Actuarial and reproductive senescence in a long-lived seabird: preliminary

evidence

D.J. Andersona,*, V. Apaniusb

aDepartment of Biology, Wake Forest University, Winston-Salem, NC 27109 7325, USAbDepartment of Biological Sciences, Florida International University, Miami, FL 33199, USA

Abstract

The evolutionary theory of aging predicts that pelagic seabirds, which have low extrinsic mortality, should show exceptional longevity.

These taxa appear to show the lowest rates of actuarial senescence among birds yet display declining reproductive performance at advanced

ages. We have studied survival and reproduction of Nazca boobies (Sula granti) in the remote Galapagos Islands since 1984. We found a

slight but detectable increase in mortality rate in the oldest ($19 yrs) cohort, indicating minimal actuarial senescence. The probability of

successful reproduction (eggs or fledglings) declined from mid-life to the age of the oldest cohort. We are currently investigating the causal

relationship between physical (foraging) performance, components of reproductive success, and longevity at our pristine study site.

q 2003 Elsevier Science Inc. All rights reserved.

Keywords: Reproduction; Actuarial senescence; Seabird; Nazca booby; Sula granti

1. Introduction

Birds and bats live longer for their body size than do

non-flying mammals, probably because flight protects them

from some sources of extrinsic mortality (Calder, 1983;

Austad and Fischer, 1991; Holmes and Austad, 1995).

Pelagic seabirds show a further enhancement in lifespan,

with greatly delayed actuarial senescence compared to other

birds (Ricklefs, 1998). Here we examine data on survival

and reproduction of wild adult Nazca boobies (Sula granti),

a pelagic seabird nesting in the Galapagos Islands, for

evidence of either actuarial or reproductive senescence,

using our long-term demographic data. This population has

been monitored since 1984 and, unlike some other seabird

populations (Nisbet, 2001), appears to show little anthro-

pogenic effect on their biology.

2. Materials and methods

2.1. Actuarial senescence: longitudinal analysis

We have studied Nazca boobies (a recently identified

taxon; Friesen et al., 2002) at Punta Cevallos, Isla Espanola,

Galapagos Islands, Ecuador since 1984, focusing on

behavioral ecology (e.g. Anderson, 1989; Anderson and

Ricklefs, 1995; Clifford and Anderson, 2001). Breeding is

seasonal, with laying from September–February, and most

fledging completed by June (Anderson, 1993). Leg-banding

of adults and young of the year began in 1985, and DJA has

conducted a band-resight survey at the beginning of each

breeding season (except the 1988–1989 season) each year

since the 1985–1986 season. Details of the survey are given

by Huyvaert and Anderson (in press). Briefly, sweeps are

made through the breeding colony at night when the number

of birds present is highest and capture is easiest, and leg

bands of all adults present are recorded. Each resighted bird

receives a temporary mark to ensure sampling ‘without

replacement’. Since not all birds are present on all nights,

each area of the colony is visited on consecutive nights until

the expected number of new birds from a regression model

is less than 1. Until 1992, aluminum leg bands were used,

with an annual band retention rate of 0.982 (Huyvaert and

Anderson, in press). Beginning in the 1992–1993 breeding

season, steel bands were used instead, with an annual

retention rate of 1.00 (Huyvaert and Anderson, in press).

Between 1992 and 1995, all aluminum-banded birds also

received a steel band, so we consider band loss since 1993 to

be negligible. Resighting efficiency is 0.994 or greater

(Huyvaert and Anderson, in press). Young of the year were

not banded in the 1988–1989 through 1991–1992 breeding

seasons.

0531-5565/03/$ - see front matter q 2003 Elsevier Science Inc. All rights reserved.

doi:10.1016/S0531-5565(03)00104-9

Experimental Gerontology 38 (2003) 757–760

www.elsevier.com/locate/expgero

* Corresponding author. Tel.: þ1-336-758-5319; fax: þ1-336-758-6008.

E-mail address: da@wfu.edu (D.J. Anderson).

Almost all birds enter the breeding population by age 7

(Huyvaert, 1999). We conducted a longitudinal analysis of

survival of two known-age cohorts banded as nestlings in

the 1984–1985 and 1985–1986 breeding seasons, respect-

ively, by calculating the proportion of each cohort alive in

each band-resight survey relative to the number that was

alive at age 7 yrs.

