Table 4 Hypothesis test outcomes from the best model predicting presence and abundance

From: Seed dispersal by Martu peoples promotes the distribution of native plants in arid Australia

 

Diversiflorum

Centrale

Eragrostis

Scaevola

Presence

Abundance

Presence

Abundance

Presence

Abundance

Presence

Abundance

Dispersal proxies

 Site distance

Strong negative**

Negative**

Weak negative under high land use**

No effect

Negative under high land use**

Negative near ephemeral water**

Weak positive

Weak positive

 Land use

Strong positive**

No effect

Weakly Positive under high use**

No effect

Weak Negative

No effect

No effect

No effect

 Water permanence

Strong positive**

Strong positive**

No effect

No effect

No effect*

Negative close to sites**

No effect

No effect

 Site type Major

Negative**

No effect

No effect

No effect

No effect*

No effect**

Negative

Positive

Engineering proxies

 Season Winter

No effect*

No effect

No effect

Negative

Negative

Positive**

No effect

No effect

 Fire frequency

No effect

No effect

Strong positive*

Weak negative

No effect

No effect

No effect

No effect

 TSF diversity

No effect*

No effect

No effect

No effect

No effect*

No effect

No effect

No effect

  1. No effect = coefficient not in best model: dropping coefficient reduced AIC value of the model during model selection and ranking. Positive effect  =  covariate increased presence or abundance. Negative effect  =  covariate reduced presence or abundance. *The predictor was significant in the global model but not in the best model. **Support for anthropogenic dispersal or anthropogenic engineering on presence or abundance.