Authors: Santangeli, Andrea; Antão, Laura;Weigel, Benjamin;Vanhatalo, Jarno;Hällfors, Maria;Kaarlejärvi, Elina;Ovaskainen, Otso;Roslin, Tomas;Saastamoinen, Marjo;

Most studies evaluating impacts of conservation, e.g. protected areas, are taxonomically restricted, with a large bias towards vertebrates. Here, we apply a robust matching of protected and unprotected sites, and model species occurrence trends over several decades through joint species distribution modeling using long-term systematic survey data on forest understory plants, phytoplankton from lakes, birds and mammals collected across a whole country, Finland. Effect of protected areas (PAs) on species occurrence trends varied strongly between and within the higher taxonomic groups considered. Effect of PAs was broadly positive for mammals and phytoplankton, with a large proportion of species showing more positive trends inside than outside of PAs. For plants, we found no measurable effect of PAs, with most species showing no difference in trends within and outside of PAs. Conversely, birds had more negative occurrence trend inside than outside of PAs. Ecology and life-history strategies of the contrasting species and higher groups likely drive the observed patterns. These contrasting responses to conservation highlight the importance of considering multiple and contrasting taxonomic groups when assessing the potential effects of PAs on biodiversity.