Dataset: Oyster spat elemental ratios
Deployment: Fodrie_SpatStrings

Means and standard errors for field-collected larval and settlement shell Sr:Ca and Ba:Ca ratios were calculated and plotted by site to assess spatial variation in geochemical signatures among collection sites. Signatures from larval shells were used to examine possible temperature and salinity gradients present among natal sites. Additionally, contour plots were used to explore how settler shell elemental concentrations of Mn, Sr, Ba, and Pb varied with temperature and salinity. Contour plots were created using the graphics package in R (version 3.0.3). Multiple regression models were then used to quantitatively assess the relationship between salinity, temperature and shell signatures in a natural environment. Because some collection sites did not produce any spat over a given collection period, spat were grouped only by site to increase the sample size and statistical power of our results. A logarithmic transformation of elemental ratio was used as the response variable.

Linear Discriminate Function Analysis (DFA) was used to examine spatial variability in settler shell geochemistry and to determine the viability of using geochmical fingerprints to assess connectivity in oyster populations. All 23 sites were used in preliminary DFAs, however the classification success was low, directing us toward independent examination of PS sites from the BBCS sites. Because adjacent sites often experienced similar temperature/salinity gradients, PS sites were then grouped by geographic quadrant within PS: Northwest (NW; WC, EH, StP), Northeast (NE; RD, HT), Southeast (SE; OK, CI, WB), and Southwest (SW; OR, SoP, SQ). BBCS sites were similarly broken up into 5 groups based on geomorphology and site location: Bay (JB, WM), Creek (WH, TC), Newport (NeU, NeM, NeL), North (NoU, NoM, NoL), and Sound (BoS, BaS). Jack-knifed classification matrices, without sample replacement, were compared to expected classification matrices, based on random chance, to assess classification success. Sites were additionally grouped based on similar temperature and salinity profiles, however classification success did not improve significantly over geomorphological quadrants so analysis did not continue with these groupings. Because natal origins are unknown and modeled dispersal pathways for the area (e.g. Haase et al. 2012) have not been empirically validated, no DFA was performed on larval signatures.

BCO-DMO Data Manager Processing Notes:
* added a conventional header with dataset name, PI name, version date
* modified parameter names to conform with BCO-DMO naming conventions
* rounded LR params to 5 decimal places
* replaced parameter name \s with _
* removed parameter name chars (LR) as data submitter indicated it is irrelevant and part of post-processing.
* removed #DIV/0! values
* removed "Name" column as data submitter indicated it is irrelevant and part of post-processing.
* Added Site Code, Lat, and Lon from site info dataset