63504 – 103461 Albatross Threatened- All 0 0 - 0 84721 67299 – 104929

Mean Projected Mean Projected Threat Status Initial IUCN Quasi- Quasi- Mean Final Common IUCN / Modeled Population extinction extinction Risk Mean Final Abundance Name NZTCS Abundance Trajectory Scenario risk 95% CI Abundance 95% CI Nationally Vulnerable Chatham Petrel Vulnerable 1614 Increasing Baseline 0 0 - 0 1964 1580 – 2407 Threatened- Nationally Vulnerable All 0 0 - 0 2219 1691 – 2847 Cook's Petrel Vulnerable 972388 Increasing Baseline 0 0 - 0 868105 661240 - 1114388 At Risk- 73 Relict All 0 0 - 0 900751 691315– 1148930 Grey- headed Endangered 390756 Decreasing Baseline 0 0 - 0 272031 189185 - 376325 Albatross Threatened- Nationally Vulnerable All 0 0 – 0 278014 196559 – 379534 Hutton's Endangered 325002 Stable Baseline 0 0 – 0 130081 80832 – 196431 Shearwater Threatened- Nationally Vulnerable All 0 0 – 0 149758 91864 – 228358 Pycroft's Petrel Vulnerable 24834 Increasing Baseline 0 0 - 0 20874 14187 – 29409 At Risk- Recovering All 0 0 – 0 21280 14670 – 29645 Salvin's Albatross Vulnerable 128728 Unknown Baseline 0 0 – 0 99629 63218 – 148075

Mean Projected Mean Projected Threat Status Initial IUCN Quasi- Quasi- Mean Final Common IUCN / Modeled Population extinction extinction Risk Mean Final Abundance Name NZTCS Abundance Trajectory Scenario risk 95% CI Abundance 95% CI Threatened- Nationally Critical All 0 0 – 0 176702 139623 – 219853 Southern Royal Vulnerable 46676 Stable Baseline 0 0 - 0.01 17360 11242 – 25402 Albatross At Risk- Naturally Uncommon All 0 0 - 0.02 18739 12568 – 26682 Westland Petrel Endangered 16132 Unknown Baseline 0 0 - 0 10171 7012 – 14171

74 At Risk- Naturally Uncommon All 0 0 – 0 15602 12047 – 19799 White- chinned Vulnerable 4396776 Decreasing Baseline 0 0 – 0 439490 204017 – 816423 Not Petrel Threatened All 0 0 – 0 527908 261130 – 939283 White- necked Vulnerable 143886 Increasing Baseline 0 0 – 0 124631 92546 – 163434 At Risk- Petrel Relict All 0 0 - 0 135699 99641 – 179609

*Refers to subspecies. New Zealand Threat Classification System 6.5.0. IUCN Red List 7 Birdlife 2020.

Changes in viability under intervention scenarios

We simulated 83 conservation action scenarios. One third of these scenarios represented baseline (no intervention) conditions averaging a mean projected quasi- extinction risk of 0.32 spanning all possible outcomes (0-1). When considering the difference between baseline and prescribed intervention scenarios, there was little compelling evidence that significant conservation gains were associated with the priority assigned to scenarios (Figure 4.2). Within species, there was congruence between decrease in mean projected quasi-extinction risk, regardless of credible difference, and the hierarchical order of priority assigned to interventions scenarios in

52% of instances. For mean final abundance, population trends tracked scenario priorities 84% of the time (Supplemental Materials Table 4.A.5).

75

Figure 4.2. Decrease in mean projected quasi-extinction risk by intervention scenario priority (Baseline - Prescribed Scenario).

Next, we excluded delineations between scenario priorities and instead present only the results for baseline (no intervention) and all (comprehensive intervention) scenarios for each species as this captures maximum potential viability gains. Only nine species (Antipodean Albatross, Auckland Island Shag, Black-fronted Tern, Fairy

Tern, Rough-faced Shag, Northern Royal Albatross, Pitt Island Shag, Stewart Island

Shag, Yellow-eyed Penguin) saw mean credible (95% certainty) reductions in

76 projected quasi-extinction risk when all prescribed interventions were simulated. For these nine species, mean credible decreases in projected quasi-extinction risk ranged from 0.04-0.76 (Table 4.2). The highest ameliorations in risk were 0.76 and 0.23 for

Black-fronted Tern and Fairy Tern, respectively.

Table 4.2. Mean credible (95% certainty) decrease in projected quasi-extinction risk and proportional increase in mean final abundances under all (comprehensive intervention) scenarios relative to baseline (no intervention).

Mean decrease in projected Proportional increase quasi-extinction risk and 95% in abundance and 95% Common name CI CI 173 Black-fronted Tern 0.76 (0.75-0.76) .76 (115.27-261.94) 3.7 Northern Royal Albatross 0.23 (0.22-0.24) 6 (3.57-3.96) 20. Fairy Tern 0.15 (0.12-0.19) 07 (16.57-24.31) 1.5 Pitt Island Shag 0.14 (0.1-0.18) 9 (1.45-1.76) Rough-faced Shag 0.07 (0.06-0.09) 2.2 (2.03-2.39) 4.4 Stewart Island Shag 0.07 (0.05-0.09) 3 (3.89-5.05) 2.0 Auckland Island Shag 0.06 (0.05-0.08) 4 (1.88-2.21) 2.4 Antipodean Albatross 0.04 (0.03-0.05) 6 (2.32-2.62) 1.6 Yellow-eyed Penguin 0.04 (0.03-0.06) 2 (1.54-1.71)

Difference of means in log transformed estimated abundance increases under all scenarios relative to baseline scenarios ranged from 0.02-5.16. Ten species were projected to more than double in abundance in the all scenario relative to baseline scenario (Black-fronted Tern, Fairy Tern, Stewart Island Shag, Northern Royal

Albatross, Black Petrel, Antipodean Albatross, Snare’s Crested Penguin, Rough-faced

77 Shag, Chatham Island Shag, Auckland Island Shag) with 95% confidence

(Supplemental Materials Table 4.A.6).

We examined proposed intervention actions for trends between frequency of prescription and viability change. For the all intervention actions, we calculated the weighted average of credible viability gains (relative to baseline scenarios) across all species (Supplemental Materials Table 4.A.7). Invasive species removal was prescribed with the highest frequency (24 scenarios, eight credible decreases in quasi- extinction risk) but ranked second for the greatest weighted-average decrease in mean projected quasi-extinction risk (6%) with credible log transformed increases in abundances ranging from -0.04 to 3. The second most frequently prescribed action was translocation, which was simulated in 20 scenarios (six credible decreases in quasi-extinction risk) and was ranked third in weighted average decrease in mean projected quasi-extinction risk (5%) with credible log transformed increases in abundances ranging from -0.06 to 1.49. Interventions that were intended to provide direct improvements to a specific vital rate (such as the construction of wind shields to increase hatching success) were prescribed nine times (eight credible decreases in quasi-extinction risk) but yielded the highest viability gains in mean projected quasi- extinction risk (32%) with credible log transformed increases in abundances ranging from 0.03 to 5.16. Bycatch mitigation was recommended in 13 scenarios (four credible decreases in quasi-extinction risk) and yielded 4% decrease in projected quasi-extinction risk with credible log transformed increases in abundances ranging from 0.02 to 1.01

78 Discussion

Under no-intervention (baseline) scenarios, the seabird mPVA projected seven of the 27 species to have relatively high mean projected quasi-extinction risk (> 0.80) and 13 species to have relatively low mean projected quasi-extinction risk (0). The

IUCN Red List conservation status of the species with high mean projected quasi- extinction risk was variable, with four species considered vulnerable, the lowest threat category in this study. Meanwhile, the New Zealand Threat Classification

System (NZTCS) ranked these species largely as highly threatened. These findings suggest that closer examination, beyond global and national threatened species lists, is required to ensure that resources are properly apportioned within threatened taxa.

Species with low (0) mean projected quasi-extinction risk generally tracked Red List assessments with 77% considered vulnerable and only one species (Salvin’s

Albatross) considered threatened-nationally critical by the NZTCS.

When comparing viability measures between baseline and all prescribed actions scenarios, the seabird mPVA projected mean credible decreases in projected quasi-extinction risk for nine species, with the Black-fronted Tern and Fairy Tern presenting the greatest differences between baseline and all scenarios (0.76 and 0.23, respectively). The greatest credible log transformed increases in abundances were seen in the Black-fronted Tern (5.16), Fairy Tern (3), and Stewart Island Shag (1.54).

Considering that all prescribed interventions involved modeling 67 conservation actions across 27 species, our results suggest that alternative actions might also be worth considering to reduce extinction risk for most of the focal species. To meet

79 conservation goals, scenarios should be simulated to estimate the relative effect of novel combinations of actions, various reasonable targets for trialed actions (e.g. increasing translocation sites, increasing bycatch mitigation), and coordinated actions executed beyond national borders.

Beyond examining baseline and comprehensive interventions, we also considered how the priority assigned to conservation plans related to viability gains.

The NZ Action Plan based intervention priority on perceived urgency. We found limited correlations between intervention priority and modeled gains in mean projected quasi-extinction risk or mean final abundance. We found trends in viability gains (without regard for credible differences between measures) tracked scenario priority in 84% and 52% of cases for mean final abundance and mean projected quasi-extinction risk, respectively. One limitation to this analysis was the uneven distribution of priority counts, ranging from 3-17. These results indicate that there may be a relationship between perceived urgency of prescribed actions and viability gains; however, it may not be equal between viability metrics and further research is needed to confirm this.

We further examined the relationship between prescribed actions and conservation gains. Effectiveness of conservation actions varied between viability measures. Interventions that involved direct manipulation of vital rates to model actions such as artificially incubating eggs intended to increase fledging success, had the greatest weighted-mean credible reduction on projected quasi-extinction risk

(32%) and greatest log transformed increase in abundance (mean 5.16, Black-fronted

80 Tern) while being prescribed with moderate frequency (nine scenarios). Actions that were recommended with the highest frequency, invasive species removal and translocation appeared to have less promise for conservation gain. While these two actions were recommended most often, they had limited impact on viability measures.

These results suggest that conservation actions affect species in nuanced and unexpected ways, and reinforces the potential benefits of modeling the projected impact of actions before implementation.

An advantage to using the seabird mPVA to predict quasi-extinction risk under various scenarios is the transparent reporting of uncertainty across all species and viability measures (quasi-extinction risk and increase in abundance). Notably, the

Pitt Island Shag and the Black-fronted Tern were predicted to have high uncertainty in baseline mean projected quasi-extinction risk. There are several possible explanations for these results that merit further investigation before a recovery plan for these species is implemented. Uncertainty in parameter values and low starting populations are likely to account for the high uncertainty demonstrated by these two species. The Pitt Island Shag has a relatively low initial population size, which could lead to more uncertainty in population trajectories due to stochasticity. The Black- fronted Tern also has fairly low initial population size and is a single island endemic, but other species with similar initial population sizes or limited ranges demonstrated high confidence in viability predictions. This may be because similar species were projected with high certainty to either fall below quasi-extinction thresholds or steadily increase in abundance due to either unfavorable or advantageous vital rates

81 and current population trends. Indeed, the 13 species with a mean projected quasi- extinction risk of 0, had the narrowest confidence intervals. These results are critical to pinpoint knowledge gaps and emphasize which species require further research before reliably making recovery plan predictions. In particular, dispersal rates data and adult and subadult survival rates should be improved. The seabird mPVA is designed to quickly incorporate improved data for future reevaluation of viability.

This study does not aim to be a precise predication of seabird populations into time. For one, the model does not approximate every ecosystem interaction, for instance interspecies competition for nesting sites or the effects of climate change on food availability. Nor does the model incorporate the feasibility, costs, and competing interests of stakeholders relevant to any of these interventions. Rather, this should be considered a first-round decision-making framework that provides a systematic analysis across species to assess the potential conservation gains of suites of actions.

A key benefit to this approach is estimating the relative benefit of conservation scenarios. The caveats bounding the results of this study are not unique and should be considered alongside other guidance on appropriate applications of PVAs (Morris et al., 1999; Brook et al., 2000; Coulson et al., 2001).

The framework we present here provides a means to quickly and consistently project relative quasi-extinction risk across a highly threatened taxon. With our approach we are also able to evaluate conservation gains across a hierarchal national strategic plan before implementation. Performing these assessments before enactment provides the opportunity to consider which species and actions should be prioritized

82 to more efficiently allocate limited conservation resources. This is timely given a recent assessment that endemic New Zealand birds extinction risk trends are higher today than 40 years ago (Garcia-R and Di Marco, 2020).

Beyond relevance to seabirds, this study outlines how a standardized approach might be applied to assessing conservation options for large groups of species belonging to a taxonomic group. This data-driven method uniformly evaluates recovery plans, returning relative viability metrics that managers can compare across and within species to match conservation goals. Thus, managers may find efficiencies in recovery plans and compare potential costs and gains. With limited time and resources available for conservation, there is a need to rapidly maximize resources and understand confidence in recovery plans. This mPVA approach could be replicated for other threatened taxa to guide conservation efforts.

