Message-ID: <20191120183828.Horde.ODqg-bVGjjkuebziwJThc4u@webmail.uni-bremen.de>
Date: 2019-11-20T17:38:28Z
From: Tim Richter-Heitmann
Subject: How to subsample a community dataset
In-Reply-To: <CAFXQFgM_70uXuER+PoYbwN2mBJsds3DtMgdh+0hVXRsGQkFQGg@mail.gmail.com>
Hi Alexandre,
i am not sure if i understood you correctly, but if you want to
normalize your samples to a common sample size (e.g., to the lowest
amount
of observations across all localities), there is vegan's "rrarefy" for
this purpose:
https://www.rdocumentation.org/packages/vegan/versions/2.4-2/topics/rarefy
Note that this way of subsampling may remove some (rare) species
entirely, so the presence/absence matrix of a rarefied dataset may
(slightly) differ
from the one you are working with now.
If you happen to work with microbial data, there are plenty more ways
to normalize your data set.
Hope this helps, Tim
Zitat von "Alexandre F. Souza" <alexsouza.cb.ufrn.br at gmail.com>:
> Hi everyone,
>
> I have two community matrices, one which is presence-absence and the other
> which contains abundances of hundreds of species in hundreds of localities.
> Because localities had different sizes, I would like to standardize the
> number of species per locality before performing beta-diversity
> compositional analyses. Are you aware of any packages or functions that
> perform this task?
>
> Thank you very much in advance,
>
> Sincerely,
>
> Alexandre
--
Tim Richter-Heitmann
Universit?t Bremen