Random median sampling to enhance the precision of population estimates

We describe the characteristics of a sampling procedure called random median sampling that was proposed to enhance the precision of population estimates. In performing random median sampling, we first select a sampling item at random from the sampling area. We roughly compare the abundance of individuals in the selected item with that of the adjacent two items in order to identify the item that has median abundance, i.e., the item that has the second largest abundance among the three items. We count the number of individuals of the item having the median abundance. This procedure is repeated n times in the sampling area (i = 1, 2, ..., n). Let m _i be the ith median abundance. The estimates of the mean abundance per sampling item and the variance of estimates are given by Σm _i /n and Σ(m _i –Σm _i /n)²/n(n – 1), respectively. This method is a local application of the median ranked set sampling that was proposed by Muttlak (J Appl Stat Sci 6:245–255, 1997). Random median sampling is effective when the correlation coefficient between adjacent items is small. If the correlation coefficient is close to zero, random median sampling reduces the variance of estimates to 45 or 32% of that in simple random sampling when the distribution follows a normal distribution or a Laplace distribution, respectively. The sample size required to achieve a given precision of estimate decreases accordingly. The effectiveness of random median sampling, however, is small if the correlation coefficient is large. The condition in which random median sampling is superior to simple random sampling is also discussed.     

Estimating the encounter rate variance in distance sampling

Summary .  The dominant source of variance in line transect sampling is usually the encounter rate variance. Systematic survey designs are often used to reduce the true variability among different realizations of the design, but estimating the variance is difficult and estimators typically approximate the variance by treating the design as a simple random sample of lines. We explore the properties of different encounter rate variance estimators under random and systematic designs. We show that a design-based variance estimator improves upon the model-based estimator of Buckland et al. (2001, Introduction to Distance Sampling. Oxford: Oxford University Press, p. 79) when transects are positioned at random. However, if populations exhibit strong spatial trends, both estimators can have substantial positive bias under systematic designs. We show that poststratification is effective in reducing this bias.     

Evaluation of different observational sampling regimes for use in zoological parks

Observational sampling methods provide clearly-defined guidelines for collection and analysis of behavioral data. In some situations, use of formal sampling regimes may be impractical or impossible. A case in point is data collection conducted by animal care staff at zoological parks and aquaria. Often, time is sufficiently limited that data collection is perceived as a task that cannot be accomplished given the normal constraints of the day. Here, we explore the efficacy and validity of using more variable and abridged sampling regimes in an effort to identify the appropriateness of such observation schemes for systematic monitoring of behavior. We describe the results of studies on three species (two polar bears, an Atlantic bottlenose dolphin calf, and two brown bears), conducted over a period of several years at the Brookfield Zoo, Brookfield, Illinois, USA. Data collection schemes varied both within and across groups in order to provide a basis of comparison. In all cases, there were significant differences based on sampling regime for rare behaviors (those that individually comprised <15% of the activity budget), but not for common behaviors. Subsampling from larger data sets indicated that data reliability increases with increasing observation number. We discuss the strengths and weaknesses of such sporadic sampling methods, and suggest that, in many instances such limited data collection may yet yield an accurate picture of animal activity and should not be overlooked as a viable management tool.     

Modifications of the Ekman Sampler

Many types of sediment samplers have been designed for specific purposes and for sampling in different aquatic environments. The objective of this technical note has been to introduce two specific improvements of the traditional Ekman grab sampler, which is a well-known apparatus for sampling of soft sediments, especially in bottom faunistic contexts. The two new modifications concern the automatic closing mechanism. The jointed quadrant and the jointed, hooked axis, both made of brass, provide a much handier and more reliable operation than, e.g., the traditional automatic mechanism.     