83 Supplemental Materials

Table 4.A.1 Threatened Species List and Overlap between New Zealand Department of Conservation Action Plan and seabird mPVA. IUCN status abbreviations: VU=Vulnerable, EN=Endangered, CR=Critically Endangered Included in Included in Included IUCN mPVA NZ Action in this Scientific Name Common Name Status database Plan study Comments Antipodean Diomedea antipodensis Albatross EN Yes Yes Yes Gibson's Albatross is a Diomedea antipodensis Gibson's subspecies of Antipodean gibsoni Albatross N/A No Yes No Albatross Auckland Islands 84 Phalacrocorax colensoi Shag VU Yes Yes Yes Larus bulleri Black-billed Gull EN No Yes No non-migratory Black-fronted Chlidonias albostriatus Tern EN Yes Yes Yes Phalacrocorax Bounty Islands Unable to model proposed ranfurlyi Shag VU Yes Yes No actions Buller's Ardenna bulleri Shearwater VU Yes Yes Yes Campbell Thalassarche impavida Albatross VU Yes Yes Yes Phalacrocorax Campbell Island Unable to model proposed campbelli Shag VU Yes Yes No actions

Included in Included in Included IUCN mPVA NZ Action in this Scientific Name Common Name Status database Plan study Comments Chatham Thalassarche eremita Albatross VU Yes Yes Yes Chatham Islands Phalacrocorax onslowi Shag CR Yes Yes Yes Pterodroma axillaris Chatham Petrel VU Yes Yes Yes Pterodroma cookii Cook's Petrel VU Yes Yes Yes Erect-crested No Unable to model proposed Eudyptes sclateri Penguin EN Yes Yes actions

85 Sternula nereis Fairy Tern VU Yes Yes Yes Eudyptes Fiordland pachyrhynchus Crested Penguin VU Yes Yes Yes Thalassarche Grey-headed chrysostoma Albatross EN Yes Yes Yes Hutton's Puffinus huttoni Shearwater EN Yes Yes Yes Pterodroma magentae Magenta Petrel CR Yes Yes Yes Phalacrocorax New Zealand carunculatus King Shag VU Yes Yes Yes Northern Royal Diomedea sanfordi Albatross EN Yes Yes Yes Parkinson's Procellaria parkinsoni Petrel VU Yes Yes Yes

Included in Included in Included IUCN mPVA NZ Action in this Scientific Name Common Name Status database Plan study Comments Phalacrocorax featherstoni Pitt Island Shag EN Yes Yes Yes Pterodroma pycrofti Pycroft's Petrel VU Yes Yes Yes Salvin's Thalassarche salvini Albatross VU Yes Yes Yes Snares Crested Eudyptes robustus Penguin VU Yes Yes Yes Southern Royal Diomedea epomophora Albatross VU Yes Yes Yes 86 Phalacrocorax Stewart Island chalconotus Shag VU Yes Yes Yes Procellaria westlandica Westland Petrel EN Yes Yes Yes Whenua Hou Diving Petrel was classified as unique species split from Pelecanoides georgicus in 2018. Whenua Hou Diving Pelecanoides Whenua Hou was first classified as whenuahouensis Diving Petrel CR Yes No No threatened in 2019. Procellaria White-chinned aequinoctialis Petrel VU Yes Yes Yes

Included in Included in Included IUCN mPVA NZ Action in this Scientific Name Common Name Status database Plan study Comments White-flippered Penguin is a subspecies of the Little Penguin which has Least Concern status and therefore Eudyptula minor White-flippered not included in the PVA albosignata Penguin N/A No Yes No database White-necked Pterodroma cervicalis Petrel VU Yes Yes Yes Yellow-eyed Megadyptes antipodes Penguin EN Yes Yes Yes 87 Fulmar Prion currently listed as Least Concern, but was previously listed as Vulnerable updating data could allow simulations to Pachyptila crassirostris Fulmar Prion LC No Yes No be run Buller's currently listed as Near Thalassarche bulleri Albatross NT No Yes No Threatened Thalassarche cauta White-capped currently listed as Near steadi Albatross NT No Yes No Threatened New Zealand Subspecies of White-fronted White-fronted Tern which is listed as Near Sterna striata striata Tern N/A No Yes No Threatened

Included in Included in Included IUCN mPVA NZ Action in this Scientific Name Common Name Status database Plan study Comments Subspecies of White-fronted Sterna striata Southern White- Tern which is listed as Near aucklandorna fronted Tern N/A No Yes No Threatened Eastern Considered subspecies of Eudyptes chrysocome Rockhopper Southern Rockhopper filholi Penguin N/A No Yes No Penguin White-bellied Storm-petrel White-bellied currently listed as Least Fregetta grallaria Storm-petrel LC No Yes No Concern

88 Subspecies of the Little Puffinus assimilis Kermadec Little Shearwater which is listed kermadecensis Shearwater N/A No Yes No as Least Concern Subspecies of the Little Puffinnis assimilis North Island Shearwater which is listed haurakiensis Little Shearwater N/A No Yes No as Least Concern Sula dactylatra fullagari Masked Booby LC No Yes No Listed as Least Concern Phalacrocorax varius varius Pied Shag LC No Yes No Listed as Least Concern Sterna vittata bethunei Antarctic Tern LC No Yes No Listed as Least Concern Subspecies of Sooty Tern Sterna fuscata New Zealand currently listed as Least kermadeci Sooty Tern N/A No Yes No Concern

Table 4.A.2 Confirmed and Probable Breeding Site List by Species and Country

Breeding Global Populations Common Breeding within New Name Populations Zealand Breeding Sites by Country Antipodean New Zealand: Auckland, Campbell, Disappointment, Adams, Pitt, Chatham Albatross 7 7 (Rekohu), Antipodes. Auckland Island Shag 4 4 New Zealand: Enderby, Auckland, Rose, Ewing. Black Petrel 2 2 New Zealand: Great Barrier (Aotea), Little Barrier. Black-Fronted Tern 1 1 New Zealand: South Island.

89 Buller's New Zealand: Archway, Poor Knights 1, Tawhiti Rahi, Aorangi, Aorangaia, Motu Shearwater 7 7 Purihi, Motu Kapiti. Campbell Albatross 2 2 New Zealand: Jeanette Marie, Campbell. Chatham Albatross 1 1 New Zealand: The Pyramid. Chatham New Zealand: Star Keys (Motuhope), North East Reef, Pitt, Rabbit, Chatham Island Shag 5 5 (Rekohu). Chatham Petrel 3 3 New Zealand: Chatham (Rekohu), South East (Rangatira), Pitt. Cook's Petrel 3 3 New Zealand: Little Barrier, Codfish, Great Barrier (Aotea). Australia: Goat, Teal, White, Sandy, Little Eyre, Eyre, Island B, Webb, West Wallabi, Clonmel, Unnamed 2, Unnamed 1, Cowrie, Egret, Leo, West Cattle, English, Pelsaert, North, Keru, Lancelin, Uncle Margie (Mangrove), Lagoon, Cow, Roberts, Coronation, Hibbertia, Byone, East Wallabi, Rat, Cohen, Dick, Helms, Fairy tern 86 1 Post Office, Wild Dog, Sandy, Gun, Gregory, Gilbert, Garden, Fairy Tern,

Fairbridge, Saint Peter, South Reef, Beacon, East Diamond Islet, Murray, Morley (Crake), Swan, Little North, Flanagan, Crescent, Long, Third Sister, Boullanger, Troubridge, The Coral Patches, Tapani, Stick, Square, Shearwater, Serventy, Flinders, Seagull Reef, Mellor, Whitlock, Plate, Burnett, Newman, King, Tasmania, Mushroom, Raven, Ram, Campbell. New Caledonia: Mbe, Ua, Mouillage, Amedee, Redika, Uatio, New Caledonia (Grand Terre), Kae, Atire, Passage. New Zealand: North Island. New Zealand: Shelter West, Catherine, Eleanor, Fanny, Pigeon, George Sound 1, Passage South (Outer), Passage North (Inner), Breaksea, Coal, Indian, Chalky, Fiordland Codfish, Entry, Hawea, Shelter East, South Island, Long (Weka), Solander, Steep- Crested To, Seymour, Stewart (Rakiura), Small Craft Harbour (Small), Taumaka, Putaihina, Penguin 27 27 Johns, Rolla. Australia: Marquarie. Chile: Norte, Mendoza, Bartolome, Schlatter, Gonzalo, Grande, Ester, Martinez, Grande Terre, Santander. French Southern Territories: 90 Croy. South Africa: Marion, Prince Edward. South Georgia and the South Sanwich Islands: Trinity, Jomfruene Island 1, Paryadin Peninsula North Island 2, Paryadin Peninsula South Island 2, Bird, Paryadin Peninsula North Island 3, Paryadin Grey-headed Peninsula South Island 3, Jomfruene Island 2, Main, Hall, South Georgia, Albatross 28 1 Jomfruene Island 3, Bernt. New Zealand: Campbell. Hutton's Shearwater 1 1 New Zealand: South Island. Magenta Petrel 1 1 New Zealand: Chatham (Rekohu). New Zealand New Zealand: White Rocks, Duffer's Reef, Rahuinui, Huinia Rock, Sentinel Rock, King Shag 8 8 North Trio, Stewart (Te Kuru Kuru), South Island. Northern Royal New Zealand: Little Sister, Forty-Fours (Motuhara), Enderby, South Island, Big Albatross 5 5 Sister. Pitt Island 11 11 New Zealand: Big Sister, Middle Sister, Chatham (Rekohu), Mangere, Little

Shag Mangere, Pitt, South East (Rangatira), Star Keys (Motuhope), Forty-Fours (Motihara), The Castle, Rabbit. New Zealand: Coppermine, Korapuki, Lady Alice (Mauimua), Whatupuke Pycroft's (Mauiroto), Red Mercury (Whakau), Stanley (Kawhitu), Stephenson (Ririwha), Petrel 12 12 West Chicken (Mauitaha), Cuvier, Double (Moturehu), Taranga (Hen), Aorangi. Salvin's French Southern Territories: Penguin. New Zealand: Rima, Depot, Proclamation, Albatross 11 10 Molly Cap, Funnel, Ruatara, Toru Islet, Spider, Tunnel, The Pyramid. Snares Crested Penguin 4 4 New Zealand: Rima, North-East, Toru Islet, Broughton. Southern Royal Albatross 5 5 New Zealand: Campbell, Auckland, Adams, South Island, Enderby. New Zealand: Dog (Papa-kaha), Wharekakahu, Green, Kinakina, South Island, 91 Stewart Island Omaui, Zero Rock, Rarotoka (Centre), Passage (Whero), Pig (Blumine), Maukiekie Shag 11 11 (Moeraki). Westland Petrel 1 1 New Zealand: South Island. Falkland Islands: New, Bottom, Kidney. French Southern Territories: L'Est, Longue, Antares, Penguin (Pingouins), Petrel, Suhm, Briggs, Inskip, Cochons, Chat, Blakeney, Cimetiere, Penn, Possession, Mayes, Grande Terre. South Africa: Marion, Prince Edward. South Georgia and the South Sandwich Islands: The Guides island 1, Wirik Bay island 1, Main, Dorada, Sigma, Verdant Island island 2, Austin, Tanner, Tidespring, Saddle, Pickersgill Island 2, Henningsen Glacier island, Pickersgill Island 1, Right Whale Rocks island 1, Kupriyanov island 3, Mollymawk, McCarthy, Poncet, Skua, Invisible, Inner Lee, Smaaland Cove Island, Harcourt, Albatross, Anvil Stacks island 1, Cooper, East Skerry, Green, Grass, White- Crescent, Annekov, Bjornstadt Bay Island 2, Bernt, Bird, Cape Vakop island 1, chinned Petrel 73 10 Pillow Rock, Proud, Hauge Reef island 1, Trollhul Island 1, Prion, Nunez

Peninsula Island 7, South Georgia. New Zealand: Auckland, Cossack Rock, Dent, Disappointment, Ewing, Enderby, Jacquemart, Monowai, Adams, Antipodes. White-necked Petrel 2 2 New Zealand: Macauley. Norfolk Island: Philip. New Zealand: South Island, Tommy, Auckland, Crayfish, Codfish, Bravo, Weka, Yellow-eyed Rose, Noble, Bench, Ewing, Campbell, Enderby, Stewart (Rakiura), Anchorage, Penguin 16 16 Adams 92

Table 4.A.3. Scenario treatment by species and priority

Common Name Essential Priority Action High Priority Action Medium Priority Action Low Priority Action invasive species Antipodean mitigate bycatch: -62 removal: pigs, cats from Albatross annual at-sea deaths Auckland Island n/a n/a invasive species Auckland Island removal: pigs, cats from Shag n/a n/a Auckland Island n/a invasive species removal: cats, rats (Polynesian and black) mitigate bycatch: -449 from Great Barrier

