Peer Review #1 of "Fundamental niche unfilling and potential invasion risk of the slider turtle Trachemys scripta (v0.1)"

1 Departamento de Ecología de la Biodiversidad, Instituto de Ecología, Universidad Nacional Autónoma de México, Ciudad Universitaria, Ciudad de México, México 2 Posgrado en Ciencias Biológicas, Universidad Nacional Autónoma de México, Coyoacán, Ciudad de México, México 3 Grupo de Ecología y Evolución de Vertebrados, Instituto de Biología, Universidad de Antioquia, Medellín, Colombia 4 American Museum of Natural History, New York, NY, United States of America

225 Australia) and around the world (areas that have not been invaded); and (ii) draw points 226 depicting the estimated realized niche of the native and non-native ranges. This procedure 227 randomizes the position of the kernel density surface of one of the two species within the 228 environmental space available for it, allowing to evaluate whether the realized niches (kernel 229 density surface) occupied in native and non-native ranges are more or less similar to the 230 distribution of similarities under a null model.

231
As a complementary test, we compared climatic covariance matrices (native versus non-232 native ranges) with a Common Principal Components Analysis (CPCA; Phillips & Arnold, 1999), 233 which compares two or more matrices considering their eigenvectors and eigenvalues in a 234 hierarchical fashion, to describe their structure in relation to the size, shape and orientation of 235 the matrices. We used the latter to establish principal components that were common between 236 matrices (i.e. progressive differences in shape, orientation and size) and to test hypotheses 237 about equality (identical eigenvectors and eigenvalues, i.e. identical size, shape and 238 orientation), proportionality (equal eigenvectors, but eigenvalues differing in a scalar amount, 239 i.e. same shape and orientation, different but proportional size), and unrelated structure 246 multivariate analysis to calibrate the niche and the occurrence density, performed with 247 ECOSPAT, and the estimation of niche overlap using Schoener's D metric (Schoener, 1970), an 248 index ranging from 0 (no overlap) to 1 (total overlap) (

313
The CPCA results showed that the best model that explained the differences between 314 the structure of the climatic matrices of the native and non-native realized niches was the one 315 with dissimilar eigenvectors and eigenvalues (AIC=42.0; Supplemental Table S2); that is, 316 matrices have different shape, orientation and size, indicating that even when there are similar 317 climatic conditions at both the native and non-native ranges, non-native individuals of T. scripta 318 occupy the available environment in a different fashion. In fact, comparing T. scripta's native 319 realized niche with its non-native one revealed incomplete niche stability (70%), while there was 320 evidence of expansion (30%) into climates available worldwide (Fig. 2).

321
Results of the six-dimensional climatic space analysis showed that the total hypervolume 322 overlaps by 29% and that the non-native realized niche is 28.5% greater than the native one 323 (Supplemental Fig. S1). The contribution of the climatic variables to the hypervolume differed 324 between range distributions: regarding the native, mean temperature of warmest quarter was 325 the most important variable, whereas for the non-native it was annual precipitation 326 (Supplemental Table S3). 342 Finally, the climatic suitability obtained for Mexico exhibits widespread areas with high potential 343 risk of invasion by T. scripta (Fig. 6).

371
Indeed, when compared with the range that T. scripta actually occupies, we evidence 372 that it is not experiencing the entire range of optimal conditions along its native distribution; 373 specifically when compared based on the warmest conditions, where suitable climate is 374 unoccupied along its higher spectrum of tolerances (Fig. 3). The latter might reflect biotic 375 features (absent from the climatic models), namely the fact that the distribution of its congener 376 species, Trachemys cataspila, starts at the southern limit of the slider turtle's, occupying warmer 377 environmental conditions; these two species do not occur sympatrically, likely due to 378 competition (Seidel, 2002). Interestingly, we also found that although non-native areas include  Manuscript to be reviewed    Comparison between the optimal temperature tolerances (thick black line) and the mean annual temperature (BIO01) ranges actually occupied by the slider turtles at their native (green line) and non-native (blue line) ranges. The red line represents the mean of the total temperature range occupied by T. scripta around the World (i.e., native+non-native ranges).
Occurrence probabilities are shown by dots for native (green) and non-native (blue) records.
(c) and (d) depict the comparison between the optimal temperature tolerances and the occupied ranges based on the mean temperature of the warmest quarter (BIO10) and the coldest quarter (BIO11), respectively. x-axis: temperature (degrees Celsius); y-axis: occurrence likelihood (scale 0 to 1) based on the cumulative frequency of records occurring at a particular temperature (graphics were performed in R).