In Defense of Indices: The Case of Bird Surveys

Abstract: Indices to population size have come under increasing criticism in recent years, on the grounds that indices might not faithfully represent the entire population. Most criticisms involve surveys of birds, particularly those based on point counts, which is my focus here. A variety of quantitative methods have been developed to reduce the bias of point counts, such as distance sampling, multiple-observer surveys, and time-of-detection methods. I argue that these developments are valuable, in that they enhance understanding of the detection process, but that their practical application may well be limited, likely to intensive studies focusing on a small number of species. These quantitative methods are not generally applicable to extensive, multiple-species surveys. Although criticism of the thoughtless use of indices is welcome, their wholesale rejection is not.     

Methods of estimating ungulate populations in tropical forests

Pellet group counts, drive counts and track counts were used to estimate population abundances of one small and four large species of duiker (Cephalophus sp.) in a moist evergreen forest in northeastern Zaire. Procedures to develop estimates based on line transect counts of pellet groups are described. The Fourier Series estimator proved to be useful for estimating densities of pellet groups. Comparisons were made between population abundance in heavily and lightly hunted areas, and between small and large species. The three census methods were consistent in showing no significant difference between animal abundance in the hunted populations or between abundances of the smaller species relative to the larger species. All methods showed potential as population indices for assessing trends of rare and elusive species in forest environments. Drive counts and pellet group counts may also be useful for estimating densities.     

Towards a standardized index of European rabbit abundance in Iberian Mediterranean habitats

European rabbits Oryctolagus cuniculus are a keystone species in Iberian Mediterranean ecosystems. However, the reliability of methods for estimating rabbit abundance, particularly when at low numbers, is not well understood. Further, better standardization of these methodologies would allow abundance estimates to be more reliably compared between areas and periods. Consequently, we compared several frequently used methods of estimating rabbit abundance and assessed their advantages and disadvantages. During the summers of 2008 and 2009, in 11 localities of central-southern Spain we undertook (a) driving transect counts of rabbits, either at dusk or at night, (b) linear transects on foot recording rabbit signs, (c) cleared-plot pellet counts at permanent plots, and (d) standing crop counts, both with and without habitat stratification. Density estimated at night from driving transects using the Distance Sampling method (the reference method against which all other indices were compared) varied from 0 to 2.69 rabbits ha⁻¹. Most pellet-count indices were significantly related to the reference method. In particular, cleared-plot pellet counts in permanent plots corrected for pellet persistence showed the best correlation with the reference method. In contrast, latrine counts were not related to the reference method index, and we recommend against their use. A standard methodology based on cleared-plot pellets counts could be used to monitor rabbit abundance on a large scale.     

Estimating red deer abundance in a wide range of management situations in Mediterranean habitats

When human interventions interfere with the natural regulation of wildlife populations by favouring some species, overabundance can emerge. We evaluated different methods of estimating red deer abundance in a wide range of population densities from southern Spain. Distance sampling estimates were used as the reference method across 22 localities and were compared with two kilometric abundance indices (KAIs), four indices based on pellet group counts and two browsing indices (BWIs). The average red deer density estimated by distance sampling was 19.51±3.19 deer per 100 ha, showing a wide range across the study area (0.04–66.77). Distance sampling estimates correlated with the KAIs, pellet group-based index and the BWI. The agreement with distance sampling improved when groups were used instead of individuals in the KAIs, when the minimum pellet group size was fixed at 20 pellets in the dropping counts, and when only highly palatable species were used in the BWI. Thus, several direct and indirect methods can estimate red deer abundance in Mediterranean habitats from Southern Spain with appropriate modifications.     