93 Black Petrel annual at-sea deaths Island n/a n/a vital rate adjustment: vital rate adjustment: fledging success adult survival increased increased to 75.4% and to 99% and 98.8% from 65.4% from 55.4% for Black-Fronted 93.8% for high and low high and low scenarios, Tern scenarios, respectively respectively n/a n/a establish a new colony: 400 juveniles Buller's translocated to Motoura Shearwater n/a n/a Island n/a Campbell mitigate bycatch: -111 Albatross annual at-sea deaths n/a n/a n/a

mitigate harvest: increased fledging establish a new colony: success to 71.2% from 400 juveniles Chatham 71% or approx. 10 translocated to Mangere Albatross chicks per year 1 n/a Island n/a invasive species removal: pigs, cows, Chatham Island goats from Chatham Shag n/a Island n/a n/a translocation: 400 adults to Chatham Island from South East Island; Invasive species invasive species removal: All from removal: cats from Pitt 94 Chatham Petrel Chatham Island 2 Island 3 n/a n/a Establish a new colony: invasive species translocated 400 removal: all from Great juveniles to Mokoia Cook's Petrel n/a Barrier Island 4 n/a Island increase hatching success: fledging success increased to 95% and 86% from 76% for high and low scenario, respectively; Invasive species removal: all predators Fairy tern from all breeding island 5 n/a n/a n/a Fiordland Crested n/a n/a mitigate bycatch: -4 n/a

Penguin annual at-sea deaths Grey-headed mitigate bycatch: -5 Albatross annual at-sea deaths n/a n/a n/a establish a new colony: invasive species 400 juveniles Hutton's removal: pigs, deer from translocated to Mt Shearwater South Island n/a Fyffe 6 n/a invasive species removal: all from Magenta Petrel Chatham Island 7 n/a n/a n/a establish new colony: 400 juveniles New Zealand bycatch mitigation: -1 translocated to imagined 8

95 King Shag Shag n/a annual at-sea death island n/a vital rate adjustment: fledging success increased to 91.9% and Establish a new colony: 81.9% from 71.9% for 400 juveniles Northern Royal high and low scenarios, translocated to Mangere Albatross respectively n/a n/a Island. invasive species removal and livestock- proof fencing installation: cats from Pitt Island and sheep, cattle, and pigs from Pitt Island Shag n/a n/a Chatham Island 9 n/a invasive species establish a new colony: Pycroft's Petrel n/a removal: rats 400 juveniles n/a

(Polynesian) from West translocated to Burgess Chicken Island Island, Invasive species removal: rats (Polynesian and Brown) from Ririwha Island bycatch mitigation: - 2110 annual at-sea Salvin's Albatross n/a n/a deaths n/a invasive species removal: rats Snares Crested (unspecified) from Penguin n/a n/a North-East Island n/a Southern Royal mitigate bycatch: -22 Albatross annual at-sea deaths n/a n/a n/a 96 mitigate bycatch: -33 annual at-sea deaths, invasive predator establish a new colony: removal: omnivores and 400 juveniles Stewart Island carnivores from South translocated to imagined Shag n/a Island 10 island 11 n/a invasives species removal: all from South Westland Petrel Island 12 n/a n/a n/a establish a new colony: invasives species 400 juveniles White-chinned mitigate bycatch: -1700 removal: pigs, cats from translocated Mt Paris, Petrel annual at-sea deaths Auckland Island n/a assumed area of 2 km 2 White-necked establish a new colony: Petrel n/a n/a 400 juveniles n/a

translocated Raoul Island invasive species Yellow-eyed removal: pigs from Penguin n/a n/a Auckland Island n/a 1Reliable harvest data is unavailable. However, harvest is widely considered to occasionally occur, be prosecuted when discovered, and target chicks (Robertson, 1991). Fledging success adjustment is based on most recent nest counts. 2Invasive species removed from Chatham Island were: black rat, Brown rat, Polynesian rat, common brushtail possum, Western European hedgehog, wild boar, cat, dog, cow, sheep, goat, house mouse. 3Action Plan additionally called for the removal of weka from Pitt Island. However, weka impact is not currently included in seabird mPVA. 4Invasive species removed from Great Barrier were: black rat, Polynesian rat, cat, dog, European rabbit, house mouse, and wild boar. 5Invasive species removed by island; Rat Island: house mouse; Cohen Island: red fox; from Garden Island: house mouse, 97 domestic cat, red fox; Saint Peter Island: house mouse, heath monitor lizard; North Island: house mouse, black rat, brown/Norway rat, unspecified mouse, domestic cat, ferret, least weasel, stoat, brushtail possum, west European hedgehog, Polynesian rat; North: house mouse; Swan: Black rat, brown/Norway rat, house mouse, domestic cat, red fox; New Caledonia: black rat, brown/Norway rat, Polynesian rat, house mouse, cat, domestic dog, unspecified mongoose, unspecified mammals omnivore, wild boar; Boullanger Island: house mouse; Atire: unspecified rats; Flinders: black rat, brown/Norway rat, house mouse, domestic cat, domestic dog, wild boar; Whitlock: house mouse; King Island: Black rat, house mouse, cat; Tasmania: black rat, Brown/Norway rat, house mouse, domestic cat, domestic dog, European polecat, red fox, wild boar; Ram Island: unspecified rats, unspecified mice, domestic cat, domestic dog; Great Barrier Island: black rat, Polynesian rat, cat, dog, European rabbit, house mouse, wild boar. 6Mt Fyffe is presumed to be predator-free and constrained to 2 km 2 for this scenario. 7Invasive species removed from Chatham Island were: Black rat, Brown rat, Polynesian rat, Common brushtail possum, west European headgehog, wild boar, cat, dog, cow, sheep, goat, house mouse. The Action Plan additionally called for weka removal, however, weka are currently controlled for in Magenta Petrel nesting sites on Chatham Island. The Action Plan additionally called for increasing fledging success, however, baseline fledging success is reported at 97% and cannot be increased.

8The imagined island is presumed to have no invasive species, area of 2 km 2, located at -40.90 (lat) and 174.03 (long), which is the midpoint between the two largest breeding colonies. 9Action Plan also called for the removal of weka, but weka impact is currently not included in seabird mPVA. 10 Invasive predators removed from South Island were: black rat, brown/Norway rat, common brushtail possum, red-necked wallaby, west European hedgehog, wild boar, domestic cat, domestic dog, ferret, least weasel, stoat, house mouse. 11 The imagined island is presumed to have no invasive species, an area of 2 km 2, located at -45.36 (lat) and 170.86 (long), which is the midpoint between the two largest populations. 12 Invasive species removed from South Island were: black rat, brown rat, possum, pig, cat dog, ferret, least weasel, stoat, mouse, cow, sheep, tahr, red deer.

98

Table 4.A.4. Database and literature review of attempted seabird colony movements (direct translocation or attraction by decoy or acoustic playback). N=131, average individuals moved=373

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Miskelly et al 2009, Table 1, Serventy et al 1989, Fledgling Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1954 Translocation Island 50 chick seabird ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1960 Translocation Island 16 Juvenile seabird, Oecologia ART, Serventy et al 1989 Fledging 99 Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1962 Translocation Island 13 Juvenile seabird, Oecologia ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1964 Translocation Island 19 Juvenile seabird, Oecologia ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1965 Translocation Island 21 Juvenile seabird, Oecologia ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1966 Translocation Island 17 Juvenile seabird, Oecologia ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1967 Translocation Island 21 Juvenile seabird, Oecologia Ardenna Short-tailed 1968 Translocation Fisher 20 Juvenile ART, Serventy et al 1989 Fledging

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference tenuirostris Shearwater Island translocation and philopatry in a seabird, Oecologia ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1969 Translocation Island 19 Juvenile seabird, Oecologia ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1970 Translocation Island 20 Juvenile seabird, Oecologia ART, Serventy et al 1989 Fledging Ardenna Short-tailed Fisher translocation and philopatry in a tenuirostris Shearwater 1971 Translocation Island 20 Juvenile seabird, Oecologia 10 Jones and Kress 2012, A review of the world's active seabird restoration Fratercula Atlantic projects, Supplemental Materials arctica Puffin 1979 Translocation Ile Bono 200 chick Table 2 ART, Kress et al 1988, Re- Eastern Egg estabilishment of the Atlantic Rock Puffins at a former breeding site in Fratercula Atlantic (Muscongus the Gulf of Maine. Journal of Field arctica Puffin 1973 Translocation Bay) 5 chick Ornithology ART, Kress et al 1988, Re- Eastern Egg estabilishment of the Atlantic Rock Puffins at a former breeding site in Fratercula Atlantic (Muscongus the Gulf of Maine. Journal of Field arctica Puffin 1974 Translocation Bay) 54 chick Ornithology Fratercula Atlantic 1975 Translocation Eastern Egg 91 chick ART, Kress et al 1988, Re-

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference arctica Puffin Rock estabilishment of the Atlantic (Muscongus Puffins at a former breeding site in Bay) the Gulf of Maine. Journal of Field Ornithology Eastern Egg Kress et al 1988, Re-estabilishment Rock of the Atlantic Puffins at a former Fratercula Atlantic (Muscongus breeding site in the Gulf of Maine. arctica Puffin 1976 Translocation Bay) 98 chick Journal of Field Ornithology Eastern Egg Kress et al 1988, Re-estabilishment Rock of the Atlantic Puffins at a former Fratercula Atlantic (Muscongus breeding site in the Gulf of Maine. 10 arctica Puffin 1977 Translocation Bay) 99 chick Journal of Field Ornithology Eastern Egg Kress et al 1988, Re-estabilishment Rock of the Atlantic Puffins at a former Fratercula Atlantic (Muscongus breeding site in the Gulf of Maine. arctica Puffin 1978 Translocation Bay) 91 chick Journal of Field Ornithology Eastern Egg Kress et al 1988, Re-estabilishment Rock of the Atlantic Puffins at a former Fratercula Atlantic (Muscongus breeding site in the Gulf of Maine. arctica Puffin 1979 Translocation Bay) 92 chick Journal of Field Ornithology Eastern Egg Kress et al 1988, Re-estabilishment Rock of the Atlantic Puffins at a former Fratercula Atlantic (Muscongus breeding site in the Gulf of Maine. arctica Puffin 1980 Translocation Bay) 100 chick Journal of Field Ornithology Fratercula Atlantic Eastern Egg Kress et al 1988, Re-estabilishment arctica Puffin 1981 Translocation Rock 100 chick of the Atlantic Puffins at a former

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference (Muscongus breeding site in the Gulf of Maine. Bay) Journal of Field Ornithology ART, Kress et al 1988, Re- Eastern Egg estabilishment of the Atlantic Rock Puffins at a former breeding site in Fratercula Atlantic (Muscongus the Gulf of Maine. Journal of Field arctica Puffin 1982 Translocation Bay) 100 chick Ornithology ESTIMATE Kress, S, 1997, Using Animal Behavior for Conservation, Journal of the Yamashina Institute for Ornithology, "954 [chicks] were 10 brought to Eastern Egg Rock in the years 1973-86", Kress et al 1988 details 830 chicks translocated between 1973-1982 and Kress et al 1987 details 94 translocated in 1985. Therefore 954-924=30 chicks Eastern Egg translocated between 1983, 1984, Rock 1986, averaging 10 chicks per year, Fratercula Atlantic (Muscongus assuming chicks were translocated arctica Puffin 1983 Translocation Bay) 10 chick annually ESTIMATE Kress, S, 1997, Using Eastern Egg Animal Behavior for Conservation, Rock Journal of the Yamashina Institute Fratercula Atlantic (Muscongus for Ornithology, "954 [chicks] were arctica Puffin 1984 Translocation Bay) 10 chick brought to Eastern Egg Rock in the

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference years 1973-86", Kress et al 1988 details 830 chicks translocated between 1973-1982 and Kress et al 1987 details 94 translocated in 1985. Therefore 954-924=30 chicks translocated between 1983, 1984, 1986, averaging 10 chicks per year, assuming chicks were translocated annually Eastern Egg ART, Kress et al 1987, Transplanted Rock Puffins Return to Seal Island, 10 Fratercula Atlantic (Muscongus Fratercula Fund of the National arctica Puffin 1985 Translocation Bay) 94 Juvenile Aududon Society ESTIMATE Kress, S, 1997, Using Animal Behavior for Conservation, Journal of the Yamashina Institute for Ornithology, "954 [chicks] were brought to Eastern Egg Rock in the years 1973-86", Kress et al 1988 details 830 chicks translocated between 1973-1982 and Kress et al 1987 details 94 translocated in 1985. Eastern Egg Therefore 954-924=30 chicks Rock translocated between 1983, 1984, Fratercula Atlantic (Muscongus 1986, averaging 10 chicks per year, arctica Puffin 1986 Translocation Bay) 10 chick assuming chicks were translocated

