Effects of Geolocation Archival Tags on Reproduction and Adult Body Mass of Sooty Shearwaters (Puffinus Griseus)

Home , Sooty shearwater

New Zealand Journal of Zoology, 2009, Vol. 36: 355-366 355 1175-8821 (Online); 0301-4223 (Print)/09/3603-0355 © The Royal Society of New Zealand 2009

Effects of geolocation archival tags on reproduction and adult body mass of sooty shearwaters (Puffinus griseus)

JOSH ADAMS1,* chick groups was not significantly different after DARREN SCOTT2 controlling for skeletal size. Effects on chicks were 2 more pronounced in six pairs where both parents SAM McKECHNIE carried tags. Chick mass was negatively related to GRANT BLACKWELL2 the duration that adults carried tags. In this study, SCOTT A. SHAFFER3 none of the chicks reared by pairs where both parents 2 were tagged, 54% of chicks reared by pairs where HENRIK MOLLER one parent was tagged, and 83% of chicks reared 1 United States Geological Survey by non-handled and non-tagged parents achieved Western Ecological Research Center a previously determined pre-fiedging mass thresh- Moss Landing Marine Laboratories old (564 g; Sagar & Horning 1998). Body mass 8272 Moss Landing Road of adults carrying tags and returning from trans- Moss Landing equatorial migration the following year were 4% California, 95039, USA lighter on average than non-tagged birds, but this 2Kia Mau Te Tītī Mo Ake Tōnu Atu Research Team difference was not statistically significant. Reduced Centre for Study of Agriculture mass among chicks reared by adults carrying tags Food & Environment during the chick-provisioning period indicated that University of Otago adults altered "normal" provisioning behaviours to PO Box 56, Dunedin 9054, New Zealand maintain their own body condition at the expense 3 of their chicks. Population-level information de- Ecology & Evolutionary Biology rived from telemetry studies can reveal important University of California habitat-linked behaviours, unique aspects of sea- Santa Cruz, California, 95060, USA bird foraging behaviours, and migration ecology. *Present address: Centre for Study of Agriculture, Information for some species (e.g., overlap with Food & Environment and Department of Zoology, fisheries) can aid conservation and marine ecosystem University of Otago, PO Box 56, Dunedin 9054, management. We advise caution, however, when New Zealand. [email protected] interpreting certain data related to adult provisioning behaviours (e.g., time spent foraging, provisioning

Downloaded by [24.94.86.228] at 22:26 20 September 2013 rates, etc.). If effects on individuals are of concern, Abstract We attached 11 g (1.4% body-mass we suggest shorter-term deployments, smaller and equivalent) global location sensing (GLS) archival lighter tags, and alternative attachment techniques, tag packages to tarsi of 25 breeding sooty shearwa- especially when investigating threatened or endan- ters (Puffinus griseus, titi) on Whenua Hou (Cod- gered species. fish Island), New Zealand during the chick-rearing period in 2005. Compared with chicks reared by non-handled adults that did not carry tags, deploy- Keywords breeding success; foraging; geoloca- ment of tags on one or both adult parents ultimately tion and telemetry; shearwaters; sooty shearwaters; resulted in 35% reduction in chick body mass and tag effects significantly reduced chick skeletal size preceding fledging (19 April). However, body mass between INTRODUCTION Radio telemetry, satellite telemetry (Argos 2008), Z09007; Online publication date 3 September 2009 global positioning systems (GPS), and global lo- Received 7 April 2009; accepted 25 July 2009 cation sensing systems (GLS or geolocation) are 356 New Zealand Journal of Zoology, 2009, Vol. 36

widely used methods for determining movements of that devices attached to breeding procellariiform free-ranging seabirds (Burger & Shaffer 2008). most seabirds should not exceed c. 3% of body mass. ARGOS satellite platform terminal transmitters despite the increasing use of GlS tags (especially (PTTs) and GPS devices are expensive (c. US$2,500 among small- to medium-sized seabirds <1500 g), per device), the former require additional data fees few studies have addressed the effects that these and until recently, packages were relatively heavy devices may have on breeding success, or condition (>20 g). in contrast, recent advances in microproces- of tagged birds or their offspring (but see igual et al. sor-based electronics enable production of relatively 2005). herein, we investigated the effects of GlS inexpensive GlS archival tags (hereafter "tags") tags applied to provisioning adult sooty shearwaters. that are small, lighter (minimum size <5 g), and Specifically, we assessed effects on chick size (com- enable long-term (>1 year) deployments because bination of skeletal morphometrics) and chick mass they are less constrained by power requirements. by comparing two groups: chicks reared by parents GlS tags record light-levels and some also record where one or both parents were fitted with tags ver- temperature, and pressure (wilson et al. 2002; Shaf- sus chicks reared by non-handled and non-tagged fer et al. 2006). Once integrated with an accurate parents. We also compared mass at first landfall clock, these measures are used to calculate daily among returning, post-wintering adults with and latitudinal (potentially adjusted using sea surface without tags. Results detailing aspects of the forag- temperature) and longitudinal fixes (± c. 200 km ing ecology, diving behaviours, and trans-equatorial accuracy) based on day-length and timing of sunrise migration of sooty shearwaters based primarily on and sunset (Shaffer et al. 2005). hydrostatic pressure the deployments of tags in this study are presented also can be measured to derive dive-depth data that by Shaffer et al. (2006,2009). link foraging behaviour to oceanic conditions and location (Shaffer et al. 2009). in seabird studies, tags most often are attached to leg bands to minimise loss (Phillips et al. 2004; Methods Shaffer et al. 2006; Gonzáles-Solís et al. 2007; but see Tremblay et al. 2003 for an alternate attachment). study species Added mass, increased aerodynamic and hydrody- Sooty shearwater (Puffinus griseus, tītī) is a me- namic drag, and modification of a bird's centre of dium-sized (c. 750-1100 g) burrow-nesting petrel mass caused by externally attached devices can in- that departs its colonies in April-may in the South- crease energy expenditure and alter normal behav- ern hemisphere after breeding to spend the austral iours (Wilson & Culik 1992; Hull 1997; Phillips et winter in high-productivity oceanic regions in the al. 2003). in species such as sooty shearwaters (Puffi- Northern hemisphere (Shaffer et al. 2006). indi- nus griseus) where diving can be frequent, pro- viduals return to their colonies to initiate breeding longed, and deep (Shaffer et al. 2009), effects in October-November (Richdale 1944,1945,1963; imposed on adults by carrying tags may translate to warham et al. 1982; Spear & Ainley 1999). Sooty effects on breeding success. for example, adult shearwaters are long-lived (>30 years) and breed

