Widespread extinction is a critical threat to biodiversity and is largely caused by human overexploitation of habitat and populations. A widely used and hence well studied organism group for indication of biodiversity is birds. In Europe especially, farmland specialists have suffered from intensified agricultural practices such as increase of monoculture, use of pesticides, and heavy machinery. This has been shown to be partly caused by an EU legislation called the Common Agricultural Policy (CAP). A specific type of farmland, termed High Nature Value (HNV) farmland, seems to be particularly advantageous for farmland specialist birds and makes up an important conservation target. The current study was done in the Natura2000 reserve Târnava Mare, Romania, to find out which habitat types play an essential role for occurrence of farmland species. Farmlands in Târnava Mare are highly diverse in structure, characterizing a mosaic of grassland, meadows and fields, and low-intensity farming practices. With bird point count survey data from 2015 to 2019, I evaluated the effect of different habitat types on five species listed in the Farmland Bird Indicator (FBI) and as farmland specialists: red-backed shrike (Lanius collurio), yellowhammer (Emberiza citrinella), Eurasian skylark (Alauda arvensis), Eurasian tree sparrow (Passer montanus), and common whitethroat (Sylvia communis). I compared habitat proportion in presence and absence of the species for 2019’s data with Mann-Whitney tests. They all showed significant results for meadow proportion. All species except the common whitethroat showed significant results for crop proportion, while only two species (red-backed shrike and yellowhammer) showed significant results for scrub. Independent of which habitat type was tested (meadow, crop or scrub), all species with significant result – except for Eurasian skylark with a negative relationship in crop habitat - showed a positive response to a higher proportion of the tested habitat. The same species except Eurasian tree sparrow were modelled with the generalized N-mixture model of Dail and Madsen (2011) to evaluate what is influencing abundance, recruitment rates, survival probabilities and detectability over five years. The day of the season affected the detectability of almost all species. The effect of habitat on recruitment rate and survival probability, however, could only be shown for yellowhammer. For the latter, proportion of meadow affected recruitment and proportion of reed affected survival. In conclusion, the presence of species seems to be generally higher in habitats associated with low-intensity farming on the single season scale. Over time, however, a significant effect on population dynamic parameters for the same species could not be shown for most species. This could be a result of insufficient data for each year, too few years of data, or that the tested habitat types are not affecting these parameters over time.