Predictors of elephant poaching in a wildlife crime hotspot: The Ruvuma landscape of southern Tanzania and northern Mozambique

Understanding the spatial distribution of elephant carcasses in relation to ecological characteristics and human activities is critical to developing targeted management strategies for reducing poaching. We employ a spatial modelling approach to quantify the relative contribution of multiple climatic, ecological, human and protected area management predictors of the number of elephant carcasses in a recognized poaching hotspot: the Ruvuma landscape of northern Mozambique and southern Tanzania. This includes the Niassa Reserve in the south and the Selous Game Reserve in the north. In Mozambique, the number of elephant carcasses is positively associated with State-managed protected areas such as Niassa Reserve, but particularly with environmental variables including low rainfall and high temperatures. In Tanzania, elephant carcasses are positively associated with community-managed sites. A strong focus on effective management of protected areas in the Ruvuma landscape is crucial to reducing the killing of elephants.     

A population survey of forest elephants (Loxodonta africana cyclotis) in northern Congo

From February to April of 1989 transects were run in four forested areas of northern Congo (Brazzaville) to census elephant populations using dung counts. A total of 401.0 km of line transects was completed during the survey. An overall extrapolated density of 0.7 elephants/km² was obtained with a variation from 0.3 to 0.9 elephants/km². This study shows for the first time that elephants occur in high densities in a number of sites in northern Congo. When elephant dung density is plotted against distance from the nearest village for each transect a significant positive linear correlation results. Based on this correlation we hypothesize that the population of elephants in much of northern Congo is high. Local reports and our own observations indicate that elephant poaching is heavy in northern Congo and that the CITES ban on trade in African elephant ivory has had a limited effect on the level of poaching.     

Comparison of sample and total counts of elephant and buffalo in Masai Mara, Kenya

Both aerial transect sample counts and total counts of elephant and buffalo were conducted in the study area during the wet season. The results from the two counting methods were tested for significant difference. The test showed that the results were not significantly different for both the elephant ( P > 0.05) and buffalo ( P > 0.05).     

Diet and distribution of elephant in the Maputo Elephant Reserve, Mozambique

The distribution and diet of the elephants of the Maputo Elephant Reserve were studied using dung counts, satellite tracking and faecal analysis. The results were compared with earlier data from before the civil war in Mozambique. The elephant population decreased during the civil war, but 180 animals still remain. Earlier studies described the elephants as preferring the grass plains. Currently, the elephants prefer the dense forest patches over the high quality forage found in the grass plains. Water salinity affected distribution; elephant dung piles were found closer to fresh water in the dry season. A total of 95 different plant species were identified in the faeces. The percentage of grass was relatively low compared with other studies, increasing at the beginning of the rainy season. At the end of the dry season, elephants concentrated on the few available browse species with young leaves, but generally preferred grass species to browse species. Diet composition was mainly affected by season and less by habitat. The elephants have changed their habitat preference in reaction to poaching, and probably increased the contribution of browse species in the diet. The presence of forest patches has been vital for the survival of the elephants.     

New evidence for hybrid zones of forest and savanna elephants in Central and West Africa

The African elephant consists of forest and savanna subspecies. Both subspecies are highly endangered due to severe poaching and habitat loss, and knowledge of their population structure is vital to their conservation. Previous studies have demonstrated marked genetic and morphological differences between forest and savanna elephants, and despite extensive sampling, genetic evidence of hybridization between them has been restricted largely to a few hybrids in the Garamba region of northeastern Democratic Republic of Congo (DRC). Here, we present new genetic data on hybridization from previously unsampled areas of Africa. Novel statistical methods applied to these data identify 46 hybrid samples – many more than have been previously identified – only two of which are from the Garamba region. The remaining 44 are from three other geographically distinct locations: a major hybrid zone along the border of the DRC and Uganda, a second potential hybrid zone in Central African Republic and a smaller fraction of hybrids in the Pendjari–Arli complex of West Africa. Most of the hybrids show evidence of interbreeding over more than one generation, demonstrating that hybrids are fertile. Mitochondrial and Y chromosome data demonstrate that the hybridization is bidirectional, involving males and females from both subspecies. We hypothesize that the hybrid zones may have been facilitated by poaching and habitat modification. The localized geography and rarity of hybrid zones, their possible facilitation from human pressures, and the high divergence and genetic distinctness of forest and savanna elephants throughout their ranges, are consistent with calls for separate species classification.     