Efficiency of sampling methods for accurate estimation of species richness of corticolous microlichens in the Atlantic rainforest of northeastern Brazil

Three different sampling methods were compared with regard to accurate estimation of species richness of crustose and microfoliose lichens in Atlantic rainforest remnants of northeastern Brazil. Quantitative transect sampling yielded three times as many species as non-quantitative opportunistic sampling, even if the number of sampled phorophyte trees was higher on average for the latter method. Repetitive non-quantitative opportunistic sampling of a single site resulted in twice the number of species than one-time sampling, but only two thirds of the number recovered by means of quantitative transect sampling, even of the number of sampled phorophyte trees was highest for this method. The analysis showed that opportunistic sampling fails to detect rare, inconspicuous, sterile, and/or cryptic species, usually neglected or overlooked even by the experienced collector upon visual inspection in the field. On the other hand, quantitative sampling forces even the inexperienced collector to sample lichen thalli or pieces of bark which only after detailed morphological, anatomical, and chemical study in the laboratory reveal themselves as distinct species. Accordingly, of the 456 lichen species collected and identified across the 22 studied localities, no less than 76 were unique to the quantitatively sampled transect, and the lichen species sampled with this method had a significantly higher proportion of rare, inconspicuous, sterile, and/or cryptic species. It is also apparent that it is not the higher number of specimens collected via quantitative sampling that results in a higher number of species, but the method of selection of the specimens, which is subjective and biased towards abundant, conspicuous, fertile and/or distinctive species in opportunistic sampling, but objective and unbiased in quantitative sampling.     

Molecular cytogenetics of Androctonus scorpions: an oasis of calm in the turbulent karyotype evolution of the diverse family Buthidae

After years of qualitative and subjective study, quantitative colour science is now enabling rapid measurement, analysis and comparison of colour traits. However, it has not been determined how many replicates one needs to accurately quantify a species' colours for studies aimed at broad cross‐species trait comparison. We address this major methodological knowledge gap. We first quantified and assessed the variance in colour within and between species. Reflectance spectra of flowers from ten plant species and plumage of 20 bird species were measured using a spectrometer, and reflectance (i.e. brightness) and tetrahedral colour‐space coordinates were calculated. analysis of variance (ANOVA) analyses indicate that there is far more variation in the colours of birds and flowers between species (> 77%) than within species. A Mean Absolute Deviation from the Mean test was applied to indicate the sampling replication required for each species. Tetrahedral coordinates were sampled precisely with only one individual per species. Greater replication was needed to sample reflectance with the desired precision, particularly for darker coloured species. Our findings will allow researchers to allocate their sampling effort in a way that maximises the precision of their colour data collection. The fact that only a few replicates per species are necessary will greatly facilitate broad cross‐species comparisons of colour in the future. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 114 , 69–81.     

三叶斑潜蝇幼虫种群空间格局及抽样技术研究

三叶斑潜蝇Liriomyza trifolii是我国重要的园艺及蔬菜害虫,研究三叶斑潜蝇种群的空间格局和抽样技术,可为该虫的危害调查与防治提供理论依据。应用Iwao m*-m回归分析法、Taylor的幂法则及6个聚集指标,对三叶斑潜蝇幼虫在番茄和豇豆上的空间分布型和抽样技术进行了研究,并做了影响因素分析。结果表明:三叶斑潜蝇幼虫在番茄和豇豆上均呈聚集分布,分布的基本成分是以个体群形式存在,通过分布型参数,采用Kc法、Iwao法及Taylor幂法计算出了在不同精度下三叶斑潜蝇田间的理论抽样数。     

A comparison of pump sampling systems for live zooplankton collection

Simultaneous and sequential tests of pump sampling equipment showed that sampling mortality of Lake Michigan zooplankton collected from the cooling water intake of a nuclear power plant was dependent on the sampling method used. Two sampling devices and two pump types were tested to determine which combination resulted in the lowest sampling mortality. An in-line filter trap located on the suction side of the pump provided the lowest sampling mortality. Higher sampling rates were achieved with a centrifugal pump without increasing sampling mortality.     

Magic bullets and golden rules: data sampling in molecular phylogenetics

Data collection for molecular phylogenetic studies is based on samples of both genes and taxa. In an ideal world, with no limitations to resources, as many genes could be sampled as deemed necessary to address phylogenetic problems. Given limited resources in the real world, inadequate (in terms of choice of genes or number of genes) sequences or restricted taxon sampling can adversely affect the reliability or information gained in phylogenetics. Recent empirical and simulation-based studies of data sampling in molecular phylogenetics have reached differing conclusions on how to deal with these problems. Some advocated sampling more genes, others more taxa. There is certainly no ‘magic bullet’ that will fit all phylogenetic problems, and no specific ‘golden rules’ have been deduced, other than that single genes may not always contain sufficient phylogenetic information. However, several general conclusions and suggestions can be made. One suggestion is that the determination of a multiple, but moderate number (e.g., 6–10) of gene sequences might take precedence over sequencing a larger set of genes and thereby permit the sampling of more taxa for a phylogenetic study.     

