Abundance-diversity relationship as a unique signature of temporal scaling in the fossil record
Data files
Jun 06, 2024 version files 591.13 KB
-
README.md
2.29 KB
-
Table_S1.txt
430.83 KB
-
Table_S2.txt
112.30 KB
-
Table_S3.txt
9.61 KB
-
Table_S4.txt
31.80 KB
-
Table_S5.txt
4.30 KB
Jun 24, 2024 version files 593.32 KB
Abstract
Species diversity increases with the temporal grain of samples according to the species-time relationship, impacting paleoecological analyses because the temporal grain (time averaging) of fossil assemblages varies by several orders of magnitude. We predict a positive relation between total abundance and sample size-independent diversity (ADR) in fossil assemblages because an increase in time averaging, determined by a decreasing sediment accumulation, should increase abundance and depress species dominance. We demonstrate that, in contrast to negative ARDs of non-averaged living assemblages, the ARD of Holocene fossil assemblages is positive, unconditionally or when conditioned on the energy availability gradient. However, the positive fossil ADR disappears when conditioned on sediment accumulation, suggesting that ADR can be a signature of diversity scaling induced by variable time averaging. Conditioning ADR on sediment accumulation can identify and remove the scaling effect caused by time averaging, providing an avenue for unbiased biodiversity comparisons across space and time.
README: Abundance-diversity relationship as a unique signature of temporal scaling in the fossil record
https://doi.org/10.5061/dryad.fttdz0903
Description of the data and file structure
Table S1 – Diversity and abundance of living (non-averaged) molluscan assemblages in the northern Adriatic Sea, with source references. Data columns correspond to the reference, dataset ID (optional), latitude, longitude and water depth (m) of the assemblage, raw sample size and total abundance/m2. Diversity indices correspond to the effective number of species based on the PIE-based and Simpson index, the effective number of species based on the Shannon index, and evenness values based on the Pielou J and Bulla O. NA - data not available.
Table S2 – Diversity and abundance of fossil (time-averaged) molluscan assemblages collected in sediment cores in the northern Adriatic Sea. Individual fossil assemblages are in rows, data columns correspond to the core ID, sediment depth (cm), increment thickness (cm), systems tract (HST – highstand systems tract, MFZ – maximum flooding zone, TST – transgressive systems tract), facies association/environment, sediment accumulation rate (cm/y), sample size, fossil abundance/dm3, Shannon H, Gini-Simpson index, Probability of interspecific encounter, the effective number of species (Simpson and PIE), evenness values based on the Pielou J and Bulla O, and the location of the assemblages along the first NMDS axis.
Table S3 - Abundance-diversity relations in 30 geographic datasets with benthic foraminiferal living assemblages (LARD), with references. NA - data not available.
Table S4 – Abundance-diversity relations in 73 benthic foraminiferal fossil assemblages (FARD) in local stratigraphic series and in regional datasets, with references. NA - data not available.
Table S5 – Input chronological data for Bacon function, used in the computation of sediment accumulation rates (cm/y) that are available in Table S2.
Software Files Hosted by Zenodo
R_script_for_ADR_ms_-_figure_1.R - visualizing the species-time relationship in Figure 1
1. STR function is used for sample accumulation and the resulting computation of diversity under increasing time averaging
2. Neutral model /Hubbell 2001/ consists of metacommunity simulation /theta=20, metacommunity size=100000/ and local community dynamic sourced from metacommunity and
subjected to dispersal limitation /m=0.1/, using two lifespans /1 and 10 years/. Local community is set to n=300.
3. Function for non-neutral models /Thompson et al. 2020/, simulating 25 patches, with dispersal limitation, with two scenarios for niche width and two scenarios for competition intensity.
4. Model outputs - non-averaged living and time-averaged fossil assemblages - are subsampled to the same sample size, and plotted as average species-time relationship in the left column and as rank-abundance distributions not affected by time averaging and expected to occur when time averaging is 1000 years.
R_script_for_ADR_ms_-_figures2-6.R
Input data for figures 2-6 are in Tables S1-S5.
1. 3*3 plots are generated for the conceptual figure 2, using three initial types of abundance-diversity relation in
living assemblages (random, negative, positive) for 25 samples (LADR). LADR is randomly-generated, and each sample is subjected to random time averaging.
Assuming that the scaling exponent is 0.1 and lifespan is one year, diversity and abundance is predicted for the corresponding
time averaging, and abundance-diversity correlation is computed for fossil assemblages (FADR).
2. Objects are generated to keep information on diversity and abundance of living and fossil molluscan assemblages, their water depth
/indicator of energy or resource availability/, and sediment accumulation rate of fossil assemblages. These objects are used in lme models
and in figures.
3. Linear-mixed effect models and AIC values for models where variability in abundance and diversity is predicted by variability in sediment accumulation rate, and testing whether the positive abundance-diversity relation disappears if conditioned by energy availability and sediment accumulation rate
4. Reproduction of figures 3-5
5. Figure 6 is based on living and fossil assemblages with benthic foraminifers (Table S3 and Table S4)
6. Effects of variability in sediment accumulation on abundance-diversity relationship