As you correctly mentioned when assessing the certainty of evidence in a systematic review and meta-analysis of studies that are observational you should start the GRADE from low level of evidence. then you should assess whether could you level up the evidence certainty or not.

For example finding a big effect size is one of the reasons for leveling up the evidence. Also presence of publication bias against the founded Effect size can be interpreted that if there was no publication bias the effect size would have been greater so again you can level up the evidence.

However when performing systematic review and meta-analysis on the prevalence or incidence there is actually no effect size to conclude what I mentioned earlier. Therefore there will not be a reason for leveling up the GRADE certainty of evidence in systematic reviews and meta-analysis of the prevalence or incidence. Therefore the level of evidence in your study will result in low or very low level of evidence. if there are any downgrading criteria present.

Also regarding your question about the reason for observed publication bias, one probable reason is that articles that found lower or higher prevalence was not accepted to be published. another reason lies in the epidemiology of the disease or medical condition that you are researching on which means authors who have found high or low prevalence found it interesting and aimed to report it as an article which again can result in publication bias.