2.2. Actuarial senescence: cross-sectional analysis

To conduct a cross-sectional analysis of annual survival,

we calculated the annual survival probability from the

beginning of the 2001–2002 breeding season to the

beginning of the 2002–2003 season. We used the known-

age adults and two other groups banded as adults in 1984–

1985 and 1985–1986 (assumed to be $4 yrs old at banding

(Huyvaert, 1999)) and so aged a minimum of 20 and 21 yrs,

respectively, in 2001–2002. For these two groups (the

‘oldest adults’) we pooled sightings to increase the sample

size and the calculated transition probability is for birds

$20 yrs of age.

2.3. Actuarial senescence: parameter estimation

Following Ricklefs (1998), we calculated the Weibull

parameter ðv ¼ a=ðb2 1Þ; units of 1/time) based on the

equation mx ¼ m0 þ axb that relates age ðxÞ to age-specific

mortality ðmxÞ and the minimum mortality ðm0Þ rate. The

latter can be interpreted as the mortality rate of young adults

before senescence, which largely reflects mortality from

extrinsic factors (Ricklefs, 1998). It was not possible to

estimate m0 directly from the data because the non-linear

estimation procedure did not converge or calculated a

negative value. Therefore, we used the minimum mortality

rate observed in each analysis, 0.0218, 0.0316, and 0.0667

for the 1984–1985, 1985–1986 and cross-sectional groups,

respectively.

The non-linear estimation procedure (PROC NLIN) of

SAS (SAS Institute Inc, Cary, NC; v. 8.2) was used for

fitting a and b parameters.

2.4. Reproductive senescence

Breeding of these same groups of birds was monitored

daily during the entire 2000–2001 breeding season, and

each bird was scored for attempted reproduction (reached

the egg stage, yes/no) and successful fledging (produced an

offspring in juvenile plumage; yes/no). Breeding perform-

ance of birds known to be alive in that year was related to

age using logistic regression (PROC LOGISTIC of SAS) to

compare the fit of sequential models containing the intercept

only, sex, age, and the sex £ age interaction. We conserva-

tively considered age as a categorical factor, especially

because the oldest adult group was not known-age. We

divided the analysis into early life (4 – 8 yrs) when

reproductive success was expected to improve and late

life (13–16, $19) for evidence of senescent decline in

reproductive performance. In this article, we only consider

the analysis of the latter group but show the data for

comparison of reproductive success across the lifespan.

3. Results

3.1. Actuarial senescence

Survival probability showed little variation with age in

both longitudinal and cross-sectional analyses (Fig. 1). The

Weibel shape parameter was 0.137 ða ¼ 1:22 £ 1021 ^

2:80 £ 1021; b ¼ 25:99 £ 1022 ^ 9:29 £ 1021Þ for the

1984–1985 cohort, 0.076 ða ¼ 4:90 £ 1024 ^ 1:26 £ 1023;

b ¼ 1:97 ^ 9:78 £ 1021Þ for the 1985–1986 cohort, and

0.065 ða ¼ 2:71 £ 1023 ^ 8:99 £ 1023; b ¼ 1:17 ^ 1:22Þ

in the cross-sectional analysis. The median of these three

values (0.076) is plotted against the median m0 for the

three groups in Fig. 2, with the nine species that showed

model convergence in Ricklefs (1998). Over 20% of the

birds alive at age 7 were still alive at age 18.

3.2. Reproductive senescence

Reproductive performance improved with age for both

sexes at the beginning of the reproductive lifespan (Fig. 1).

After age 12, the probability of breeding was higher for

females (x2 ¼ 10:61 df ¼ 1; P ¼ 0:0011) than for males.

After controlling for differences between sexes, breeding

probability decreased with age (x2 ¼ 10:69 df ¼ 4;

Fig. 1. Age-specific survival and reproduction of Nazca boobies.

Longitudinal survival histories of two cohorts (solid grey lines) and

cross-sectional transition probabilities between age classes (dotted grey

line) are associated with the left Y-axis. Reproductive performance of males

(dotted lines, 95% confidence intervals) and females (solid lines, 95%

confidence intervals) in the 2000–2001 breeding season is shown as the

proportion of living birds that attempted to reproduce (open circles) and the

proportion of breeding birds that fledged an offspring (closed circles).

Reproductive data are associated with the right axis. Reproductive data for

9–12 yr-olds are missing because young were not banded in the

corresponding years in the past. Males showed lower per capita

reproductive success than females did at most ages because the adult sex

ratio is strongly male-biased (unpub. data).

D.J. Anderson, V. Apanius / Experimental Gerontology 38 (2003) 757–760758

P ¼ 0:030). The interaction between sex and age was not

significant (x2 ¼ 5:06; df ¼ 4; P ¼ 0:28).