Downloaded by [24.94.86.228] at 22:26 20 September 2013 seabirds may alter foraging behaviours to maintain many times during their lifetimes. their own condition at the expense of their offspring Sooty shearwaters nest on a few remaining main- (drent & dann 1980; croxall 1982; Saether et al. land sites and in large colonies on offshore islands 1993; Navarro & González-Solís 2007). Tracking in New Zealand (warham & wilson 1982; hamilton devices deployed on seabirds have been shown to et al. 1997; hamilton 1998; Scott et al. 2008) and in disrupt food consumption (wanless et al. 1989), southern South America (everett & Pittman 1993; decrease adult mass and colony attendance (Söhle Reyes-Arriagada et al. 2007). females lay a single 2003), prolong foraging trips and alter blood chem- egg between mid November and early december, istry (Navarro & González-Solís 2006), and alter and pairs take turns incubating for c. 53 days (Rich- diving behaviour (healy et al. 2004). These effects dale 1944; warham et al. 1982). eggs hatch in Janu- can cause reduced chick growth rates (Ackerman et ary and chicks remain in their burrow for 86-106 al. 2004), abandonment of breeding attempts (wan- days. They emerge to exercise their wings in mid to less et al. 1988), and decreased survival of chicks to late April, and desert their natal breeding colony by fledging (Phillips et al. 2003; Ackerman et al. 2004). mid may (falla 1934; Richdale 1954; warham & To decrease the likelihood of extended trip durations Wilson 1982). Among seabirds, fledglings with the and nest desertions, Phillips et al. (2003) proposed greatest mass have the highest probability of survival Adams et al.—Archival tag effects 357

to recruitment (Perrins et al. 1973; coulson & Porter tags. we attached metal bands to the right tarsi and 1985; Miskelly & Taylor 2004). Richdale (1954) first determined mass (±5.0 g; 1500 g Pesola® spring proposed a threshold pre-fledging mass of 455 g for scale, Pesola Ag, Baar, Switzerland). sooty shearwater chicks, below which post-fledging we attached multi-sensor (light, temperature, and survival and subsequent recruitment was not ob- pressure) tags (Lotek LTD 2400, Lotek Wireless, served. data presented by Sagar & horning (1998) Newmarket, Ontario, Canada) to an acrylic saddle indicated that post-fledging survival and recruitment secured to a plastic darvic band with epoxy (loc- required a minimum chick mass of 564 g. Parents tite®, Henkel Technologies, Düsseldorf, Germany) may reduce investment in chick rearing when breed- and two cable ties (fig. 1). each darvic band was ing conditions are unfavourable (clucas et al. 2008; pre-shaped to fit the flattened contour of the tarsus Newman et al. 2009). Parents are expected to forego and all rough edges were filed smooth. Each darvic or abandon breeding to maintain their own body band was compressed to prevent full rotation around condition when faced with sub-optimal foraging the tarsus but could move freely along the length conditions and in studies where researcher manipu- of the tarsus. The combined tag package weighed lations have increased energetic expenditure during 11 g (1.4% of an individual's body mass) and was foraging (drent & dann 1980; duriez et al. 2000). placed on the left tarsus of each adult. At the time of tagging, we collected a feather sample and a study site drop of blood in a micro-capillary tube by pricking we studied sooty shearwaters nesting on whenua each adult's medial tarsal vein with a sterile needle. Hou (Codfish Island; 46.8°S, 167.7°e) during the we determined sex for 21 of 25 adults equipped 2004-05 breeding season (mid November 2004 to with tags: 10 males and 11 females (Griffiths et al. mid May 2005) and upon the adults' return to the 1998; dawson et al. 2001). during the initial capture colony in October-december 2005 after completion period in January-february 2005, six adults were re- of trans-equatorial migrations. whenua hou is a captured a second time to re-program and re-deploy 1396 ha nature reserve located 3 km west from the tags. Upon each tag deployment, all birds were held north-west coast of Rakiura (Stewart Island), south- for an additional c. 15 min to allow the epoxy to cure ern New Zealand (Scott et al. 2009, this issue). in before returning them to their breeding burrows. for early January 2005, before capturing birds to attach each bird, we recorded total handling time, number tags, we identified nesting burrows using an infrared of times handled, and tag deployment duration. we burrowscope (lyver et al. 1998) to inspect all burrow arbitrarily selected a comparison group of chicks entrances in three discrete sub-colonies (Alphonse (i.e., chicks reared by non-tagged breeding adults) A, B, and c; n = 180, 100, and 170, respectively; that were located within burrows in Alphonse A, B, see clucas et al. 2008; Newman et al. 2009). when and c sub-colonies. we did not handle adults within necessary, we created sealable inspection hatches our comparison chick group. over nests sites to aid capture of adults and chicks. Measurement of chicks during the breeding capture, handling, and tag attachment season and adults returning after post-breeding

Downloaded by [24.94.86.228] at 22:26 20 September 2013 Between 28 January and 15 february 2005 (early migration chick-rearing stage), we captured 25 breeding adults All chicks reared by 19 pairs where at least one (from 19 breeding burrows) at night using one-way adult member carried a tag, and all chicks reared trapdoors attached to burrow entrances. we random- by a representative sample of 42 non-tagged pairs, ly selected breeding burrows with single entrances were weighed and measured on 19 April 2005. we and discrete nesting chambers at the Alphonse A measured chick mass (±5.0 g; 1000 g Pesola® spring sub-colony. we excluded from consideration any scale) and obtained seven measures of skeletal size: entrances that led to more than one nesting chamber tarsus-length, mid-toe (to tip of claw) length, bill- and nesting chambers with more than one burrow depth, culmen-length, culmen plus head-length (i.e., entrance. This maximised the chances that all tagged total head-length), nares to culmen tip length (all adults could be recaptured and ensured that they skeletal measurements ±0.1 mm; vernier callipers) were associated with only one chick per entrance. and flattened wing chord (±1 mm; stopped rule). Once removed from their burrows, individual birds we estimated percentage coverage of down on the were held in a cotton pillowcase to limit visual stim- ventral and dorsal sides combined to the nearest ulation. we retained birds in the pillowcase while 10% following Hunter et al. (2000). All chick mea- banding, obtaining measurements, and attaching surements were conducted by dS. At the Alphonse 358 New Zealand Journal of Zoology, 2009, Vol. 36

