A test of statistical techniques for detecting density dependence in sequential censuses of animal populations

Summary Principal and reduced major axes, and Bulmer's (1975) tests have been suggested as methods for detecting the presence of density dependence in a series of population censuses that are unsuitable for analysis by alternative means e.g. by k-factor analysis. These alternative methods are tested using census data, some of which are previously unpublished, from natural populations known from independent evidence to be subject to density dependent processes. All the methods fail to detect density dependence reliably, irrespective of sample size and the dynamics of the population. We conclude that none of the methods tested is sufficiently reliable to be useful as a test of density dependence in sequential censues of animal populations.     

A Numerical Comparison of Some BAYESian Methods of Estimating a Binomial Parameter

This note considers the optimal and suboptimal sequential and fixed sample size estimation of the unknown binomial parameter, p, for a beta prior distribution for p and under quadratic loss and constant observation cost. A numerical comparison of the methods is presented.     

Population Density and Home Range Size of <Emphasis Type="Italic">Indri indri</Emphasis> in a Protected Low Altitude Rain Forest

We report the results of a census of Indri indri conducted in Betampona Nature Reserve, a lowland rain forest in eastern Madagascar. In addition, we conducted a year-long study of the ranging behavior of 3 groups in the southwestern region of the reserve. We used 2 methods to calculate population density and home range size, and to provide minimum and maximum estimates. Population density of Indri indri ranged from 6.9–13.2 individuals/km² in Betampona. Mean home range size is 27 ha. The values for population density and home range size are intermediate between values for Indri indri in selectively logged and undisturbed montane rain forest. Our results suggest a relationship between habitat disturbance, population density and home range size for the species. Recent increases in habitat disturbance appear to cause an increase in population density and a decrease in home range size. The results are consistent with ones for other folivorous primate populations. Further research on habitat requirements of Indri and availability in Betampona is necessary to investigate the possibility of translocating Indri from nearby forest fragments into Betampona.     

Precision and accuracy of visual census technique for fish assemblages on coral patch reefs

Synopsis A visual census technique is described in which the results of three separate enumerations of fish at a site are combined to produce a best estimate of the fish fauna present. Its precision and accuracy are examined, and compared to those of censuses obtained by modifications of the technique. Visual censuses can display high repeatability, but they seldom (if ever) completely sample the fish present at a site. Accuracy varies with technique used. In our tests, the preferred method yielded 82% of species and 75% of individuals known to be present and potentially censurable at the time the observations were made. Visual censuses are of comparable accuracy to ichthyocide collections of unenclosed sites, but the two methods sample different components of the total fish fauna. It is important when using visual censuses to remember that their accuracy is not 100%.     

Are Plant Censuses Carried Out on Small Quadrats More Reliable than on Larger Ones?

Plant censuses are known to be significantly affected by observers’ biases. In this study, we checked whether the magnitude of observer effects (defined as the % of total variance) varied with quadrat size: we expected the census repeatability (% of the total variance that is not due to measurement errors) to be higher for small quadrats than for larger ones. Variations according to quadrat size of the repeatability of species richness, Simpson equitability and reciprocal diversity indices, Ellenberg indicator values, plant cover and plant frequency were assessed using 359 censuses of vascular plants. These were carried out independently by four professional botanists during spring 2002 on the same 18 forest plots, each comprising one 400-m² quadrat, four 4-m² and four 2-m² quadrats. Time expenditure was controlled for. General Linear Models using random effects only were applied to the ecological indices to estimate variance components and magnitude of the following effects (if possible): plot, quadrat, observer, plant species and two-way interactions. High repeatability was obtained for species richness and Ellenberg indicator values. Species richness and Ellenberg indicator values were generally more accurate but also more biased in large quadrats. Simpson reciprocal diversity and equitability indices were poorly repeatable (especially equitability) probably because plant cover estimates varied widely among observers, irrespective of quadrat size. Grouping small quadrats usually increased the repeatability of the variable considered (e.g. species richness, Simpson diversity, plant cover) but the number of plant species found on those pooled 16 m² was much lower than if large plots were sampled. We therefore recommend to use large, single quadrats for forest vegetation monitoring.     

