Doing some analysis of the gender and ethnicity breakdown of different ethnobotanical knoweldge in Pakistan for the MS Traditional use of medicinal plants among the Kalasha, Ismaeli and Sunni ethnic groups in Khyber Pakhtunkhwa provine, Pakistan.

Crude Descriptive Clustering (uses lumped within species)

(Ignoring frequency of reports produces clusters by ethnicity, Kalasha are outliers.)

Just the tree

Tree+species matrix

Taking into account frequency give Ismaeli as weak outliers

Just the tree

Tree with significance values

## Bootstrap (r = 0.49)... Done.
## Bootstrap (r = 0.6)... Done.
## Bootstrap (r = 0.69)... Done.
## Bootstrap (r = 0.8)... Done.
## Bootstrap (r = 0.89)... Done.
## Bootstrap (r = 1.0)... Done.
## Bootstrap (r = 1.09)... Done.
## Bootstrap (r = 1.2)... Done.
## Bootstrap (r = 1.29)... Done.
## Bootstrap (r = 1.4)... Done.

Tree with species matrix (colours represent %age of informants within category who reported this species)

Conclusions at this point

At least by some measures of knowlege-distance, ethnicities cluster together.

At the rough level (ignoring plant part and use); there are no gender similarities that cut across ethnic differences. I suspect this could change if, e.g., we coded out uses.

There is only a small group of species with gender differences within ethnicicity: 1. Isatis 2. Morchella 3. Withania 4. Diospyros

A more detailed look at the data (broken out by use)

Code Ethnobotanical Category
C1 Blood & Circulatory System
C2 Dental
C3 Digestive System
C4 Endocrine System
C5 Reproductive Systems
C6 Urinary System
C7 Immune System
C8 Metabolic system, fevers and nutrition
C9 Muscular-skeletal Systems
C10 Nervous System and Mental Health
C11 Respiratory System
C12 Sensory Systems
C13 Skin and related tissues
C14 Infections, parasites and fungi
C15 Inflammation and pain
C16 Poisons and antidotes
C17 Zoopharmacognosy
C18 ‘Humors’
C19 Belief Systems
C20 General, nonspeicfic
C21 Other
C22 Custom category: wounds

Updated Venn Diagram is already a little more interesting

Download the pdf

Tree including plant uses

Species tree

Tree with species matrix (colours represent %age of informants within category who reported this species)

Ordination

Ordination at the the presence-absence level shows a similar thing as the Venn Diagram (just uglier)

## Run 0 stress 9.415094e-05 
## Run 1 stress 0.03624801 
## Run 2 stress 9.407555e-05 
## ... New best solution
## ... procrustes: rmse 0.0002128876  max resid 0.0002732872 
## *** Solution reached
## Warning in metaMDS(p, distance = "binomial"): Stress is (nearly) zero - you
## may have insufficient data

Oridnation using the number of reports is a bit more complex, but is very similar to the tree+grid above (just uglier)

## Run 0 stress 0.09979933 
## Run 1 stress 0.1380712 
## Run 2 stress 0.07994761 
## ... New best solution
## ... procrustes: rmse 0.2365892  max resid 0.4218658 
## Run 3 stress 0.07994761 
## ... New best solution
## ... procrustes: rmse 2.902687e-06  max resid 4.546932e-06 
## *** Solution reached