between log-transformed chick mass among birds reared by non-tagged adults. first we conducted principal components analyses (PcA) on three body-size measures (tarsus length, total head length, and wing chord) to account for multicollinearity among these morphometric variables (Graham 2003). We controlled for chicks' size in our AN- COVA by including the first principal component (PC 1) as a covariate. After pooling chicks reared by non-tagged adults in the three sub-colonies, we conducted a similar ANcOVA to determine differences between the tagged and non-tagged comparison groups. Secondly, we tested for differences in all body measurements between the two comparison groups using two sample t-tests. All tests were two- tailed with alpha = 0.05. We used a backward, stepwise, linear regression (Systat 2007) to evaluate which of the following three explanatory variables affected log-transformed chick mass (dependent variable) on 19 April 2005: ( 1 ) Total number of times adult(s) were handled per nest-site (number of times handled); (2) Total time adult(s) handled per nest-site (total handling time: handling required for initial attachment of a transmitter was 15 min [does not include 15 min non-manipulation restraint period during epoxy curing] + subsequent handling durations for removal or refitting of a transmitter); and (3) Total number of days adult(s) carried tags per nest-site (tracking duration). Tracking duration reflected the summed number of days one or both adults carried tags to better account for combined effects when both adult members of a pair were tagged. we included Pc1 in our regres- Fig. 1 light, temperature, depth GlS tags were attached sion to control for the effects of chick skeletal size. Downloaded by [24.94.86.228] at 22:26 20 September 2013 to the left tarsi of sooty shearwaters captured during the Before evaluating the regression model, we assessed 2004-05 breeding season on Whenua Hou (Codfish Island), potential multicollinearity among each of these New Zealand (A). close-up image of a tag before it was three variables with simple pair-wise Pearson's removed from a shearwater which had just made first land- correlations. fall after completing a trans-equatorial migration (B). Because mass at fledging can affect post-fledging survival and subsequent recruitment, we evaluated the effects of carrying tags on chick mass in the A sub-colony between 21 October and 4 November context of Sagar & horning (1998) who predicted 2005, we captured tagged and non-tagged adult a minimum threshold mass of 564 g for subsequent birds at their first landfall, and measured their body survival (i.e., recruitment to the population return- masses. ing to the colony). According to Sagar & horning (1998) the proportion of birds surviving increases Analyses with mass above 564 g, but the authors do not pres- Before pooling data from the three sub-colonies to ent a function describing this increase. All values compare group chick masses and skeletal morpho- reported throughout are mean ± standard error metrics, we used ANcOVA to test for differences (Se). Adams et al.—Archival tag effects 359

RESULTS chick mass and size PC1 accounted for 79% (λ = 2.37) of the total vari- tag deployments ance explained by the components and all three in total, 25 adults from 19 breeding burrows were skeletal variables loaded heavily and positively on fitted with GLS tags between 28 January and 15 this PC (all loadings >0.83). There was a significant february 2005. in six of 19 breeding burrows both difference in log-transformed mass among chicks members of the pair were fitted with tags. In 13 reared by non-tagged adults and surviving to 19 April breeding burrows only one member of a pair was fit- among the three sub-colonies (Alphonse A, B, and ted with a tag. Subsequent death of one chick (reared c; ANcOVA; sub-colony: F2,38 = 3.53, P = 0.039; by a pair where both members were tagged) reduced Pc1 : F1,3 8 = 3.53, P < 0.001 ; Tukey's Honestly-Sig- our sample size for the following results to 18 pairs. nificant-Difference Test: Alphonse A versus B, P = The number of times an individual adult was handled 0.040). Among chicks reared by non-tagged parents ranged from one to four times. handling time for in the three sub-colonies, there were no significant pairs (individuals summed when both parents were differences among the seven skeletal measures and tagged) equipped with tags was 24.2 ±2.1 min (n = percentage down (arcsine-transformed percentage). 18 pairs). Tracking duration varied during the chick Accordingly, we pooled the chick masses from two provisioning period (97 ± 8 shearwater days, n=18 sub-colonies (Alphonse A and C) and skeletal data pairs, individuals within a pair carried tags for 77 ± and percentage down from all the three sub-colonies 1 days, n = 25 individuals). (Alphonse A, B and c) to form the non-tagged comparison groups (Table 1). Chicks surviving to 19 Breeding success of tagged birds April reared by non-tagged pairs (Alphonse A and Eighteen of 19 chicks (95%) reared by tagged par- C) were heavier (769.7 ± 36.1 g; n = 31) and larger ents survived to 19 April. Two of the study chicks than chicks reared by adults fitted with tags in Al- failed to fledge: both parents tagged at the same bur- phonse A (498.3 ± 51.0; n = 18), but this difference row abandoned their chick within the first week of was not significantly different when analyses were this study (one parent deserted immediately after tag controlled for skeletal-size among chicks (ANCOVA log-transformed mass: F , = 2.87, P = 0.097; Pc1 : attachment; the second parent deserted two nights 1 46 F1, = 40.62, P < 0.001). The average mass and size later after being recaptured to remove and replace 4 6 of chicks reared by pairs where only one adult was its tag) and a second tagged pair reared a chick to 19 tagged (584.6 ± 52.4 g, n = 13) were greater than April (200 g), but this chick died before fledging.

