Discrepancies between Global Wood Density Database and data in BIOMASS

[fabeit]

Hi Maxime,

I have noticed discrepancies with the original data from https://datadryad.org/stash/dataset/doi:10.5061/dryad.234

In particular, some species in BIOMASS are only recognized at the genus level, so with "sp" in species, but this is not the case in the GWDD. Why these differences?

Also I have noticed that Canarium schweinfurtii is also listed under Canarium schweinfurthii, but the first one is the correct spelling. Maybe you can use the new WFO package to recheck the taxonomy? I am doing it already but cannot update the package.

[MaximeRM]

Hi Fabio,

thank for this info. Indeed, one occurrence of Canarium schweinfurthii in the DRYAD data is mispelled as Canarium schweinfurtii (the correct spelling is with a h). I will use the BIOMASS correctTaxo function to track any other potential error in the DRYAD database (even if a new large database will soon be published by the groupe of J. Chave).

Regarding your first question, can you please reformulate it (I do not understand the "so with "sp"") and provide an example?

Thank again

[MaximeRM]

Just for info, the tnrs service wrongly considers that Canarium schweinfurtii is correct (as well as Canarium schweinfurthii...). In fact, I realize that this species is misspelled in several websites (even in the Paris MNHN database..., https://science.mnhn.fr/taxon/species/canarium/schweinfurtii). Thus I cannot really modify it in the database without taking the risk that other people will misspell it without being able to identify the error. For the remaining taxa, since the last update of the correctTaxo, some taxa names have been changed so I will update the wdData in the package.

[fabeit]

Thank you Maxime for your response, according to WFO the spelling is without H http://www.worldfloraonline.org/taxon/wfo-0000583735;jsessionid=FA868E295B3FD14D402F92C29622B0ED
indeed plan taxonomy is complicated :-) . At the moment I am using World Flora Online and their R package to check for misspellings and synonyms.

In regards to the "sp" question, in the original DRYAD database all entries are identified at the species level. Instead, in the BIOMASS database there are 357 records which are only identified to the genus level and have "sp" in the species column. a few examples here

>          family        genus species    wd
>   1:    Fabaceae      Afzelia      sp 0.689
>   2:    Fabaceae      Afzelia      sp 0.670
>   3:   Meliaceae       Aglaia      sp 0.723
>   4:    Fabaceae      Albizia      sp 0.450
>   5:    Fabaceae      Albizia      sp 0.520
>  ---                                       
> 353: Monimiaceae      Wilkiea      sp 0.684
> 354:   Myrtaceae Xanthomyrtus      sp 0.870
> 355:  Annonaceae      Xylopia      sp 0.500
> 356:    Rutaceae  Zanthoxylum      sp 0.470
> 357:    Rutaceae  Zanthoxylum      sp 0.904

[fabeit]

Can I ask you a general question in regards to assigning wood density? I typically go from species > Genus > Family starting from the region of the plot data. I am working with central African species at the moment.

In the case where the species is not present in the wood density database African tropical region, would it be better to use the average genus from the African tropical region or use the matching species from another tropical region? Of course most of the species are only found in their region, but there are some common species across tropical regions.

I would appreciate any advice.

[fabeit]

just in case you are interested I found about 230 problematic taxa, but perhaps you have already corrected most of them. I'm glad I'm not a botanist :)
match wd species.xlsx

[MaximeRM]

Hi Fabio,

ok thank for spotting the "sp" problem in the dataset. I will put an issue to control for that in the next version, as well as to check your xls file.

Regarding the spelling of C. schein.... I guess that botanists do not agree as the CJB database, which is in my opinion the most updated checklist for African plants, spell it with a h... This explains why correctTaxo returns both as correct.

Just a though, If you intend to correct typos to match the dryad data in the BIOMASS package, I recommend you to use the correctTaxo function as it is this one that is applied to the dryad data in the package.

Regarding your question, I strongly recommend you to read those papers:

1. Chave, J., Muller-Landau, H. C., Baker, T. R., Easdale, T. A., Steege, H. T., & Webb, C. O. (2006). Regional and phylogenetic variation of wood density across 2456 neotropical tree species. Ecological applications, 16(6), 2356-2367.
2. Flores, O., & Coomes, D. A. (2011). Estimating the wood density of species for carbon stock assessments. Methods in Ecology and Evolution, 2(2), 214-220.

Looking at both papers, you will see that wood density is poorly conserved at the family level. Thus, I prefer to stop at the genus level and then assign the mean stand level estimate that may give a better estimate given that it partly account for the successional stage or the soil type etc. That said, this has been never tested and many people either use the Family or stand level without any clear background on what is the best strategy. For the biogeographic constraint, I personnally do not use this option anymore and prefer to rely on the taxonomy, rather than on the geography... You can find some background on that in the above-mentioned Chave et al. paper.

Best