Extinct Pleistocene carnivores were diurnal and highly active - fresh off the press 🐯 Only BMR and diurnality are robust predictors of extinction and this also stands for extant species 🦨🐅🦝 With @johnalroy.bsky.social we use an exhaustive sample and account for phylogenetic and trait uncertainty.

Human impacts on large mammals went well beyond triggering late Quaternary mass extinctions. A new paper by Brook et al. showing that biogeographic patterns were erased by the spread of domesticated species:

royalsocietypublishing.org/doi/10.1098/...

A related paper is in press. Stay tuned.

Phylogenetic relatedness predicts spatial co-occurrence in Neotropical bats and birds. A new paper by Anikó Tóth with Kate Lyons, Drew Allen, and me. Includes Drew's great new method of inferring interaction processes based on co-occurrence data.

royalsocietypublishing.org/doi/10.1098/...

Hubbell's neutral model of biodiversity is founded on tree inventory data for Barro Colorado Island – which don't fit the model. A better one is the compound abundance distribution involving the geometric series. A preprint of the newly submitted BCI paper is here:

ecoevorxiv.org/repository/v...

Based on a cross-validation test, CEGS outperforms all other key distributions when fitted to the new paper's data set. Support is stronger in richer communities. When CEGS holds, all other richness estimators are inaccurate. Another completed manuscript soon to be submitted.

CEGS arises from a two-parameter population dynamics model assuming that recruitment rates are per-species and death rates are per-individual. CEGS fit to the Barro Colorado Island tree inventory is near-perfect, so nothing more needs to be assumed. Completed manuscript soon to be submitted.

Back that up. What exactly is wrong, how is it not addressed in the 16 section-long discussion, and how is my main argument affected? Take another look at Fig. 5A, for example. "A better alternative is the Poisson log-normal distribution" (O'Hara 2005), but assuming it invalidates H and D.

Shannon's H, Simpson's D, and Pielou's J are useless in ecology. Fitting data with a compound abundance distribution based on the geometric series is a better way to quantify variation – and to estimate species richness. New paper in Ecology Letters.

onlinelibrary.wiley.com/doi/10.1111/...

Um, they're a long way from sucking this year. The Mets might not have swept that series last week but they did beat the Cubs solidly (18 runs to 10).

Fisher's alpha goes down, not up, as more stems are randomly drawn from across the BCI plot (upper green line). But it goes up if stems are drawn in order going SW-to-NE across the plot (lower). Chao 1 rises steeply (red). CEGS doesn't (upper blue) while capturing the area signal (lower blue).

Methods matter! BCI tree diversity partitioned by size class (5 cm DBH running window). CEGS (blue): flat, then a sudden drop ~40 cm. Chao 1: lower, different, and higher variance. CBR/alpha/theta (orange/purple/pink): ~50% increase on the way up to ~25 cm. But diversity should always decline.

Likelihood differencing is here: the new release of the richness R package for diversity analysis includes an LD version of the CEGS function, which fits species abundance distributions and estimates richness. Check this out if you found my recent preprints interesting.

github.com/johnalroy/ri...

The BCI manuscript is now done. The log series and ZSM fit the overall 50 ha data poorly. They predict the 1 ha subplot data just as poorly (last post) and underestimate beta diversity (new graph). CEGS is of course best. Manuscript available upon request: comments would be very much appreciated.

CEGS wins again. Predictive test based on the 50 ha BCI tree plot dataset of Hubbell/Foster/Condit (thanks). CEGS fits to species counts in 1 ha plots better predict counts in 9 plots to the east (blue). PLN wins once (red). Log series loses every time. Likelihood ratio cutoff is 10:1.

Rarefaction is also dead – and CEGS gets the job done again. It recovers richness when common species are set aside or when samples are drawn from composite ecological communities. Rarefaction, Shannon's H, and Simpson's D are way off in both cases.

www.authorea.com/users/363764...

The article conflates fitting and description with "training" and "prediction". The figure shows a saturated descriptive model, not a predictive one. Cross-validation is barely mentioned and saturation not at all. Parsimonious models work: they make robust predictions about new data.

A preprint of a full paper explaining the CEGS distribution model and the way it estimates species richness is now available on EcoEvoRxiv:

doi.org/10.32942/X2W...

It draws comparisons with other models and estimators including the Poisson log normal, the log series + Fisher's alpha, and Chao 1.

A manuscript about the global consequences of the Late Pleistocene megafaunal mass extinctions, which is now in review.

ecoevorxiv.org/repository/v...

I have just submitted a short paper about the compound exponential-geometric series (CEGS). CEGS predicts count distributions and species richness with high accuracy. A preprint is available upon request.

No problem. It's CEGS. Here's a "subsample at random and estimate richness" analysis of the classic BCI data set of Condit et al. (1996). CEGS asymptotes almost immediately. OPR is much the same. ACE and Chao 1 are way off, consistent with your results. Great paper, I totally agree with it.

Australians suffer from the illusion that there's more to the news than Trump. Kind of.

Probably not. Among other things, they have to play the Mets six times.

Evenness is dead. Long live species richness.

If you want to use an index, try this R calculation. Where n is a list of species counts:

length(n) / mean((n %o% (n - 1))^0.5 / (n %o% n)^0.5)

advance.sagepub.com/users/363764...

Seriously, it's a sign of the ongoing apocalypse. Hill numbers are everywhere!