Modeling species-abundance relationships in multi-species collections

S. Peng; Z. Yin; H. Ren; Q. Guo

Home/Archive/USGS

USGS · 1001825

Modeling species-abundance relationships in multi-species collections

S. Peng·Z. Yin·H. Ren·Q. Guo

View original source ↗Export metadata ↓

Abstract

Species-abundance relationship is one of the most fundamental aspects of community ecology. Since Motomura first developed the geometric series model to describe the feature of community structure, ecologists have developed many other models to fit the species-abundance data in communities. These models can be classified into empirical and theoretical ones, including (1) statistical models, i.e., negative binomial distribution (and its extension), log-series distribution (and its extension), geometric distribution, lognormal distribution, Poisson-lognormal distribution, (2) niche models, i.e., geometric series, broken stick, overlapping niche, particulate niche, random assortment, dominance pre-emption, dominance decay, random fraction, weighted random fraction, composite niche, Zipf or Zipf-Mandelbrot model, and (3) dynamic models describing community dynamics and restrictive function of environment on community. These models have different characteristics and fit species-abundance data in various communities or collections. Among them, log-series distribution, lognormal distribution, geometric series, and broken stick model have been most widely used.

Explore related subjects

Acta Ecologica Sinica Article

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

S. Peng, Z. Yin, H. Ren, Q. Guo. 2003. Modeling species-abundance relationships in multi-species collections. https://pubs.usgs.gov/publication/1001825

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2008

Modeling potential habitats for alien species Dreissena polymorpha in continental USA

The effective measure to minimize the damage of invasive species is to block the potential invasive species to enter into suitable areas. 1864 occurrence points with GPS coordinates and 34 environmental variables from Daymet datasets were gathered, and 4 modeling methods, i.e., Logistic Regression (LR), Classification and Regression Trees (CART), Genetic Algorithm for Rule-Set Prediction (GARP), and maximum entropy method (Maxent), were introduced to generate potential geographic distributions for invasive species Dreissena polymorpha in Continental USA. Then 3 statistical criteria of the area under the Receiver Operating Characteristic curve (AUC), Pearson correlation (COR) and Kappa value were calculated to evaluate the performance of the models, followed by analyses on major contribution variables. Results showed that in terms of the 3 statistical criteria, the prediction results of the 4 ecological niche models were either excellent or outstanding, in which Maxent outperformed the others in 3 aspects of predicting current distribution habitats, selecting major contribution factors, and quantifying the influence of environmental variables on habitats. Distance to water, elevation, frequency of precipitation and solar radiation were 4 environmental forcing factors. The method suggested in the paper can have some reference meaning for modeling habitats of alien species in China and provide a direction to prevent Mytilopsis sallei on the Chinese coast line.

Acta Ecologica Sinica↗

USGS2007

Natural restoration of degraded rangeland ecosystem in Heshan hilly land

This study examined the 20-yr trend of natural restoration of a degraded rangeland ecosystem after disturbance in Heshan hilly land. The results showed that herbs and shrubs were the dominant plants in the community and only a small number of the shade-intolerant tree species had invaded, showing the characteristics of assembly of pioneer communities. The organic matter content, soluble nitrogen, available phosphorus and available potassium had recovered to the level of the local climax community. Part of the ecological functions such as water and soil conservation had also recovered. While the functions of water and soil conservation recovered first, more time was needed for productivity and other functions to completely recover, suggesting the idiosyncratic nature of different ecosystem variables in response to time and microclimate change. Particularly, nutrient cycling recovered very slowly by natural restoration and artificial plantation may be necessary to accelerate the restoration process. ?? 2007 Ecological Society of China.

Acta Ecologica Sinica↗

USGS2007

Spatial and temporal patterns of net primary productivity in the duration of 1981-2000 in Guangdong, China

The knowledge of net primary production (NPP) dynamics at regional scale will help to understand terrestrial carbon cycling, especially with respect to land use and global climate change. Guangdong province has high plant growth potential because of plenty of light, heat, and water resources in this region. Forest coverage increased significantly from less than 30% in the early l980s to approximately 60% in 2000 owing to the launching of the "Greening Guangdong in 10 years", a provincial afforestation and reforestation project started in 1985. Meanwhile, economy growth has been fast in Guangdong province during the past 20 years. Long-term spatial and temporal NPP dynamics in Guangdong province are not well-known. To fill this knowledge gap, the spatial and temporal patterns of annual NPP from 1981 to 2000, derived from the global production efficiency model (GLO-PEM), were analyzed in this study. NPP patterns were compared at three spatial scales (i. e. , province, region, and city) and among three major forest types (i. e. , broadleaf, coniferous, and mixed). The results showed that for the entire province annual NPP varied between (1360 ±431) and (1626 ± 471) g/(m^2•a), with a mean value of (1480 ±407)g/(m^2•a). NPP increased to the maximum value (1534 ±121 g/(m^2•a)) in late 1980s (1986~1990) while decreased in early 1990s (1991~1995), and then recovered slightly in late 1990s (1996~2000). NPP differed distinctly across geographic regions, with the highest in the southwest coastal region, followed by the southeast coastal region, and the lowest in the inner land region. The differences were probably caused by vegetation composition, heat and water resources, and the distribution of the cropland. NPP dynamics of 21 cities were divided into three types. NPP kept stable in 12 cities including Shaoguan, Qingyuan, and Meizhou etc. NPP increased in Chaozhou, Shanwei, Zhanjiang and Jieyang, and decreased significantly (p<0.05) in 5 cities (i. e. , Foshan, Zhongshan, Shenzhen, Dongguan and Zhuhai). The decrease of NPP in these 5 cites can partly be explained by land cover and land use changes (e. g. , urbanization) driven by the economy development in the Pearl River Delta. NPP varied among the three major forest types. The mixed forest had the highest NPP, followed by the broadleaf forest and the conifer forest. Long-term mean NPP were (1364 ± 390) g/(m^2•a), (1391 ± 372) g/(m^2•a), and (1704 ± 450) g/(m^2•a) in the conifer, the broadleaf, and the mixed forest, respectively.

Acta Ecologica Sinica↗