Limits...
What ecological factors shape species-area curves in neutral models?

Cencini M, Pigolotti S, Muñoz MA - PLoS ONE (2012)

Bottom Line: While general features of species-area curves are quite universal across ecosystems, some quantitative aspects can change significantly.Within the framework of spatially explicit neutral models, here we scrutinize the effect of varying the local population size (i.e. the number of individuals per site) and the level of habitat saturation (allowing for empty sites).We conclude that species-area curves become shallower when the local population size increases, while habitat saturation, unless strongly violated, plays a marginal role.

View Article: PubMed Central - PubMed

Affiliation: Istituto dei Sistemi Complessi, Consiglio Nazionale delle Ricerche, Rome, Italy.

ABSTRACT
Understanding factors that shape biodiversity and species coexistence across scales is of utmost importance in ecology, both theoretically and for conservation policies. Species-area relationships (SARs), measuring how the number of observed species increases upon enlarging the sampled area, constitute a convenient tool for quantifying the spatial structure of biodiversity. While general features of species-area curves are quite universal across ecosystems, some quantitative aspects can change significantly. Several attempts have been made to link these variations to ecological forces. Within the framework of spatially explicit neutral models, here we scrutinize the effect of varying the local population size (i.e. the number of individuals per site) and the level of habitat saturation (allowing for empty sites). We conclude that species-area curves become shallower when the local population size increases, while habitat saturation, unless strongly violated, plays a marginal role. Our findings provide a plausible explanation of why SARs for microorganisms are flatter than those for larger organisms.

Show MeSH

Related in: MedlinePlus

Exponent z for the SSM as a function of  at fixed .The left panel shows the case  for  with NN-kernel. The shaded area indicates the value  (including the estimated error) of the exponent for the MVM with . For  statistical errors increase because a smaller number of realizations was used as simulations become very costly. The right panel shows the case  for  with NN-kernel. The shaded areas display the value of the exponent for the MVM with both the NN- and square-kernel (K = 7) and the SSM with square-kernel (K = 7) for M = 100 and  (i.e. ).
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3366919&req=5

pone-0038232-g003: Exponent z for the SSM as a function of at fixed .The left panel shows the case for with NN-kernel. The shaded area indicates the value (including the estimated error) of the exponent for the MVM with . For statistical errors increase because a smaller number of realizations was used as simulations become very costly. The right panel shows the case for with NN-kernel. The shaded areas display the value of the exponent for the MVM with both the NN- and square-kernel (K = 7) and the SSM with square-kernel (K = 7) for M = 100 and (i.e. ).

Mentions: Figure 2a shows the dependence of the exponent z on the speciation rate v ( for SSM) for the MVM and SSM; MCP was excluded because as seen in Fig. 1 no reasonable power-law range exists for close to . Let us start comparing the two models with NN dispersal. As for the case (Fig. 1), the exponents are different and the curves produced by the SSM are consistently shallower than those generated by the voter model in the explored range of v-values. In this figure we can see that the exponents for the SSM with NN-dispersal appear to be close to (but not coincident with) those of the MVM with the square-kernel (K = 7). However, when comparing the exponents of the SSM and MVM when the square-kernel (K = 7) is employed for both, we still observe that the former is shallower (see also Fig 3 and its discussion in the next section). Notice that increasing further K in the MVM does not produce further changes in the exponent [26], [27]. Therefore, as the comparison with the same dispersal kernel reveals, the decrease in the exponent z due to the increase of the local population size is a genuine effect. We also observe that the function z(v) is remarkably similar in the two models (independently of the dispersal kernel employed), as demonstrated in Fig. 2b where 1/z is shown as a function of . In particular, both models are fairly well described by the fitting formula [27].


What ecological factors shape species-area curves in neutral models?

Cencini M, Pigolotti S, Muñoz MA - PLoS ONE (2012)

Exponent z for the SSM as a function of  at fixed .The left panel shows the case  for  with NN-kernel. The shaded area indicates the value  (including the estimated error) of the exponent for the MVM with . For  statistical errors increase because a smaller number of realizations was used as simulations become very costly. The right panel shows the case  for  with NN-kernel. The shaded areas display the value of the exponent for the MVM with both the NN- and square-kernel (K = 7) and the SSM with square-kernel (K = 7) for M = 100 and  (i.e. ).
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3366919&req=5

pone-0038232-g003: Exponent z for the SSM as a function of at fixed .The left panel shows the case for with NN-kernel. The shaded area indicates the value (including the estimated error) of the exponent for the MVM with . For statistical errors increase because a smaller number of realizations was used as simulations become very costly. The right panel shows the case for with NN-kernel. The shaded areas display the value of the exponent for the MVM with both the NN- and square-kernel (K = 7) and the SSM with square-kernel (K = 7) for M = 100 and (i.e. ).
Mentions: Figure 2a shows the dependence of the exponent z on the speciation rate v ( for SSM) for the MVM and SSM; MCP was excluded because as seen in Fig. 1 no reasonable power-law range exists for close to . Let us start comparing the two models with NN dispersal. As for the case (Fig. 1), the exponents are different and the curves produced by the SSM are consistently shallower than those generated by the voter model in the explored range of v-values. In this figure we can see that the exponents for the SSM with NN-dispersal appear to be close to (but not coincident with) those of the MVM with the square-kernel (K = 7). However, when comparing the exponents of the SSM and MVM when the square-kernel (K = 7) is employed for both, we still observe that the former is shallower (see also Fig 3 and its discussion in the next section). Notice that increasing further K in the MVM does not produce further changes in the exponent [26], [27]. Therefore, as the comparison with the same dispersal kernel reveals, the decrease in the exponent z due to the increase of the local population size is a genuine effect. We also observe that the function z(v) is remarkably similar in the two models (independently of the dispersal kernel employed), as demonstrated in Fig. 2b where 1/z is shown as a function of . In particular, both models are fairly well described by the fitting formula [27].

Bottom Line: While general features of species-area curves are quite universal across ecosystems, some quantitative aspects can change significantly.Within the framework of spatially explicit neutral models, here we scrutinize the effect of varying the local population size (i.e. the number of individuals per site) and the level of habitat saturation (allowing for empty sites).We conclude that species-area curves become shallower when the local population size increases, while habitat saturation, unless strongly violated, plays a marginal role.

View Article: PubMed Central - PubMed

Affiliation: Istituto dei Sistemi Complessi, Consiglio Nazionale delle Ricerche, Rome, Italy.

ABSTRACT
Understanding factors that shape biodiversity and species coexistence across scales is of utmost importance in ecology, both theoretically and for conservation policies. Species-area relationships (SARs), measuring how the number of observed species increases upon enlarging the sampled area, constitute a convenient tool for quantifying the spatial structure of biodiversity. While general features of species-area curves are quite universal across ecosystems, some quantitative aspects can change significantly. Several attempts have been made to link these variations to ecological forces. Within the framework of spatially explicit neutral models, here we scrutinize the effect of varying the local population size (i.e. the number of individuals per site) and the level of habitat saturation (allowing for empty sites). We conclude that species-area curves become shallower when the local population size increases, while habitat saturation, unless strongly violated, plays a marginal role. Our findings provide a plausible explanation of why SARs for microorganisms are flatter than those for larger organisms.

Show MeSH
Related in: MedlinePlus