table 1 Differences in body measurements, including, mass, flattened wing chord, total head-length (THL), percentage down coverage, nares to tip of culmen-length, bill-depth, culmen-length, tarsus-length, and mid-toe to tip of claw length between pre-fledging (19 April 2005) sooty shearwater chicks reared by parents on Whenua Hou (Codfish Island), Downloaded by [24.94.86.228] at 22:26 20 September 2013 New Zealand with and without GLS tags. Data presented are mean ± SE, and sample sizes are shown in parentheses. Also shown are Student's two-sample t-values (two-tailed) and corresponding P-values. Chicks reared by non-tagged adults Chicks reared by tagged adults (42) (18) Parameter mean Se mean Se t P mass1 769.7(31) 36.1 498.3 51.0 1.69 0.097 flat wing chord (mm) 254.4 4.8 220 8.0 3.81 < 0.001 Thl (mm) 93.4 0.6 89.5 0.8 3.50 < 0.001 down (%) 90.8 2.4 98.0 0.6 -2.87 0.006 Nares-length (mm) 33.3 0.2 32.5 0.4 1.93 0.059 Bill-depth (mm) 13.6 0.1 13.1 0.2 2.62 0.011 culmen-length (mm) 40.8 0.3 39.8 0.4 2.20 0.031 Tarsus-length (mm) 57.7 0.4 55.8 0.4 3.07 0.003 mid-toe length (mm) 69.1 0.4 67.6 0.6 2.27 0.027 1for mass, n = 31 (Alphonse A and C sub-colonies only), ANCOVA with skeletal size as covariate (PC1); t = 6.37, P< 0.001. 360 New Zealand Journal of Zoology, 2009, Vol. 36

those of the five chicks reared by pairs when both are prone to desertions during incubation following adults were tagged (274.0 ± 31.7 g, n = 5; ANcOVA disturbance (warham 1990). furthermore, because

log-transformed mass: Fus = 7.52, P = 0.015; and we lacked the appropriate control groups and data

skeletal size PC1: F115 = 7.85, P = 0.013). Chicks to assess desertion, we can not make comparisons reared by non-tagged adults had significantly greater with the "normal" rate of desertion that occurred in skeletal measurements (except nares-length; Table 2005. An important aspect of the lore and Traditional 1). Backward stepwise regression revealed that log- ecological Knowledge of the Rakiura Māori and transformed chick mass on 19 April was negatively Hauraki Māori muttonbirders (harvesters) stipulates related to tracking duration (t=-2.44, P=0.028) and strict avoidance of disturbance to adults (Kitson & Pc1 (t = 2.71, P = 0.016). By 19 April 2005, none moller 2008; lyver et al. 2008; moller & lyver in of the five chicks reared by pairs where both parents press). when adult short-tailed shearwaters (Puffi- were tagged, and seven of 13 chicks (54%) reared nus tenurostris) were handled, Klomp & Schultz by a pair where one adult was tagged exceeded the (2000) observed greater rates of nest desertion dur- 564 g threshold (Sagar & horning 1998). in contrast, ing incubation than during post-guard-chick-rearing 35 of 42 of chicks (83%) reared by non-tagged pairs periods. Söhle (2003) found adult sooty shearwaters exceeded 564 g. that were handled near the time of hatching, but not subjected to carrying 26 g imitation satellite trans- Mass of adults returning from migration mitter packages (ISTs), reduced colony attendance Fourteen of 25 adults (56%) fitted with tags in Jan- by 40% (this was not statistically different from a uary-february 2005 (six of 10 males and eight of third group of adults that were outfitted with ISTs). 11 females) were recaptured in their burrows in Among sooty shearwaters fitted with ISTs, Söhle the Alphonse A colony between 21 October and 4 (2003) found colony attendance was reduced by November 2005, after completing 5—6 month long, 26% from early march to mid April (hatching- trans-equatorial migrations (see Shaffer et al. 2006). early chick-rearing periods). Grey-faced petrels All tagged adults appeared in good condition with (Pterodroma macroptera gouldi) abandoned eggs only light abrasions on their tarsi caused by the dar- after attaching satellite transmitters, but this effect vic bands which held the tags (fig. 1B). mean mass was not measured in comparison with any control or of 37 adults without tags (789.2 ± 10.1 g, n = 37) was comparison groups (macleod et al. 2008). common greater than that of adults that carried tags (758.9 ± murres (Uria aalge) sometimes deserted after being 16.1 g, n = 14), but this difference was not significant fitted with VHF radio transmitters during incuba- (log transformation: t = 1.68, P = 0.10). tion, but this effect was minimised by reducing the 49 handling time (wanless et al. 1985). Mass among chicks from pairs where adults carried tags was reduced by 35% compared with chicks dIscussIon raised by non-tagged parents. we found that longer tracking durations were correlated with lighter chicks Impact of tags on chick mass, size, on 19 April, just before fledging. Although we were

Downloaded by [24.94.86.228] at 22:26 20 September 2013 and post-fledging survival not able to control for initial chick age between the Breeding success among sooty shearwaters in this two groups, sooty shearwaters are considered to have study was reduced for pairs where one or both mem- relatively synchronous egg laying and thus hatching bers were fitted with tags. Because we used a com- periods (Richdale 1944; warham et al. 1982), and parative approach rather than a designed experiment the cumulative effects of altered chick provisioning for this study, we can not fully partition the summed on chick growth probably reflected real differences effects of capture, handling, attachment, and deploy- in our comparison. Together with the lack of a sig- ment duration. Nevertheless, our measured effects of nificant effect of handling frequency and handling fitting provisioning adults with tags on chick growth, duration during capture, we suggest that the impacts survival, and potential post-fledging recruitment in- on chicks measured here were most likely to have dicate that adult foraging or provisioning of chicks, been caused by the increased costs of provisioning or both were compromised. imposed on parents carrying tags rather than by the Two sooty shearwaters deserted their chick within disturbance caused by capture and handling during the first week of this study. It is unclear if desertion attachments. Similarly proportioned radio transmit- resulted from handling, tag attachment, or a com- ters (1.6% adult body mass) attached to tufted puf- bination of stressors. Other species of shearwaters fins (Fraturcula cirrhata) reduced mean growth rates Adams et al.—Archival tag effects 361

