Potential biological removal of albatrosses and petrels with minimal demographic information

Abstract

Seabirds such as albatrosses and petrels are frequently caught in longline and trawl fisheries, but limited demographic data for many species creates management challenges. A method for estimating the potential biological removal (the PBR method) for birds requires knowledge of adult survival, age at first breeding, a conservation goal, and the lower limit of a 60% confidence interval for the population size. For seabirds, usually only the number of breeding pairs is known, rather than the actual population size. This requires estimating the population size from the number of breeding pairs when important demographic variables, such as breeding success, juvenile survival, and the proportion of the adult population that engages in breeding, are unknown. In order to do this, a simple population model was built where some demographic parameters were known while others were constrained by considering plausible asymptotic estimates of the growth rate. While the median posterior population estimates are sensitive to the assumed population growth rate, the 20th percentile estimates are not. This allows the calculation of a modified PBR value that is based on the number of breeding pairs instead of the population size. For threatened albatross species, this suggests that human-caused mortalities should not exceed 1.5% of the number of breeding pairs, while for threatened petrel species, mortalities should be kept below 1.2% of the number of breeding pairs. The method is applied to 22 species and sub-species of albatrosses and petrels in New Zealand that are of management concern, of which at least 10 have suffered mortalities near or above these levels.

Introduction

Albatrosses and petrels (the Procellariiformes) are naturally long-lived birds, killed on land and at sea by a variety of sources of anthropogenic origin, likely at rates beyond those they can sustain. An important mortality source is through interactions with fisheries, with high levels of mortalities associated with numerous commercial fisheries (Weimerskirch and Jouventin, 1987, Brothers, 1991, Weimerskirch et al., 1997, Gales, 1998, Gales et al., 1998, Ryan and Boix-Hinzen, 1999, Sagar et al., 2000, Inchausti and Weimerskirch, 2001, Tuck et al., 2001, Baker et al., 2002, Baker et al., 2007, Nel et al., 2002, Ryan et al., 2002, Lewison and Crowder, 2003, Croxall, 2008, Moore and Žydelis, 2008, Ryan and Watkins, 2008, Zador et al., 2008, Žydelis et al., 2009). While it is difficult to estimate fisheries related mortalities accurately (Uhlmann et al., 2005, Miller and Skalski, 2006), and mitigation measures have drastically reduced bycatch rates in some fisheries (SC-CAMLR, 2006, Gilman et al., 2008, Delord et al., 2010), hundreds of thousands of seabirds are killed each year (Baker et al., 2007, Žydelis et al., 2009, Brothers et al., 2010). On land, there are risks including mortality and habitat degradation associated with alien species (Seto and Conant, 1996, Imber et al., 2000, Imber et al., 2003, Baker et al., 2002, Jouventin et al., 2003, Igual et al., 2006, Jones et al., 2008). Seabird demographic parameters, such as breeding success and survival, are also linked to environmental conditions (Baduini et al., 2001, Thompson and Ollason, 2001, Weimerskirch et al., 2003, Doherty et al., 2004, Ainley et al., 2005, Grosbois and Thompson, 2005, Jenouvrier et al., 2005, Votier et al., 2005, Crespin et al., 2006, Delord et al., 2008), and climate change must be considered a threat (Croxall, 2004, Boyce et al., 2006, Sutherland et al., 2006, Barbraud et al., 2008). Baker et al. (2002) provide a good overview of additional threats that seabirds face, such as over-extraction of prey-species, and chemical and physical pollution.

The population-level impact of bycatch and other sources of human-caused mortality on albatrosses and petrels is not well understood, with many difficulties caused by limited demographic data. These populations are sensitive to changes in adult survival (Crespin et al., 2006), which means that their ability to sustain mortalities beyond natural mortality is limited. While many species are thought to be in decline (Gales, 1998), definitive population trends are lacking for most species (Baker et al., 2002), and population estimates are based on rule-of-thumb multipliers from estimates of the number of breeding pairs (Gales, 1998, Taylor, 2000, Brooke, 2004a, Brooke, 2004b). This means that there is a need to develop tools to help guide managers when there is minimal demographic data, and to develop ways to assess the impacts or likely impacts from human activities.

