Petersen Estimate Research Articles

Abundance, Length Composition, and Migration of Spawning Inconnu in the Selawik River, Alaska

Abstract Inconnus Stenodus leucichthys in the Selawik River, Alaska, were sampled during spawning migrations in the years 1993–1996. Modified Petersen estimates of spawning inconnu were 5,190 (95% confidence interval = 3,690–7,272) for 1995 and 5,157 (3,038–12,983) for 1996. Fork lengths of migrating fish ranged from 52 to 120 cm. Minor differences in length-frequency distributions were observed among the 3 years sampled. Migrating inconnu reached holding areas 25 km below the spawning area in the first week of July and resided there for up to 1 month. Further upstream movement in late August was associated with precipitation. A short reach of river, approximately 12 km long, was identified as the spawning area by use of radiotelemetry. Postspawning migrations occurred as early as September 27 and were completed by October 19. Dispersal during the winter, as measured by tag returns, showed mixing with Kobuk River inconnu within the Selawik Lake–Hotham Inlet complex. Lack of observed interdrainage migratio...

Australian fur seals at Seal Rocks, Victoria: pup abundance by mark - recapture estimation shows continued increase

The abundance of Australian fur seal pups was determined at Seal Rocks, Westernport, Victoria in late December 1997 using a mark–recapture procedure with repeated recapture sessions. Pups (n = 1291) were marked by clipping the black guard hair on the head to reveal lighter underfur. Recaptures from the whole colony were made on eight occasions 1–3 days later. In the recapture sessions, a mean of 32% of sighted pups had been marked. Estimates of pup numbers over the eight recapture sessions were calculated using the Petersen estimate and then combined by taking their arithmetic mean. The combined estimate was 4024 (95% confidence range 3908–4141). In 1991–92, a similar procedure led to an estimate of pup numbers of 2817 (95% confidence range 2703–2930). For both breeding seasons, estimates of pup numbers from each recapture session were also combined assuming a joint hypergeometric distribution; there was little difference in the results from the two procedures, although the confidence intervals for the hypergeometric mean were smaller than those for the arithmetic mean. From 1991–92 to 1997–98, pup numbers increased by 43%, at an exponential rate of 0.059 (95% confidence range 0.0526–0.0664), equivalent to 6.1% per annum (5.4–6.9%). This is greater than the rate of increase of pups at the colony between 1968–69 and 1991–92, which was 0.023 (95% confidence range 0.0198–0.0268), equivalent to 2.4% per annum (2.0–2.7%). Because of the rapid rate of increase of the Australian fur seal colony at Seal Rocks and the importance of the locality for tourism, we recommend that the abundance of pups there be determined every 3–5 years using a mark–recapture technique.

The stratified Petersen estimator with a known number of unread tags.

The Petersen estimator estimator of abundance can be biased when the assumption of homogeneous capture probability or homogeneous recapture probability is violated. Often this heterogeneity is related to the time or place of capture or recapture, and if these can be stratified, the stratified Petersen estimator reduces the bias caused by this heterogeneity. In some experiments, not all the recovered tagged animals can be examined, and only a subsample has its stratum of release and recovery determined. We develop methods for this modified experiment and apply them to estimate the number of salmon returning to spawn in a river in British Columbia, Canada.

Mark–Recapture Estimation of Escapement of Masu Salmon Oncorhynchus masou with a Comparison to a Fence Count

Mark–recapture estimates of the spawning escapement of masu salmon Oncorhynchus masou (also known as cherry salmon) were compared with a complete fence count in the Shokanbetsu River, northern Japan, in 1997. A trap was installed in a fishway at a dam 4.5 km upstream from the river mouth, and 255 adult masu salmon were marked with an opercular punch. Subsequently, spawners and carcasses were sampled on the spawning grounds; 27 marked fish were recaptured. Maximum likelihood (ML) Darroch and pooled Petersen estimates (95% confidence intervals in parentheses) were 492 fish (362–622) and 511 fish (387–635), respectively, and the fence count was 460 fish. The ML Darroch estimate, although a little less precise than the pooled Petersen estimate, was closer to the complete fence count, and it permitted us to estimate each release stratum size. Advantages of the stratified estimator were less bias and the ability to estimate individual strata sizes.