Abundance and Distribution of Sympatric Gibbons in a Threatened Sumatran Rain Forest

Agile gibbons (Hylobates agilis) and siamangs (Symphalangus syndactylus) are sympatric small apes inhabiting threatened forests of Sumatra, Indonesia. We censused both species in the 3,568-km² Bukit Barisan Selatan National Park, at the southern limit of their ranges, over a 7-mo period in 2001. First, we monitored daily calling rates from known populations to develop probabilities of calling during a specified number of days and used the probability of calling at 1 time during 3 days to convert calling rates to abundance. Next, we used 3-day calibrated call count censuses (n=31) stratified by distance from forest edge and across a range of elevations to estimate species-specific group densities. We used group size from the known populations as well as data collected ad libitum during the census to convert group density to individual density. Agile gibbon group density averaged 0.67 km^–2 (SE = 0.082) and group size averaged 2.6 (SE = 0.73) for a population estimate of 4,479 (SE = 1,331) individuals. Siamang group density averaged 2.23 km^–2 (SE = 0.245), and group size averaged 3.9 (SE = 1.09) for a population estimate of 22,390 (SE = 8,138). Agile gibbon and siamang densities are negatively correlated, with agile gibbons more abundant in mid-elevation forests and siamangs most abundant in lowland and submontane forests. The small group sizes of agile gibbons indicate potential survival problems in infant and juvenile size classes. Although neither species is presently threatened by direct human disturbance, continued deforestation will jeopardize the long-term viability of both species in Bukit Barsian Selatan National Park and on Sumatra.     

Conducting animal censuses amongst abrupt topography; a GIS‐based alternative to Distance

Management often bases decisions on estimates of animal density and population size. Aerial sampling is expensive whilst current ground methods, noticeably Distance sampling, assume a single detection function for each habitat and that visibility in a given habitat type declines in a smooth, increasing manner with distance from the observer. If the visibility within a habitat varies widely or is suddenly cut off, as amongst abrupt topography, these assumptions are invalidated. We present an affordable, accessible and accurate method for conducting repeated annual censuses on large herbivores in such terrain. We used a GIS habitat‐based approach. Monthly, over a two year period, we conducted repeated road transects in Ithala Game Reserve, South Africa, using GPS to record the geographic position of each sighting. These records (n = 8742) were then imported into a GIS and overlaid onto an existing habitat‐type GIS layer. With the sampled area thus defined as an irregular polygon in the GIS encompassing all records, we calculated densities of each herbivore species by habitat type and extrapolated total population size estimates. Estimates of maximum population size for wildebeest and zebra correlated (±15%) with management's estimates based on aerial surveys, walked transects and experience. White rhino are individually counted in the reserve and our estimates (year 1: 52; year 2: 57) matched these known numbers (year 1: 50; year 2: 53). Our method also yielded realistic numbers for impala, but unrealistic numbers for kudu and warthog. Our GIS‐based census technique produced realistic maximum population estimates for abundant grazing and mixed‐feeder mesoherbivores and for scarce, but highly visible, megaherbivores, but not for cryptic species. With the increasing availability of GIS data, the technique is recommended to those working in abrupt terrain frustrated by the inability of current ground census techniques, principally Distance sampling, to produce realistic population estimates.     

Volatility in the effective size of a freshwater gastropod population

Despite the utility of gastropod models for the study of evolutionary processes of great generality and importance, their effective population size has rarely been estimated in the field. Here, we report allele frequency variance at three allozyme‐encoding loci monitored over 7 years in a population of the invasive freshwater pulmonate snail Physa acuta (Draparnaud 1805), estimating effective population size with both single‐sample and two‐sample approaches. Estimated N_e declined from effectively infinite in 2009 to approximately 40–50 in 2012 and then rose back to infinity in 2015, corresponding to a striking fluctuation in the apparent census size of the population. Such volatility in N_e may reflect cryptic population subdivision.     

Using a Normal Approximation to Test for the Binomial Distribution

The commonly used method to test for the binomial distribution is the x²-test. In this paper, we introduce an alternative method to test for the binomial distribution. Suppose N is the number of sample groups with n individuals each, x_ij is the jth sample in ith group, a Bernoulli variable with parameter and VVI=s²/[m(1 - m)/n]. Then it is well know that the asymptotic distribution of the statistic (N - 1) VVI is x²(N - 1) under the hypothesis p₁ = p₂ = … = p_N. Here we find that VVI has an asymptotic normal distribution N(1, 2(1 - 1/n)/(N - 1)). Unlike the x²-statistic, the variance of the normal test statistic is a function of n. This method is convenient in detecting spatial patterns and dispersion in the study of diseased organisms (e.g., plants) in field samples.     