Demographic data are often limited to adult survival (s), age at first breeding (α), and the number of breeding pairs (B). An important conservation challenge is assessing the ability of albatrosses and petrels to sustain mortalities from interactions with fisheries and other sources with this limited data. One approach to assessing the potential for populations to sustain additional mortalities is given by the PBR method:

$$\mathit{PBR}=\frac{1}{2}{R}_{\max }{N}_{\min }f$$

where R_max is the maximum annual recruitment rate, N_min is a conservative estimate of population size (Wade (1998) recommended the 20th percentile) and f is a recovery factor between 0.1 and 1 (Wade, 1998, Taylor et al., 2000, Hunter and Caswell, 2005, Niel and Lebreton, 2005, Dillingham and Fletcher, 2008). A value of f = 0.1 is suggested for threatened species, f = 0.3 for near-threatened species, and f = 0.5 for all other species due to the potential for bias in population estimates (Wade, 1998, Dillingham and Fletcher, 2008); values can be more or less conservative than these suggestions depending on individual circumstances (e.g. expected future risks, knowledge of the population resiliency, or the management objective could be used to raise or lower f). The calculated value can be used to set mortality limits provided all mortality sources are considered, or to determine that mortality levels are likely sustainable, unsustainable, or that sustainability cannot be determined without further data (Dillingham and Fletcher, 2008). For albatrosses and petrels, mortality estimates are often inaccurate or unavailable. Even for regulated fisheries with observer programs, insufficient observer coverage and data quality hinders accurate mortality estimates, especially for rare species (Uhlmann et al., 2005, Miller and Skalski, 2006). Further, up to half of mortalities in longline fisheries may be unobserved due to drop-out between hooking and retrieval (Brothers et al., 2010), while substantial numbers of seabirds have been observed deliberately cut from hooks to avoid detection (Gales et al., 1998). Unaccounted mortalities are also a substantial problem in trawl fisheries (Moore and Žydelis, 2008, Ryan and Watkins, 2008). Therefore, this method is more likely, at present, to help assess the possible impact from individual sources, rather than the overall sustainability.

While the PBR method does not require much data, the data that are required are not directly present for albatrosses and petrels. However, allometric relationships can be used to calculate the maximum annual growth rate (λ_max = R_max + 1) for birds (Niel and Lebreton, 2005) from s and α. Unfortunately, population estimates are typically based on limited data and educated guesses, extrapolated from the number of breeding pairs. If measured at all, the number of breeding pairs in a given year may be measured accurately (e.g. a large albatross that spends a great deal of time on a nest is relatively easy to observe) or with difficulty (e.g. burrowing, nocturnal petrels). However, all species have a sizable portion of the population that does not return to the colony in a given year, which makes it challenging to estimate the total population size. In order to use the PBR method for albatrosses and petrels, an approach to population estimation given s, α, and the number of breeding pairs (B) is developed. These estimates are combined with the allometric calculation of R_max to create a modified PBR method specific to albatrosses and petrels.

The approach is based on building a simple population model and considering the distribution of demographic parameter values that lead to plausible asymptotic estimates of the growth rate λ. These estimates determine the population size per breeding pair. Combined with an estimate of the number of breeding pairs, these can be used to estimate the population size, and provide quantiles of the estimate. Estimates are based on asymptotic results assuming constant values for s, α and breeding parameters, and will perform better for populations with limited temporal variation in these parameters. Combined with the PBR method, this allows at least a rough assessment of the potential for albatrosses and petrels to sustain additional mortalities given only s and α, and the number of breeding pairs.

Procellariiformes begin life as well-tended chicks, spend several years away as juveniles, return either to their natal colony or elsewhere as pre-breeders, and, if successful at acquiring a mate, become breeders. Breeding pairs attempt to raise, at most, one chick annually (with the exception of a few small tropical species not considered here; Brooke, 2004a). However, not all adult birds are breeders, and breeders may skip breeding for a year, so there is also a group of non-breeding adults. For many populations, little is known about the numbers of juveniles, pre-breeders, and non-breeding adults.

