The formation of CaCO3 in the ocean is an important process affecting the exchange of CO2 with the atmosphere, especially on longer timescales. Formation of CaCO3 in the largest part of the open ocean is mostly done by planktonic organisms, coccolithophores, foraminifera and to lesser degree pteropods. CaCO3 that is formed near the ocean surface mostly sinks through the water column, often together with organic debris and dissolves at depth where the solubility product of CaCO3 in its two main crystal forms (calcite and aragonite) increases due to the increase in pressure. Formation and dissolution of CaCO3 affect both dissolved inorganic carbon and total alkalinity and hence CO2. For this reason, biogeochemical ecosystem models often include the cycling of CaCO3. Here the production of CaCO3 is analyzed in the global biogeochemical model REcoM. In REcoM, biogenic CaCO3 production is limited to phytoplankton only and is strictly proportional to the growth of the small phytoplankton model compartment. It is assumed that, a constant fraction of the small phytoplankton consists of coccolithophores and that these have a constant ratio between CaCO3 and organic carbon. When the model is integrated over a thousand years under a monthly climatological forcing, REcoM output shows that with these assumptions, the alkalinity begins to deviate strongly with respect to the available observations. For the standard set of parameters of the ecosystem model, a transfer of alkalinity is identified from the ocean surface to the deep ocean, mostly in the depth range between 1700 m to 2000 m. Changes in the freshwater or salinity distribution but more likely an unrealistic distribution of CaCO3 production and dissolution in the model are possible explanations. In this study, it is tested that whether changes in the parameterization of the CaCO3:POC ratio in the model can improve the alkalinity distribution. A simple reduction in the ratio leads to a much improved alkalinity distribution but with an export of CaCO3 of 0.343 Pg C yr^-1 which is slightly lower than the estimation of Berelson et al.,2007. Furthermore, a more complicated dependency of the CaCO3 : POC production rate on temperature, nutrients and biomass suggested by Aumont and Bopp,2006 does not lead to a significant improvement on alkalinity distribution. Formation of CaCO3 in the form of aragonite and dependency of the dissolution of CaCO3 on seawater saturation state are neglected in REcoM.