Traditional models for counting species in ecology don't work well when dealing with rare species or repeated measurements. A new method was developed to handle this, using zero-inflated count data. By comparing different models, researchers found that the zero-inflated negative binomial model performed better than traditional models in most cases. However, all models had trouble predicting accurately at high species densities. It's important to choose the right model and predictors when estimating species abundance and occupancy.