among puffin chicks by 50%, and decreased fledging et al. 2005), we found no evidence that this oc- success from 84% among non-manipulated pairs to curred to an effective degree among tagged sooty 33% among tagged pairs (whidden et al. 2007). shearwaters in our study. Effects on chick mass and Based on recaptures of sooty shearwaters band- skeletal development were especially pronounced ed as chicks during the pre-fledging period Söhle among chicks reared by parents who were both (2003) found that 69% of sooty shearwater chicks tagged. Although telemetry of mated pairs provides emerged between 24 April and 4 may 2000. Sagar unique sex-specific information regarding allocation & horning (1998) found no difference in mass be- of parental roles when chick-rearing, we suggest tween chicks captured during 15-23 April versus 29 researchers should avoid attaching devices to both April-7 may and determined that there was a 564 g members of a pair. threshold on 19 April, chicks less than 564 g failed with respect to handling effects, there may also to survive after fledging to recruit into the breed- be differences between species and between sexes. ing population. Although not formally analysed, female wandering albatrosses (Diomedea exulans) they suggested that the likelihood of recruitment took longer to recover after being handled (Weim- increased with mass above this threshold. Although erskirch et al. 2002) and male thick-billed murres 95% of chicks reared by tagged adults in our study (Uria lomvia) increased their foraging trip durations survived to 19 April, a lesser proportion reached the (Paredes et al. 2005). Radio transmitters attached to predetermined threshold pre-fledging mass: zero Cassin's auklets (Ptychoramphus aleuticus) caused chicks reared by tagged pairs and 54% of chicks reduced mass gain among chicks and lowered fledg- reared where one parent was tagged. in contrast, ing rates, with the effect more pronounced when 83% of chicks reared by undisturbed, non-tagged males rather than females carried Vhf radio trans- parents reached the threshold pre-fledging mass. mitters (Ackerman et al. 2004). Studies of sex dif- Although igual et al. (1995) found little evidence of ferences in foraging behaviours that target males and statistically significant effects of GLS tags on Cory's females from different pairs may be less detrimental shearwater in terms of overall breeding success, it is to breeding success and still provide useful sex- not clear from their methods how they determined specific information (Ackerman et al. 2004; Adams this parameter. Although they detected a significant et al. 2004; Paredes et al. 2005). effect on returning adult masses, data for chick masses are not presented. we consider the difference Adults returning from trans-equatorial in the percentages of chicks achieving the survival migration threshold mass reared by tagged parents to indicate Although no shearwaters initially fitted with tags a biologically significant effect. were recaptured missing tags, explanations for the The difference in pre-fledging chick mass and nine failed recoveries in November 2005 include the proportion of birds achieving the mass threshold possibility that returning tagged adults avoided de- between nests where both parents were tagged com- tection because they either failed to return or delayed pared with those where only one parent was tagged, their timing of return (at least one bird remained in and the negative relationship between tracking dura- the Northern hemisphere until November 2005,

Downloaded by [24.94.86.228] at 22:26 20 September 2013 tion and chick mass, indicates that shorter-duration and did not return to the colony until late december tag deployments might impact provisioning adults 2005; the tag was subsequently recovered during and their chicks less. Einoder & Goldsworthy (2005) November 2006, the following breeding season; made a similar suggestion regarding satellite-tracked SAS unpubl. data), or they did not survive the post- short-tailed shearwaters (Puffinus tenuirostris), but breeding migration (Shaffer et al. 2006). Return effects were not measured. Shorter-term (c. 33 days) ratios of both sexes were similar to departure ratios, attachments of iSTs presumably had no effect on indicating no greater effect on apparent survival chick condition (Söhle 2003), however mass data according to sex. mean body mass of returning are not reported by those authors. tagged adults in October 2005 was 4% lighter, but Other studies of seabirds with bi-parental care not significantly different, from untagged adults have found that non-manipulated partners compen- from within the same colony. in a separate study, sated for their mate's reduced colony attendance and body mass of sooty shearwater adults carrying iSTs chick provisioning (Wanless et al. 1988; Paredes during the pre-breeding period was 8.3% lighter, but et al. 2005). Although compensatory provisioning no mass reduction was detected in a second group might act to maintain breeding success in some fitted during the mid-breeding (i.e., chick rearing) instances when one parent is compromised (Paredes period (Söhle 2003). 362 New Zealand Journal of Zoology, 2009, Vol. 36

In a similar GLS tagging study of Cory's shearwa- the most proficient divers within the Order Procel- ter (Calonectris diomedea), adults that carried tags lariiformes; they possess unique skeletal adaptations, and were recaptured after returning from migration such as flattening of the tarsus and the narrowing were 1.6% lighter (statistically significant) than non- of the pelvis that are probably advantageous when tagged adults (igual et al. 2005). The results from diving and manoeuvring under water (Kuroda 1954). our study suggest that during migration, adults had it is possible that the position of attachment of leg- recovered from handling and may have overcome mounted devices can reduce swimming speeds and or adjusted for detrimental costs incurred while manoeuvrability and therefore hinder a bird's ability carrying tags during the previous breeding season to catch prey during pursuit diving. and throughout their post-breeding, trans-equatorial healy et al. (2004) determined that placement migration. in the breeding season following experi- of packages on little penguins (Eudyptula minor) mental trials, Söhle (2003) found no difference in nearest to the birds' centre of mass was critical for number of returns of adult sooty shearwaters fitted maintaining balance and minimising behavioural with iSTs in the previous season compared with a effects on diving. Maintaining balance for flying group that did not carry iSTs. Navarro & González- seabirds must also be important, as species such as Solís (2006) found that the proportion of Cory's sooty shearwater and other Procellariiformes must shearwaters returning and breeding the following constantly manoeuvre adroitly to extract energy year, after being fitted with GLS tags and subjected from surface winds during long-distance foraging to increased flight costs (trimmed primary feathers) trips and migration (Pennycuick 2002). We do not during the breeding season, did not differ from a know to what degree the addition of 11 g packages control group. igual et al. (2005) found that return placed at the maximum posterior (and lateral) loca- rates were consistently 6-20% greater among non- tion affected flight or diving performance of the GLS-tagged adult Cory's shearwater groups than birds in our study. Potential effects of this placement tagged adults, but this difference was not statistically on balance could be minimised by an alternative significant. attachment closer to the shearwaters' centre of mass (i.e., ventral breast feathers; Tremblay et al. 2003). transmitter mass and placement Such attachment, however, would not be feasible for for procellariiform seabirds, Phillips et al. (2003) long-term (c. 6 month) tracking of the shearwaters' suggested a maximum mass for attached devices post-breeding trans-equatorial migrations because equivalent to 3% of an individual's body mass. Some body feathers are moulted at this time, but may be studies have found that even light-weight devices of more desirable for shorter-term deployments during 0.7-3.0% of body mass (Wanless et al. 1988; Acker- the breeding season. man et al. 2004; Paredes et al. 2005) have caused Our results signal a need for caution when reduction of body mass, offspring attendance, pro- interpreting results from individual seabirds carrying visioning rates, and frequency of foraging trips. in tracking or sensing devices. Some behaviours mea- our study, the combined mass of the tag and attach- sured among equipped birds may not be representa- ment was <2% of adult body mass, yet we measured tive of the species or population investigated. we do