Population recovery in the moustached tamarin (Saguinus mystax): Management strategies and mechanisms of recovery

In 1978, 66% of the individuals of Saguinus mystax and 9.5% of Saguinus fuscicollis were cropped from a population at the Yarapa river, Peru. The effects of cropping on the remaining tamarin population were evaluated by conducting censuses in 1981 and 1982 and by trapping and release of Saguinus mystax in 1981. Three hundred kilometers of trail were covered in the censuses, and all the groups within the 1.9-km² study area were located. Within three years after cropping, the population of S. mystax had increased by 124%, more than double the size left in 1978. Increased reproductive rate, early breeding, and reduced infant mortality contributed to the recovery. On the other hand, S. fuscicollis had decreased by 12% in the three years following the cropping but had increased in the fourth year to a level slightly below the precropping density. The cropping of more S. mystax than S. fuscicollis might have contributed to the decline of the latter. The cropping of a sizable percentage of S. mystax from a natural population does not seem to impair its recuperative powers. It may take longer than four years for a population exhibiting high density, such as that at the Yarapa site, to recover completely.     

Five Confidence Intervals of the Closed Population Size in the Capture‐recapture Problem under Inverse Sampling with Replacement

In the capture‐recapture problem for two independent samples, the traditional estimator, calculated as the product of the two sample sizes divided by the number of sampled subjects appearing commonly in both samples, is well known to be a biased estimator of the population size and have no finite variance under direct or binomial sampling. To alleviate these theoretical limitations, the inverse sampling, in which we continue sampling subjects in the second sample until we obtain a desired number of marked subjects who appeared in the first sample, has been proposed elsewhere. In this paper, we consider five interval estimators of the population size, including the most commonly‐used interval estimator using Wald's statistic, the interval estimator using the logarithmic transformation, the interval estimator derived from a quadratic equation developed here, the interval estimator using the χ²‐approximation, and the interval estimator based on the exact negative binomial distribution. To evaluate and compare the finite sample performance of these estimators, we employ Monte Carlo simulation to calculate the coverage probability and the standardized average length of the resulting confidence intervals in a variety of situations. To study the location of these interval estimators, we calculate the non‐coverage probability in the two tails of the confidence intervals. Finally, we briefly discuss the optimal sample size determination for a given precision to minimize the expected total cost. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Studies on the spatial distribution pattern of larvae of the mosquito,Anopheles sinensis,in rice fields

The spatial distribution patterns of the population of Anopheles sinensis larvae were studied in the rice field area in the suburb of Urawa city in Japan, during the summer seasons in 1973 and 1974. The distribution pattern of the larval population within the field, analysed by the m−m regression method, indicated that the basic component of larval distribution was not a group of individuals but a single individual and such components were distributed contagiously over the field. This basic pattern did not change significantly according to developmental stage, census date or field. Therefore, we could describe the distribution pattern of the population in a rice field by the single linear regression, x=0.021+1.339x(r²−0.912). Also, the relation for the whole population in the field area including the five fields could be shown by the linear regression, x=0.049+1.749x(r²−0.959). The value of α remained to be nearly equal to zero, but the value of β became larger than the value for the single-field relation. Such a change in distribution pattern seemed to reflect the greater heterogeneity in conditions among the fields than within individual field. Using the information on the distribution patterns mentioned above, some considerations were given on the sampling plans for mosquito larvae, including samplesize determination and application of sequential methods to estimate population size as well as to classify population level.     

Effects of sampling time,species richness and observer on the exhaustiveness of plant censuses

Question: How may sampling time affect exhaustiveness of vegetation censuses in interaction with observer effect and quadrat species richness? Location: French lowland forests. Methods: Two data sets comprised of 75 timed, one‐hour censuses of vascular plants carried out by five observers on 24 400‐m² forest quadrats were analysed using mixed‐effect models. Results: The level of exhaustiveness increased in a semi‐logarithmic way with sampling time and decreased with quadrat species richness. After one hour, 20 to 30% of the species remained undetected by single observers. This proportion varied among observers and the discrepancy increased with increasing sampling time. Fixing the sampling time may make richness estimates vary less between observers but the time limit should be at least 30 min to reduce the bias in exhaustiveness between rich and poor quadrats. Conclusions We advocate the use of sampling methods based on spatially or temporally‐replicated censuses and statistical analyses that correct for the lack of census exhaustiveness in vegetation studies.     