Evaluation of a Mark-Resighting Technique for Woodland Caribou in Newfoundland

We evaluated properties of the Petersen mark-resight technique for estimating the population size of 6 woodland caribou (Rangifer tarandus) herds in Newfoundland. Our objective was to determine the robustness and efficiency of a novel marking technique for estimating population size of caribou where sightability bias is marginal. We marked caribou with pressurized oil-alkyde paint applied from a helicopter. Resighting surveys were conducted 2-3 weeks later. Data from 15 radiocollared animals indicated populations were closed and marks were not lost. The probability of resighting a marked individual was independent of group size, which indicated heterogeneity in sighting probabilities had a marginal effect on the reliability of population estimates. Thus, most of the fundamental assumptions of the Petersen estimate were supported. Although we could not assess accuracy of this method, the degree of precision was dependent primarily on the number of marked animals resighted. Increasing the initial number of marked individuals did not significantly reduce the width of confidence intervals unless the total number of animals resighted also was increased. However, due to the efficiency of marking many animals in a short period, more effort can be allocated to resighting surveys without increasing costs. We believe this technique should provide a cost-effective method for obtaining precise population estimates of woodland caribou and other large mammals where sightability bias is low.

Alternative Approaches to Aerial Censusing of Elk

When animals occur in groups, aerial counts usually miss more of the smaller groups. Previous studies produced corrections for this problem that were based on estimates of the probability of sighting various group sizes. In the present study, extensive ground relocations of groups containing radiotagged elk (Cerous elaphus) demonstrated that group size is very fluid in this species during winter. We thus explored estimation of total numbers by simulations using the Petersen method and by using field data collected on the Madison-Firehole Elk Herd of Yellowstone National Park (YNP). Comparisons with estimates based on a logistic regression of estimated sighting probabilities indicated the Petersen estimate may provide a suitable alternative method. The new approach avoids the need to repeatedly use the same correction factors.

Efficacy of brodifacoum to control house mice, Mus domesticus, in wheat crops in Southern Australia

A field trial to test the efficacy of brodifacoum pellets against house mice ( Mus domesticus) inhabiting wheat crops in the Victorian Wimmera was conducted in spring 1996. The density of mice in wheat crops was approximately 50 mice per hectare. There were two treatments (1 kg ha −1 and 2 kg ha −1) applied aerially using pellets (10 mm in length) with 0.005% brodifacoum. We examined three population indices, minimum number known to be alive, adjusted trap success and population density using the Petersen estimate. In two out of three plots, 1 kg ha −1 was equally as effective as 2 kg ha −1, but in a third it failed for no apparent reason. There were significant reductions in these population indices by up to 99% 7 days after treatment. An application rate of 1 kg ha −1 of brodifacoum is sufficient for effective control of mice in wheat crops which are at maximum tillering and when densities of mice are < 80 ha −1. There was substantial bait unconsumed after 3 days which suggests that a reduced application rate could have sufficed. Nineteen mice were fitted with radio-collars prior to baiting to monitor their fate and location of death at one site treated at 2 kg ha −1. Three disappeared without trace. Of the remainder, 25% ( n = 4) were located dead on the surface or half submerged in burrows and 75% ( n = 12) died underground. All mice died at least 3 days after baiting. Mice were observed on the surface during the day indicating that ingestion of brodifacoum may have changed their behaviour, especially their activity patterns. This may increase the risk of predation and hence the risk of secondary poisoning. No bodies of non-target species were found in the treated crops up to 9 days after application of brodifacoum.

不完全報告を含む標識実験における推定問題について

The estimation of population size by a tagging experiment including incomplete reports of the number of recaptured tagged fishes is considered. It is assumed that the sample is classified into two classes. In class 1 the reporting rate of recaptured tagged fish is 1, and in class 2 the rate may be less than 1. In this case the Petersen estimator and the Chapman type estimator based on the observation in class 1 (CTE) have been proposed, although large sample in class 1 is not expected. In this paper, we examine the effects of the observation in class 2 as additional information for the estimation of population size. Some estimators which include the observations in both classes, such as the modified maximum likelihood estimator (MMLE), are evaluated in senses of the bias and the mean squared error. As a result, if the sample size in class 1 is small and true reporting rate is high, then the MMLE is preferable to the CTE. In the other case, however, the CTE is superior to the estimators based on the observations in both classes.

Visibility Bias on Counts of Nesting Canada Geese

The apparent high visibility of goose nests in arctic habitats often has led researchers to assume that single ground counts approach complete inventories of the nesting geese in a given area. No studies to date have attempted to measure the visibility bias in these counts. Annual nest counts of Eastern Prairie Population (EPP) Canada geese (Branta canadensis interior) were conducted from 1976 to 1994 near Cape Churchill in northern Manitoba. Portions of the study area were searched by 2 independent crews in 1993-94, and a modification of the Petersen estimator was used to estimate both visibility bias and the unbiased nest density. Individual crews located 77% (±2.9%) of active and 39% (±5.3%) of destroyed nests, or only 72% (±2.4%) of all nests present. Single counts seem to be reliable indicators of trends in breeding parameters in our study, but we recommend visibility bias be estimated where accurate nest density estimates are required.