Comparison of taxonomic, colony morphotype and PCR-RFLP methods to characterize microfungal diversity

We compared three methods for estimating fungal species diversity in soil samples. A rapid screening method based on gross colony morphological features and color reference standards was compared with traditional fungal taxonomic methods and PCR-RFLP for estimation of ecological indices of soil microfungal community composition. Normalized counts of colony morphotypes on dichloran rose bengal medium were used to estimate species richness (S) and evenness (J) and to calculate Shannon's diversity (H) and Simpson's (SI) dominance indices. Isolates were obtained by dilution plating techniques from litter and soil layer samples taken from Douglas-fir forest and clear-cut areas at two locations in the Cascade Mountains. The highest correspondence (97%) was observed between taxonomic identification and RFLP patterns (32:33). Cladistic analyses of PCR-RFLP patterns indicated an 81% correspondence between RFLP patterns:colony morphotypes (33:41). A correspondence of 78% was observed between traditional taxonomic identification:colony morphotypes (32:41). Statistical analyses of ecological indices based on quantitative application of the colony morphotyping method indicated significant differences (P < 0.05) in fungal community composition between forested and clear-cut areas at the Toad Road site but not at the Falls Creek site. Comparisons of ecological indices based on traditional identification of taxa by microscopic characterization on defined culture media resulted in identical findings of statistical significance. The colony morphotyping approach is proposed as a screening method to identify potential effects of land management practices, edaphic factors and pollutants on microfungal diversity.     

Sampling skinks and geckos in artificial cover objects in a dry mixed grassland-shrubland with mammalian predator control

Introduced mammalian predators threaten populations of endemic New Zealand lizards but their effects on lizard populations have not been quantified on the mainland. We trialled the use of artificial cover objects (ACOs) for sampling small terrestrial lizards (the skinks Oligosoma maccanni, O. nigriplantare polychroma and O. inconspicuum, and gecko Hoplodactylus maculatus) in three experimental mammal-management treatments: a mammal-proof fence, two sites in an intensive mammal-removal area, and an experimental control site with no mammal removal. These predator control regimes were established in 2005-2006 to protect endangered grand skinks (O. grande) and Otago skinks (O. otagense) at Macraes Flat, North Otago. We (1) counted skinks and geckos found under ACOs on a single day and compared these counts between treatments, and (2) estimated lizard population sizes (N) based on capture-mark-recapture (CMR) of lizards under ACOs in daily and weekly sampling sessions. Our results provide baseline data of the abundance of the small lizard species shortly after implementation of predator management. Single-day counts of skinks were significantly higher inside the mammal fence compared with the experimental control. No consistent differences were found between the other treatments. Significantly more skinks were counted in gully habitats than in ridge habitats. N?, counts of skinks from the first day of CMR, and the total number of individuals caught were correlated, but these relationships must be validated with independent data. Few geckos were caught unless ACOs were placed near rock outcrops. Only two skinks but 25% of geckos moved between adjacent ACOs (5-m spacing). The recapture rate of skinks was low; captures declined when ACOs were checked daily but not when they were checked weekly. Because of potential biases of these methods, we propose to compare counts in ACOs and N? based on CMR in ACOs with N? based on CMR in pitfall traps with 3-m spacing.     

Estimating roe deer abundance from pellet group counts in Spain: An assessment of methods suitable for Mediterranean woodlands

Despite of the generalized expansion of wild ungulates in Europe, roe deer (Capreolus capreolus) is experiencing contrasted population trends; it is expanding in some regions while declining in others likely due to the expansion of other deer species. In both extremes, reliable methods to estimate population abundance are required. We evaluated different methods of estimating deer abundance in Mediterranean woodlands based on pellet group counts. Distance Sampling applied to pellet counts and a new easier and cost-effective method based on strip-variable transect counts (FST) were assessed comparing their estimates (pellet group density) with the abundance indices obtained from traditionally used reference methods (faecal standing crop) in 61 localities (n = 183 surveys). The average roe deer density estimated from faecal standing crop was 5.56 ± 0.75 (range 0.01–20.74) deer per 100 ha. Distance Sampling and FST estimates correlated with reference methods. As a first conclusion it may be noted that all indirect methods used here can be used to estimate roe deer abundance. The selection of a given method based on pellet counts to estimate roe deer population abundance should take into account the specific objectives of the research, resources available, and the timescale in which the information is required. Among them, Distance Sampling may be used when human resources and skills are enough but FST is a rapid and efficient alternative to estimate pellet group density when they are not.     