Orangutan Nesting Behavior in Disturbed Forest of Sabah,Malaysia: Implications for Nest Census

High concentrations of orangutans remain in the multiple-use forests of the Lower Kinabatangan, Sabah, Malaysia. Compared to primary forest, the habitat is highly fragmented, characterized by a low tree density (332 stems/ha), small tree size (83.6% of trees are <20 m high), low basal area (18 m²/ha), abundance of canopy gaps and high level of soil disturbance. The forest structure and composition influence orangutan nesting patterns, and thus directly influence the results of nest surveys used to determine orangutan population size. In logged forests, tall and large trees are the preferred nesting sites of orangutans. The scarcity of suitable nesting sites in the logged-over forests of Kinabatangan, could partly explain the lower daily rate of nest construction (r = 1.00) versus those of other orangutan populations. The nest decay rate t recorded at the study site (average ± SD = 202 ± 151 days) strongly depends on the species of tree in which a nest is built. Our results illustrate that the nest-related parameters used for orangutan censuses fluctuate among habitat types and emphasize the need to determine specific values of r for specific orangutan populations and of t for different tree species in order to achieve accurate analysis of census data.     

When and where to count? Implications of migratory connectivity and nonbreeding distribution to population censuses in a migratory bird population

Migratory connectivity is a metric of the co-occurrence of migratory animals originating from different breeding sites, and like their spatial dispersion, can vary substantially during the annual cycle. Together, both these properties affect the optimal times and sites of population censusing. We tracked taiga bean geese (Anser fabalis fabalis) during 2014–2021 to study their migratory connectivity and nonbreeding movements and determine optimal periods to assess the size of their main flyway population. We also compared available census data with tracking data, to examine how well two existing censuses covered the population. Daily Mantel's correlation between breeding and nonbreeding sites lay between 0 and 0.5 during most of the nonbreeding season, implying birds from different breeding areas were not strongly separated at other times in the annual cycle. However, the connectivity was higher among birds from the westernmost breeding areas compared to the birds breeding elsewhere. Daily Minimum Convex Polygons showed tracked birds were highly aggregated at census times, confirming their utility. The number of tracked birds absent at count sites during the censuses however exceeded numbers double-counted at several sites, indicating that censuses might have underestimated the true population size. Our results show that connectivity can vary in different times during the nonbreeding period, and should be studied throughout the annual cycle. Our results also confirm previous studies, which have found that estimates using marked individuals usually produce higher population size estimates than total counts. This should be considered when using total counts to assess population sizes in the future.     

Estimating the density and biomass of moray eels (Muraenidae) using a modified visual census method for hole-dwelling reef fauna

Synopsis We developed a modified visual census technique suitable for fairly sedentary, hole-dwelling species and used it to determine the diversity, density, biomass, and habitat use of morays at four sites including fringing, patch, and bank reefs in Barbados, West Indies. The method involved a careful search during two passages over transects by day and by night, noting size and position of morays seen, and using the number of new observations on the second passage to estimate the proportion of morays not seen. First passage counts were generally higher than numbers of morays recorded in previous visual censuses. These estimates were raised when we used the time period (day or night) with the highest average density visible for each species, added different individuals from the two passages, and corrected for individuals not seen. We saw more goldentail morays (Gymnothorax miliaris) during the day but more spotted (G. moringa), viper (Enchelycore nigricans), and chestnut (E. carychroa) morays at night. Counts of chain morays (Echidna catenata) did not differ significantly between night and day. Relative abundance and size of moray species varied among sites and to a lesser extent among habitat types within sites, but spotted morays were generally the most important in numbers and biomass. Mean estimated moray density and biomass were much higher than in most previous visual censuses, but similar to rotenone samples on other reefs and comparable to those of other families of predatory fishes. Our findings confirm that morays are abundant and potentially important predators that should be the focus of more attention in ecological studies of coral reefs.     