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference annually ESTIMATE Kress, S, 1997, Using Animal Behavior for Conservation, Seal Island Journal of the Yamashina Institute National for Ornithology, "950 [chicks] were Wildlife taken to Seal Island between 1984- Refuge 89", Kress et al 1985 details 626 outer translocated 1985-1989. Therefore Fratercula Atlantic Penobscot 950-626=324 chicks translocated in arctica Puffin 1984 Translocation Bay 324 chick 1984 Seal Island 10 National Wildlife Refuge outer ART, Kress et al 1985, 1985 Puffin Fratercula Atlantic Penobscot Transplant, Fratercula Fund of the arctica Puffin 1985 Translocation Bay 99 Juvenile National Aududon Society Seal Island National Wildlife Refuge outer ART, Kress et al 1985, 1985 Puffin Fratercula Atlantic Penobscot Transplant, Fratercula Fund of the arctica Puffin 1986 Translocation Bay 149 Juvenile National Aududon Society Fratercula Atlantic Seal Island ART, Kress et al 1985, 1988 Puffin arctica Puffin 1988 Translocation National 188 Juvenile Transplant, Fratercula Fund of the

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Wildlife National Aududon Society Refuge outer Penobscot Bay Seal Island National Wildlife Refuge outer ART, Kress et al 1985, 1989 Puffin Fratercula Atlantic Penobscot Transplant, Fratercula Fund of the 10 arctica Puffin 1989 Translocation Bay 190 Juvenile National Aududon Society Miskelly et al 2009, Table 1, Kress Eastern Egg 1980, Petrel transplant, Egg Rock Rock update 1980, Newsletter of the Hydrobates Leach's (Muscongus Seabird Restoration Program of leucorhous Storm-petrel 1980 Translocation Bay) 20 chick National Audubon Society 3 Roby et al 2002, Effects of Colony Relocation on Diet and Productivity Hydroprog Decoys/Acous East Sand of Caspian Terns, The Journal of ne caspia Caspian Tern 1999 tic Playback Island 2800 Adult Wildlife Management Roby et al 2002, Effects of Colony Relocation on Diet and Productivity Hydroprog Decoys/Acous East Sand 1700 of Caspian Terns, The Journal of ne caspia Caspian Tern 2000 tic Playback Island 0 Adult Wildlife Management Hydroprog Caspian Tern 2001 Decoys/Acous East Sand 1780 Adult Roby et al 2002, Effects of Colony

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference ne caspia tic Playback Island 0 Relocation on Diet and Productivity of Caspian Terns, The Journal of Wildlife Management Miskelly and Gummer, 2013, Attempts to anchor pelagic fairy Pachyptila prions (Pachyptila turtur) to their turtur Fairy Prion 2002 Translocation Mana Island 40 Juvenile release site on Mana Island, Notornis Miskelly and Gummer, 2013, Attempts to anchor pelagic fairy Pachyptila prions (Pachyptila turtur) to their turtur Fairy Prion 2003 Translocation Mana Island 100 Juvenile release site on Mana Island, Notornis 10 Miskelly and Gummer, 2013, Attempts to anchor pelagic fairy Pachyptila prions (Pachyptila turtur) to their turtur Fairy Prion 2004 Translocation Mana Island 100 Juvenile release site on Mana Island, Notornis Gummer et al, 2016, Report on the 2016 supplementary translocation of Fairy Prion (titiwainui) chicks from Stephens Island (Takapourewa) to Pachyptila Mana Island, Friends of Mana Island turtur Fairy Prion 2015 Translocation Mana Island 100 chick Inc. Gummer et al, 2016, Report on the 2016 supplementary translocation of Fairy Prion (titiwainui) chicks from Pachyptila Stephens Island (Takapourewa) to turtur Fairy Prion 2016 Translocation Mana Island 100 chick Mana Island, Friends of Mana Island

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Inc. DOC, 2019, Tiny seabirds transferred 800 km from Chathams Pelagodro White-faced to aid conservation, DOC Media ma marina Storm-petrel 2019 Translocation Mana Island 50 chick Releases 2019 Miskelly et al 2004, Establishment of a colony of common diving Pelecanoid Common petrels by chick transfers and es urinatrix Diving-petrel 1997 Translocation Mana Island 90 chick acoustic attraction, Emu Miskelly et al 2004, Establishment of a colony of common diving 10 Pelecanoid Common petrels by chick transfers and es urinatrix Diving-petrel 1998 Translocation Mana Island 100 chick acoustic attraction, Emu Miskelly et al 2004, Establishment of a colony of common diving Pelecanoid Common petrels by chick transfers and es urinatrix Diving-petrel 1999 Translocation Mana Island 49 chick acoustic attraction, Emu Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Pelecanoides, Pachyptila, and Pelecanoid Common Motuora Puffinus: Family Procellerariidae), es urinatrix Diving-petrel 2007 Translocation Island 25 chick Biological Conservation Miskelly et al 2009, Translocation of Pelecanoid Common Motuora eight species of burrow nesting es urinatrix Diving-petrel 2008 Translocation Island 66 chick seabirds (genera Pterodroma,

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Pelecanoides, Pachyptila, and Puffinus: Family Procellerariidae), Biological Conservation ART, Brown, Joseph, 1983, The Pelecanus Return of the Brown Pelican, Book. occidentali Brown (EASTERN Brown Pelican, s Pelican 1968 Translocation Grand Terre 25 Juvenile subspecies) ART, Brown, Joseph, 1983, The Pelecanus Return of the Brown Pelican, Book. occidentali Brown Rockerfeller (EASTERN Brown Pelican, s Pelican 1968 Translocation Refuge 21 Juvenile subspecies) 10 ART, Brown, Joseph, 1983, The Pelecanus Return of the Brown Pelican, Book. occidentali Brown (EASTERN Brown Pelican, s Pelican 1969 Translocation Grand Terre 30 Juvenile subspecies) ART, Brown, Joseph, 1983, The Pelecanus Return of the Brown Pelican, Book. occidentali Brown Rockerfeller (EASTERN Brown Pelican, s Pelican 1969 Translocation Refuge 25 Juvenile subspecies) ART, McNease et al 1984, The Brown Pelican Restocking Program in Louisianna, Proceedings Annual Conference Southeast Association of Pelecanus Fish and Wildlife Agencies occidentali Brown Rockerfeller (EASTERN Brown Pelican, s Pelican 1970 Translocation Refuge 100 Juvenile subspecies)

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference ART, McNease et al 1984, The Brown Pelican Restocking Program in Louisianna, Proceedings Annual Conference Southeast Association of Pelecanus Fish and Wildlife Agencies occidentali Brown (EASTERN Brown Pelican, s Pelican 1971 Translocation Grand Terre 65 Juvenile subspecies) ART, McNease et al 1984, The Brown Pelican Restocking Program in Louisianna, Proceedings Annual Conference Southeast Association of 10 Pelecanus Fish and Wildlife Agencies occidentali Brown (EASTERN Brown Pelican, s Pelican 1972 Translocation Grand Terre 100 Juvenile subspecies) ART, McNease et al 1984, The Brown Pelican Restocking Program in Louisianna, Proceedings Annual Conference Southeast Association of Pelecanus Fish and Wildlife Agencies occidentali Brown (EASTERN Brown Pelican, s Pelican 1973 Translocation Grand Terre 100 Juvenile subspecies) ART, McNease et al 1984, The Brown Pelican Restocking Program Pelecanus in Louisianna, Proceedings Annual occidentali Brown Conference Southeast Association of s Pelican 1974 Translocation Grand Terre 100 Juvenile Fish and Wildlife Agencies

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference (EASTERN Brown Pelican, subspecies) ART, McNease et al 1984, The Brown Pelican Restocking Program in Louisianna, Proceedings Annual Conference Southeast Association of Pelecanus Fish and Wildlife Agencies occidentali Brown (EASTERN Brown Pelican, s Pelican 1975 Translocation Grand Terre 101 Juvenile subspecies) ART, McNease et al 1984, The Brown Pelican Restocking Program 11 in Louisianna, Proceedings Annual Conference Southeast Association of Pelecanus Fish and Wildlife Agencies occidentali Brown (EASTERN Brown Pelican, s Pelican 1976 Translocation Grand Terre 99 Juvenile subspecies) ART, Brown, Joseph, 1983, The Pelecanus Return of the Brown Pelican, Book. occidentali Brown (EASTERN Brown Pelican, s Pelican 1977 Translocation North Island 95 Juvenile subspecies) ART, Brown, Joseph, 1983, The Pelecanus Return of the Brown Pelican, Book. occidentali Brown Isle aux (EASTERN Brown Pelican, s Pelican 1978 Translocation Pitre 101 Juvenile subspecies) Pelecanus Brown Isle aux ART, Brown, Joseph, 1983, The occidentali Pelican 1979 Translocation Pitre 122 Juvenile Return of the Brown Pelican, Book.

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference s (EASTERN Brown Pelican, subspecies) ART, Brown, Joseph, 1983, The Pelecanus Return of the Brown Pelican, Book. occidentali Brown Isle aux (EASTERN Brown Pelican, s Pelican 1980 Translocation Pitre 191 Juvenile subspecies) Phoebastri Fisher, Harvey 1981, Experiments in a Laysan Eastern homing in Laysan Albatross, The immutabilis Albatross 1961 Translocation Island 991 Juvenile Condor Phoebastri Fisher, Harvey 1981, Experiments in a Laysan Green homing in Laysan Albatross, The 11 immutabilis Albatross 1961 Translocation Island 112 Juvenile Condor Phoebastri Fisher, Harvey 1981, Experiments in a Laysan Lisianksi homing in Laysan Albatross, The immutabilis Albatross 1962 Translocation Island 2021 Juvenile Condor Phoebastri Fisher, Harvey 1981, Experiments in a Laysan Eastern homing in Laysan Albatross, The immutabilis Albatross 1965 Translocation Island 97 chick Condor ART, Imber et al 2003, Post- Hauturu fledging migration, age of first (Little return and recruitment, and results of Procellaria Barrier inter-colony translocation of Black parkinsoni Black Petrel 1986 Translocation Island) 46 chick Petrel, Notornis Hauturu ART, Imber et al 2003, Post- Procellaria (Little fledging migration, age of first parkinsoni Black Petrel 1987 Translocation Barrier 60 chick return and recruitment, and results of

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Island) inter-colony translocation of Black Petrel, Notornis ART, Imber et al 2003, Post- Hauturu fledging migration, age of first (Little return and recruitment, and results of Procellaria Barrier inter-colony translocation of Black parkinsoni Black Petrel 1988 Translocation Island) 40 chick Petrel, Notornis ART, Imber et al 2003, Post- Hauturu fledging migration, age of first (Little return and recruitment, and results of Procellaria Barrier inter-colony translocation of Black 11 parkinsoni Black Petrel 1989 Translocation Island) 49 chick Petrel, Notornis ESTIMATE, Grummer, H, 2003, Chick translocation as a method of establishing new surface-nesting seabird colonies, DOC Science Internal Series, further cites McHalick 1999. "249 chicks from Great Barrier [to Little Barrier Island]…between 1986-1990". Hauturu Imber et al 2003 details 195 chicks (Little translocated between 1986-1989. Procellaria Barrier Therefore, 54 chicks translocated in parkinsoni Black Petrel 1990 Translocation Island) 54 chick 1990 Pterodrom Chatham Gummer et al 2014, Best practice a axillaris Petrel 2002 Translocation Pitt Island 41 chick techniques for the translocation of

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Chatham petrels (Pterodroma axillaris), Cook's petrels (P. cookii) and Pycroft's petrels (P. pycrofti), DOC, Table 2 Gummer et al 2014, Best practice techniques for the translocation of Chatham petrels (Pterodroma axillaris), Cook's petrels (P. cookii) Pterodrom Chatham and Pycroft's petrels (P. pycrofti), a axillaris Petrel 2003 Translocation Pitt Island 49 chick DOC, Table 2 Gummer et al 2014, Best practice 11 techniques for the translocation of Chatham petrels (Pterodroma axillaris), Cook's petrels (P. cookii) Pterodrom Chatham and Pycroft's petrels (P. pycrofti), a axillaris Petrel 2004 Translocation Pitt Island 55 chick DOC, Table 2 Gummer et al 2014, Best practice techniques for the translocation of Chatham petrels (Pterodroma axillaris), Cook's petrels (P. cookii) Pterodrom Chatham and Pycroft's petrels (P. pycrofti), a axillaris Petrel 2005 Translocation Pitt Island 55 chick DOC, Table 2 Sweetwater Gummer et al 2014, Best practice Conservatio techniques for the translocation of Pterodrom Chatham n Covenant, Chatham petrels (Pterodroma a axillaris Petrel 2008 Translocation Chatham 47 chick axillaris), Cook's petrels (P. cookii)