Sampling for Regional Monitoring of Nematode Communities in Agricultural Soils

Regional assessment of nematode communities to monitor the condition or ecological health of agricultural soils requires sampling programs with measures of known reliability and the ability to detect differences over time. Numbers of fields sampled in a region, samples taken per field, and subsamples assayed per sample must be balanced with cost to provide the best sampling scheme. We used components of variance from statewide surveys in North Carolina (1992) and Nebraska (1993) to estimate number of (i) fields to be sampled; (ii) 20-core, composite soil samples to be obtained for each field; and (iii) subsamples to be assayed for each composite sample to detect a specified amount of change in index values within a geographic region. Variances for these three components were used to estimate the degree of reliability for five ecologically based indices (four measures of maturity and one of diversity) of nematode communities. Total variance for maturity and diversity indices, based upon communities of free-living nematodes, was greater in North Carolina than in Nebraska; the opposite was true for indices based strictly upon maturity of communities of plant-parasitic nematodes or of all nematodes in soil. Variability within samples was greater in North Carolina than in Nebraska, especially for maturity indices based only upon free-living nematodes. We identified two possible sampling strategies for a regional survey: Option 1, with two independent samples per field and a single subsample assayed per sample, which would provide a reliability ratio value ≥0.6 for most indices; and Option 2, with three independent samples per field and two subsamples assayed per sample, which would provide a reliability ratio value ≥0.7 for several indices. When cost was considered, Option 1 was the better strategy. Number of fields to be sampled within a region or state varied with the index chosen; with specific indices, however, a 10% change in mean index value could be detected with a sample of 50 to 100 fields.     

Toward Reliable Estimates of Abundance: Comparing Index Methods to Assess the Abundance of a Mammalian Predator

Due to time and financial constraints indices are often used to obtain landscape-scale estimates of relative species abundance. Using two different field methods and comparing the results can help to detect possible bias or a non monotonic relationship between the index and the true abundance, providing more reliable results. We used data obtained from camera traps and feces counts to independently estimate relative abundance of red foxes in the Black Forest, a forested landscape in southern Germany. Applying negative binomial regression models, we identified landscape parameters that influence red fox abundance, which we then used to predict relative red fox abundance. We compared the estimated regression coefficients of the landscape parameters and the predicted abundance of the two methods. Further, we compared the costs and the precision of the two field methods. The predicted relative abundances were similar between the two methods, suggesting that the two indices were closely related to the true abundance of red foxes. For both methods, landscape diversity and edge density best described differences in the indices and had positive estimated effects on the relative fox abundance. In our study the costs of each method were of similar magnitude, but the sample size obtained from the feces counts (262 transects) was larger than the camera trap sample size (88 camera locations). The precision of the camera traps was lower than the precision of the feces counts. The approach we applied can be used as a framework to compare and combine the results of two or more different field methods to estimate abundance and by this enhance the reliability of the result.     

Comparison of Methods for Estimating Bird Abundance and Trends From Historical Count Data

Abstract: The use of bird counts as indices has come under increasing scrutiny because assumptions concerning detection probabilities may not be met, but there also seems to be some resistance to use of model-based approaches to estimating abundance. We used data from the United States Forest Service, Southern Region bird monitoring program to compare several common approaches for estimating annual abundance or indices and population trends from point-count data. We compared indices of abundance estimated as annual means of counts and from a mixed-Poisson model to abundance estimates from a count-removal model with 3 time intervals and a distance model with 3 distance bands. We compared trend estimates calculated from an autoregressive, exponential model fit to annual abundance estimates from the above methods and also by estimating trend directly by treating year as a continuous covariate in the mixed-Poisson model. We produced estimates for 6 forest songbirds based on an average of 621 and 459 points in 2 physiographic areas from 1997 to 2004. There was strong evidence that detection probabilities varied among species and years. Nevertheless, there was good overall agreement across trend estimates from the 5 methods for 9 of 12 comparisons. In 3 of 12 comparisons, however, patterns in detection probabilities potentially confounded interpretation of uncorrected counts. Estimates of detection probabilities differed greatly between removal and distance models, likely because the methods estimated different components of detection probability and the data collection was not optimally designed for either method. Given that detection probabilities often vary among species, years, and observers investigators should address detection probability in their surveys, whether it be by estimation of probability of detection and abundance, estimation of effects of key covariates when modeling count as an index of abundance, or through design-based methods to standardize these effects.     

