Project Description
NMML has conducted pup branding and observational studies of California sea lions for 20 years and has developed a very important long-term database to document the demography of this population through rapid growth from the 1970s through the 1990s. Recently, the growth has slowed and the population may be reaching carrying capacity. The study is at a critical juncture because for the first time in the last 3 decades (with the exception of El Nino events) pup production has stabilized. With a continuation of this long term data collection program we have the opportunity to understand the factors that limit population growth in otariids which has not been previously documented. At San Miguel Island we measure pup production, early pup mortality, and survival and natality of branded sea lions. We have generated estimates of age specific survival and natality for animals up to 19 years of age. We have documented lower survival and reproductive senescence in the older ages but still do not have sufficient data to define the right hand side of the survival and natality curves. Thus, it is important to continue the brand re-sight effort to describe lifetime patterns in survival and reproduction, which are needed for a species-specific population growth model.
Issues & Justification
In 2002 we observed the first indication of density dependent effects on pup survival, with mortality in excess of 50% through age 4 months and this trend continued in 2003 and 2004. Pup mortality studies indicated that a substantial amount of mortality was associated with hookworm infections. It appears that pup mortality induced by a density-dependent parasitic disease may be the determining factor in population growth but it is also important to identify other factors that may be contributing to the slowed population growth, such as emigration. Branded animals from San Miguel Island have been sighted opportunistically during the breeding season at San Nicolas, San Clemente, and Santa Barbara Islands. Emigration is an important factor in population dynamics and in 2005 we began a re-sight program at San Nicolas Island to investigate the possibility that emigration from San Miguel Island is occurring in response to competition for breeding space. The focused re-sight effort at San Nicolas Island resulted in the observation of some branded reproductive females at San Nicolas Island that had never been seen at San Miguel Island, indicating that emigration is occurring but is currently at a low level. Continued re-sight effort at San Nicolas Island is important to monitor the trend in emigration rates as the San Miguel Island population fluctuates around carrying capacity. Re-sight effort at San Nicolas will also provide the opportunity to identify adult male California sea lions branded in Washington and Oregon that breed at San Nicolas Island.
Goals
- Conduct live and dead pup surveys to estimate pup production
- Re-sight marked animals to estimate age-specific survival and natality rates (density dependent changes in first year survival and survival and natality of older females) and to document emigration.
- Evaluate causes of first year mortality.
- Brand a sample from 2008 cohort of pups for study continuation.
Methods
Census:
Live pups will be counted at the end of July. Pup mortality surveys will be conducted at 15 day intervals beginning at the end of June. The data will be combined and modeled to produce a pup production estimate.
Marking:
A sample of 300 pups will be branded and 200 pups will be tagged in October 2008. The sample sex ratio is skewed to provide a larger sample size for females (with approximately 300 females and 200 males) to increase the precision of female survival estimates and to provide a larger sample size for natality estimation. Before branding, each pup is weighed to nearest 0.1 kg, sexed, its length and girth is measured and it is tagged in the fore-flippers with yellow roto-tags. The tags facilitate location of branded animals in large groups and provide a return address for animals found dead on beaches or in nets. The potential effect of branding on the health of pups is evaluated each January by taking morphometric measurements on branded and unbranded pups.
Re-sighting:
On San Miguel Island, re-sighting surveys will be conducted intensively from May through August to observe all age and sex classes during the breeding season. At San Nicolas Island, monthly surveys lasting one week will be conducted between May and August. Surveys at San Clemente and Santa Barbara Islands will be conducted during 2 weeks in July to evaluate the level of permanent emigration from San Miguel Island to the other breeding colonies in the Channel Islands. In addition, major concentrations of juveniles and males at Ano Nuevo Island will be monitored between July and August. At San Miguel Island and San Nicolas Islands, one week surveys will be conducted in the fall and winter to observe females with pups and yearlings. During the breeding season, a minimum of 8 hours/day of observation effort will be dedicated to locating marked animals at at the breeding colonies and recording their reproductive status. This same level of sampling was conducted in past years has resulted in high re-sighting rates that has enabled precise estimation of survival rates (CV between 1-10%) and a relatively complex model with important covariates (age, year, weight) to explain variability in survival and re-sighting probability. The primary determinants of the information content of the data are the number of years of re-sight effort and the probability of re-sighting.
Survival estimation:
The re-sighting data has been analyzed with the computer software MARK using the Cormack-Jolly-Seber model to estimate survival and based on covariates such as initial weight, age and El Nino. Further analysis will be conducted as additional data are collected that will enable the full age-specific survivorship curve to be described as the cohorts age through their entire life span.
Natality Estimation:
Estimation of natality is complicated by the uncertainty in determining whether a female has a pup. To date, analyses have been conducted using POPAN with methods described by Schwarz and Stobo (1997) that use within year observations of females with and without their pup to estimate the probability that a female has a pup. A method described recently by Kendall et al (2004) based on state-misclassification will also be investigated as well as a technique similar to that proposed by Payne et al (1990).