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Island and Pycroft's petrels (P. pycrofti), DOC, Table 2 Gummer et al 2014, Best practice Sweetwater techniques for the translocation of Conservatio Chatham petrels (Pterodroma n Covenant, axillaris), Cook's petrels (P. cookii) Pterodrom Chatham Chatham and Pycroft's petrels (P. pycrofti), a axillaris Petrel 2009 Translocation Island 44 chick DOC, Table 2 Gummer et al 2014, Best practice Sweetwater techniques for the translocation of Conservatio Chatham petrels (Pterodroma 11 n Covenant, axillaris), Cook's petrels (P. cookii) Pterodrom Chatham Chatham and Pycroft's petrels (P. pycrofti), a axillaris Petrel 2010 Translocation Island 39 chick DOC, Table 2 Gummer et al 2014, Best practice Sweetwater techniques for the translocation of Conservatio Chatham petrels (Pterodroma n Covenant, axillaris), Cook's petrels (P. cookii) Pterodrom Chatham Chatham and Pycroft's petrels (P. pycrofti), a axillaris Petrel 2011 Translocation Island 70 chick DOC, Table 2 Gummer 2003, Chick translocation as a method of establishing new surface-nesting seabird colonies Pterodrom Chatham Cuvier DOC (H. Gummer pers. Obs) No a axillaris Petrel N/A Translocation Island 90 chick date provided. Pterodrom Chatham N/A Translocation Cuvier 90 chick Gummer 2003, Chick translocation

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference a axillaris Petrel Island as a method of establishing new surface-nesting seabird colonies DOC (H. Gummer pers. Obs) No date provided. Madeiros J, 2004, 2004 Translocation of Cahow chicks to Pterodrom Bermuda Nonsuch Nonsuch Island, Bermuda Audubon a cahow Petrel 2004 Translocation Island 15 chick Society 50th Anniversary Report Calile et al 2012, Establishment of a new, secure colony of Endangered Bermuda Petrel Pterodorma cahow 11 by translocation of near-fledged Pterodrom Bermuda Nonsuch nestlings, Bird Consevation a cahow Petrel 2005 Translocation Island 21 chick International Calile et al 2012, Establishment of a new, secure colony of Endangered Bermuda Petrel Pterodorma cahow by translocation of near-fledged Pterodrom Bermuda Nonsuch nestlings, Bird Consevation a cahow Petrel 2006 Translocation Island 21 chick International Calile et al 2012, Establishment of a new, secure colony of Endangered Bermuda Petrel Pterodorma cahow by translocation of near-fledged Pterodrom Bermuda Nonsuch nestlings, Bird Consevation a cahow Petrel 2007 Translocation Island 25 chick International

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Calile et al 2012, Establishment of a new, secure colony of Endangered Bermuda Petrel Pterodorma cahow by translocation of near-fledged Pterodrom Bermuda Nonsuch nestlings, Bird Consevation a cahow Petrel 2008 Translocation Island 23 chick International ESTIMATE Madeiros J, 2004, 2004 Translocation of Cahow chicks to Nonsuch Island, Bermuda Audubon Society 50th Anniversary Report, "During the last two years, a total of 11 36-near fledged Cahow chicks have been translocated to this second colony [on Nonsuch Island]" Pterodrom Bermuda Nonsuch Therefore, 36/2=13 average annual a cahow Petrel 2012 Translocation Island 13 chick translocations ESTIMATE Madeiros J, 2004, 2004 Translocation of Cahow chicks to Nonsuch Island, Bermuda Audubon Society 50th Anniversary Report, "During the last two years, a total of 36-near fledged Cahow chicks have been translocated to this second colony [on Nonsuch Island]" Pterodrom Bermuda Nonsuch Therefore, 36/2=13 average annual a cahow Petrel 2013 Translocation Island 13 chick translocations

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Gummer et al 2014, Best practice Cape's techniques for the translocation of Sanctuary Chatham petrels (Pterodroma (Hawke's axillaris), Cook's petrels (P. cookii) Pterodrom Bay), North and Pycroft's petrels (P. pycrofti), a cookii Cook's Petrel 2010 Translocation Island 50 chick DOC, Table 2 Gummer et al 2014, Best practice Cape's techniques for the translocation of Sanctuary Chatham petrels (Pterodroma (Hawke's axillaris), Cook's petrels (P. cookii) Pterodrom Bay), North and Pycroft's petrels (P. pycrofti), 11 a cookii Cook's Petrel 2011 Translocation Island 102 chick DOC, Table 2 Gummer et al 2014, Best practice Cape's techniques for the translocation of Sanctuary Chatham petrels (Pterodroma (Hawke's axillaris), Cook's petrels (P. cookii) Pterodrom Bay), North and Pycroft's petrels (P. pycrofti), a cookii Cook's Petrel 2012 Translocation Island 80 chick DOC, Table 2 Gummer et al 2014, Best practice Cape's techniques for the translocation of Sanctuary Chatham petrels (Pterodroma (Hawke's axillaris), Cook's petrels (P. cookii) Pterodrom Bay), North and Pycroft's petrels (P. pycrofti), a cookii Cook's Petrel 2013 Translocation Island 115 chick DOC, Table 2 Pterodrom Grey-faced Mount Miskelly et al 2009, Translocation of a gouldi Petrel 1999 Translocation Maunganui, 30 chick eight species of burrow nesting

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference North Island seabirds (genera Pterodroma, Pelecanoides, Pachyptila, and Puffinus: Family Procellerariidae), Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Matakohe Pelecanoides, Pachyptila, and Pterodrom Grey-faced (Limestone) Puffinus: Family Procellerariidae), a gouldi Petrel 2004 Translocation Island 40 chick Biological Conservation Miskelly et al 2009, Translocation of 11 eight species of burrow nesting seabirds (genera Pterodroma, Matakohe Pelecanoides, Pachyptila, and Pterodrom Grey-faced (Limestone) Puffinus: Family Procellerariidae), a gouldi Petrel 2005 Translocation Island 31 chick Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Matakohe Pelecanoides, Pachyptila, and Pterodrom Grey-faced (Limestone) Puffinus: Family Procellerariidae), a gouldi Petrel 2006 Translocation Island 40 chick Biological Conservation Miskelly et al 2009, Translocation of Matakohe eight species of burrow nesting Pterodrom Grey-faced (Limestone) seabirds (genera Pterodroma, a gouldi Petrel 2007 Translocation Island 22 Juvenile Pelecanoides, Pachyptila, and

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Puffinus: Family Procellerariidae), Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Matakohe Pelecanoides, Pachyptila, and Pterodrom Grey-faced (Limestone) Puffinus: Family Procellerariidae), a gouldi Petrel 2008 Translocation Island 41 Juvenile Biological Conservation Pterodrom a White-winged Cabbage ART, Priddel 2001, A trial leucoptera Petrel 1995 Translocation Tree Island 30 chick translocation fo Gould's Petrel, Emu 11 ART, Priddel et al 2004, Establishment of a an additional breeding colony of burrow-nesting Pterodrom seabirds by translocation of a White-winged Boondelbah fledglings: a successful case study, leucoptera Petrel 1999 Translocation Island 100 chick unpublished ART, Priddel et al 2004, Establishment of a an additional breeding colony of burrow-nesting Pterodrom seabirds by translocation of a White-winged Boondelbah fledglings: a successful case study, leucoptera Petrel 2000 Translocation Island 100 chick unpublished Sweetwater Miskelly et al 2009, Translocation of Pterodrom Magenta Conservatio eight species of burrow nesting a magentae Petrel 2006 Translocation n Covenant 1 chick seabirds (genera Pterodroma,

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Pelecanoides, Pachyptila, and Puffinus: Family Procellerariidae), Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Sweetwater Pelecanoides, Pachyptila, and Pterodrom Magenta Conservatio Puffinus: Family Procellerariidae), a magentae Petrel 2007 Translocation n Covenant 8 chick Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting 12 seabirds (genera Pterodroma, Sweetwater Pelecanoides, Pachyptila, and Pterodrom Magenta Conservatio Puffinus: Family Procellerariidae), a magentae Petrel 2008 Translocation n Covenant 13 chick Biological Conservation Gummer et al 2014, Best practice techniques for the translocation of Chatham petrels (Pterodroma axillaris), Cook's petrels (P. cookii) Pterodrom Pycroft's Cuvier and Pycroft's petrels (P. pycrofti), a pycrofti Petrel 2001 Translocation Island 30 chick DOC, Table 2 Gummer et al 2014, Best practice techniques for the translocation of Chatham petrels (Pterodroma Pterodrom Pycroft's Cuvier axillaris), Cook's petrels (P. cookii) a pycrofti Petrel 2002 Translocation Island 100 chick and Pycroft's petrels (P. pycrofti),

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference DOC, Table 2 Gummer et al 2014, Best practice techniques for the translocation of Chatham petrels (Pterodroma axillaris), Cook's petrels (P. cookii) Pterodrom Pycroft's Cuvier and Pycroft's petrels (P. pycrofti), a pycrofti Petrel 2003 Translocation Island 102 chick DOC, Table 2 Gummer et al 2014, Best practice techniques for the translocation of Chatham petrels (Pterodroma axillaris), Cook's petrels (P. cookii) 12 Pterodrom Pycroft's Motuora and Pycroft's petrels (P. pycrofti), a pycrofti Petrel 2013 Translocation Island 70 chick DOC, Table 2 Bell et al 2005, Translocation of Fluttering Shearwaters: developing a Puffinus Fluttering method to re-establish seabirds, gavia Shearwater 1991 Translocation Maud Island 101 Various Notornis Bell et al 2005, Translocation of Fluttering Shearwaters: developing a Puffinus Fluttering method to re-establish seabirds, gavia Shearwater 1992 Translocation Maud Island 46 chick Notornis Bell et al 2005, Translocation of Fluttering Shearwaters: developing a Puffinus Fluttering method to re-establish seabirds, gavia Shearwater 1993 Translocation Maud Island 102 chick Notornis Puffinus Fluttering 1994 Translocation Maud Island 41 chick Bell et al 2005, Translocation of

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference gavia Shearwater Fluttering Shearwaters: developing a method to re-establish seabirds, Notornis Bell et al 2005, Translocation of Fluttering Shearwaters: developing a Puffinus Fluttering method to re-establish seabirds, gavia Shearwater 1995 Translocation Maud Island 27 chick Notornis Bell et al 2005, Translocation of Fluttering Shearwaters: developing a Puffinus Fluttering method to re-establish seabirds, gavia Shearwater 1996 Translocation Maud Island 17 chick Notornis 12 Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Pelecanoides, Pachyptila, and Puffinus Fluttering Puffinus: Family Procellerariidae), gavia Shearwater 2006 Translocation Mana Island 40 chick Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Pelecanoides, Pachyptila, and Puffinus Fluttering Puffinus: Family Procellerariidae), gavia Shearwater 2007 Translocation Mana Island 91 chick Biological Conservation Miskelly et al 2009, Translocation of Puffinus Fluttering eight species of burrow nesting gavia Shearwater 2008 Translocation Mana Island 94 chick seabirds (genera Pterodroma,

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Pelecanoides, Pachyptila, and Puffinus: Family Procellerariidae), Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Pelecanoides, Pachyptila, and Puffinus Hutton's Kaikoura Puffinus: Family Procellerariidae), huttoni Shearwater 2005 Translocation Peninsula 10 chick Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting 12 seabirds (genera Pterodroma, Pelecanoides, Pachyptila, and Puffinus Hutton's Kaikoura Puffinus: Family Procellerariidae), huttoni Shearwater 2006 Translocation Peninsula 86 chick Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Pelecanoides, Pachyptila, and Puffinus Hutton's Kaikoura Puffinus: Family Procellerariidae), huttoni Shearwater 2007 Translocation Peninsula 95 chick Biological Conservation Miskelly et al 2009, Translocation of eight species of burrow nesting seabirds (genera Pterodroma, Puffinus Hutton's Kaikoura Pelecanoides, Pachyptila, and huttoni Shearwater 2008 Translocation Peninsula 100 chick Puffinus: Family Procellerariidae),

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Biological Conservation Kilauea Point ART, Byrd 1984, A cross fostering Wildlife experiement with Newell's race of Puffinus Newell's Administrati Manx shearwater, Journal of newelli Shearwater 1978 Translocation ve Site 9 Eggs Wildlife Management Kilauea Point ART, Byrd 1984, A cross fostering Wildlife experiement with Newell's race of Puffinus Newell's Administrati Manx shearwater, Journal of newelli Shearwater 1979 Translocation ve Site 13 Eggs Wildlife Management 12 ART, Byrd 1984, A cross fostering experiement with Newell's race of Puffinus Newell's Mokuaaeae Manx shearwater, Journal of newelli Shearwater 1979 Translocation Island 25 Eggs Wildlife Management Kilauea Point ART, Byrd 1984, A cross fostering Wildlife experiement with Newell's race of Puffinus Newell's Administrati Manx shearwater, Journal of newelli Shearwater 1980 Translocation ve Site 43 Eggs Wildlife Management ESTIMATE Gummer, H, 2003, Chick translocation as a method of establishing new surface-nesting seabird colonies, DOC Science Puffinus Manx Cardigan Internal Series, cites further Brooke puffinus Shearwater 1980 Translocation Island 50 chick 1990. "250 chicks were moved from