Adjusting for missed tags in salmon escapement surveys

One of the assumptions in mark-recapture studies is that all tags recovered are correctly identified. In some cases, tags are overlooked, and a subsample of fish previously classified as untagged is examined to look for missed tags. In this paper, a method for estimating the salmon escapement using a mark-recapture methodology after correcting for missed tags is developed. The variance of the revised estimator is derived and the loss of efficiency compared with a Petersen estimate, where no tags are overlooked, is found. Optimal allocation of effort between the initial and second examination is considered. Finally, a numerical example using actual data from the 1994 Fraser River sockeye salmon (Oncorhynchus nerka) season is used to illustrate the results.

Marsh deer Blastocerus dichotomus population estimate in the Paraná River, Brazil

This paper reports the results of an aerial survey of a marsh deer Blastocerus dichotomus population occupying an ara to be flooded by the construction of a hydroelectric power plant (Pôrto Primavera) in the Paraná River, Brazil. Forty transects totalling 143 km were flown during the dry and flood seasons. Population size was estimated using the modified Petersen estimator, the line transect method, and two variations of the strip transect method. The results revealed mean densities of 0·50–0·54 deer/km 2 and a population of approximately 950 individuals, the second largest population reported for this vulnerable species in Brazil. Individuals were most commonly alone, but group sizes of up to five deer were observed. Newborn fawns were observed from September to November. The effectiveness of a rescue operation is discussed relative to the possible benefits for the species' conservation. Deer translocations have a high failure rate. Public pressure to translocate marsh deer is not likely to aid in their conservation but may actually assist to precipitate their decline locally. Instead, acquisition of suitable marsh habitat in which the deer already reside could be much more successful and less costly.

Direct estimates of multi-mesh gillnet selectivity to Oreochromis shiranus chilwae

The selectivity of multi-mesh monofilament nylon gillnets to Oreochromis shiranus chilwae was estimated directly by gillnetting in a small reservoir containing a population marked by size class. Numbers in each size class were estimated using the adjusted Petersen estimate. Pearson Type I curves were fitted to plots of selectivity against the fish length/mesh perimeter ratio, for two modes of capture: wedging and tangling. The generality of the selectivity curves is tested and discussed. The tangling component of the selectivity was unimportant, and for many applications the selectivity curves may be approximated by normal curves obtainable through indirect methods. However, the increase in the amplitude of the curves with mesh size, deserves further study.

Direct estimates of multi‐mesh gillnet selectivity to Oreochromis shiranus chilwae

The selectivity of multi‐mesh monofilament nylon gillnets to Oreochromis shiranus chilwae was estimated directly by gillnetting in a small reservoir containing a population marked by size class. Numbers in each size class were estimated using the adjusted Petersen estimate. Pearson Type I curves were fitted to plots of selectivity against the fish length/mesh perimeter ratio, for two modes of capture: wedging and tangling. The generality of the selectivity curves is tested and discussed. The tangling component of the selectivity was unimportant, and for many applications the selectivity curves may be approximated by normal curves obtainable through indirect methods. However, the increase in the amplitude of the curves with mesh size, deserves further study.

The abundance and distribution of bottlenosed dolphins (Tursiops truncatus) in Doubtful Sound, New Zealand

Photographic identification and standardized zigzag surveys were used to study the distribution and abundance of bottlenose dolphins in Doubtful Sound, New Zealand. Forty individuals were identified from nicks and markings on their dorsal fins. Chapman's and Bailey's modifications to the Lincoln–Petersen estimate and the computer program CAPTURE were used to estimate the number of marked individuals in the population, which ranged from 30 to 40 animals for three survey periods (two summers and a winter). The proportion of identifiable individuals in the population was assessed empirically as 65.5%. These data resulted in a total population estimate of approximately 58 individuals. The majority of the dolphins in the catalogue were found in each survey period, indicating their residency in the fiord. Movements of dolphins around the sound were not predictable, but there were areas in which dolphins were more likely to be found and other areas in which dolphins were never seen. However, there were no obvious seasonal or daily movement patterns.

Seasonal Utilization and Movement of Black Sea Bass on a South Carolina Artificial Reef

Abstract The principal objectives of this study were to estimate the average standing stock and utilization rate of black sea bass Centropristis striata on a heavily fished artificial reef. Secondary objectives were to define age and size compositions of the reef and hard-bottom populations of black sea bass and evaluate movements of the fish on reef and hard-bottom habitats. Fish of legal size (>20 cm long) were tagged and released at Capers Reef off Charleston, South Carolina, and at four hard-bottom areas within 10 km of the artificial reef. The total numbers offish tagged and released were 636 at the reef and 615 on hard bottom. A single-census Petersen estimate of the initial winter standing stock of legal-sized black sea bass on Capers Reef was 3,227 fish. The estimated utilization rate for the 6-month study period was 36%. All fish recaptured at the artificial reef had been released there. Twenty-five fish from the hard-bottom release groups were recaptured, only two of which had moved from the rel...

