The relatively recent development of high-throughput sequencing (HTS) techniques has revealed a wealth of novel sequences found in very low abundance: the rare biosphere. Today, most studies of diversity of microorganisms are carried out almost exclusively with HTS techniques. However, culturing seems indispensable for diversity studies especially if the aim is exploring the rare biosphere. We have carried out a deep (1 million sequences per sample) pyrosequencing analysis of two marine bacterial samples and isolated a culture collection from one of them. We have shown that the collectors curves of the pyrosequencing data were close to reaching an asymptote and estimated the total sample richness and sequencing effort necessary to obtain given fractions of the total richness experimentally. Comparing the pyrosequencing data and the isolate sequences we found that isolation retrieves some of the rarest taxa and that the composition of rare taxa follows an annual succession. We have also shown that increasing the number of tags sequenced would slowly uncover the isolated bacteria however, even if the whole bacterial diversity were found by sequencing, culturing would still be essential for the study of marine bacterial communities, especially if the target is the rare biosphere.