Dung pellets used to identify the distribution and density of dik‐dik

Estimates of the population density of elusive animals based on counts of dung deposits are often inaccurate. We report a refinement of this technique for territorial ungulates which exploits variation in pellet characteristics between individuals. In a study of Kirk's dik‐dik ( Madoqua kirki ) we found that pellet size and shape were less variable within individuals than between individuals. These findings were used to census a Kenyan population entirely on the basis of uniquely identifiable pellet groups, without any prior knowledge of the identity and location of individuals. The estimates were later compared with the results of an extensive capture and radio telemetry study. Population density estimated by the pellet‐matching method and by telemetry differed by less than 5%. In addition, only one border of 18 territories was incorrectly assigned, and only one territory was not detected. This represents a significant improvement in accuracy over indiscriminate dung counting techniques. We suggest that this method may also provide an accurate and non‐intrusive way of estimating population density in other solitary ungulates.     

Rainfall and the population dynamics of elephant dung‐piles in the forests of southern Ghana

Decaying dung‐piles were observed at three sites in the lowland forests of southern Ghana. Time to disappearance was inversely related to rainfall in the month of deposition. The monthly survival of each age‐class decreased with age and was inversely proportional to rainfall in the month of deposition. A simulation revealed that if elephant numbers and defaecation rates are constant, then dung density in a given month will be governed by rainfall in the two preceding months. Changes in annual rainfall and its monthly distribution in southern Ghana mean that dung‐pile densities will vary from one month to the next, and between the same months in different years, even if elephant numbers are constant. Thus dung counters must take rainfall into account. A new method of censusing elephants in Ghana is presented, and suggestions are made for improving elephant counts in central African forests.     

Survivorship of a declining population of southern elephant seals, Mirounga leonina, in relation to age, sex and cohort

This study quantified both the age- and sex-specific survival rates of juveniles and adults, and tested for interannual differences in age-specific survival rates of the southern elephant seal population at Marion Island. Pups were tagged on an annual basis from 1983 onwards at Marion Island, and a consistent recapture program yielded data that was analysed using the software package MARK to obtain maximum-likelihood estimates of survival and capture probability. On average, 1st-year survival was 0.58 and 0.62, and survival rate averaged over the first 3 years of life, 0.69 and 0.74 for males and females, respectively. From years 4 to 9, the average survival rate was 0.66 and 0.75 for males and females, respectively. Survival estimates for elephant seals in their 10th–13th year are also presented, although these are based on very small sample sizes. Averages of age-specific survival estimates from the earlier (mostly 1983–1987 cohorts) and later (mostly 1988–1992 cohorts) periods were compared and considerable reductions were observed in 4th- and 5th-year male survival, and 4th-year female survival. The comparatively low adult survival is suggested as the proximate cause, and food limitation as deduced from the decline in survival of elephant seals with comparatively high energetic demands as the ultimate cause behind the population decline at Marion Island. Although not tied in with the decline of the population, 1987, 1990 and 1993 were identified as high-mortality years. Received: 14 December 1998 / Accepted: 14 June 1999     

Estimating elephant densities from wells and droppings in dried out riverbeds

In this paper, we present a new method for estimating elephant densities by counting elephant wells and dung boli within dry seasonal flooding rivers. A combination of aerial and ground counts of elephant wells and dung boli in the Ewaso Ngiro River were related to elephant numbers, obtained from an on‐going monitoring program of individually identified elephants in Samburu and Buffalo Spring National Reserves, Kenya. The total number of elephant observations was highly correlated with both densities of wells and dung boli at a spatial resolution of 4‐km river‐section. This indicates that both wells and droppings can be used for estimating relative densities at such spatial resolution. The method can be used as a quick and reliable way of estimating relative elephant densities in semiarid regions but is sensitive to differences in the time when different parts of the river dry out and will be unreliable in areas with secondary un‐censused water sources. A short 4‐week period between the river dry out and the count is recommended, because of an error induced by a level of well reuse and the difficulties in counting areas of high well densities from the air.     