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference Skomer Island to Cardigan Island between 1980 and 1984" Therefore averaging 50 chicks per year ESTIMATE Gummer, H, 2003, Chick translocation as a method of establishing new surface-nesting seabird colonies, DOC Science Internal Series, cites further Brooke 1990. "250 chicks were moved from Skomer Island to Cardigan Island Puffinus Manx Cardigan between 1980 and 1984" Therefore 12 puffinus Shearwater 1981 Translocation Island 50 chick averaging 50 chicks per year ESTIMATE Gummer, H, 2003, Chick translocation as a method of establishing new surface-nesting seabird colonies, DOC Science Internal Series, cites further Brooke 1990. "250 chicks were moved from Skomer Island to Cardigan Island Puffinus Manx Cardigan between 1980 and 1984" Therefore puffinus Shearwater 1982 Translocation Island 50 chick averaging 50 chicks per year ESTIMATE Gummer, H, 2003, Chick translocation as a method of establishing new surface-nesting Puffinus Manx Cardigan seabird colonies, DOC Science puffinus Shearwater 1983 Translocation Island 50 chick Internal Series, cites further Brooke

Method of Scientific Colony No. Stage name English Name Year Movement Target Site Birds Class Reference 1990. "250 chicks were moved from Skomer Island to Cardigan Island between 1980 and 1984" Therefore averaging 50 chicks per year ESTIMATE Gummer, H, 2003, Chick translocation as a method of establishing new surface-nesting seabird colonies, DOC Science Internal Series, cites further Brooke 1990. "250 chicks were moved from Skomer Island to Cardigan Island 12 Puffinus Manx Cardigan between 1980 and 1984" Therefore puffinus Shearwater 1984 Translocation Island 50 chick averaging 50 chicks per year ART, Grummer 2003, Chick translocation as a method of establishing new surface-burrowing Red-footed Sea Life seabird colonies: a review, Dept of Sula sula Booby 2000 Translocation Park 19 chick Conservation, New Zealand

Table 4.A.5. Instances of viability measures (mean projected quasi-extinction risk or mean final abundance) and hierarchical intervention scenario priority (Baseline ‰‰‰Low ‰‰‰Medium ‰‰‰High ‰‰‰Essential ‰‰‰All), regardless of confidence interval or statistical difference between scenarios. Mean projected Mean projected quasi-extinction risk Mean final quasi-extinction trend tracks scenario abundance Mean final abundance Common name Scenario rank risk rank trend tracks scenario rank Chatham Island Baseline 1 N/A 1 N/A Shag High 0.99 TRUE 2 TRUE Magenta Petrel Baseline 0.01 N/A 305 N/A Essential 0.01 FALSE 371 TRUE Antipodean Baseline 0.95 N/A 13 N/A Albatross High 0.91 TRUE 34 TRUE 12 Essential 0.91 FALSE 37 TRUE All 0.91 FALSE 37 FALSE Black-Fronted Baseline 0.63 N/A 51 N/A Tern High (High) 0.16 TRUE 737 TRUE Essential (High) 0.02 TRUE 2819 TRUE All (High) 0.01 TRUE 6389 TRUE Grey-headed Baseline 0 N/A 272031 N/A Albatross Essential 0 N/A 278014 TRUE Hutton's Baseline 0 N/A 130081 N/A Shearwater Medium 0 N/A 125342 FALSE Essential 0 N/A 137114 TRUE All 0 N/A 149758 TRUE Northern Royal Baseline 0.51 N/A 534 N/A Albatross Low 0.48 TRUE 759 TRUE

Mean projected Mean projected quasi-extinction risk Mean final quasi-extinction trend tracks scenario abundance Mean final abundance Common name Scenario rank risk rank trend tracks scenario rank Essential (High) 0.3 TRUE 1870 TRUE All (High) 0.3 FALSE 1936 TRUE Pitt Island Shag Baseline 0.6 N/A 183 N/A Medium 0.48 TRUE 294 TRUE Westland Petrel Baseline 0 N/A 10171 N/A Essential 0 N/A 15602 TRUE Yellow-eyed Baseline 0.82 N/A 139 N/A Penguin Medium 0.78 TRUE 225 TRUE Fairy tern Baseline 12 0.98 N/A 6 N/A Essential (High) 0.82 TRUE 185 TRUE Auckland Island Baseline 0.92 N/A 38 N/A Shag Medium 0.85 TRUE 81 TRUE Buller's Baseline 0 N/A 544989 N/A Shearwater Medium 0 N/A 611759 TRUE Campbell Baseline 0 N/A 81858 N/A Albatross Essential 0 N/A 84721 TRUE Chatham Baseline 0.17 N/A 1513 N/A Albatross Medium 0.2 FALSE 1795 TRUE Essential 0.19 FALSE 1761 FALSE All 0.17 FALSE 2058 TRUE Chatham Petrel Baseline 0 N/A 1964 N/A High 0 N/A 1898 FALSE

Mean projected Mean projected quasi-extinction risk Mean final quasi-extinction trend tracks scenario abundance Mean final abundance Common name Scenario rank risk rank trend tracks scenario rank Essential 0 N/A 2185 TRUE All 0 N/A 2219 TRUE Cook's Petrel Baseline 0 N/A 868105 N/A Low 0 N/A 900336 TRUE High 0 N/A 882012 N/A All 0 N/A 900751 TRUE Fiordland Crested Baseline 0.18 N/A 2375 N/A Penguin Medium 0.2 FALSE 2957 TRUE Rough-faced Shag Baseline 0.93 N/A 30 N/A 12 Medium 0.86 TRUE 61 TRUE High 0.88 FALSE 51 FALSE All 0.85 TRUE 67 TRUE Black Petrel Baseline 0 N/A 2163 N/A High 0.02 FALSE 2226 TRUE Essential 0 N/A 4870 TRUE All 0 N/A 5746 TRUE Pycroft's Petrel Baseline 0 N/A 20874 N/A Medium 0 N/A 21250 TRUE High 0 N/A 21141 FALSE All 0 N/A 21280 TRUE Salvin's Albatross Baseline 0 N/A 99629 N/A Medium 0 N/A 176702 TRUE

Mean projected Mean projected quasi-extinction risk Mean final quasi-extinction trend tracks scenario abundance Mean final abundance Common name Scenario rank risk rank trend tracks scenario rank Snares Crested Baseline 0.01 N/A 18006 N/A Penguin Medium 0.01 FALSE 40073 TRUE Southern Royal Baseline 0 N/A 17360 N/A Albatross Essential 0 N/A 18739 TRUE Stewart Island Baseline 0.97 N/A 10 N/A Shag Medium 0.93 TRUE 31 TRUE High 0.9 TRUE 58 TRUE All 0.9 FALSE 56 FALSE White-chinned Baseline 0 N/A 439490 N/A 13 Petrel Low 0.01 FALSE 513442 TRUE High 0.01 FALSE 512983 FALSE Essential 0 N/A 527056 TRUE All 0 N/A 527908 TRUE White-necked Baseline 0 N/A 124631 N/A Petrel Medium 0 N/A 135699 TRUE

Table 4.A.6. Log transformed increase in abundance under All scenarios relative to Baseline scenario

Name Status Log transformed increase in abundance and 95% CI Black-fronted Tern Endangered 5.16 (4.75-5.57) Fairy Tern Vulnerable 3 (2.81-3.19) Stewart Island Shag Vulnerable 1.49 (1.36-1.62) Northern Royal Albatross Endangered 1.32 (1.27-1.38) Black Petrel Vulnerable 1.01 (0.92-1.1) Antipodean Albatross Endangered 0.9 (0.84-0.96) Snare's Crested Penguin Vulnerable 0.84 (0.79-0.9) Rough-faced Shag Vulnerable 0.79 (0.71-0.87) Chatham Island Shag Critically Endangered 0.74 (0.57-0.9) Auckland Island Shag Vulnerable 0.71 (0.63-0.79) 13 Salvin's Albatross Vulnerable 0.62 (0.55-0.69) Yellow-eyed Penguin Endangered 0.48 (0.43-0.53) Pitt Island Shag Endangered 0.47 (0.37-0.56) Westland Petrel Endangered 0.45 (0.41-0.49) Chatham Albatross Vulnerable 0.32 (0.3-0.33) Fiordland Penguin Vulnerable 0.23 (0.21-0.26) Magenta Petrel Critically Endangered 0.21 (0.18-0.23) White-chinned Petrel Vulnerable 0.2 (0.18-0.21) Hutton's Shearwater Endangered 0.14 (0.11-0.16) Buller's Shearwater Vulnerable 0.11 (0.1-0.13) Chatham Petrel Vulnerable 0.11 (0.09-0.13) Southern Royal Albatross Vulnerable 0.08 (0.07-0.09) White-necked Petrel Vulnerable 0.08 (0.07-0.09) Campbell Albatross Vulnerable 0.04 (0.03-0.05)

Cook's Petrel Vulnerable 0.04 (0.03-0.04) Grey-headed Albatross Endangered 0.02 (0.02-0.03) Pycroft's Petrel Vulnerable 0.02 (0.02-0.03)

Table 4.A.7. Prescribed intervention actions by species and action and associated viability gains (decrease in mean quasi-extinction risk or increase in mean final abundance).

Difference of means decrease Difference of means log Common in quasi-extinction risk + 95% transformed increase in Name Action Intervention(s) CI abundance + 95% CI All Vital Rate Improvementx2 0.76 (0.75-0.76) 5.16 (4.75-5.57) Black-fronted Essential Vital Rate Improvement 0.75 (0.74-0.76) 4.2 (3.87-4.52) Tern High Vital Rate Improvement 0.63 (0.59-0.67) 2.58 (2.36-2.8)

13 Invasives Removal + Northern All Translocation 0.23 (0.22-0.24) 1.32 (1.27-1.38) Royal Essential Invasives Removal 0.22 (0.21-0.23) 1.29 (1.24-1.35) Albatross Low Translocation 0.03 (0.03-0.04) 0.37 (0.35-0.39) Invasives Removal + Vital Fairy Tern Essential Rate Improvement 0.15 (0.12-0.19) 3 (2.81-3.19) Pitt Island Shag Medium Invasives Removal 0.14 (0.1-0.18) 0.47 (0.37-0.56) Invasives Removal + All Translocation 0.07 (0.05-0.09) 1.49 (1.36-1.62) Stewart Island High Invasives Removal 0.07 (0.05-0.09) 1.54 (1.4-1.68) Shag Medium Translocation 0.04 (0.03-0.05) 0.92 (0.81-1.04) Translocation +Bycatch All Mitigation 0.07 (0.06-0.09) 0.79 (0.71-0.87) Rough-faced High Bycatch Mitigation 0.05 (0.04-0.06) 0.53 (0.46-0.61) Shag Medium Translocation 0.06 (0.05-0.08) 0.7 (0.62-0.77)

Auckland Island Shag Medium Invasives Removal 0.06 (0.05-0.08) 0.71 (0.63-0.79) Invasives Removal + All Bycatch Mitigation 0.04 (0.03-0.05) 0.9 (0.84-0.96) Antipodean Essential Bycatch Mitigation 0.04 (0.03-0.05) 0.9 (0.84-0.96) Albatross High Invasives Removal 0.04 (0.03-0.05) 0.83 (0.77-0.89) Yellow-eyed Penguin Medium Invasives Removal 0.04 (0.03-0.06) 0.48 (0.43-0.53) Translocation + Harvest All Mitigation 0.01 (-0.01-0.02) 0.32 (0.3-0.33) Chatham Essential Harvest Mitigation -0.01 (-0.02-0) 0.16 (0.15-0.18) Albatross Medium Translocation -0.02 (-0.03--0.01) 0.17 (0.16-0.19) Vital Rate Improvement + All Bycatch Mitigation 0 (0-0) 1.01 (0.92-1.1) 13 Essential Bycatch Mitigation 0 (0-0) 0.85 (0.76-0.94) Black Petrel High Vital Rate Improvement -0.01 (-0.01--0.01) 0.03 (0.02-0.04) Snares Crested Penguin Medium Invasives Removal 0 (0-0) 0.84 (0.79-0.9) Chatham Invasives Removal Island Shag High 0 (0-0.01) 0.74 (0.57-0.9) Salvin's Bycatch Mitigation Albatross Medium 0 (0-0) 0.62 (0.55-0.69) Westland Petrel Essential Invasives Removal 0 (0-0) 0.45 (0.41-0.49) Magenta Invasives Removal Petrel Essential 0 (0-0) 0.21 (0.18-0.23) White- Translocation +Invasives chinned Petrel All Removal + Bycatch 0 (0-0) 0.2 (0.18-0.21)

Mitigation Essential Bycatch Mitigation 0 (0-0) 0.19 (0.18-0.21) High Invasives Removal 0 (0-0) 0.17 (0.15-0.18) Low Translocation 0 (0-0) 0.17 (0.15-0.18) Invasives Removal + All Translocation 0 (0-0) 0.14 (0.11-0.16) Hutton's Essential Invasives Removal 0 (0-0) 0.05 (0.02-0.08) Shearwater Medium Translocation 0 (0-0) -0.06 (-0.1--0.01) Buller's Shearwater Medium Translocation 0 (0-0) 0.11 (0.1-0.13) Invasives Removal + All Translocation 0 (0-0) 0.11 (0.09-0.13) Chatham Essential Translocation 0 (0-0) 0.1 (0.08-0.12) Petrel High Invasives Removal 0 (0-0) -0.04 (-0.04--0.03) 13 Southern Royal Albatross Essential Bycatch Mitigation 0 (0-0) 0.08 (0.07-0.09) White-necked Petrel Medium Translocation 0 (0-0) 0.08 (0.07-0.09) Campbell Bycatch Mitigation Albatross Essential 0 (0-0) 0.04 (0.03-0.05) Invasives Removal + All Translocation 0 (0-0) 0.04 (0.03-0.04) High Invasives Removal 0 (0-0) 0.02 (0.01-0.02) Cook's Petrel Low Translocation 0 (0-0) 0.04 (0.03-0.04) Grey-headed Bycatch Mitigation Albatross Essential 0 (0-0) 0.02 (0.02-0.03) Pycroft's Invasives Removalx2 + Petrel All Translocation 0 (0-0) 0.02 (0.02-0.03)

High Invasives Removal 0 (0-0) 0.01 (0.01-0.02) Invasives Removal + Medium Translocation 0 (0-0) 0.02 (0.01-0.02) Fiorland Crested Penguin Medium Bycatch Mitigation -0.01 (-0.03-0.01) 0.23 (0.21-0.26)

13

CHAPTER 5

Conclusion

Biodiversity and ecosystem service protection are foundational to conservation science. As obstacles become increasingly restrictive, novel approaches are required to direct conservation planning. In this dissertation I applied different analyses to answer strategic conservation questions in two key areas: water conservation and seabird conservation.