Abundance-based similarity indices and their estimation when there are unseen species in samples

A wide variety of similarity indices for comparing two assemblages based on species incidence (i.e., presence/absence) data have been proposed in the literature. These indices are generally based on three simple incidence counts: the number of species shared by two assemblages and the number of species unique to each of them. We provide a new probabilistic derivation for any incidence-based index that is symmetric (i.e., the index is not affected by the identity ordering of the two assemblages) and homogeneous (i.e., the index is unchanged if all counts are multiplied by a constant). The probabilistic approach is further extended to formulate abundance-based indices. Thus any symmetric and homogeneous incidence index can be easily modified to an abundance-type version. Applying the Laplace approximation formulas, we propose estimators that adjust for the effect of unseen shared species on our abundance-based indices. Simulation results show that the adjusted estimators significantly reduce the biases of the corresponding unadjusted ones when a substantial fraction of species is missing from samples. Data on successional vegetation in six tropical forests are used for illustration. Advantages and disadvantages of some commonly applied indices are briefly discussed.     

Monitoring butterfly abundance: beyond pollard walks

Most butterfly monitoring protocols rely on counts along transects (Pollard walks) to generate species abundance indices and track population trends. It is still too often ignored that a population count results from two processes: the biological process (true abundance) and the statistical process (our ability to properly quantify abundance). Because individual detectability tends to vary in space (e.g., among sites) and time (e.g., among years), it remains unclear whether index counts truly reflect population sizes and trends. This study compares capture-mark-recapture (absolute abundance) and count-index (relative abundance) monitoring methods in three species (Maculinea nausithous and Iolana iolas: Lycaenidae; Minois dryas: Satyridae) in contrasted habitat types. We demonstrate that intraspecific variability in individual detectability under standard monitoring conditions is probably the rule rather than the exception, which questions the reliability of count-based indices to estimate and compare specific population abundance. Our results suggest that the accuracy of count-based methods depends heavily on the ecology and behavior of the target species, as well as on the type of habitat in which surveys take place. Monitoring programs designed to assess the abundance and trends in butterfly populations should incorporate a measure of detectability. We discuss the relative advantages and inconveniences of current monitoring methods and analytical approaches with respect to the characteristics of the species under scrutiny and resources availability.     

How accurate are single site transect data for monitoring butterfly trends? Spatial and temporal issues identified in monitoring Lasiommata megera

Multiple transect counts following Butterfly Monitoring Scheme (UKBMS) guidelines and Jolly–Seber estimates of population size were used to monitor the abundance of second generation Lasiommata megera on a single site in southern England. The two methods resulted in different patterns of emergence being detected. The proportion of the population (estimated by Jolly–Seber) recorded with transect counts depended on the time of day and weather with afternoon transect counts best recording the trend in abundance over the flight period, but even then counts recorded a variable fraction of the population (6.2–51.3%). Increasing the frequency with which transect counts are carried out per week reduced variation and increased the fit of transect counts to Jolly–Seber generated population estimates. However, indices of abundance generated from randomly selected transect counts for L. megera within sampling weeks varied 4-fold and indices for other butterfly species were also highly variable. For L. megera, transect count variability is attributed to non-representative placement of the transect route and changes in the behaviour and spatial distribution in relation to population size and season. We suggest that transect counts need to be fully validated before the data are used to monitor changes of butterfly populations at individual sites.     