The influence of sample size on the determination of population trends in the vulnerable Lesser Kestrel Falco naumanni overwintering in South Africa

Reliable measures of population sizes of endangered and vulnerable species are difficult to achieve because of high variability in population sizes and logistic constraints on sample sizes, yet such measures are crucial for the determination of the success of conservation and management strategies aimed at curbing declines in population size. In this study, we use bootstrap confidence intervals to assess the effects of sample size on the reliable determination of roost size in the vulnerable Lesser Kestrel Falco naumanni overwintering at a large communal roost in South Africa. Using results from a unique five-season daily survey, we show that the currently recommended number of replicate sample counts (international action plan for the Lesser Kestrel) of three censuses per season is too low to reliably assess roost size. Bootstrap confidence intervals asymptote at 40–60 censuses per season, with 50 being the modal value. We recommend 50 censuses per season as the optimal sample size. We were also able to show that the population size of Lesser Kestrels at this roost significantly increased over each of the first four seasons of the study period and decreased slightly in the final season.     

Land use history, hurricane disturbance, and the fate of introduced species in a subtropical wet forest in Puerto Rico

Tropical forests are suffering from increasing intensities and frequency of disturbances. As a result, non-native species accidentally introduced or intentionally planted for farming, plantations, and ornamental purposes may spread and potentially invade undisturbed native forest. It is not known if these introduced species will become invasive, as a result of recurrent natural disturbances such as hurricanes. Using data from three censuses (spanning 15 years) of a 16-ha subtropical wet forest plot, we investigated the impact of two hurricanes on populations of plant species that were planted in farms and plantations that were then abandoned and from the natural spread of species introduced into Puerto Rico in the past. The populations of four species (Citrus paradis, Mangifera indica, Musa sp., and Simarouba glauca) changed little over time. Six species (Artocarpus altilis, Calophyllum calaba, Genipa americana, Hibiscus pernambucensis, Syzygium jambos, and Swietenia macrophylla) declined between the first two censuses after Hurricane Hugo, then increased again in Census 3 after Hurricane Georges. Spathodea campanulata gradually increased from census to census, while Coffea arabica declined. These introduced species represent only a small part of the forest basal area and few show signs of increasing over time. The number of stems per plant, new recruits, and the growth rates of these introduced species were within the ranges of those for native plant species. The mortality rates over both census intervals were significantly lower for introduced species (<5% year⁻¹) than for native ones (15% year⁻¹). Many new recruits established after Hurricane Hugo (prior to this study) had opened the forest canopy and these high mortality rates reflect their death as the canopy recovered. Only Swietenia macrophylla and Syzygium jambos showed an increase in stem numbers in the closed canopy area of forest that had suffered limited human disturbance in the past. A future increase in frequency of disturbance may enable greater stem numbers of introduced species to establish, while lower-mortality rates compared to native species, may allow them to persist during inter-hurricane intervals. An increase in the population of introduced species, especially for those that grow into large trees, may have an impact on this tropical forest in the future.     

BOOTSTRAP CORRECTION FOR DIURNAL ACTIVITY CYCLE IN CENSUS DATA FOR ANTARCTIC SEALS1

Antarctic phocid seals and particularly the crabeater (Lobodon carcinophagus) have been observed to display a diurnal cycle in their propensity to haul out on pack ice where they are visible for census. The fact that they are not visible for much of the 24-h period means that density estimates made over broad geographic areas at various times of the day statistically confound this cycle with geographic variability. Limitation of census observations to times of peak haulout results in extreme logistical difficulties and/or considerable reduction in sample size upon which to base population estimates. Reduced sample size results in high variability in population estimates and broad confidence bands. To develop a model with which to correct density estimates for variability due to diurnal cycle, a series of stationary censuses at fixed locations in the Antarctic continental ice pack was made over significant fractions of several days. A unimodal polynomial model for the observed density variation in any one location was statistically significant; a similar model combining multiple locations with densities standardized to peak daily values was also significant. The latter model was used to make corrections for time of day to density estimates in three test data sets taken over broad geographic areas of the Antarctic. Statistical simulation (bootstrap) methods were used to determine if variances of corrected density estimates were lower than those based on uncorrected observations taken only during the peak haulout times of the day. Results were that 95% interval estimates for corrected densities were narrowed to between 40% and 61% of the uncorrected estimates. While there are additional possible sources of variation in haulout tendency, pending further data collection and analyses, the model developed represents a considerably more precise methodology than either averaging over haulout variability or limiting observations to peak daily periods.