Water conservation has broad ecological importance. Increasing drought frequency and intensity is expected to effect ecological health within the U.S.

(Georgakakos et al., 2014). In Chapter 2, I evaluated the potential efficacy of conventional water conservation actions to alleviate water stress at the county-level in the continental U.S. Finding potential benefits largely insufficient; I then identified dominant water use sectors. This analysis suggests that consumer-driven water conservation has the greatest potential to alleviate water stress within the agriculture sector (generally the dominant water user). This conclusion supports existing studies of agricultural water footprints and emerging campaigns to direct consumer towards reducing dietary water use (Hoekstra and Mekonnen, 2012).

Groundwater withdrawal was included in calculating water stress and dominant water withdrawal sectors. Groundwater depletion is an important consideration for drought tolerant plants and sensitive habitats. The finding that agriculture dominates water withdrawal for the majority of counties, informs how groundwater is diverted. This study could be expanded by tracking where

136 groundwater is redistributed and the downstream effects of agricultural water use

(e.g. runoff). As climate change continues to drive increased drought density and frequency across the western U.S., studies related to effectively alleviating this pressure are crucial. This study could be replicated in other countries expected to experience increased drought and with other actions proposed to alleviate water stress.

Biodiversity protection similarly requires deliberate prioritization of target species and interventions. In Chapters 3 and 4, I scale up the use of meta-population viability analysis to guide seabird conservation. In Chapter 3, I identified threatened species with the highest relative extinction risk. I also contrast baseline extinction risk for 99 species with current IUCN Red List assessments. These results help inform prioritizations of threatened seabird species. Mismatches with current listings prioritize species for reassessment and highlight the different outcomes attributed to various assessment parameters. Current listings are based on criteria related to proxies for extinction (e.g. population decline and structure). The quantitative modeling approach presented here generally resulted in less precautionary assessments (i.e. lower predicted extinction risk), supporting previous research (Gärdenfors, 2000).

This suggests that to improve consistency among assessment criteria, thresholds for the quantitative analysis, criterion E, should be reevaluated and potentially reduced.

Future applications of the mPVA could use relative extinction risk as a reference to distinguish how individual assessment criteria A-D affect listings.

137 In Chapter 4, I outlined how PVA might be applied to national conservation planning. New Zealand is a hotspot for seabird diversity, hosting 86 of the world’s

346 seabird species, far more than any other country (IUCN, 2013; Croxall et al.,

2012b). Thirty-three of these species are endemic to New Zealand, making it the world leader for seabird endemism by a factor of 3 compared to it closest competitor

(Croxall et al., 2012b). Unfortunately, 40% of New Zealand seabird species are threatened according to the IUCN (IUCN, 2020). Using the New Zealand national action plan for threatened seabirds as a case study, I determined that most prescribed conservation scenarios would not yield significant benefits to viability. Given the importance of the region to seabird conservation, the seabird mPVA is a useful tool for rapid evaluation of conservation plans prior to implementation. This process should be replicated to project whether alternative conservation plans are likely to meet targets.

It is additionally informative to understand how prescribed actions relate to perceived efficacy. To examine this, I evaluated the effect of proposed actions against assigned priority and frequency of prescription. In general, I found a positive correlation between benefits to viability and ascribed priority, but a poor association between benefits and frequency of prescribed actions. This suggests that managers should challenge intervention expectations. This analysis provides a template to test the relative effect of interventions. Future applications of the mPVA could systematically compare viability gains of individual actions at various thresholds (e.g.

1 or 2 translocations of 200 or 400 individuals), suites of actions, and optimize island-

138 based interventions (e.g. eradication of black rats on Chatham Island could benefit up to 5 threatened seabird species).

Taken together, these chapters demonstrate the guidance that varied approaches can provide within ecosystem services and biodiversity conservation planning. Though simple, the adjusted water stress analysis exploits existing data to refocus conservation campaigns at a high-level. The mPVA analysis is complex, but likewise harnesses available information to direct investment towards the greatest opportunities for benefit. As shown here, an mPVA approach can identify species at high relative extinction risk, evaluate risk within groups, and project viability under various conservation scenarios. At a broader-level, quantitative analysis can fill in data gaps, estimate relative intervention benefits, and provide a benchmark to examine assessment criteria. If extended or replicated for other systems, these analyses could substantively inform effective conservation by guiding where effort is focused and how it is fulfilled.

139 BIBLIOGRAPHY

Akçakaya, RH, Butchart, SHM, Mace, GM, Stuart, SN, Hilton-Taylor, C. 2006. Use and misuse of the IUCN red list criteria in projecting climate change impacts on biodiversity. Glob. Chang. Biol. 12: 2037–2043.

Alcamo, J, Döll, P, Henrichs, T, Kaspar, F, Lehner, B, Rösch, T, Siebert, S. 2003. Global estimates of water withdrawals and availability under current and future “business-as-usual” conditions. Hydrol. Sci. J. 48: 339–348.

Alliance for Water Efficiency. 2016. Indoor Water Use: Dish Washer. Home Water Work. Org.

Almond, REA, M., G, Petersen, T (Eds). 2020. Living Planet Report 2020. 43-7 p.

Arizona Department, OWR. 2011. State of Arizona DEPARTMENT OF WATER RESOURCES ANNUAL REPORT FY 2011.

Averyt, K, Fisher, J, Huber-Lee, A, Lewis, A, Macknick, J, Madden, N, Rogers, J, Tellinghuisen, S. 2011. Freshwater Use by U.S. Power Plants: Electricity’s Thirst for a Precious Resource. Energy Water a Warm. World Initiat. - Union Concerned Sci.: 1–62.

Bates, B, Kundzewicz, ZW, Wu, S, Palutikof, J. 2008. Climate Change and Water. 210 p.

Brook, BW, O’Grady, JJ, Chapman, a P, Burgman, M a, Akçakaya, HR, Frankham, R. 2000. Predictive accuracy of population viability analysis in conservation biology. Nature 404: 385–387.

Caswell, H. 2001. Matrix Population Models: construction, analysis, and interpretation, 2nd Edition. Sinauer Associates, Sunderland, MA.

Catlin, DH, Zeigler, SL, Brown, MB, Dinan, LR, Fraser, JD, Hunt, KL, Jorgensen, JG. 2015. Metapopulation viability of an endangered shorebird depends on dispersal and human-created habitats: Piping plovers (Charadrius melodus) and prairie rivers. Mov. Ecol. 4: 1–15.

Chen, L, Roy, SB, Goldstein, R a. 2013. Projected Freshwater Withdrawals Under Efficiency Scenarios for Electricity Generation and Municipal Use in the United States for 2030. J. Am. Water Resour. Assoc. 49: 231–246.

Collen, B, Dulvy, NK, Gaston, KJ, Gärdenfors, U, Keith, DA, Punt, AE, Regan, HM, Böhm, M, Hedges, S, Seddon, M, Butchart, SHM, Hilton-Taylor, C, Hoffmann, M,

140 Bachman, SP, Akçakaya, HR. 2016. Clarifying misconceptions of extinction risk assessment with the IUCN Red List. Biol. Lett. 12.

Coulson, T, Mace, GM, Hudson, E, Possingham, HP. 2001. The use and abuse of population viability analysis. Trends Ecol. Evol. 16: 219–221.

Croll, DA, Maron, JL, Estes, JA, Danner, EM, Byrd, G V. 2005. Introduced Predators Transform Subarctic Islands from Grassland to Tundra. Science (80-. ). 307: 1959– 1961.

Croxall, JP, Butchart, SHM, Lascelles, B, Stattersfield, AJ, Sullivan, B, Symes, A, Taylor, P. 2012a. Seabird conservation status, threats and priority actions: a global assessment. Bird Conserv. Int. 22: 1–34.

Croxall, JP, Butchart, SHM, Lascelles, BEN, Alison, J, Sullivan, BEN, Symes, A, Taylor, P, Croxall, JP, Butchart, SHM, Lascelles, BEN, Alison, J, Sullivan, BEN, Symes, A, Seabird, PT, Stattersfield, AJ, Sullivan, BEN, Symes, A. 2012b. Seabird conservation status , threats and priority actions: a global assessment. Bird Conserv. Int.: 1–34.

Department of Natural Resources, SC. 2014. Water Usage.

Desholm, M. 2009. Avian sensitivity to mortality: Prioritising migratory bird species for assessment at proposed wind farms. J. Environ. Manage. 90: 2672–2679.

Dobson, A, Lodge, D, Alder, J, Cumming, GS, Keymer, J, McGlade, J, Mooney, H, Rusak, JA, Sala, O, Wolters, V, Wall, D, Winfree, R, Xenopoulos, MA. 2006. Habitat loss, trophic collapse, and the decline of ecosystem services. Ecology 87: 1915–1924.

Dulvy, NK, Sadovy, Y, Reynolds, JD. 2003. Extinction vulnerability in marine populations. Fish Fish. 4: 25–64.

El-Sheikh, S. 2018. IUCN ’ s position on review of progress and the post-2020 biodiversity framework. Conv. Biol. Divers.

EnergyStar. 2003. Clothes Washers. EnergyStar.

Environmental Protection Agency, U. 2015. Leak Facts. Epa.gov.

EPA. 2013a. Bathroom Faucet Fact Sheet.

EPA. 2016a. Frequently Asked Questions: Watersense Labeled Showerheads.

EPA. 2016b. Green Building.

141

EPA. Save Water and Energy by Showering Better.

EPA. 2013b. Watersense Labeled Irrigation Controller.

EPA. 2013c. Watersense Labeled Toilets Fact Sheet.

Estes, JA, Duggins, DO. 1995. Sea Otters and Kelp Forests in Alaska: Generality and Variation in a Community Ecological Paradigm. Ecol. Monogr. 65: 75–100.

Finkelstein, ME, Doak, DF, Nakagawa, M, Sievert, PR, Klavitter, J. 2010. Assessment of demographic risk factors and management priorities: Impacts on juveniles substantially affect population viability of a long-lived seabird. Anim. Conserv. 13: 148–156.

Garcia-R, JC, Di Marco, M. 2020. Drivers and trends in the extinction risk of New Zealand’s endemic birds. Biol. Conserv. 249.

Gärdenfors, U. 2000. Population Viability Analysis in the Classification of Threatened Species : Problems and Potentials. Ecol. Bull.: 181–190.

Georgakakos, A, Fleming, P, Dettinger, M, Peters-Lidard, C, Richmond, T (T. C., Reckhow, K, White, K, Yates, D. 2014. Ch. 3: Water Resources. Climate Change Impacts in the United States: The Third National Climate Assessment. Washington, DC. 69-112 p.

GoogleTrends. 2018. Available: https://trends.google.com/trends/explore?q=water savings.

Greenwald, N, Suckling, KF, Hartl, B, Mehrhoff, LA. 2019. Extinction and the U.S. Endangered Species Act. PeerJ 2019: 1–9.

Hoekstra, AY, Mekonnen, MM. 2012. The water footprint of humanity. Proc. Natl. Acad. Sci. 109: 3232–3237.

Honig, SE, Mahoney, B. 2016. Evidence of seabird guano enrichment on a coral reef in Oahu, Hawaii. Mar. Biol. 163: 1–7.

Hooper, DU, Adair, EC, Cardinale, BJ, Byrnes, JEK, Hungate, BA, Matulich, KL, Gonzalez, A, Duffy, JE, Gamfeldt, L, Connor, MI. 2012. A global synthesis reveals biodiversity loss as a major driver of ecosystem change. Nature 486: 105–108.

