Rachel,
Method you can utilize for this question would be either using Generalized Linear Regression (GLR)'s binary model (logistic regression) or using Forest-Based Classification and Regression (random forest). Since you have land-use types (categories), random forest would be a great fit as it can model explanatory variables as categories directly. Currently, GLR toolset only allow continuous variables as inputs.
A shortcoming of Exploratory Regression tool for this problem will be the binary prediction as it only supports predicting a target variable that is continuous. Also, the Logistic Regression tool does not work with categorical variables meaning you will need to encode your categories using methods such as one-hot encoding prior to regression.
In short, I recommend using Forest Based Classification and Regression in Train Mode to understand the drivers behind presence and absence. In this case, you might find variable importance plot particularly useful to understand drivers behind presence/absence.
Hope this helps,
Orhun