Simultaneous confidence intervals for comparing biodiversity indices estimated from overdispersed count data

Diversity indices might be used to assess the impact of treatments on the relative abundance patterns in species communities. When several treatments are to be compared, simultaneous confidence intervals for the differences of diversity indices between treatments may be used. The simultaneous confidence interval methods described until now are either constructed or validated under the assumption of the multinomial distribution for the abundance counts. Motivated by four example data sets with background in agricultural and marine ecology, we focus on the situation when available replications show that the count data exhibit extra‐multinomial variability. Based on simulated overdispersed count data, we compare previously proposed methods assuming multinomial distribution, a method assuming normal distribution for the replicated observations of the diversity indices and three different bootstrap methods to construct simultaneous confidence intervals for multiple differences of Simpson and Shannon diversity indices. The focus of the simulation study is on comparisons to a control group. The severe failure of asymptotic multinomial methods in overdispersed settings is illustrated. Among the bootstrap methods, the widely known Westfall–Young method performs best for the Simpson index, while for the Shannon index, two methods based on stratified bootstrap and summed count data are preferable. The methods application is illustrated for an example.     

Sampling communities of ground-foraging ants: Pitfall catches compared with quadrat counts in an Australian tropical savanna

The relative abundances of ant species captured in pitfall traps was compared with those obtained by direct counts in quadrats at a savanna site in Kakadu National Park, Northern Territory. Two measures of abundance in traps were used, one based on total numbers of ants, the other on species frequency of occurrence. All species commonly recorded in quadrats were collected in traps, and their relative abundances were highly correlated on all occasions. Of the 20 most common species in quadrats, five occurred with a significantly different (in all cases lower) frequency in pitfall traps, but these species represented only 1.8–3.1% of total quadrat counts. Results from quadrats and pitfall traps were particularly similar (r > 0.8) when species-were classified into functional groups. Frequency data from traps may sometimes overestimate the abundance of widespread, solitary foraging species (e.g. ‘Chelaner’ and Tetramorium spp.) and underestimate species with large colony sizes (e.g. Iridomyrmex spp.). Data based on total numbers of ants in traps may be more prone to distortion arising from species differences in locomotor behaviour. Species counts in traps could be scaled to reduce these distortions. The finding that pitfall traps gave results comparable with those from quadrat counts provides support for the use of pitfall traps in studies of Australian ant communities in open habitats.     

Method for Sampling Meat Surfaces

A new method for sampling meat surfaces was developed. Bacterial counts of beef carcasses by the cotton swab technique and by the new method showed that the latter gave higher counts. These counts were closely correlated with data obtained by using the swab method. Advantages of the new method are its simplicity, rapidity, and adaptability to routine use on any type of carcass.     

Using time‐of‐detection to evaluate detectability assumptions in temporally replicated aural count indices: an example with Ring‐necked Pheasants

ABSTRACT The validity of treating counts as indices to abundance is based on the assumption that the expected detection probability, E(p), is constant over time or comparison groups or, more realistically, that variation in p is small relative to variation in population size that investigators seek to detect. Unfortunately, reliable estimates of E(p) and var(p) are lacking for most index methods. As a case study, we applied the time‐of‐detection method to temporally replicated (within season) aural counts of crowing male Ring‐necked Pheasants (Phasianus colchicus) at 18 sites in southern Minnesota in 2007 to evaluate the detectability assumptions. More specifically, we used the time‐of‐detection method to estimate E(p) and var(p), and then used these estimates in a Monte Carlo simulation to evaluate bias‐variance tradeoffs associated with adjusting count indices for imperfect detection. The estimated mean detection probability in our case study was 0.533 (SE = 0.030) and estimated spatial variation in E(p) was 0.081 (95% CI: 0.057–0.126). On average, both adjusted (for) and unadjusted counts of crowing males qualitatively described the simulated relationship between pheasant abundance and grassland abundance, but the bias‐variance tradeoff was smaller for adjusted counts (MSE = 0.003 vs. 0.045, respectively). Our case study supports the general recommendation to use, whenever feasible, formal population‐estimation procedures (e.g., mark‐recapture, distance sampling, double sampling) to account for imperfect detection. However, we caution that interpreting estimates of absolute abundance can be complicated, even if formal estimation methods are used. For example, the time‐of‐detection method was useful for evaluating detectability assumptions in our case study and the method could be used to adjust aural count indices for imperfect detection. Conversely, using the time‐of‐detection method to estimate absolute abundances in our case study was problematic because the biological populations and sampling coverage could not be clearly delineated. These estimation and inference challenges may also be important in other avian surveys that involve mobile species (whose home ranges may overlap several sampling sites), temporally replicated counts, and inexact sampling coverage.     