Estimates of the number of breeding pairs usually refer to per annum breeding pairs and omit skipping breeders, rather than the total number of pairs that sometimes breed. In particular, breeding pairs belonging to Diomedea and Phoebetria species skip breeding the year after breeding successfully yet are still part of a breeding pair. Non-breeding adults are then potentially composed of two groups of birds; the first group are those that bred successfully in the previous year and are therefore obligate non-breeders, and the second group are the other non-breeders, which includes both mature birds who are not currently members of a breeding pair as well as members of a breeding pair who skip breeding even though they are not obligate non-breeders.

In the primary literature, population size of seabirds is often used interchangeably or near-interchangeably with the number of breeding pairs (e.g. Woehler and Croxall, 1997, Baker et al., 2002, Elliott and Walker, 2005, Delord et al., 2008). This is because many standard methods of estimating animal abundance (Schwarz and Seber, 1999) are not applicable; while breeding birds appear at colonies, other birds may not (Baker et al., 2002). While information limited to the number of breeding pairs may be helpful in looking at population trends, there may be substantial variability in the proportion of birds breeding in a given year (Chastel et al., 1995, Cam et al., 1998, Jenouvrier et al., 2005). Thus, short-term trends in the number of breeding pairs may not be related to trends in the population size, while the long-term relationship between the number of breeding pairs and population size may not be linear.

Chastel et al. (1995) found that body condition could influence both breeding success and the proportion of experienced breeders not breeding, and that some long-lived species may prefer to reduce reproductive effort to increase survival probability, while Elliott and Walker (2005) observed a c. 50% increase in the number of nests of the Antipodean albatross (Diomedea antipodensis) on Antipodes I. between 2001 and 2004 that cannot be explained by intrinsic population growth. Further, the proportion of non-breeding experienced breeders varies considerably between species (Chastel et al., 1995) and within species (Nevoux et al., 2010). This suggests that counts of the short-term number of breeding pairs, especially for long-lived birds such as albatrosses and large petrels, could be influenced by environmental conditions rather than a change in population size; additionally, it is possible that the average proportion of non-breeding adults is related to population density (i.e. is density-dependent). In order to assess the ability of a population to sustain additional mortalities, an actual population estimate is needed. It is therefore important to not only develop a rule-of-thumb multiplier (as in Brooke (2004b)), but to have some understanding of the potential variability in the multiplier. This means that any rule-of-thumb multiplier relating the number of observed breeding pairs to the total population size needs to account for uncertainty in the proportion of non-breeding adults.

In order to develop the modified PBR method for albatrosses and petrels, the primary challenge is in estimating the population size from α, s, and B to calculate N_min for a range of demographic parameter values. Since the number of breeding pairs is estimated rather than known, uncertainty in B should also be incorporated. The secondary step is to combine these estimates with allometric estimates of R_max to get an equation of the form,

$$\mathit{PBR}=\tau f\widehat{B}$$

where τ is a coefficient that incorporates a species’ maximum growth rate and a species-appropriate population multiplier, and also includes uncertainty in the estimate of the number of breeding pairs, and $\widehat{B}$ is the estimated number of breeding pairs.

This methodology is applied to 22 species or sub-species in New Zealand deemed by the New Zealand Ministry of Fisheries to be vulnerable to fishing (Fletcher et al., 2008). Many of the over 40 Procellariiform species in the New Zealand region are threatened (BirdLife International, 2010), with most breeding at limited numbers of locations on a few offshore islands (Gales, 1998, Brooke, 2004a). There is a dearth of data for most of these species for many demographic parameters (Brooke, 2004a, ACAP, 2010), and the inaccessibility of colonies make many species difficult and costly to study. While these populations and fisheries bycatch were the motivation for developing this method, it could be applied to most albatross and petrel species found globally, where many of the same threats and data limitations occur.