Downloaded by [24.94.86.228] at 22:26 20 September 2013 significant effects on chick-growth and breeding not know the degree to which tags affected data on success. Attachment of devices on sooty shearwaters locations, movements, and behaviours of the sooty probably affects both flight and underwater foraging. shearwaters in our study (Shaffer et al. 2006,2009) Sooty shearwaters' foraging trips are lengthy (thou- although frequency of provisioning or meal masses sands of kilometres and 1-2 weeks in duration) and delivered to chicks likely were reduced. Further- diving to >15 m depth is common (Weimerskirch & more, it is unclear whether tagged birds attempted Sagar 1996; Shaffer et al. 2009). to compensate for added costs of carrying tags by Wilson & Culik (1992) concluded that the attach- diving more or less frequently; diving depths and ment of devices may adversely affect the foraging durations may also have been affected. further data behaviour and reproductive energetics for penguins, are required to determine the full range of effects and considered that the streamlining of a device is and the costs associated with carrying leg-mounted more crucial than its mass). carrying an external de- tags. effects may vary according to species, size and vice will increase the drag on a bird underwater, and morphology, and foraging mode. streamlining and position of transmitters should be Telemetry and bird-borne sensing of the marine of utmost importance (Bannasch et al. 1994; healy environment have contributed important information et al. 2004). Sooty shearwaters are probably among to studies of marine ecology and seabird biology Adams et al.—Archival tag effects 363

(Burger & Shaffer 2008; Phillips et al. 2008). Geolo- (TOPP) program and the Kia Mau Te Tïfï Mo Ake Tönu cation tagging and telemetry of seabirds in general is Atu Research project (the latter funded mainly by the New apowerful tool to inform conservation of threatened Zealand Foundation for Research, Science & Technology). SAS received financial assistance from the Gordon and seabirds (Birdlife International 2004; Phillips et al. Betty Moore Foundation, the David and Lucile Packard 2006) and to increase our understanding of climate Foundation, the Sloan Foundation, and the National change and other global threats to marine ecosys- Science Foundation. All protocols were approved by the tems. For large-scale (both temporal and spatial) UCSC Institutional Animal Care and Use Committee and studies, leg-mounted tags offer an alternative to sat- the Animal Ethics Committee of the University of Otago. ellite tracking with an attachment technique that has Thanks to A. Lai, C. MacLeod, K.Wilson, and M. Rayner proven the ability to record year-round movements for valuable comments that improved this manuscript. The use of trade, product, or firm names in this publication patterns and migration routes of several species is for descriptive purposes only and does not imply (Weimerskirch & Wilson 2000; Croxall et al. 2005; endorsement by the United States Government. Shaffer et al. 2006; González-Solís et al. 2007). Although the mass of GLS tags as a proportion of adult body mass is low compared to other tracking techniques, interspecific responses of seabirds to tag- REFERENCES ging manipulations may depend on environmental variability, life history strategies, and individual Ackerman JT, Adams J, Takekawa JY, Carter HR, Whit- energy thresholds during breeding or migration. worth DL, Newman SH, Golightly RT, Orthmeyer Because tagging manipulations may cause birds to DL 2004. Effects of radio transmitters on the reproductive performance of Cassin's auklets. behave abnormally, we suggest that, whenever pos- Wildlife Society Bulletin 32: 1229-1241. sible, future studies should include well controlled comparison groups (e.g., non-tagged individuals, Adams J, Takekawa JY, Carter HR 2004. Foraging distance and control treatments to assess handling effects and homerange of Cassin's auklets nesting at two independent of tagging). colonies in the California Channel Islands. Condor 106: 618-637. However, despite our observations of negative Argos 2008. Argos user's manual. Available on-line: impacts on sooty shearwater breeding success, the http://www.argos-system.org/documents/userarea/ ethics and potential risks of deployment of tags argos_manual_en.pdf (accessed 11 February must be balanced against the risks of not deploying 2009). 64 p. them and thereby missing crucial information on Bannasch R, Wilson RP, Culik B 1994. Hydrodynamic as- their biology that can assist conservation. Much will pects of design and attachment of aback-mounted depend on the objectives of tag deployments, and device in penguins. Journal of Experimental Biol- caution is advised especially when deploying tags ogy 194: 83-96. on threatened species. The design and attachment of the devices can be further modified to minimise BirdLife International 2004. Tracking ocean wanderers. disruption. The placement and streamlining of tags The global distribution of albatrosses and petrels. Results from the global Procellariiform track- in future studies on shearwaters and other seabirds ing workshop, 1-5 September 2003, Gordon's Downloaded by [24.94.86.228] at 22:26 20 September 2013 should be considered to minimise efficiency penal- Bay, South Africa. Cambridge, UK, BirdLife ties incurred by the reduction in hydrodynamic drag International. Available on-line: http://www. or displacement of mass. birdlife.org/action/science/species/seabirds/track- ing_ocean_wanderers.pdf (accessed 10 February 2009). 116p. Burger AE, Shaffer SA 2008. Application of tracking and ACKNOWLEDGMENTS data-logging technology in research and conserva- We thank the tïtï research team and members of the field tion of seabirds. The Auk 125: 253-264. crew, R. Mules, B. Newton, and D. Vercoe for assistance Clucas R, Fletcher DJ, Moller H 2008. Estimates of adult with tag deployments and recovery. C. Baduini (Claremont survival rate for three sooty shearwater (Puffinus Colleges, California, USA) provided molecular sexing griseus) colonies in New Zealand. Emu 108: from blood samples. The Whenua Hou Management 237-250. Committee, RakiuraTïtï Islands Administering Body, the Zoology Department of the University of Otago, Te Whare Coulson JC, Porter JM 1985. Reproductive success of Wânanga O Otägo, Southland Department of Conservation kittiwakes Rissa tridactyla: the roles of clutch issued permits and supported this project. This research size, chick growth rates and parental quality. Ibis was conducted as part of the Tagging of Pacific Pelagics 127: 450-466. 364 New Zealand Journal of Zoology, 2009, Vol. 36