After the initial increase in probability of fledging (one

offspring by birds that attempted to breed) early in life,

females that bred were marginally more successful in

fledging offspring than males (x2 ¼ 3:28, df ¼ 1;

P ¼ 0:07). After controlling for the difference between

sexes, fledging probability decreased with age (x2 ¼ 11:11;

df ¼ 4; P ¼ 0:025). Again, the interaction between sex and

age was not significant (x2 ¼ 3:42; df ¼ 4; P ¼ 0:49).

4. Discussion

4.1. Divergent rates of senescence

Nazca boobies appear to show the low rate of actuarial

senescence expected for pelagic seabirds, at least over the

age range available to us in this study and the heterogeneity

in the data which hampered estimation of the Weibel

parameter. Extrapolation of the survival curve suggests that

approximately 5% of birds alive as 7 yr olds will still be

alive at age 30 (Fig. 1). Evidence of declining reproductive

performance before age 20 was thus a surprising finding.

Our results are preliminary in the sense that reproductive

performance has been documented for only approximately

2/3 of potential lifespan, yet provides compelling evidence

that reproductive performance increases early in life and

then wanes in this long-lived seabird species.

4.2. Causes of reproductive senescence

The reasons for curtailed fledging success with age

among breeders in 2000–2001 will be a focus of future

research, but some markers of reproductive senescence can

be analyzed with data in hand. The reproductive organs of

females showed little evidence of declining function with

age (Fig. 1), since they continued to lay eggs to age 19 þ .

Preliminary analysis of clutch size for females showed no

age-related decline, but hatching success and especially

rearing of hatched offspring fell with increasing age (unpub.

data). For males, clutch size, hatching success, and rearing

success all showed age-related declines (unpub. data). Thus,

several components of reproductive performance may be

declining with age in Nazca boobies.

4.3. Reproductive senescence in long-lived seabirds

Studies of other long-lived seabirds show diverse

patterns of reproductive performance at advanced age. In

the short-tailed shearwater (Puffinus tenuirostris), the

probability of reproduction (incubating an egg) declined

monotonically with age for males but not females (re-

analysis of Table 1 in Wooller et al., 1990), while fledgling

production from those eggs showed a step function decrease

with older age in males but not females (Wooller et al.,

1990). Fledgling production from eggs decreased in the

oldest cohort of wandering albatrosses (Diomedea exulans;

Weimerskirch, 1992), but not in kittiwake gulls (Rissa

tridactyla; Thomas and Coulson, 1988). However, hatching

rate appeared to decline in the oldest kittiwakes (Thomas

and Coulson, 1988). In the fulmar (Fulmarus glacialis),

reproductive performance was assessed in old birds prior to

their disappearance (i.e. presumed dead) compared to those

that continued returning to the colony and breeding. Birds

that did not return had decreased probability of laying eggs

and fledging offspring prior to their disappearance (Ollason

and Dunnet, 1988), suggesting that reproductive senescence

portends mortality. In contrast to these studies, egg and

fledgling production by very old common terns (Sterna

hirundo) did not decline with age, even in the oldest cohort

(Nisbet et al., 2002).

The simplest interpretation of these diverse outcomes is

that the selective factors acting on reproductive function are

independent from those acting on longevity (Ricklefs et al.,

2003). This contrasts with the prevailing paradigm (Par-

tridge, 1987) that reproductive effort accelerates senescence

and mortality in a concerted fashion as suggested in the

fulmar study (Ollason and Dunnet, 1988). The alternative

perspective, supported by other avian taxa (Ricklefs, et al.,

2003), is that reproductive senescence and longevity are

effectively decoupled. A possible exception is that the

sustained physical performance required of foraging parents

may be a causal link between reproductive success and

longevity, as suggested in the fulmar study (Ollason and

Dunnet, 1988).

Unlike previous studies, we have little reason to believe

that anthropogenic factors (e.g. environmental contami-

nants, exploitation of marine resources, direct mortality

from long-line fishing (Sagar et al., 2000)), have influenced

the demography of our population. Therefore, some of the

problems that limit inferences about age-dependent repro-

duction and survival (Nisbet, 2001) are minimized in our

system, underscoring the importance of field studies of long-

lived seabirds in protected environments in order to

Fig. 2. Estimates of coefficients of Weibull functions fitted to the

relationship between mortality rate and age for Nazca boobies and for the

nine species in Ricklefs (1998) with fitted model convergence.

D.J. Anderson, V. Apanius / Experimental Gerontology 38 (2003) 757–760 759

understand the evolutionary ecology of organisms with slow

aging.