Nest and breeding population abundance of Least Terns: assessing bias and variation in survey timing and methods

Methods commonly used to estimate the number of nests and size of the breeding population at colonies of Least Terns (Sternula antillarum) and other waterbirds include walk‐through counts of nests (ground‐nest counts) and counts of incubating adults from the colony perimeter (incubating‐adult counts). The bias and variance of different methods and the comparability of repeated surveys versus once‐annual censuses are poorly understood. Our objectives were to assess (1) the potential bias and variation of the more rapid incubating‐adult counts compared to the time‐intensive, and presumably more accurate, ground‐nest counts, and (2) how accurately a once‐annual census captured peak nesting abundance. We studied nine Least Tern colonies at Cape Lookout National Seashore (CALO), North Carolina, from April to August 2010–2012. We analyzed observer and survey method agreement with concordance correlation coefficients (ρ_c). We deployed time‐lapse cameras at 156 nests and used repeated‐measures logistic regression to determine if the proportion of time spent incubating varied with colony, time of day, or time of season. We found substantial agreement in abundance estimates of Least Tern nests and incubating adults between observers and survey methods, and among different times of day and seasons (all comparisons ρ_c > 0.97). Least Terns incubated eggs 94% of the time on average during daylight hours, irrespective of colony, nesting stage, or month. Although the nesting peak at CALO occurred during the recommended census period for Least Terns, abundance estimates for surveys conducted at different times during that period varied by as much as 39%. We recommend conducting incubating‐adult counts to estimate nest and breeding population abundance of Least Terns or other waterbirds when vegetation or dunes do not obstruct views of nesting colonies. In addition, given the variation in abundance estimates for surveys conducted at different times during the recommended survey period, incubating‐adult counts should be performed at least twice during the census period, with the maximum count reported as peak nest abundance.     

Accounting for bias from sequencing error in population genetic estimates

Sequencing error presents a significant challenge to population genetic analyses using low-coverage sequence in general and single-pass reads in particular. Bias in parameter estimates becomes severe when the level of polymorphism (signal) is low relative to the amount of error (noise). Choosing an arbitrary quality score cutoff yields biased estimates, particularly with newer, non-Sanger sequencing technologies that have different quality score distributions. We propose a rule of thumb to judge when a given threshold will lead to significant bias and suggest alternative approaches that reduce bias.     

Effects of population characteristics and structure on estimates of effective population size in a house sparrow metapopulation

Effective population size (N_e) is a key parameter to understand evolutionary processes and the viability of endangered populations as it determines the rate of genetic drift and inbreeding. Low N_e can lead to inbreeding depression and reduced population adaptability. In this study, we estimated contemporary N_e using genetic estimators (LDNE, ONeSAMP, MLNE and CoNe) as well as a demographic estimator in a natural insular house sparrow metapopulation. We investigated whether population characteristics (population size, sex ratio, immigration rate, variance in population size and population growth rate) explained variation within and among populations in the ratio of effective to census population size (N_e/N_c). In general, N_e/N_c ratios increased with immigration rates. Genetic N_e was much larger than demographic N_e, probably due to a greater effect of immigration on genetic than demographic processes in local populations. Moreover, although estimates of genetic N_e seemed to track N_c quite well, the genetic N_e‐estimates were often larger than N_c within populations. Estimates of genetic N_e for the metapopulation were however within the expected range (<N_c). Our results suggest that in fragmented populations, even low levels of gene flow may have important consequences for the interpretation of genetic estimates of N_e. Consequently, further studies are needed to understand how N_e estimated in local populations or the total metapopulation relates to actual rates of genetic drift and inbreeding.     

Some comments on anderson and pospahala's correction of bias in line transect sampling

ANDERSON and POSPAHALA (1970) investigated the estimation of wildlife population size using the belt or line transect sampling method and devised a correction for bias, thus leading to an estimator with interesting characteristics. This work was given a uniform mathematical framework in BURNHAM and ANDERSON (1976). In this paper we show that the ANDERSON-POSPAHALA estimator is optimal in the sense of being the (unique) best linear unbiased estimator within the class of estimators which are linear combinations of cell frequencies, provided certain assumptions are met.     

Estimating population density from indirect sign: track counts and the Formozov–Malyshev–Pereleshin formula

For many purposes it is often desirable to estimate animal population densities over large areas. Where total counts are not possible and sightings are relatively rare, a range of methods exists to estimate densities from indirect sign. Such methods are frequently unreliable and usually require independent calibration or confirmation. We present an analytical method for estimating population density from track counts. The method, widely known in the Russian Federation but not in the English language scientific literature, requires counts of tracks of known age, together with estimates of animal daily travel distances. We use simulations to verify the theoretical basis of the approach and to indicate potential precision that may be achieved. We illustrate application of the approach using a large data set on ungulate track counts in the Russian Far East. We suggest that under most circumstances, nonparametric bootstrapping will be the most appropriate method for deriving estimates of confidence intervals about density estimates. As with other approaches to estimating density from indirect sign, the method that we discuss is vulnerable to violations of an array of underlying assumptions. However, it is easily applied and could represent an important method by which the relationship between indices of abundance and absolute density can be understood.     