基于表型数据的辣椒核心种质构建研究

以收集保存的603份辣椒种质为材料,根据果形指数大小将其分成5组。基于28个性状的表型数据,采用简单比例、平方根比例、对数比例及遗传多样性指数比例法计算各组内取样份数,比较4种组内取样比例法、6种总体取样规模和2种取样方法在构建辣椒核心种质中的作用和效果。结果表明:(1)简单比例、平方根比例、对数比例、遗传多样性指数比例法入选的材料份数占预选核心种质份数依次为24.2%、22.2%、21.1%、17.8%,说明遗传多样性指数比例法对各组取样数量的修正能力最强,使取样更加均衡。(2)当总体取样规模为15%时,遗传多样性指数比例法构建的预选核心种质遗传多样性指数(I)达到最大,表型保留比例(RPR)超过98%;当总体取样规模超过20%时,RPR值、变异系数(CV)和极差符合率(CR)虽然平缓增加,但I值反而减小;说明15%为合适的总体取样规模。(3)利用对数比例法和多样性比例法,在15%的总体取样规模下,聚类取样构建的核心种质I值、RPR值、CV值及CR值均高于随机取样。(4)该研究根据所获得的优化方案最终在表型水平建立了包含91份种质的辣椒核心种质。     

Variance estimation for systematic designs in spatial surveys

Summary In spatial surveys for estimating the density of objects in a survey region, systematic designs will generally yield lower variance than random designs. However, estimating the systematic variance is well known to be a difficult problem. Existing methods tend to overestimate the variance, so although the variance is genuinely reduced, it is over‐reported, and the gain from the more efficient design is lost. The current approaches to estimating a systematic variance for spatial surveys are to approximate the systematic design by a random design, or approximate it by a stratified design. Previous work has shown that approximation by a random design can perform very poorly, while approximation by a stratified design is an improvement but can still be severely biased in some situations. We develop a new estimator based on modeling the encounter process over space. The new “striplet” estimator has negligible bias and excellent precision in a wide range of simulation scenarios, including strip‐sampling, distance‐sampling, and quadrat‐sampling surveys, and including populations that are highly trended or have strong aggregation of objects. We apply the new estimator to survey data for the spotted hyena (Crocuta crocuta) in the Serengeti National Park, Tanzania, and find that the reported coefficient of variation for estimated density is 20% using approximation by a random design, 17% using approximation by a stratified design, and 11% using the new striplet estimator. This large reduction in reported variance is verified by simulation.     

Evaluation of a sweep net electrofishing method for the collection of small fish and shrimp in lotic freshwater environments

A major limitation of methods for collecting small fish and shrimp is that no single method is equally effective for all species or all life history stages. We compared the efficacy of the sweep net electrofishing (SNE) method with two other commonly used sampling methods for the collection of small fish and shrimp, the standard sweep net (SW) and the point abundance electrofishing (PAE) method developed by Copp & Penaz (1988). The effectiveness of the three methods was compared under three different current speeds in an Australian lowland river. The SNE method collected an order of magnitude more shrimp and a greater size range than the other two methods, irrespective of the current speed. The SNE method captured a similar number of fish to the SW method, but significantly more individuals than the PAE method. However, the two electrofishing methods caught a greater size range of both fish and shrimp than did the SW method. The effectiveness of the SNE method appears to be due to its ability to immediately capture shocked individuals, thereby lessening the likelihood of escape. Whilst the SNE method is not quantitative, its high level of effectiveness is likely to be particularly useful in systems with low densities of fish and shrimp.     