5. Conclusions

Nazca boobies show the expected shallow decline in

survival with age, indicating a potential lifespan over

30 yrs. The probability of successful reproduction declines

with age more rapidly, suggesting the existence of a post-

reproductive lifespan in this species. Further study of the

oldest age classes will evaluate this possibility more

definitively. Nazca boobies offer an exceptional opportunity

for studying the nexus of physiological performance,

lifetime reproductive success, and aging because the

reproductive histories and age-structures are known from

our long-term studies.

Acknowledgements

Supported by National Geographic and NSF grants

(DEB 9304579, DEB 9629539, DEB 9726444) to D.J.A.

We thank the Galapagos National Park Service for

permission to work in the Park, the Charles Darwin

Research Station, TAME airlines, Galapagos Network,

and Ecoventura for logistical support, our many students,

technicians, and spouses for their contributions to the

fieldwork, and R.E. Ricklefs and A.A. Schuerlein for

comments on a previous draft.

References

Anderson, D.J., 1989. The role of hatching asynchrony in siblicidal brood

reduction of two booby species. Behav. Ecol. Sociobiol. 25, 363–368.

Anderson, D.J., 1993. In: Poole, A., Gill, F. (Eds.), Masked Booby (Sula

dactylatra), The Birds of North America, vol. 73. The Birds of North

America, Inc, Philadelphia, pp. 1–16.

Anderson, D.J., Ricklefs, R.E., 1995. Evidence of kin-selected tolerance by

nestlings in a siblicidal bird. Behav. Ecol. Sociobiol. 37, 163–168.

Austad, S.N., Fischer, K.E., 1991. Mammalian aging, metabolism, and

ecology: evidence from the bats and marsupials. J. Gerontol. 46,

B47–B53.

Calder, W.A. III, 1983. Body size, mortality, and longevity. J. Theor. Biol.

102, 135–144.

Clifford, L.D., Anderson, D.J., 2001. Experimental demonstration of the

insurance value of extra eggs in an obligately siblicidal seabird. Behav.

Ecol. 12, 340–347.

Friesen, V.L., Anderson, D.J., Steeves, T.E., Jones, H., Schreiber, E.A.,

2002. Molecular support for the species status of the Nazca Booby. Auk

119, 820–826.

Holmes, D.J., Austad, S.N., 1995. Birds as animal models for the

comparative biology of aging: a prospectus. J. Gerontol. 50A,

B29–B66.

Huyvaert, K.P., 1999. Post-fledging survival and dispersal in Galapagos

masked boobies (Sula dactylatra). MSc Thesis, Wake, Forest

University, Winston-Salem, NC

Huyvaert, K.P., Anderson, D.J., 2003. Limited dispersal in Nazca boobies.

J. Avian Biol. in press.

Nisbet, I.C.T., 2001. Detecting and measuring senescence in wild birds:

experience with long-lived seabirds. Exp. Gerontol. 36, 833–843.

Nisbet, I.C.T., Apanius, V., Friar, M.S., 2002. Breeding performance of

very old common terns. J. Field Ornithol. 73, 117–124.

Ollason, J.C., Dunnet, G.M., 1988. Variation in breeding success in

fulmars. In: Clutton-Brock, T.H., (Ed.), Reproductive Success,

University of Chicago Press, Chicago, pp. 268–278.

Partridge, L., 1987. Is accelerated senescence a cost of reproduction? Funct.

Ecol. 1, 317–320.

Ricklefs, R.E., 1998. Evolutionary theories of aging: confirmation of a

fundamental prediction, with implications for the genetic basis and

evolution of life span. Am. Nat. 152, 24–44.

Ricklefs, R.E., Schuerlein, A.A., Cohen, A., 2003. Age-related patterns of

fertility in captive populations of birds and mammals. Exp. Gerontol.

this issue.

Sagar, P.M., Molloy, J., Weimerskirch, H., Warham, J., 2000. Temporal

and age-related changes in survival rates of southern Buller’s

albatrosses (Thallasarche bulleri bulleri) at the Snares, New Zealand,

1948–1997. Auk 117, 699–708.

Thomas, C.S., Coulson, J.C., 1988. Reproductive success of kittiwake gulls

Rissa tridactyla. In: Clutton-Brock, T.H., (Ed.), Reproductive Success,

University of Chicago Press, Chicago, pp. 268–278.

Weimerskirch, H., 1992. Reproductive effort in long-lived seabirds: age-

specific patterns of condition, reproduction and survival in the

wandering albatross. Oikos 64, 464–473.

Wooller, R.D., Bradley, J.S., Skira, I.J., Serventy, D.L., 1990. Reproductive

success of short-tailed shearwaters Puffinus tenuirostris in relation to

their age and breeding experience. J. Anim. Ecol. 59, 161–170.

D.J. Anderson, V. Apanius / Experimental Gerontology 38 (2003) 757–760760