InbreedingDepressionInPlants.xlsx

Inbreeding depression data associated with:

Alice A. Winn, Elizabeth Elle, Susan Kalisz, Pierre-Olivier Cheptou, Christopher G. Eckert,
Carol Goodwillie, Mark O. Johnston, David A. Moeller, Richard H. Ree, Risa D. Sargent, and Mario Vallejo-Marin.  2011.  
Analysis of inbreeding depression in mixed-mating plants provides evidence for selective interference and stable mixed mating.  Evolution 65(12):3339-3359.

Data were collected as part of the NESCent working group, The Paradox of mixed mating in flowering plants.  

The group searched for inbreeding depression data for populations of species for which outcrossing rates were available 
(see Goodwillie et al 2005, Ann. Rev. Ecol. Evol. Syst 36: 47-79).  The spreadsheet includes some data that were not used 
for the analyses reported in the paper. For example, if not enough life stages were measured by the original authors for us 
to calculate 3- or 4-stage inbreeding depression, we could not use the data for our analyses.  Our decision rules and a table 
that includes all data used in our analyses are included in our publication (Table 2).

Questions can be directed to Elizabeth Elle, Simon Fraser University (eelle@sfu.ca).

Column headings as follows.  Also see notes tab in the spreadsheet, and there is a tab for refs that includes all references compiled.

Genus

Species

Population:  location name used in original publication

Year: when crosses were completed, if performed on field-grown plants.  If seeds were collected in the field, 
then brought to a greenhouse and plants from the seeds were subsequently used as parents in the self- and out-crosses, 
we used the year of collection in this field rather than year of crosses.  

Sample	: N = natural environment, A = altered (e. g. trees in seed farms)

Environment: Location where fitness measurements were taken.  F = field, C = Common garden, G = Greenhouse, M = multiple (plants moved during lifetime)

Ref: original reference.  See reference tab.

Group of six columns labelled Design: Aspects of the experimental design of the original study
* Experimental detail: included for more complex designs, e.g. density was varied
* Design. Coded:  1 = female plant crossed to another plant and selfed.  2= mini-diallel.  Pairs of plants were reciprocally crossed, 
    each acting as both maternal and paternal plant.  3= mini-nested.  Two mothers nested in one father, and all of them also selfed.  4= unknown or other.
* Parents: F = field collected (level of outcrossing unknown), O = outcrossed (for one generation), S = selfed (one generation), 
     M = multigenerational (multiple generations of breeding prior to the self- and out-crosses used for the experiment), U = unknown
* #families: when indicated by authors, number of families (crosses) with offspring measured
* Pollen: S = single donor, M = mixed, U = unknown
* n/fam:  number of offspring measured per family (cross), if indicated by the authors

For each stage of inbreeding depression, we have groups of columns (usually in 3s) including all data measured by the authors.  
The stages are early/seeds, germination, survival, and fecundity.  For stages 1 and 4, many different variables were measured in the original studies, 
and so our group ranked the variables according to how we thought they were related to fitness and used the highest ranked variable to calculate inbreeding 
depression (later columns in the spreadsheet).  In the spreadsheet, the variables are listed in order of utility as we ranked them, so seed number per cross (column N)
 was considered a better estimate of fitness than proportion of total flowers that set fruit (column Q), which was better than weight of seeds (which could be total 
or individual seed weights; column T).  All reported data were included here regardless of how we ranked variables for our own analysis.

We report:  the average values from the paper for selfed (SELF column) and outcrossed (OUT column) offspring for each trait, 
followed by the relative fitness (REL W column) calculated in the standard manner (self/outcross).  If only the relative fitness 
was provided in the paper, we report that value.  Because relative fitness is always a proportion, we did not standardize the measurements 
of the different studies, but include only what was presented in the paper.

For stage 4 (Fecundity), columns include average SELF, OUT, and REL W as well as a Type" column, as follows:
*Fecundity could be SD (total seed production), FR (Fruit number) FL (Flower number), PFL (proportion flowering), OV (Ovule number).
*Male fitness (rarely measured) could be PP (pollen production), PV (Pollen viability), PPV (PP*PV) ,SS (Siring success), FL (flower number, 
    only if authors suggested this was related to pollen production and thus male function). 
*Veg size:  some estimate of vegetative biomass.  B (Biomass), H (height), L (Leaf number), V (Volume), LW (leaf width). 
 
Finally, several variables were calculated in the spreadsheet, included in the final columns:
Stage 1  Stage 4:  relative fitnesses repeated from earlier in the spreadsheet, using our ranking rules to determine which variables provided the 
best estimates of fitness.  That is, if multiple variables were measured within a stage, we used only the best estimate rather than averaging them.  
Other users of these data may of course decide to do things otherwise.

Inbreeding depression calculated as in Husband and Schemske 1996.  Evolution 50(1): 54-70.   
IBD is 1-(w1*w2*w3*w4), that is, 1 minus the product of the relative fitness at each stage.
Three-stage IBD includes stages 1, 2, and 4 (seed production, germination, and fecundity).  
Four-stage IBD also includes stage 3, survival, which was measured less often.  
Because we have formulas in this spreadsheet, cases where not all stages are measured come up as undefined values (#VALUE! in spreadsheet). 
 
Finally, we have a notes column with additional information.

Missing data are recorded as empty cells for all variables.