A Quantitative Comparison of Two Zooplankton Sampling Methods,a Plankton Trap and a Towed Net,in the Baltic

A comparison between a large plankton trap with a capacity of 23 litres and a modern towed net was made in the course of zooplankton production studies in the Baltic. On the average, the net efficiency was 75% of that of the trap. Both methods were equally efficient in catching naupliar stages of copepods. The net efficiency was especially low during the zooplankton maximum for adult copepods and cladocerans (41 and 51%, respectively), probably due to the active avoidance of the net by fast swimming species. The net was only about 66% as efficient as the trap for catching rotifers. This difference is probably caused by the loss of small and softbodied forms through the mesh. Only about 65% of the total zooplankton biomass retained in the trap was collected by the net. Net sampling is not recommended for quantitative zooplankton studies.     

Analysing the representativeness of local‐scale palaeodiversity measurements: a case from the Lower Cretaceous plant assemblage of Hautrage (Mons Basin,Belgium)

Two plant fossil‐bearing beds from the middle Barremian of Belgium were analysed to ascertain how experimental designs affect conclusions regarding palaeodiversity at a local scale. We analysed eight lateral samples per bed taken regularly every 3 m using an exhaustive sub‐sampling method. The Clench equation was used to evaluate the completeness of the taxonomic inventory of the samples and the sampling effort needed to obtain a reliable representation of diversity. The number of replicates needed to obtain the same representation of diversity from different nearby lateral samples of the same bed ranged from 5 to 19. Richness (S), Evenness (J) and the number of equiprobable taxa (2^H’) greatly varied between samples from the same bed, even over short distances. Only one of the studied samples was representative of the taxonomic inventory of its bed. Our study shows that 1) the selection bias of the sampling area is reduced by increasing the number of lateral samples taken in a bed, enabling more reliable conclusions about local‐scale diversity; 2) intense sub‐sampling methods are needed to account for statistically independent observations of detailed lateral variation; and 3) sampling methods in palaeodiversity analyses must look for a similar degree of representativeness in samples rather than a homogeneous sample size. Using a sampling effort analysis provides evidence for the completeness of the data set, adjusting the amount of work required. Implementing the Clench equation in palaeodiversity analyses improves the performance of data acquisition in palaeoecological studies and provides a quality test of the data sets derived from them.     

不同采样设计评估鱼类群落效果比较

鱼类群落生态学研究结果的准确性很大程度上依赖于采样设计的合理性和准确性,正确的采样调查设计不仅可以降低调查成本,其结果也对渔业资源的评估或者管理起到相当重要的作用.本文利用计算机模拟定点采样、简单随机采样和分层采样,比较了3种采样设计的采样效果、相对误差及相对偏差.结果表明： 定点采样设计的采样效果 (采样效果平均值为3.37)要弱于简单随机采样和分层随机采样 (采样效果平均值为0.961).3种采样设计中,分层采样设计在鱼类群落丰富度评估时表现最好,其采样效果、相对误差和相对偏差表现最佳.随着采样数的增加,分层采样设计的采样效果有所下降,但其采样精度提高.     

Composition and diversity of the spider fauna in the canopy of a montane forest in Tanzania

Spiders were sampled using insecticide knockdown in an African montane forest in the Uzungwa Mountains of Tanzania. The results are used to discuss the faunal composition at the site and in comparison to other sites, and the implications of the results for estimating spider diversity in Africa are discussed. A total of 5233 adults comprising 149 species were collected from 11 samples covering a total of 906 m² of projected area. Three species contributed 45% of the sample. Previous insecticide knockdown studies of tropical lowland forest canopies have shown a dominance of Theridiidae, Salticidae and Araneidae. In the present study Linyphiidae dominated in abundance and were the second most diverse in terms of species richness. Other abundant families were Oonopidae and Pholcidae, while Theridiidae, Salticidae and Araneidae were rich in species. This supports a previous study, which indicated that the importance of linyphiids increases with altitude. Species richness was predicted using a number of estimators, which produced relatively similar results. Using the abundance-based estimator, Chao 1, the predicted richness for the total area sampled is 183 ± 15 species. This indicates that at least 20% of the area's spider community remains unsampled. A high ratio of undescribed species (approximately 80%) and a relatively high species turnover compared to a site 20 km away within the same forest complex suggests that the number of spiders in Africa could well be much higher than the current, published estimate of 20000 species.