The use of standard and inverse Leslie experiments to estimate the abundance of the coconut crab ( Birgus latro L.) in Vanuatu

By the removal and addition of coconut crabs ( Birgus latro L.) within small areas of the Republic of Vanuatu we were able to ascertain the relationship between catch per unit effort (crabs caught per bait) and density. These standard and inverse Leslie experiments were done in conjunction with a Petersen mark recapture study at one of the sites. The predicted increase in density from changes in CPUE could be compared with the known value (how many crabs were added) and also with the change in the Petersen estimate. This information enabled a catchability coefficient to be calculated allowing the actual density of coconut crabs in any area to be estimated with only CPUE data. This technique is now being used as a tool in the management of coconut crab stocks.

INTER‐OBSERVER COUNT DISCREPANCIES IN A SHORE‐BASED CENSUS OF GRAY WHALES (ESCHRICHTIUS ROBUSTUS)

AbstractEstimations of gray whale abundance have generally assumed that shore‐based observers record all whales migrating through the viewing area during periods uncompromised by visibility. We tested the repeatability of data collected at the standard gray whale census site at Granite Canyon Marine Laboratory in central California by using pairs of observers maintaining independent sighting records. Proximal shore sites were occupied 6 d (60 h) in January 1986 where one team counted 845 whales in 427 groups while the other team counted 990 whales in 477 groups. A comparison of the records showed that the first team missed 290 whales seen by the second team, and the second team missed 204 whales seen by the first team. The total number of whales in the viewing area was calculated for each team by the Petersen estimate, using mutually sighted whale groups as “recaptures”. On average, observers recorded only 79% of the whales. More whales (68%) were missed when entire groups of whales were not seen rather than when groups were undercounted (32%). Visibility did not appear to affect observed rates of missed whales. Whales migrating at intermediate distances from the shore were less often missed than were those &gt; 6 km or &lt; 1 km offshore. This count discrepancy test confirms that an uncorrected calculation of population size for gray whales based on sighting records from solitary observers will be underestimated.

Using Radio-Telemetry for Mark-Recapture Studies with Edge Effects

(1) Radio-telemetry provides a technique for estimating population abundance of species that move on and off a study plot during a survey period. (2) Monte Carlo simulations indicate that either a bias-corrected version of the 'mean Petersen' or a Petersen estimate based on the means of the observations may be suitable for estimating abundance from radio-telemetry data. (3) The proposed bias correction may make it possible to obtain useful abundance estimates for very small populations, given sufficiently high probabilities of marking and recapture.

Estimation of Small‐Mammal Population Size

Two methods of estimating small—mammal population size from multiple capture—mark—recapture occasions, the Lincoln—Petersen estimator and program CAPTURE, were evaluated and compared using computer simulation. Comparisons were made for population sizes of 50, 75, and 100, trapping periods of 5, 7, and 10 d and numerous patterns of capture probabilities. Program CAPTURE often failed to provide useful estimates of population size due to the poor performance of its model selection procedure. The Lincoln—Petersen estimator provides estimates of population size with low biases and standard errors except when uniform behavioral response occurs or when there is high heterogeneity in individual capture probabilities. We suggest that researchers use CAPTURE as a screening technique to check for evidence of uniform trap responses and/or high individual heterogeneity in capture probabilities. If these influences are not present, the Lincoln—Petersen estimator may be the best method for estimation of small—mammal population size.

Failure of Statistical Tests to Detect Assumption Violations in the Mark–Recapture Population Estimation of Brook Trout in Adirondack Ponds

Misleading mark–recapture population estimates were obtained in three field comparisons of the performance of the Assinica and Temiscamie strains of brook trout Salvelinus fontinalis. Three and possibly four of the six assumptions required for the Jolly–Seber method of population estimation were violated. First, violation of the assumption that marked and unmarked fish have homogeneous probability of being captured was detected by comparisons of the proportion of marked trout present in seining samples with the proportion present when the ponds were drained. Second, violation of the assumption of homogeneous probability of recapture was detected by the Leslie test. Third, violation of the assumptions of homogeneous survival both between marked and unmarked fish and among marked fish occurred and was probably the primary cause of the misleading estimates. Petersen estimates based on recapture of marked fish at pond draining were calculated for comparative purposes and also failed to accurately gauge the size of these brook trout populations. Inaccurate population estimates, deviating by as much as 41% from the actual numbers, might have been accepted if the study had relied solely on data from sampling for the statistical tests of the assumptions. Assumptions were violated even though our sampling covered 45–75% of the pond areas and samples contained substantial portions (20–30%) of the total population. Careful validation of the assumptions of mark-recapture models is stressed as prerequisite to the acceptance of numerical estimates of wild populations.