Developing indices of relative abundance for monitoring cave and ground wētā (Orthoptera) in southern beech forest,New Zealand

We sampled populations of forest-floor dwelling cave and ground wētā using footprint tracking tunnels and spotlight transect counts in southern beech forest, New Zealand. Samples were compared to estimates of wētā density based on mark–recapture estimates from 25?m² enclosures. Both activity indices captured variability in cave wētā in time and space, were strongly correlated with each other, and have the potential for monitoring cave wētā activity levels. Comparisons between indices and cave wētā density estimates were equivocal, as recapture rates were too low to calculate high-resolution density estimates. We also found that cave wētā counts had a curved relationship increasing with temperature, and a negative relationship with increasing shrub and woody debris cover. Based on these preliminary results, tracking tunnels could be a viable method of monitoring cave wētā as they appear more efficient than transect counts and are relatively inexpensive. However, further calibration trials are needed to determine if indices mirror robust population density estimates.     

Sampling method affects estimates of population density and reproductive maturity of summer‐form pistachio psyllids

Abstract

In this study, the effects of using three different sampling methods (yellow sticky traps, beating tray, organdy bags) on estimating the distribution through the canopy of the summer‐form adults of the pistachio psyllid (Agonoscena targionii (Licht.)) was investigated throughout the vegetative period. The traps or bags were hung at two heights in the tree canopy, 3.5 m above ground and 2 m above ground. Psyllids were counted weekly, and their sex determined. Adult abundance varied during the study, but all three sampling methods indicated that densities were higher in the upper canopy than the lower canopy. The percentage of females having mature eggs was lowest in late May, and highest in June and September. Sex ratios were highly variable among sampling dates and among sampling methods, at both canopy heights. Sex ratios in bags tended to be more female‐biased in the lower canopy than in the upper canopy. Sticky trap counts were significantly male biased, more so in the upper canopy than in the lower canopy. Mean sex ratios differed significantly between bag and sticky trap samples at both heights, but did not differ significantly between tray and bag samples at either height. In addition, for a given tray or bag count, sticky trap catch of males was larger than catch of females, and was larger in the upper canopy than in the lower canopy. Bag counts (psyllids per leaflet) strongly increased with tray count.

Of the three sampling methods described here, the bag samples provided direct estimates of psyllid numbers per leaflet, but this method is extremely time consuming. Bag counts (psyllids per leaflet) increased with tray count, suggesting that beat trays can be used to estimate absolute densities.     

Correlation of cell counts and indices in testicular FNAC with histology in male infertility.

OBJECTIVE: To evaluate the value of percentage cell counts and cell indices in testicular fine needle aspiration cytology (FNAC) in male infertility and their correlation with histologic categories as seen in open testicular biopsies. STUDY DESIGN: Differential cell counts were performed, and cell indices, including spermatic index, Sertoli cell index and sperm-Sertoli cell index, were calculated in testicular fine needle aspiration (FNA) smears in 30 azoospermic males whose open testicular biopsies were classified as normal spermatogenesis in 10 cases, maturation arrest in 5, hypospermatogenesis in 6, Sertoli cell only syndrome in 5 and tubular/peritubular sclerosis in 4. RESULTS: In normal spermatogenesis, FNA smears showed up to 40% Sertoli cells, and spermatozoa were the predominant spermatogenetic cell type. There was a progressive increase in Sertoli cell percentage and Sertoli cell index and reduction in spermatozon percentage, spermatic index and sperm-Sertoli cell index with increasing severity of reduction in spermatogenesis in different histologic categories. The differences between mean counts and indices in normal spermatogenesis and other histologic categories were statistically significant (P < .01). CONCLUSION: The percent cell counts and cell indices in testicular FNAC correlate with histologic categories and are useful in evaluating male infertility.