Comparison of Ecological Communities: The Problem of Sample Representativeness

Cao, Yong; Williams, D. Dudley; Larsen, David P.

doi:10.2307/3100084

Cited by 27 publications

(43 citation statements)

References 60 publications

(81 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The effect of fragment area on species presence and richness is a matter of intense debate (Cao et al 2002;Oertli et al 2002). In our study area correlated with several other spatial attributes, as has been frequently found (Riffell et al 2001;Tsao 2000).…”

Section: Discussionmentioning

confidence: 85%

Habitat and landscape factors associated with neotropical waterbird occurrence and richness in wetland fragments

Guadagnin

Maltchik

2006

Biodivers Conserv

View full text Add to dashboard Cite

Habitat destruction alters the structure and spatial arrangement of remnants and creates a novel matrix, affecting the structure of biological communities. This study evaluated the contribution of patch, class and landscape level attributes of 42 wetland fragments and surrounding landscapes on the richness, abundance and presence of waterbird species. Twelve monthly censuses identified 55 species--ranging from 3 to 40 among fragments, and a total of 84,286 birds--ranging from 15 to 19,322. Wetland areas varied from 0.2 ha to 145.2 ha. The presence of 37 out of 42 species could be predicted from the structure of the fragments and the landscapes. The most important predictors were the wetland area (22 models), the rice field matrix permeability (13 models), the microhabitat richness (7 models), the wetland connectivity (5 models), and the wetland isolation (5 models). Discriminant analysis showed that fragments richer in species than expected from their areas, are found in landscapes with greater connectivity and matrix permeability than the species-poor fragments. The total area of rice fields was not related to the richness, abundance or presence of waterbirds in the wetlands fragments. The results suggest that the maintenance of the connectivity among fragments and the matrix permeability through rice-fields could be managed to foster the waterbird conservation at the landscape scale.

show abstract

Section: Discussionmentioning

confidence: 85%

Habitat and landscape factors associated with neotropical waterbird occurrence and richness in wetland fragments

Guadagnin

Maltchik

2006

Biodivers Conserv

View full text Add to dashboard Cite

show abstract

“…There was no significant difference in the proportion of taxa recorded in RBA samples at each site, although the significance level of the test was 0.06, and we expect that with a greater number of sites and replicates, this difference is likely to be significant. Indeed, a difference in the recovery of taxa from site to site was expected because similar sized samples may differentially represent the communities from which they are drawn when collected from different sites (Cao et al 2002).…”

Section: Site Comparisonmentioning

confidence: 99%

What do qualitative rapid assessment collections of macroinvertebrates represent? A comparison with extensive quantitative sampling

Gillies

Hose

Turak³

2008

Environ Monit Assess

View full text Add to dashboard Cite

It is a fundamental tenet of Rapid Biological Assessments (RBA) that the samples collected reflect the community from which they are drawn. As with any biological sampling, RBA collections are subject to sampling error resulting in the omission of some taxa. The aim of this study is to compare the composition of RBA samples with an estimate of community structure based on extensive quantitative sampling. We used logistic regression to explore the relationships between the frequency of a taxon being collected in an RBA sample and its biological and ecological traits, namely its abundance, distribution, body size and habit. RBA samples and quantitative estimates of community structure were made in riffles in the Kangaroo and Nepean Rivers, New South Wales, Australia. Single RBA samples may collect up to 63% of the taxa that are collected by extensive quantitative sampling at a site. The frequency of a taxon being recorded in an RBA sample was significantly and positively related to all traits tested indicating a bias in the collection methods towards large, abundant and widely distributed taxa. Accordingly, taxa missed by RBA sampling were generally small, narrowly distributed or rare. These findings enhance our understanding of what RBA samples represent, and the bias and source of errors associated with RBA sampling. This study also quantifies the utility of RBA methods for biodiversity assessment.

show abstract

“…from 1 to 62 sample units for caddisflies and 1 to 125 sections each 100 m long for fishes). Sampling unit size could be characterized either by the number of elementary units pooled or more generally by the area (caddisflies) or length (fish) of the pooled sample (Cao et al 2002), and here we adopt the second possibility. Functional diversity measures (FAD, APWD and MFAD), the number of functional units and species richness were then calculated for each size.…”

Section: Monotonicitymentioning

confidence: 99%

“…Initially, each sample unit (Surber sample for the caddisflies or capture for a 100 m river section for fishes) was considered separately. These elementary units were then pooled successively and randomly to generate data at increasing sample unit sizes according to Cao et al (2002) up to the point where all units were pooled into a single unit (i.e. from 1 to 62 sample units for caddisflies and 1 to 125 sections each 100 m long for fishes).…”

Section: Monotonicitymentioning

confidence: 99%

A measure for assessing functional diversity in ecological communities

2007

View full text Add to dashboard Cite

Functional diversity is regarded as a key in understanding the link between ecosystem function and biodiversity, but its measurement is rather problematic. The two widely used continuous measures are the dendrogram-based measure (DBM) and the functional attribute diversity (FAD). In contrast to DBM, FAD does not require the knowledge of the entire species pool before the analysis, and hence FAD is a more ideal tool for measuring functional diversity. However, the original form of FAD and its variants have several undesirable properties. Here, we suggest a modified FAD (denoted by MFAD), which-as illustrated by artificial and actual data sets-allows calculating functional diversity without violating the twinning and monotonicity criteria such that the number of species collected is compensated for. These requirements are met by replacing the original species by so-called functional species and then by dividing FAD by the number of functional units. Accordingly, MFAD measures the dispersion of species in the functional traits space so that MFAD values for different communities can directly be compared if the same set of functional traits is used. Finally, using data of two freshwater communities (caddisfly and riverine fish), we evaluate the change of species richness and functional diversity in relation to sampling effort (sample unit size). We found that functional diversity is a better and more reliable community descriptor than species richness in a sense that it converges to the maximum faster in the function of sampling effort.

show abstract

Comparison of Ecological Communities: The Problem of Sample Representativeness

Cited by 27 publications

References 60 publications

Habitat and landscape factors associated with neotropical waterbird occurrence and richness in wetland fragments

Habitat and landscape factors associated with neotropical waterbird occurrence and richness in wetland fragments

What do qualitative rapid assessment collections of macroinvertebrates represent? A comparison with extensive quantitative sampling

A measure for assessing functional diversity in ecological communities

Contact Info

Product

Resources

About