I have zero-inflated data in my research about arthropods abundance. During my analysis with NMDS, the Bray-Curtis analysis can not handle for 0 data. Which one is better?

1. I am using R for analysis, and then I can use the 0 elimination function. Is it OK to do?

2. Or should I transform it first? what kind of transformation for 0 data? as if I transform log(x+1) it will be keep 0.

Thank you for helping me with the answer and discussion

Best regards,

Dwi Harya