Assessing Predictive Ability of Species-Area Relationship Models
| dc.contributor.advisor | Houlahan, Jeff | |
| dc.contributor.author | Yoo, Philip J. | |
| dc.date.accessioned | 2024-09-05T12:59:18Z | |
| dc.date.available | 2024-09-05T12:59:18Z | |
| dc.date.issued | 2022-04 | |
| dc.description.abstract | The species-area relationship (SAR) is a well-established concept but there is still limited understanding of its predictive ability. I investigated the predictive ability of four SAR models (species/area, log-species/log-area, species/log-area, and log-species/area) to new data using linear regression models. I collected 84 SAR datasets and broke them into training and test sets. For each training set I estimated the slope and intercept for each of the 4 SAR models and then used these estimates to predict species richness in each of the 84 test sets. The predictive ability of SAR models for 68 of the 84 was more accurate than the mean. Whether I was assessing the transferability of SAR models in space (i.e., a model built using data from one geographical location in order to predict onto another location) or non-spatially (i.e., model built using randomized sites to predict richness to sites nearby) the predictive ability was similar. | |
| dc.format.extent | ii, 21 | |
| dc.format.medium | electronic | |
| dc.identifier.uri | https://unbscholar.lib.unb.ca/handle/1882/38095 | |
| dc.language.iso | en | |
| dc.publisher | University of New Brunswick | |
| dc.rights | http://purl.org/coar/access_right/c_abf2 | |
| dc.subject.discipline | Biology | |
| dc.title | Assessing Predictive Ability of Species-Area Relationship Models | |
| dc.type | bachelor thesis | |
| oaire.license.condition | http://creativecommons.org/licenses/by/4.0/ | |
| thesis.degree.discipline | Biology | |
| thesis.degree.grantor | University of New Brunswick | |
| thesis.degree.level | bachelors | |
| thesis.degree.name | B.Sc. |