Growth and miniaturization among alvarezsauroid dinosaurs

Abstract

(Current Biology 31, 3687–3693.e1–e5; August 23, 2021) After publication, two errors were discovered in the original paper. First, we neglected to log-transform body mass information when estimating rate heterogeneity. In the “Phylogenetic body size macroevolutionary analysis” section of the Method details, an additional point is needed in the third paragraph: log-transforming the body size measurements is essential in ancestral reconstruction. We have added a sentence to clarify this point. In the “Body size evolution” section of the Results, there is an error in the sentence “Our test for rate heterogeneity of body size as a trait on our phylogenetic tree suggested a rate decrease after node 27 (the common ancestor of Qiupanykus zhangi and its sister group; colored in blue in Figure 3A; Figure S3).” In the reanalyzed dataset, using log-transformed data, with reasonable AIC thresholds, we return no inference of rate shifts. We have changed this sentence to “Our test for rate heterogeneity of body size as a trait on our phylogenetic tree suggested no distinct rate shift (Figure S3).” The different outputs from log- and non-log-transformed datasets are mainly caused by the very tiny size of late-branching alvarezsauroids. The evolutionary rate did decrease in absolute values, but when considering their tiny body size, the changes are insignificant. Finally, a sentence in the Discussion, “Upon reaching the inferred optimal small body mass of less than 5 kg (Figure S3; after node 27 in the phylogenetic tree), later alvarezsauroids (including Qiupanykus and its sister group) showed relatively little innovation in body size,” should be deleted because there is no strong evidence supporting a rate shift there. In our macroevolutionary analyses, we ran the ancestral state reconstruction of body mass evolution on both log-transformed (log kg) and original (kg) mass data, but considering the need for log-transformed data, we have replaced Figure S3B based on the log-transformed dataset. The only difference between these two panels is that the mass data are log-transformed. Second, we selected a branch-length calculation method arbitrarily, and other methods return preferences for different models. In our estimation of branch lengths for the alvarezsauroid tree, we arbitrarily chose the “equal” method of the strap package, which indeed returns a preference for our data fitting an OU model (among the single-regime models we tested). However, other branch-length calculation methods return best fits to other models. For example, using the “mbl” argument to calculate branch lengths using the “palaeotree” package returns a preference for a BM model, among the single-regime models we tested. Because the fit of a given evolutionary model to our data is not robust to arbitrary differences in branch-length calculations, we can no longer robustly support selective pressure for smaller body sizes. However, our qualitative observations about the decrease in alvarezsauroid body mass remain unaffected. Increased samples, use of a variety of branch-length scaling methods, and the evaluation of multi-regime models of evolution on our observations will undoubtedly improve our ability to understand the processes by which alvarezsauroid body mass decreased. These errors do not negate our observations about patterns of body size evolution among alvarezsauroids. We are grateful to Roger Benson (University of Oxford) for pointing out the issues of concern. The authors apologize for any confusion.Figure S3BAlvarezsauroid phylogenetic tree with geological time scale, ancestral state reconstruction of body mass, trait rate heterogeneity and phylogenetic lineage diversity through time. Related to Figure 3. (original)View Large Image Figure ViewerDownload Hi-res image Download (PPT) Growth and miniaturization among alvarezsauroid dinosaursQin et al.Current BiologyJuly 6, 2021In BriefQin et al. report their broadly sampled osteohistological analyses and quantitative body size macroevolution reconstruction for a bizarre theropod dinosaur group, the Alvarezsauroidea. Highly variable growth strategies and quick miniaturization are found in alvarezsauroids, possibly associated with a dietary switch to feeding on social insects. Full-Text PDF Open Archive