Effective population size does not predict codon usage bias in mammals

Synonymous codons are not used at equal frequency throughout the genome, a phenomenon termed codon usage bias (CUB). It is often assumed that interspecific variation in the intensity of CUB is related to species differences in effective population sizes (N_e), with selection on CUB operating less efficiently in species with small N_e. Here, we specifically ask whether variation in N_e predicts differences in CUB in mammals and report two main findings. First, across 41 mammalian genomes, CUB was not correlated with two indirect proxies of N_e (body mass and generation time), even though there was statistically significant evidence of selection shaping CUB across all species. Interestingly, autosomal genes showed higher codon usage bias compared to X‐linked genes, and high‐recombination genes showed higher codon usage bias compared to low recombination genes, suggesting intraspecific variation in N_e predicts variation in CUB. Second, across six mammalian species with genetic estimates of N_e (human, chimpanzee, rabbit, and three mouse species: Mus musculus, M. domesticus, and M. castaneus), N_e and CUB were weakly and inconsistently correlated. At least in mammals, interspecific divergence in N_e does not strongly predict variation in CUB. One hypothesis is that each species responds to a unique distribution of selection coefficients, confounding any straightforward link between N_e and CUB.     

Decadal changes in habitat characteristics influence population trajectories of southern elephant seals

Understanding divergent biological responses to climate change is important for predicting ecosystem level consequences. We use species habitat models to predict the winter foraging habitats of female southern elephant seals and investigate how changes in environmental variables within these habitats may be related to observed decreases in the Macquarie Island population. There were three main groups of seals that specialized in different ocean realms (the sub‐Antarctic, the Ross Sea and the Victoria Land Coast). The physical and climate attributes (e.g. wind strength, sea surface height, ocean current strength) varied amongst the realms and also displayed different temporal trends over the last two to four decades. Most notably, sea ice extent increased on average in the Victoria Land realm while it decreased overall in the Ross Sea realm. Using a species distribution model relating mean residence times (time spent in each 50 × 50 km grid cell) to 9 climate and physical co‐variates, we developed spatial predictions of residence time to identify the core regions used by the seals across the Southern Ocean from 120°E to 120°W. Population size at Macquarie Island was negatively correlated with ice concentration within the core habitat of seals using the Victoria Land Coast and the Ross Sea. Sea ice extent and concentration is predicted to continue to change in the Southern Ocean, having unknown consequences for the biota of the region. The proportion of Macquarie Island females (40%) utilizing the relatively stable sub‐Antarctic region, may buffer this population against longer‐term regional changes in habitat quality, but the Macquarie Island population has persistently decreased (?1.45% per annum) over seven decades indicating that environmental changes in the Antarctic are acting on the remaining 60% of the population to impose a long‐term population decline in a top Southern Ocean predator.     

Estimating densities of large herbivores in tropical forests: Rigorous evaluation of a dung‐based method

When sighting‐based surveys to estimate population densities of large herbivores in tropical dense forests are not practical or affordable, surveys that rely on animal dung are sometimes used. This study tested one such dung‐based method by deriving population densities from observed dung densities of six large herbivores (chital, elephant, gaur, muntjac, sambar, and wild pig) in two habitats, dry deciduous forests (DDF) and moist deciduous forests (MDF), within Nagarahole National Park, southern India. Using the program DUNGSURV, dung pile counts, decay rates estimated from field experiments, and defecation rates derived from literature were analyzed together by a model that allows for random events affecting dung decay. Densities of chital were the highest, followed by sambar. Wild pig densities were similar in the two habitats, sambar densities were higher in DDF, and densities of the other species were higher in MDF than in DDF. We compared DUNGSURV estimates with densities estimated using distance sampling in the same season. DUNGSURV estimates were substantially higher for all species in both habitats. These differences highlight the challenges that researchers face in computing unbiased estimates of dung decay rates and in relying on defecation rates from literature. Besides the elephant, this study is the first to rigorously test the efficacy of using a dung‐based approach to estimate densities of large herbivore species in Asia, and based on this evaluation, we provide specific recommendations to address issues that require careful consideration before observed dung densities are used to derive animal densities. Our results underline the need for an experimental study of a known population in a fenced reserve to validate the true potential of using dung‐based approaches to estimate population densities.