142 Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. 2019. Global Assessment Report on Biodiversity and Ecosystem Service. Debating Nature’s Value: 1–12.

IUCN. 2013. IUCN Red List of Threatened Species. Version 2013.1. IUCN 2013 2010: Downloaded on 22th Sept, 2013.

IUCN. 2020. The IUCN Red List of Threatened Species. Version 2020-2.

James, AN, Gaston, KJ, Balmford, A. 1999. Balancing the earth’s accounts. Nature 401: 323–324.

Jiménez Cisneros, BE, Oki, T, Arnell, NW, Benito, G, Cogley, JG, Döll, P, Jiang, T, Mwakalila, SS, Kundzewicz, Z, Nishijima, A. 2015. Freshwater resources. Clim. Chang. 2014 Impacts, Adapt. Vulnerability Part A Glob. Sect. Asp.: 229–270.

Jones, HP, Kress, SW. 2012. A review of the world’s active seabird restoration projects. J. Wildl. Manage. 76: 2–9.

Jones, HP, Tershy, BR, Zavaleta, ES, Croll, DA, Keitt, BS, Finkelstein, ME, Howald, GR. 2008. Severity of the effects of invasive rats on seabirds: A global review. Conserv. Biol. 22: 16–26.

Krug, CB, Leadley, PW, Bellard, C, Periera, HM, Alkemade, R, van Kolck, J, Sumaila, UR, Teh, LSI, Walpole, M, Newbold, T, Januchowski-Hartley, SR, Mumby, PJ. 2014. Progress towards the Aichi Biodiversity Targets: An Assessment of Biodiversity Trends, Policy Scenarios and Key Actios.

Lincoln Park Zoo. 2010. Avian Reintroduction and Translocation Database.

Lindenmayer, DB, Clark, TW, Lacy, RC, Thomas, VC. 1993. Population viability analysis as a tool in wildlife management: a review with reference to Australia. Environ. Manage. 17: 745–758.

Lofquist, D, Lugaila, T, O’Connell, M, Feliz, S. 2012. Households and families, 2010. 7 p. p.

Lundqvist, J, Fraiture, C De, Molden, D. 2008. Saving Water: From Field to Fork Curbing Losses and Wastage in the Food Chain. 5–29.

Maupin, M a., Kenny, JF, Hutson, SS, Lovelace, JK, Barber, NL, Linsey, KS. 2014. Estimated Use of Water in the United States in 2010.

143 Mazdiyasni, O, AghaKouchak, A. 2015. Substantial increase in concurrent droughts and heatwaves in the United States. Proc. Natl. Acad. Sci. 112: 11484–11489.

Mekonnen, MM, Hoekstra, AY. 2012. A Global Assessment of the Water Footprint of Farm Animal Products. Ecosystems 15: 401–415.

Miskelly, CM, Taylor, GA, Gummer, H, Williams, R. 2009. Translocations of eight species of burrow-nesting seabirds (genera Pterodroma, Pelecanoides, Pachyptila and Puffinus: Family Procellariidae). Biol. Conserv. 142: 1965–1980.

Morris, W, Doak, D, Groom, M, Kareiva, P, Fieberg, J, Gerber, L, Murphy, P, Thomson, D. 1999. A Practical Handbook for Population viability analysis. 577 p.

Morris, WF, Doak, DF. 2002. Quantitative conservation biology. Theory Pract. Popul. Viability Anal.: 480.

MPI. 2019. Aquatic Environment and Biodiversity Annual Review 2018. 538 p.

Myers, N, Mittermeler, RA, Mittermeler, CG, Da Fonseca, GAB, Kent, J. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853–858.

Oki, T, Kanae, S. 2006. Global Hydrological Cycles and Resources. Science (80-. ). 313: 1068–1072.

Perlman, H. 2015. Repairing leaky faucet Perlman. Drip Calc. How much water does a Leak. faucet waste?

Peterson, TC, Heim, RR, Hirsch, R, Kaiser, DP, Brooks, H, Diffenbaugh, NS, Dole, RM, Giovannettone, JP, Guirguis, K, Karl, TR, Katz, RW, Kunkel, K, Lettenmaier, D, McCabe, GJ, Paciorek, CJ, Ryberg, KR, Schubert, S, Silva, VBS, Stewart, BC, Vecchia, A V., Villarini, G, Vose, RS, Walsh, J, Wehner, M, Wolock, D, Wolter, K, Woodhouse, C a., Wuebbles, D. 2013. Monitoring and understanding changes in heat waves, cold waves, floods, and droughts in the United States: State of knowledge. Bull. Am. Meteorol. Soc. 94: 821–834.

Pimm, SL, Jones, HL, Diamond, J. 1988. On the Risk of Extinction. Am. Nat. 132: 757–785.

Postel, SL. 2000. Entering an Era of Water Scarcity: The Challenges Ahead. Ecol. Appl. 10: 941–948.

Postel, SL, Daily, GC, Erhlich, PR. 1996. Human Appropriation Of Renewable Fresh Water. Science (80-. ). 271: 785–787.

144 Puckett, EE, Kesler, DC, Greenwald, DN. 2016. Taxa, petitioning agency, and lawsuits affect time spent awaiting listing under the US Endangered Species Act. Biol. Conserv. 201: 220–229.

Pulwarty, RS, Jacobs, KL, Dole, RM. 2005a. Drought and Water Crises: Science, Technolgy, and Management Issues; Chapter 10. 249-286 p.

Pulwarty, RS, Jacobs, KL, Dole, RM. 2005b. The hardest working river: drought and critical water problems in the Colorado River Basin. Boca Raton, FL: CRC Press.

Pulwarty, RS, Melis, TS. 2001. Climate extremes and adaptive management on the Colorado River: Lessons from the 1997–1998 ENSO event. J. Environ. Manage. 63: 307–324.

Purvis, A, Gittleman, JL, Cowlishaw, G, Mace, GM. 2000. Predicting extinction risk in declining species. Proc. R. Soc. B Biol. Sci. 267: 1947–1952.

Robertson, CJR. 1991. Questions on the Harvesting of Toroa in the Chatham Islands.

Robertson, HA, Baird, K, Dowding, JE, Elliott, GP, Hitchmough, RA, Miskelly, CM, McArthur, N, O’Donnell, CFJ, Sagar, PM, Scofield, RP, Taylor, GA. 2017. Conservation status of New Zealand birds 2016. New Zeal. Threat Classif. Ser. 19: 26 p.

Rondinini, C, Di Marco, M, Visconti, P, Butchart, SHM, Boitani, L. 2014. Update or Outdate: Long-Term Viability of the IUCN Red List. Conserv. Lett. 7: 126–130.

Roy, SB, Ricci, PF, Summers, K V, Chung, C-F, Goldstein, R a. 2005. Evalutation of the Sustainability of Water Withdrawals in the United States, 1995 to 2025. J. Am. Water Resour. Assoc. 94303: 1091–1108.

San Francisco Public Utilities, C. 2016. Water Resources Division Annual Report FY 2015-2016.

Sanson, L. 1994. An Ecotourism Case-Study in Sub-Antarctic Islands. Ann. Tour. Res. 21: 344–354.

Schreiber, EA, Burger, J. 2001. Biology of Marine Birds.

South Staff Water. 2009. Water Use In Your Home, Tips to help save water.

Spatz, DR, Holmes, ND, Reguero, BG, Butchart, SHM, Tershy, BR, Croll, DA. 2017. Managing Invasive Mammals to Conserve Globally Threatened Seabirds in a Changing Climate. Conserv. Lett. 10: 736–747.

145

Stoutenborough, JW, Vedlitz, A. 2014. Public attitudes toward water management and drought in the United States. Water Resour. Manag. 28: 697–714.

Sun, G, McNulty, SG, Moore Myers, J a., Cohen, EC. 2008a. Impacts of multiple stresses on water demand and supply across the southeastern United States. J. Am. Water Resour. Assoc. 44: 1441–1457.

Sun, G, McNulty, SG, Moore Myers, JA, Cohen, EC. 2008b. Impacts of Climate Change , Population Growth , Land Use Change , and Groundwater Availability on Water Supply and Demand across the Conterminous U.S. Water Supply 6: 1–28.

Tauli-Corpuz, V, Alcorn, J, Molnar, A. 2018. Cornered by Protected Areas. Cornered by Prot. Areas.

Taylor, G. 2000. Action plan for seabird conservation in New Zealand. Part A, Threatened seabirds. Threat. species Occas. Publ.: 234.

TIB Partners. 2018. The Threatened Island Biodiversity Database.

Tilman, D. 1996. Biodiversity: Population Versus Ecosystem Stability. Ecology 77: 350–363.

Tilman, D. 1999. Global environmental impacts of agricultural expansion: The need for sustainable and efficient practices. Proc. Natl. Acad. Sci. U. S. A. 96: 5995–6000.

Tilman, D, Cassman, KG, Matson, PA, Naylor, R, Polasky, S. 2002. Agricultural sustainability and intensive production practices. Nature 418: 671–677.

UCSC. 2020. Seabird mPVA database developed by the Conservation Action Lab, University of California, Santa Cruz. Accessed July 20, 2020.

Vörösmarty, CJ, Green, P, Salisbury, J, Lammers, RB. 2000. Global water resources: Vulnerability from climate change and population growth. Science (80-. ). 289: 284– 288.

Vörösmarty, CJ, McIntyre, PB, Gessner, MO, Dudgeon, D, Prusevich, A, Green, P, Glidden, S, Bunn, SE, Sullivan, CA, Liermann, CR, Davies, PM. 2010. Global threats to human water security and river biodiversity. Nature 467: 555–561. de Vos, JM, Joppa, LN, Gittleman, JL, Stephens, PR, Pimm, SL. 2015. Estimating the normal background rate of species extinction. Conserv. Biol. 29: 452–462.

146 Waldron, A, Adams, V, Allan, J, Arnell, A, Asner, G, Atkinson, S, Baccini, A, Baillie, EJ, Balmford, A, Beau, JA, Brander, L, Brondizio, E, Bruner, A, Burgess, N, Burkart, K, Butchart, SHM, Button, R, Carrasco, R, Cheung, W, Christensen, V, Clements, A, Coll, M, Di Marco, M, Deguigent, M, Dinerstein, E, Ellis, E, Eppink, F, Ervin, J, Escobedo, A, Fa, J, Fernandes-Llamazares, A, Fernando, S, Fujimori, S, Fulton, B, Garnett, S, Gerber, J, Gill, D, Gopalakrishna, T, Hahn, N, Halpern, B, Hasegawa, T, Havlik, P, Heikinheimo, V, Heneghan, R, Henry, E, Humpenoder, F, Jonas, H, Jones, K, Joppa, L, Joshi, A, Jung, M, Kingston, N, Klein, C, Krisztin, T, Lam, V, Leclere, D, Lindsey, P, Locke, H, Lovejoy, T, Madgwick, P, Malhi, Y, Malmer, P, Maron, M, Mayorga, J, van Meijl, H, Miller, D, Molnar, Z, Mueller, N, Mukherjee, N, Naidoo, R, Nakamura, K, Nepal, P, Noss, R, O’Leary, B, Olson, D, Palacios Abrantes, J, Paxton, M, Popp, A, Possingham, H, Prestemon, J, Reside, A, Robinson, C, Robinson, J, Sala, E, Scherrer, K, Spalding, M, Spenceley, A, Steenbeck, J, Stehfest, E, Strassbord, B, Sumaila, R, Swinnerton, K, Sze, J, Tittensor, D, Toivonen, T, Toivonen, T, Toledo, A, Negret Torres, P, et al. 2020. Protecting 30 % of the planet for nature : costs, benefits and economic implications. Campaign Nat.

Waldron, A, Miller, DC, Redding, D, Mooers, A, Kuhn, TS, Nibbelink, N, Roberts, JT, Tobias, JA, Gittleman, JL. 2017. Reductions in global biodiversity loss predicted from conservation spending. Nature 551: 364–367.

Waldron, A, Mooers, AO, Miller, DC, Nibbelink, N, Redding, D, Kuhn, TS, Roberts, JT, Gittleman, JL. 2013. Targeting global conservation funding to limit immediate biodiversity declines. Proc. Natl. Acad. Sci. 110: 1–3.

Water Conservation Facts And Tips. 2016. Natl. Geogr. Mag.

Water Saving Tips- Food Choices. 2016. GRACE Commun. Found.

Wehner, M, Easterling, DR, Lawrimore, JH, Heim, RR, Vose, RS, Santer, BD. 2011. Projections of Future Drought in the Continental United States and Mexico. J. Hydrometeorol. 12: 1359–1377.

Worm, B, Barbier, EB, Beaumont, N, Duffy, E, Folke, C, Halpern, BS, Jackson, JBC, Lotze, HK, Micheli, F, Palumbi, SR, Sala, E, Selkoe, KA, Stachowicz, JJ, Watson, R. 2006. Impacts of Biodiversity Loss on Ocean Ecosystem Services. 314: 787–790.

147