Croxall JP 1982. Energy costs of incubation and moult in Healy M, Chiaradia A, Kirkwood R, Daan P 2004. Ba- petrels and penguins. Journal of Animal Ecology lance: a neglected factor when attaching external 63: 275-282. devices to penguins. Memoirs of National Institute of Polar Research Special Issue 58: 179-182. Croxall JP, Silk JRD, Phillips RA, Afanasyev V, Briggs DR 2005. Global circumnavigations: tracking Hull CL 1997. The effect of carrying devices on breeding year-round ranges of nonbreeding albatrosses. royal penguins. Condor 99: 530-534. Science 307: 249-250. Hunter CM, Moller H, Kitson J 2000. Muttonbirder selec- Dawson DA, Darby S, Hunter FM, Krupa AP, Jones IA, tivity of sooty shearwater (tītī) chicks harvested in Burke T 2001. A critique of avian CHD-based New Zealand New Zealand Journal of Zoology molecular sexing protocols illustrated by a Z- 27: 395-414. chromosome polymorphism detected in auklets. Igual JM, Forero MG, Tavecchia G, Gonzalez-Solis J, Molecular Ecology Notes 1: 201-204. Martinez-Abrain A, Hobson KA, Ruiz X, Oro D 2005. Short-term effects of archival tags on Drent RH, Dann S 1980. The prudent parent: energetic Cory's shearwater (Calonectris diomedea). Marine adjustments in avian breeding. Ardea 68: 225- Biology 146:619-624. 252. Kitson JK, Moller H 2008. Looking after your ground: re- Duriez O, Weimerskirch H, Fritz H 2000. Regulation of source management practice by Rakiura Māori tītī chick provisioning in the thin-billed prion: an in- harvesters. Papers and Proceedings of the Royal terannual comparison and manipulation of parents Society of Tasmania 142: 161-176. Canadian Journal of Zoology 78: 1275-1283. Klomp NI, Schultz MA 2000. Short-tailed shearwaters Einoder LD, Goldsworthy SD 2005. Foraging flights of breeding inAustralia foraging in Antarctic waters. short-tailed shearwaters (Puffinus tenuirostris) Marine Ecology Progress Series 194: 307-310. from Althorpe Island: assessing their use of ner- Kuroda N 1954. On the classification and phylogeny of itic waters. Transactions of the Royal Society of the order Tubinares, particularly the shearwaters South Australia 129: 209-216. (Puffinus). Tokyo, published by the author. Everett WT, Pittman RL 1993. Status and conservation Lyver PO'B, Hamilton S, McKenzie M, Dickson I, Dooher of shearwaters in the North Pacific. Pp. 93-100. T, Broad T, Moller H 1998. A burrowscope for In: Vermeer K, Briggs KT, Morgan KH, Siegel- examining petrel nests in burrows. Department Causey D ed. The status, ecology, and conserva- of Conservation, New Zealand. Conservation tion of marine birds of the North Pacific. Special Advisory Series Notes 209: 21. Publication. Ottawa, Canadian Wildlife Service. 263 p. Lyver PO'B, Davis J, Ngamane L, Anderson A, Clarkin P 2008. Hauraki Māori mātauranga for the conser- Falla RA 1934. The distribution and breeding habitats of vation and harvest of titi, Pterodroma macroptera petrels in northern New Zealand. Auckland Insti- gouldi. Papers and Proceedings of the Royal So- tute and Museum Bulletin Records No. 1. 251 p. ciety of Tasmania 142: 149-159. González-Solis J, Croxall JP, Oro D, Ruiz X 2007. MacLeod CJ, Adams J, Lyver P 2008. At-sea distribution Trans-equatorial migration and mixing in the of satellite-tracked grey-faced petrels, P terodroma wintering areas of a pelagic seabird. Frontiers in macroptera gouldi, captured on the Ruamaahua Downloaded by [24.94.86.228] at 22:26 20 September 2013 Ecology and the Environment 5: 297-301. (Aldermen) Islands, New Zealand. Papers and Proceedings of the Royal Society of Tasmania Graham MH 2003. Confronting multicollinearity in 142: 73-88. ecological multiple regression. Ecology 84: 2809-2825. Miskelly CM, Taylor GA 2004. Establishment of a colony of common diving petrels (Pelacanoidesurinatrix) Griffiths R, Double MC, Orr K, Dawson RJ 1998. A by chick transfers and acoustic attraction. Emu DNA test to sex most birds. Molecular Ecology 104: 205-211. 7: 1071-1075. Moller H, Lyver PO'B in press. Using traditional eco- Hamilton SA 1998. Determining burrow occupancy, fledg- logical knowledge for improved sustainability: ing success and land-based threats to mainland case studies from four customary wildlife harvests and near-shore island sooty shearwater (Puffinus by Maori in New Zealand. In: Hughes C ed. In- griseus) colonies. New Zealand Journal of Zool- digenous people and biodiversity conservation. ogy 25: 443-453. Arlington, Conservation International. Hamilton SA, Moller H, Robertson CJR 1997. A review of Navarro J, González-Solís J 2006. Experimental increase the distribution and abundance of sooty shearwa- of flying costs in a pelagic seabird: effects on ters (Puffinus griseus) in New Zealand. Notornis foraging strategies, nutritional state and chick 44: 15-25. condition. Oecologia 151: 150-160. Adams et al.—Archival tag effects 365

