To detect the small island effect (SIE) and nestedness patterns of herpetofauna of the West Indies, we derived and updated data on the presence/absence of herpetofauna in this region from recently published reviews. We applied regression‐based analyses, including linear regression and piecewise regressions with two and three segments, to detect the SIE and then used the Akaike's information criterion (AIC) as a criterion to select the best model. We used the NODF (a nestedness metric based on overlap and decreasing fill) to quantify nestedness and employed two null models to determine significance. Moreover, a random sampling effort was made to infer about the degree of nestedness at portions of the entire community. We found piecewise regression with three segments performed best, suggesting the species–area relationships possess three different patterns that resulted from two area thresholds: a first one, delimiting the SIE, and a second one, delimiting evolutionary processes. We also found that taxa with lower resource requirement, higher dispersal ability, and stronger adaptation to the environment generally displayed lower corresponding threshold values, indicating superior taxonomic groups could earlier end the SIE period and start in situ speciation as the increase of island size. Moreover, the traditional two‐segment piecewise regression method may cause poor estimations for both slope and threshold value of the SIE. Therefore, we suggest previous SIE detection works that conducted by two‐segment piecewise regression method, ignoring the possibility of three segments, need to be reanalyzed. Antinestedness occurred in the entire system, whereas high degree of nestedness could still occur in portions within the region. Nestedness may still be applicable to conservation planning at portions even if it is antinested at the regional scale. However, nestedness may not be applicable to conservation planning at the regional scale even if nestedness does exist among sampling islands from a portion.

Aim Area thresholds, at which the form of the species–area relationship (SAR) changes abruptly, have played an important role in the theoretical framework of conservation biogeography and biodiversity research. The application of piecewise regressions has been advocated as a rigorous statistical technique to identify such thresholds within SARs, but a large variety of piecewise models remains untested. We explore the prevalence and number of thresholds in SARs and examine whether the currently widely used method for detecting the small island effect (SIE) is robust. Location Global. Taxon We consider all multicellular taxa based on the criteria of datasets selection. Methods We apply 15 regression models, including linear regression and piecewise regressions with two and three segments to 68 global island datasets that are sourced from the literature. Results The number of area thresholds in SARs varied among groups and correlated positively with area range of a studied system. Under the AIC or AIC c criterion, three‐segment piecewise models were more prevalent, whereas under the BIC criterion, two‐segment piecewise models were more prevalent. From the results of Aegean Sea isopods, West Indies herpetofauna, and Australian Islands mammals, we found evidence that the traditional criteria for detection of SIEs are not robust. Main conclusions Our study demonstrates that (a) to detect an SIE, the comparison should use as many models as possible, including not only variants with and without a left‐horizontal part, but also those with two and more segments; (b) naive use of the traditional two‐segment piecewise regressions may cause poor estimations of both slope and breakpoint values; (c) the number of thresholds increases with the area range of a studied system; (d) conservation biologists and applied ecologists should determine the number of area thresholds when estimating the precise species–area patterns and making management strategies in fragmented landscapes.

The small‐island effect (SIE) is an important pattern in the research fields of island biogeography and biodiversity science. Amphibians and reptiles, while playing important roles in ecosystems, are experiencing global declines. However, to date, no study has explicitly examined the generality and processes underlying the SIE in amphibians and reptiles. Here, we complied 105 global data sets to systematically evaluate the prevalence and underlying factors determining the occurrence of SIEs, area threshold (T) and difference in area threshold (ΔT) between amphibians and reptiles. We applied 27 species–area relationship (SAR) models to the 105 global data sets of amphibians and reptiles to test for the existence of SIEs. We obtained 12 island characteristics, 5 environmental variables, 2 anthropogenic influence variables and 8 species traits that are linked to species survival, which were then tested separately and in combination to examine their roles in determining the occurrence of SIEs, area threshold and ΔT between amphibians and reptiles. We found SIEs in 20 out of 45 (44%) archipelagos for amphibians and 35 out of 60 (58%) archipelagos for reptiles. The occurrence of SIEs was affected by the number of islands, mean annual precipitation and minimum range size for amphibians, whereas only by the number of islands for reptiles. Area thresholds of amphibians were significantly higher than those of reptiles within the same true island system. Area thresholds were affected by mean island area and mean annual temperature for amphibians, whereas by mean island area, temperature seasonality and minimum range size for reptiles. ΔT was affected only by island type. Our study demonstrates that the determinants of the occurrence of SIEs and area thresholds comprised both extrinsic and intrinsic variables but differed substantially between amphibians and reptiles.

To evaluate the regional biogeographical patterns of West Indian native and nonnative herpetofauna, we derived and updated data on the presence/absence of all herpetofauna in this region from the recently published reviews. We divided the records into 24 taxonomic groups and classified each species as native or nonnative at each locality. For each taxonomic group and in aggregate, we then assessed the following: (1) multiple species–area relationship (SAR) models; (2) C‐ and Z‐values, typically interpreted to represent insularity or dispersal ability; and (3) the average diversity of islands, among‐island heterogeneity, γ‐diversity, and the contribution of area effect toward explaining among‐island heterogeneity using additive diversity partitioning approach. We found the following: (1) SARs were best modeled using the Cumulative Weibull and Lomolino relationships; (2) the Cumulative Weibull and Lomolino regressions displayed both convex and sigmoid curves; and (3) the Cumulative Weibull regressions were more conservative than Lomolino at displaying sigmoid curves within the range of island size studied. The Z‐value of all herpetofauna was overestimated by Darlington (Zoogeography: The geographic distribution of animals, John Wiley, New York, 1957), and Z‐values were ranked: (1) native > nonnative; (2) reptiles > amphibians; (3) snake > lizard > frog > turtle > crocodilian; and (4) increased from lower‐ to higher‐level taxonomic groups. Additive diversity partitioning showed that area had a weaker effect on explaining the among‐island heterogeneity for nonnative species than for native species. Our findings imply that the flexibility of Cumulative Weibull and Lomolino has been underappreciated in the literature. Z‐value is an average of different slopes from different scales and could be artificially overestimated due to oversampling islands of intermediate to large size. Lower extinction rate, higher colonization, and more in situ speciation could contribute to high richness of native species on large islands, enlarging area effect on explaining the between‐island heterogeneity for native species, whereas economic isolation on large islands could decrease the predicted richness, lowering the area effect for nonnative species. For most of the small islands less affected by human activities, extinction and dispersal limitation are the primary processes producing low species richness pattern, which decreases the overall average diversity with a large among‐island heterogeneity corresponding to the high value of this region as a biodiversity hotspot.