This file READ_ME_Performance_Surfaces.txt was generated on 30/03/2024

GENERAL INFORMATION:

1. Title of Dataset: Pollinator and flower morphology interact to influence pollen receipt
2. Author Information:
3. Date of data collection:
   March and April 2015 to 2024
4. Field locations of data collection:
   Main experimental sites:
   A. Barrydale, Western Cape, South Africa (-33.935997°S, 20.679453°E)
   B. Swartberg, Western Cape, South Africa (-33.362594°S, 22.068132°E)
   C. Nuwekloof Pass, Baviaanskloof, Eastern Cape, South africa (-33.511291°S, 23.642786°E)
   D. Swellendam, Western Cape, South Africa (-34.184574°S, 20.291919°E)
5. Funding sources:

DATA AND FILE OVERVIEW

1. Description of dataset

Our dataset investigates the interaction between flower and pollinator morphology in influencing pollen receipt surfaces in two geophytes, Nerine humilis and Tritoniopsis revoluta.

Data in these files have primarily been collected from our four field sites that represented a main 'native' site of each of the taxa, we also introduced variation from
close populations in order to increased variance in each population. This allowed us to more easily determine the form and strength of selection.

2. File list

File 1
Name:1. Proboscis_Measurements.csv
Description: Proboscis measurements of all insects from our experimental populations.

File 2
Name:2. Plant_Trait_Variation.csv
Description: Floral trait measurements associated with the standing variation and increased variation for each species, from each population.

File 3
Name:3. Swellendam_Full_Datasetcsv
Description: Full dataset associated with single visits for the short style locality, Swellendam. This dataset was specifically used to generate performance surfaces and selection coefficients for the short style Nerine humilis.

File 4
Name:4. Baviaanskloof_Full_Dataset.csv
Description: Full dataset associated with single visits for the long style locality, Baviaanskloof. This dataset was specifically used to generate performance surfaces and selection coefficients for the long style Nerine humilis.

File 5
Name:5. Barrydale_Full_Dataset.csv
Description: Full dataset associated with single visits for the long tube locality, Barrydale. This dataset was specifically used to generate performance surfaces and selection coefficients for the long tubed Tritoniopsis revoluta.

File 6
Name:6. Swartberg_Full_Dataset.csv
Description: Full dataset associated with single visits for the short tubed locality, Swartberg. This dataset was specifically used to generate performance surfaces and selection coefficients for the short tubed Tritoniopsis revoluta.

File 7
Name:6. Swartberg_Full_DatasetR60.csv
Description: Full dataset associated with single visits for the short tubed locality, Swartberg. This dataset was specifically used to generate selection coefficients for short proboscid insects associated with the short tubed Tritoniopsis revoluta. We removed one outlier from this particular dataset which gave a significant linear coefficient when present (see Figure 2F), as in the "Swartberg_Full_Dataset.csv".
This dataset is the same as file 6, but without row 60.

3. Methodological information

(A) Morphological data:
Files 1 and 2 containing raw data associated with calculating descriptive statistics, specifically means and standard deviations. File 1 is associated with proboscis measurements and File 2 contains data associated with local and increased variation at our experimental localities.

(B) Performance surfaces and coefficients
Files 3-7 contains data associated with single visits based on native and increased variance. This raw data was used to generate pollen receipt surfaces and calculate selection coefficients.

DATA SPECIFIC INFORMATION FOR EACH FILE

FILE: 1. Proboscis_Measurements.csv

(A) Number of variables: 6

(B) Number of rows: 62 (including headers)

(C) Variable list:

• Date: Date insect was caught.
• Pollinator: Pollinator species
• Locality: Locality where caught
• Caught visiting: Flower pollinator was caught on
• Extended proboscis length (mm): The extended proboscis length measurement of each insect
• Functional Body Length (mm): Body length measurement associated with affecting pollination

(D) Missing data codes: 37 cells under "Functional Body Length (mm)" starting from lines 29 to 65

(E) Abbreviations used: None

(F) Other relevant information: None

FILE: 2. Plant_Trait_Variation.csv

(A) Number of variables: 4

(B) Number of rows: 441 (including headers)

(C) Variable list:

• Species: Taxon
• Locality: Locality of taxon sampled
• Natural Variation: Natural variation associated with the trait collected for each taxon
• Extended Variation: Increased variation associated with the trait collected for each taxon.

(D) Missing data codes: None

(E) Abbreviations used: None

(F) Other relevant information: Data is only missing because of unequal sampling between different taxa for natural versus extended variation columns

FILE: 3. Swellendam_Full_Dataset.csv

(A) Number of variables: 4

(B) Number of rows: 146 (including headers)

(C) Variable list:

• Source: Source of trait variation used in single visit experiment
• Pollinator: Pollinator taxon visiting flower
• Style length (mm): Style length of flower used in single visit experiment
• Pollen: Pollen grains deposited per single visit

(D) Missing data codes: None

(E) Abbreviations used: None

(F) Other relevant information: None

FILE: 4. Baviaanskloof_Full_Dataset.csv

(A) Number of variables: 4

(B) Number of rows: 70 (including headers)

(C) Variable list:

• Source: Source of trait variation used in single visit experiment
• Pollinator_species: Pollinator taxon visiting flower
• Style length (mm): Style length of flower used in single visit experiment
• Pollen: Pollen grains deposited per single visit

(D) Missing data codes: None

(E) Abbreviations used: None

(F) Other relevant information: None

FILE: 5. Barrydale_Full_Dataset.csv

(A) Number of columns : 5

(B) Number of rows: 55 (including headers)

(C) Variable list:

• Functional_pollinator_type:Describes the functional group of pollinators associated with a single visit
• pollinator: Pollinator species
• tube_length_(mm): Tube length of flower used in single visit
• pollen: number of pollen grains recorded associated with single visit event
• visitor: code for functional pollinator group used in "by" command for generalized additive models

(D) Missing data codes: None

(E) Abbreviations used:
[SPI] - Short probscid insect
[LTF] - Long proboscid fly

(F) Other relevant information: visitor code is directly linked to functional pollinator groups defined in (E) directly above

FILE: 6. Swartberg_Full_Dataset.csv

(A) Number of columns : 5

(B) Number of rows: 100 (including headers)

(C) Variable list:

• Functional_pollinator_type:Describes the functional group of pollinators associated with a single visit
• Pollinator: Pollinator species
• tube_length_(mm): Tube length of flower used in single visit
• pollen: number of pollen grains recorded associated with single visit event
• visitor: code for functional pollinator group used in "by" command for generalized additive models

(D) Missing data codes: None

(E) Abbreviations used:

[LTF] - Long proboscid fly
[SPI] - Short probscid insect

(F) Other relevant information: visitor code is directly linked to functional pollinator groups defined in (E) directly above

FILE: 7. Swartberg_Full_DatasetR60.csv

(A) Number of columns: 5

(B) Number of rows: 99 (including headers)

(C) Variable list:

• Functional_pollinator_type: Describes the functional group of pollinators associated with a single visit
• Pollinator: Pollinator species
• tube_length_(mm): Tube length of flower used in single visit
• pollen: number of pollen grains recorded associated with single visit event
• visitor: code for functional pollinator group used in "by" command for generalized additive models

(D) Missing data codes: None

(E) Abbreviations used:
[LTF] - Long proboscid fly
[SPI] - Short probscid insect

(F) Other relevant information: visitor code is directly linked to functional pollinator groups defined in (E) directly above.

Important! Row 60 was removed in this dataset associated with the above dataset "Swartberg_Full_Dataset.csv" because of an outlier that gave a significant postive linear relationship.