Newman J, Fletcher D, Moller H, Harper G, Bragg C, Scott Sagar PM, Horning DS 1998. Mass-related survival of D, McKechnie S 2009. Estimates of productivity fledging sooty shearwaters Puffinus griseus at the and detection probabilities of breeding attempts in Snares, New Zealand. Ibis 140: 329-339. the sooty shearwater (Puffinus griseus), a burrow- Scott D, Scofield P, Hunter C, Fletcher D 2008. Decline nesting petrel. Wildlife Research 36: 159-168. of sooty shearwaters, Puffinus griseus, on The Paredes R, Jones IL, Bones DJ 2005. Reduced parental Snares, New Zealand. Papers and Proceedings of care, compensatory behaviour and reproductive the Royal Society of Tasmania 142: 185-196. costs of thick-billed murres equipped with data Scott D, Moller H, Fletcher D, Newman J, Aryal J, Bragg loggers. Animal Behaviour 69: 197-208. C, Charleton K 2009. Predictive habitat modelling Pennycuick C 2002. Gust soaring as abasis for the flight of to estimate petrel breeding colony sizes: sooty petrels and albatrosses (Procellariiformes). Avian shearwaters (Puffinus griseus) and mottled petrels Science 2: 1-12. (Pterodroma inexpectata) on Whenua Hou Island. New Zealand Journal of Zoology 36: 291-306, Perrins CM, Harris MP, Britton CK 1973. Survival of this issue. manx shearwaters Puffinus puffinus. Ibis 115: 535-548. Shaffer SA, Tremblay Y, Awkerman JA, Henry RW, Teo SLH, Anderson DJ, Croll, DA, Block BA, Costa Phillips RA, Xavier JC, Croxall JP 2003. Effects of satel- DP 2005. Comparison of light-and SST-based lite transmitters on albatrosses and petrels. Auk geolocation with satellite telemetry in free ranging 120: 1082-1090. albatrosses. Marine Biology 147: 833-843. Phillips RA, Silk JRD, Croxall JP, Afanasyev V, Briggs Shaffer SA, Tremblay Y, Weimerskirch H, Scott D, DR 2004. Accuracy of geolocation estimates for Thompson DR, Sagar PM, Moller H, Taylor GA, flying seabirds. Marine Ecology Progress Series Foley DG, Block BA, Costa DP 2006. Migratory 266: 265-272. shearwaters integrate oceanic resources across Phillips RA, Silk JRD, Croxall JP, Afanasyev V 2006. the Pacific Ocean in an endless summer. Proceed- Year-round distribution of white-chinned petrels ings of the National Academy of Sciences 103: from South Georgia. Relationships with ocean- 12799-12802. ography and fisheries. Biological Conservation Shaffer SA, Weimerskirch H, Scott D, Pinaud D, Thomp- 129: 336-347. son DR, Sagar PM, Moller H, Taylor GA, Foley Phillips RA, Croxall JP, Silk JRD, Briggs DR 2008. Forag- DG, Tremblay Y, Costa DP 2009. Spatio-temporal ing ecology of albatrosses and petrels from South habitat use of breeding sooty shearwaters (Puffinus Georgia: two decades of insights from tracking griseus). Marine Ecology Progress Series doi: technologies. Aquatic Conservation: Marine and 10.3354/meps07932. Freshwater Ecosystems 17: S6-S21. Söhle IS 2003. Effects of satellite telemetry on sooty Reyes-Arriagada R, Campos-Elrwanger P, Schlatter R, shearwater, Puffinus griseus, adults and chicks. Baduini C 2007. Sooty shearwater (Puffinus gri- Emu 103: 373-379. seus) on Guafo Island: the largest seabird colony Spear LB, Ainley DG 1999. Migration routes of sooty in the world? Biodiversity and Conservation 16: shearwaters in the Pacific Ocean. Condor 101: 913-930. Downloaded by [24.94.86.228] at 22:26 20 September 2013 205-218. Richdale LE 1944. The sooty shearwater, Puffinus griseus. SYSTAT 2007. Systat 12 for windows. Richmond, Cali- Condor 46: 93-107. fornia, USA. Richdale LE 1945. The nestling of the sooty shearwater. Tremblay Y, Cherel Y, Oremus M, Tveraa T, Chastel Condor 47: 45-62. O 2003. Unconventional ventral attachment of Richdale LE 1954. Duration of parental attentiveness in time-depth recorders as a new method for in- the sooty shearwater. Ibis 96: 586-600. vestigating time budget and diving behaviour of seabirds. The Journal of Experimental Biology Richdale LE 1963. Biology ofthe sooty shearwater Puffi- 206: 1929-1940. nus griseus. Proceedings of the Zoological Society of London 141: 1-117. Wanless SM, Harris P, Morris JA 1985. Radio-monitoring Saether BE, Reidar A, Pedersen HC 1993. Regulation of as a method for estimating time budgets of Guil- parental effort in a long-lived seabird an experi- lemots Uria aalge. Bird Study 32: 170-175. mental manipulation of the cost of reproduction Wanless S, Harris MP, Morris JA 1988. The effect of radio in the antarctic petrel, Thalassoica Antarctica. transmitters on the behaviour of common murres Behavioural Ecology and Sociobiology 33: and razor bills during chick rearing. Condor 90: 147-150. 816-823. '366 New Zealand Journal of Zoology, 2009, Vol. 36

Wanless S, Harris MP, Morris JA 1989. Behavior of alcids Weimerskirch H, Shaffer SA, Mabille G, Martin J, Bou- with tail-mounted radio transmitters. Colonial tard O, Rouanet JL 2002. Heart rate and energy Waterbirds 12: 158-163. expenditure of incubating wandering albatross: basal levels, natural variation, and the effects of Warham J 1990. The petrels: their ecology and breeding human disturbance. The Journal of Experimental systems. Chapter 8: The shearwater genus Puffi- Biology 205: 475-483. nus. P. 166. London, Academic Press Limited. 440 p. Whidden SE, Williams CT, Breton AE, Buck CL 2007. Effects of transmitters on the reproductive success Warham J, Wilson GJ 1982. The size of the sooty shearwa- of tufted puffins. Journal of Field Ornithology ter population at the Snares Islands, New Zealand. 78(2): 206-212. Notornis 29: 23-30. Wilson RP, Grémillet D, Syder J, Kierspel MAM, Garthe Warham J, Wilson GJ, Keeley BR 1982. The annual S, Weimerskirch H, Schäfer-Neth C, Scolaro JA, cycle of the sooty shearwater (Puffinus griseus) Bost CA, Plötz J, Nel D 2002. Remote-sensing at the Snares Islands, New Zealand. Notornis systems and seabirds: their use, abuse and potential 29: 269-292. for measuring marine environmental variables. Marine Ecology Progress Series 228: 241-261. Weimerskirch H, Sagar P 1996. Diving depths of sooty shearwaters Puffinus griseus. Ibis 138: Wilson RP, Culik BM 1992. Packages of penguins and 786-788. device-induced data. Pp. 573-580. In: Priede GI, Swift SM ed. Wildlife telemetry: remote moni- Weimerskirch H, Wilson RP 2000. Oceanic respite for toring and tracking of animals. Chichester, Ellis wandering albatrosses. Nature 406: 955-956. Horward. 708 p. Downloaded by [24.94.86.228] at 22:26 20 September 2013