A short note today.
Neil Brocklehurst was kind enough to run the large reptile tree through some testing that revealed weaknesses at several nodes. These weaknesses were caused by inappropriate scores. For the most part these have now been repaired.
The tree was and is still fully resolved (except at one node where a skull-only taxon nests with a skull-less taxon). Bootstrap scores are higher. Thank you, Neil.
Many scores were corrected with new data. Some were corrected with better judgement. I reinflated a crushed skull here. Several officially unnamed or misnamed taxa moved slightly. Saltoposuchus moved the most from the base of the proto-dinos to the middle of the basal crocs where it more closely resembles its new sister taxa. That taxon always seemed to be an oddball before.
Many of these mobile taxa are known from drawings only, some decades old. I removed multiples specimens of Mesenosaurus, leaving the holotype in place. When taxa resemble each other too much (fewer than 3-4 steps in difference) that leads to loss of resolution. The base of the synapsids close to Mesenosaurus was also rescored. You can see the complete tree here. Other than these few changes, the basic tree topology remains largely unchanged.
It’s always refreshing to find errors and make repairs. And, thank you again, Neil.